Hetzner Price Adjustment | Hacker News Reader

Hetzner Price Adjustment

docs.hetzner.com

282 pointsgmemstr302 comments

Its fascinating that electricity is really the biggest cost of computing. I put a server in my house which was supposed to save money, but actually the electricity cost was something I hadn't thought of and really adds up. esp when the AC has to run to remove the extra heat. So much old hardware out there but its going to get junked as new hw is more efficient.

It's like with refrigerators - when some years ago it was both cost-efficient and even 'green' to scrap working refrigerators and replace them with new ones, because the difference in improved energy efficiency meant that the electricity cost savings outweighed the cost of the new refrigerator already within a year or two and the reduced pollution from less electricity (at the time, fossil fuels) outweighed the environmental cost/waste of making that new refrigerator.

Not the first time I see someone saying this for refrigerators and I don't understand how it could be possible.

From what I can find even in extreme case it's not true. Even with a fridge from 1972 which averaged 1726kw/h (Source: https://www.annualreviews.org/doi/10.1146/annurev.energy.27....) to a more recent one which will probably be around 400kwh. And with the kwh at 0.15$ (https://fred.stlouisfed.org/series/APU000072610).

You obtain a reduction of (1726-400) * 0.15 ~= 200$ a year.

I believe the average refrigerator price in the USA is over 400$ so I can't understand where the 2 years figure would come from without even talking about trying to supposedly be "green" ... And after 1988 you are down to 800kwh so you would gain something like 60$ a year.

You can get some pretty cheap fridges https://www.bestbuy.com/site/insignia-10-cu-ft-top-freezer-r... - but as you suspect it's post-facto justification of a purchase someone wanted to make anyway.

Once you account for manufacturing and disposal, running the older is probably almost always better.

Yes and it even has a lower energy usage : ~300kw/h :). But since you probably upgraded from a similar one then you are probably more inline with the 60$ I mention at the end and not 200$.

From what I understand high end fridge are often less efficient because of features like ice fountain or freezer in the lower part with drawers instead of doors.

Anyway I'm thinking it's probably not just post-facto justification because it looks too much like a seller pitch and wondering where it comes from. I'm French and I had never heard it.

1. Electricity in other countries is more expensive. e.g. in Ireland it is €0.35/kwh (~$0.40).

2. Other countries have smaller houses, therefore smaller kitchens therefore smaller refrigerators. Between that and higher energy efficiency standards in general, new refrigerators in my local store vary from 230kwh (A energy rating) to 260kwh (F energy rating) in power usage.

For your second point it probably applied 10/20 years ago too, then it means you have an even smaller budget to make some economy.

So even when doubling electricity price I fail to see how it could be possible to reach the two years originally mentioned.

And I believe each time I have seen this 2 years figure mentioned it was probably USA centric conversation. I'm French and I had never heard of it before.

$0.15 is laughably cheap for some parts of the country... Here in California it's not uncommon to pay 3x that, so $600/yr in electricity saved by a $400 appliance is a huge deal.

Even here in expensive Seattle our top rate (when you go above 500kWh a month, which is $0.1056/kWh) is $0.1307 per kWh. If you go just north the top rate falls to $0.10/kWh, and many PUDs in the state charge around half that price.

I plugged my parents 1960 fridge into a kilowatt and it used so little power I had to leave it plugged in for several days to make sure I got a reasonable reading. Something like 2 kwhr per day. So it would take about five years for a replacement to pay for itself if the replacement ran on nothing but happy thoughts.

Folks make these same sort of claims on high-efficiency HVAC systems and I never get them to pencil out. The higher efficiency doesn’t pay for itself even over the entire lifespan of the system.

Inevitably the people making these claims sell HVAC systems.

In Germany, my rule of thumb was that a constant 1W load is 1€/year. This year, I had to update it, it is now more something like 3€/year. The increase is so important I am still not sure if my calculations are right!

   1W x 24h x 365 = 8760Wh
   My current price: 0.33€/kWh (100% renewable)
   8.76kWh x 0.33€/kWh = 2.89€

You can use Nord Pool data for the last couple of years to do estimations.

https://www.nordpoolgroup.com/en/Market-data1/Dayahead/Area-...

You have day-ahead pricing at home? Please tell me how

Spot price (+margin) contracts became common in Finland at least some years ago. Most energy companies provide them now. The margin is something like 0.30–0.40 cent/kWh.

We buy energy from any Finnish seller with contract separate from transfer of electricity (a natural monopoly).

That margin is more than many commercial and industrial customers pay in Washington State.

In the cheaper public utility districts here, the per kWh price is only double those rates, with no time of day pricing or similar shenanigans.

The poster said it's an extra 0.4 cents/kWh, not 0.4 USD/kWh.

Our contract in Sweden is NordPool market price + 0,006 USD/kWh. Typical electricity prices in Sweden is anything from 0,001 USD/kWh to 0.5 USD/kWh, it varies widely day by day. This is what enables people to be able to save by using electricity when it's almost free.

Yes, you just purchase your electricity with a market price contract (usually with added margin) and every day the next day prices are published around 3pm Finnish time (perhaps 12 UTC). Then just use the free APIs to load data and use the data to adjust when you spend electricity on what. E.g. to charge EV on cheapest hours, or load batteries, or when you fire up your electric Sauna...

These days the price is different for every hour, in near future they move to 1/4 hour pricing.

About 10% of finnish household contracts are market price contracts. Most of the contracts are kinda continuous (price changes every 12 months) or fixed (e.g. a two year fixed price contract).

This is also common in Denmark. You get the new estimated hour-by-hour rates around noon the day before.

I actually made a little calendar subscription service[1] to help our household plan when to run the dishwasher, drier etc. and we've saved about 50% this quarter despite the rising electricity costs. You can usually tell when it's cheap/almost free too just by sticking your head out the window, since it's mostly always the cheapest on especially windy days.

[1] https://el.simongray.dk/

It will depend on where you live, very common in Sweden with many companies offering it.

I use $1 USD per watt per year, with electric prices around $0.10/kWh. Good enough to inform decisions like whether to leave a light on all the time ($5/year) or whether to purchase a NAS vs. used server ($40 vs. $300 per year).

I guess my prices are actually a little over $0.10 and a watt over a year is a little under 10,000Wh as shown in your calculations, so those offset to make the estimate even more accurate (than necessary).

In sf electricity is closer to 4x that rate.

I just signed a new contract at 18.9¢/kWh + delivery. That's up from last year's contract rate of 8.1¢/kWh.

With the way the world is today, we'll all be paying SF prices soon.

California's power market is heavily distorted by the private utilities needing to make a profit, extreme cost of living, and a need to import power from other states.

if you're a PG&E customer, at top tier it's well over 4x, and closer to 5x, and since so many people WFH now, many more are paying in the top tier.

This lines up with my experience, I use ~2/3€ per year per watt

You don’t get 100% renewable in Germany. The mix is 50% renewables, the rest is nuclear, coal and natural gas.

With nuclear being phased out for ideological reasons and natural gas being scarse due to the war in Ukraine, Germany will soon burn more coal than ever.

New electricity contracts already go as high as 70 Cents/kWh thanks to the misguided energy policy in Germany.

It's possible to pay extra for only renewable electricity.

No you can't (unless you do it yourself but the price is extremely high if you want uninterrupted power). What happens is: your provider purchase "green energy certificates" which means that some green electricity was generated on a sunny and/or windy day. When it's night and the wind don't blow, you use coal and gas as everyone else.

No storage is cheap enough to store solar and wind energy on day-to-day scale, let alone seasonally.

Please don't let the brainwashing gets to you and inform yourself.

There's a massive business opportunity for someone willing to produce low density batteries for home use.

In a lot of countries customers are billed significantly less during the night than during the day for electricity, so the incentives already exist to make it financially worthwhile if a home battery pack that can provide a days typical use can be installed at a reasonable cost.

This would substantially smooth out base load requirements and mean that renewables could make up substantially more of the energy mix.

> There's a massive business opportunity for someone willing to produce low density batteries for home use.

Can you explain why, please? I'm genuinely interested. I don't understand how this would be anything other than rich prepper niche market.

Installing batteries in all homes is going to be more expensive than grid storage, since you will need to pay more in total installation costs rather than installing ginormous grid batteries (or reversible dams, where feasible) in bulk.

My thinking on this is that if you don't need to pack the energy storage into a small volume (like in a vehicle), and if weight is less critically important (as it'll occupy space in e.g. a garage), it opens up a lot more options for storage medium (as an example, iron-flow packs).

At scale the costs could come down quite substantially given that the basic chemistry is iron, salt, water and little else - at which point it's a substantial financial boon to the customer given that they could see an up to 50% reduction in their energy costs (going by tariffs in my own region) to compensate.

Couldn't the grid install these somewhere for less than every home doing their own installation?

The net effect would be the same but might be cheaper overall.

My point exactly.

Iron flow sounds great though. Thanks for mentioning, I'll look into it.

It could and it might even be more efficient. That doesn't mean home installation is not a huge market though. It also has access to different capital.

> It also has access to different capital

This is what I think the real benefit is, and also why I think we'll see both. I don't see the political will or financial commitment materialising for large scale energy storage in the short term because it's an upfront cost to a taxpayer whichever way you slice it.

Individuals are free to self-fund their own battery pack installations though, and can then decide if the upfront cost is worth the reduction in rates to shift energy usage to non-peak times. Once that sufficiently takes off it becomes easier to sell the public at large on the idea, and will also drive down costs, at which point I could see wide-scale grid storage becoming thing. At that point, renewables are scalable to almost 100% of energy usage (assuming vehicles and industry transition away from combustion machinery).

When it's night and the wind don't blow, you use coal and gas as everyone else.

Since this thread is about Germany, and not everyone on HN reads threads in scope, it's worth pointing out that this is not universal.

In most of Texas, for example, the wind blows most fiercely — and regularly — at night. Solar power contributes to the grid during the day, and wind power picks up that generation loss when the sun goes down. I once saw a graph of it from a utility regulator published in a newspaper.

Not every geography is so lucky in this manner. But because this is the global internet, it's important to point out that solar and wind aren't exclusively daytime sources on most of the planet.

What Germany needs is a source of wind that works at night. I've read that Spain imports energy from North Africa. Perhaps a more robust link in that direction is the solution.

Wind blows over the north sea day and night.

There are several power companies in Germany that are offering 100% green power after X years, in the sense that they're committing to constructing or buying new green power plants with a generating capacity at least as high as the amount sold X years ago.

The actual power from your outlet is going to be the same no matter which company you buy from, but it's still buying "green power" in a more meaningful way than a zero-sum relabeling.

> No you can't (unless you do it yourself but the price is extremely high if you want uninterrupted power). What happens is: your provider purchase "green energy certificates" which means that some green electricity was generated on a sunny and/or windy day. When it's night and the wind don't blow, you use coal and gas as everyone else.

that is bullshit. also besides that germany is small compared to other countries and dense, there is no time of the day where its nowhere windy nor sunny. so at the moment there is just no need to purchase green energy certificates for the bigger providers, they can either buy power from other providers or they can produce their green power. and also it does not matter, since if you need 3kW and they can't deliver green power at the moment they will deliver more green power later, thats how it usually works.

> No storage is cheap enough to store solar and wind energy on day-to-day scale, let alone seasonally.

also bullshit. it has nothing to do with cost, more with there is no storage for everybody.

most providers that offer 100% green power actually do generate the amount of power in green that you use (that is their premise) of course they can't create different power lines, thus it's a mix.

(in germany it's forbidden to mark your product as "100% green power" if you buy certificates, of course you can write it some where but if your contract states it, it would be illegal, but there is a difference between contract and marketing, of course... Energiewirtschaftsgesetz - EnWG §42)

I'm curious what that actually accomplishes in non marketing terms. That just means someone else is getting less renewable electricity, right?

It depends on what other market forces there are. If the utility has a government mandate to install more renewable energy than there are renewable energy customers, then it is like you say - just a book-keeping exercise.

On the other hand, if there is no government mandate, or if customer demand for renewable energy exceeds that mandate, then it provides a profit incentive for the utility to install/purchase more renewable energy.

One of the British '100% renewable' companies used to buy renewable power and then destroy their surplus resellable renewable obligation certificates instead of selling them (each supplier was obliged to hold certificates covering x% of their supply, where x << 100). That had the consequence that the renewable power generated was genuinely additional, as the other generators would have to get their renewable obligation certificates from new generation rather than the secondary market, at the margin.

They stopped, publicly reasoning that the proceeds from selling certs could be used to build their own additional renewable generation. I suspect it was just leaving too much money on the table, though.

Some providers promise to build new renewable capacity that matches the consumption of their customers (not sure about the timeframe). I'm not aware of any provider that builds the required storage too, though

Well, it makes a difference what kind of energy supplier turns a profit from your energy consumption. If you're buying renewable energy from a company that still invests into fossil fuels, it's fairly useless. If you buy it from a company that mainly builds up renewable energy production capacity, that helps the transition.

Of course, even a 100% renewable supplier needs to buy from the market to smooth out peaks in energy consumption for their customer base.

(disclaimer: i work for such a renewable energy supplier)

There isn't enough renewable energy yet to cover all consumption. Which means that you paying extra is only an administrative move. You're just swapping places with the party that would otherwise have consumed the renewable energy.

You’re financing the further development of renewables.

The rush out of nuclear fission power still shocks me.

It's a miracle fuel. It really is wonderful. It solves climate change almost by itself, or it would have, had Chernobyl and Fukushima not been reported so irresponsibly.

It seems like that might be changing again though, now that non-renewable energy prices are so high.

"The history behind Germany's nuclear phase-out": https://www.cleanenergywire.org/factsheets/history-behind-ge...

It would be miracle fuel if any country in the world would have solved the waste problem by now. Finland is close, but still not done.

Russia recycles nuclear waste into nuclear fuel.

Research what France does. They recycle the waste because it is not waste, it just needs processing and can be used again. https://www.iaea.org/newscenter/news/frances-efficiency-in-t...

They're not reducing their waste to zero, so that's not a full solution to the waste problem.

Well, my home is powered with a miracle then.

https://www.world-nuclear-news.org/Articles/BN-800-running-o...

> BN-800 running on 60% MOX

> The reactor core itself was refuelled and is now using 60% mixed uranium-plutonium oxide fuel, supplied by TVEL and manufactured at its Zheleznogorsk site. The plutonium for this was produced from uranium during the operation of other nuclear power plants and recovered from the used fuel assemblies through reprocessing.

No one knows where to put the trash. That was the major reason for the phase-out, that no one knew how to handle that. I'm just the messenger, don't shoot me.

And please do not come up with future inventions that will magically remove that issue, that does not help.

Btw, it wasn't really a rush, it was started more than 30 years ago by not building new nuclear power plants. Many of the power plants are actually at the end of their lifetime anyway.

I'm almost anxious to ask, I know very little about the topic, but I've been wondering if it wasn't a viable option to just shoot the waste into space? I know it's prohibitively expensive to do that as of now, but seeing what they're doing with Spinlaunch..

I’m sure someone here can do the math, but it takes a lot of energy to get things not just in space, but out of the Earth’s gravity. You don’t want that nuclear waste raining back down. It’s far easier, cheaper, and more fool proof to just bury it deep.

In addition to the raw costs of moving anything into space, there is the contingency cost of what happens if the launch goes wrong. A new satellite blowing up on the launchpad is bad, but a cargo of nuclear waste exploding a few seconds after liftoff is catastrophic.

True, that sounds reasonable..

That's not true. Nuclear waste is so minimal in volume nuclear power plants can just sweep it under the carpet. Or more specifically, just store it on-site. Modern nuclear reactors can even use the waste of old reactors as fuel and reduce it by another 80% or so. Dealing with nuclear waste is really no big deal. At least for medium term storage.

Storing waste so you don't have to look after it for thousands of years is admittedly more difficult. Then again, future generations are always impacted by the choices we make today and I'm pretty sure the kids 300 years from now would rather deal with the cleanup of our nuclear waste than with the consequence of global warming.

> No one knows where to put the trash

They do, though. That is where the “trash” is. They have become so adept at it, they made a museum out of one facility[0].

Two misconceptions that benefit from being clarified:

1. There is very little waste for the amount of energy produced. If we only used nuclear, it would amount for 40 grams per person per year[1].

2. The waste is not trash, it is extremely valuable. It was valuable in the ground, and it is still valuable out of the ground; just not for the particular reactor that it came from.

[0]: https://www.covra.nl/en/radioactive-waste/the-art-of-preserv...

[1]: https://whatisnuclear.com/factoids.html

You're talking about fuel, and we're talking about trash.

From refinement to the reactor and spent fuel processing, the nuclear power industry generates a huge amount of trash that is low-grade radioactive and thus cannot be just dumped in a landfill.

The nuclear industry has had over seventy years to get their act together regarding fuel recycling. They'd rather just dump it into pools on-site because it's cheapest and makes it Someone Else's Problem. It's the same attitude of the chemical industry from the 1950's: dump it in a hole and let someone else deal with it after you've long since left.

Here's an idea: let's spend the money on wind turbine blade, battery, and solar panel repurposing and recycling instead.

What is the kill count for this low-grade radioactive “trash” per TJ? What is the volume per TJ?

The coal industry has convinced the ecologists to require titanic standards from literally every green energy source, especially the one with the most stable output.

> Here's an idea: let's spend the money on wind turbine blade, battery, and solar panel repurposing and recycling instead.

I am into doing that, AND the nuclear stuff. No need to tie an anchor to our ankle while we try to metaphorically swim back to the surface.

> Here's an idea: let's spend the money on wind turbine blade, battery, and solar panel repurposing and recycling instead.

The two is not exclusive. You do need stable base energy production and none of the renewables solve it well. Hydro could do it, but the locations where it is appropriate are few and it has quite a big ecological impact in and of itself.

I think the two are not exclusive and IMHO a big part could be these small-nuclear-reactors coming up.

I think renewables are very important, but just this week was a comment about Netherlands (I think) where the comparison was 200sqm against 4400 wind turbines and somewhat 20sqkm if you compare nuclear vs. wind/solar.

If you take a look at Switzerland, I have no clue where to put 4k wind turbines. Hell, even 1000 won't have enough space (or you will waste so much forests or agrar or you'll put them into alps and waste the scenery, which will break tourism).

Burning coal on the other hand has a waste disposal mechanism built in! The carbon atoms are converted into CO2, a harmless gas that we can release into the atmosphere without any consequences.

> No one knows where to put the trash. That was the major reason for the phase-out, that no one knew how to handle that.

We know so well how to handle it that we are wondering how to keep people out of the storage places in the event of a civilization collapse. That is, where current languages are forgotten and we forget what nuclear is. That's how advanced in storage we are.

It solves climate change almost by itself, or it would have, had Chernobyl and Fukushima

Every energy source is the perfect energy source, if you choose to ignore its problems. Just like every operating system is the perfect operating system if you overlook its flaws.

Messing with atoms is better than burning fossil fuels, but let's not pretend that it's perfect.

not been reported so irresponsibly.

"Blame the messenger" is an ancient, and irresponsible game.

Blame the messenger is perfectly valid if the message is not delivered accurately. I think the qualms people had with the reporting of nuclear plant problems is that they were overblown and scared people away from nuclear.

If you ignore the radioactive waste, it really is wonderful, but we still don't have a single final depot in Germany for it and it takes some 100k years to get rid of it the natural way.

"Takes a long time to become progressively less harmful" is, if you think about it, an improvement over "never really gets any less bad".

Meanwhile France, a famously pro-nuclear country, has to shut down many of its nuclear power plants in summer because they can't cool them adequately.

(Yes, I'm sure issues can be fixed. Point is, nuclear has an awful lot of practical problems for supposedly being miraculous.)

No, they didn't fix it but is has gone better. Today 25338 MW from nuclear. A year ago 37119 MW from nuclear. So almost 1/3rd down.

Source: https://www.rte-france.com/en/eco2mix/power-generation-energ...

Yeah, I'm French and I'm a bit sad that people seems to deal with nuclear energy as if it was some kind a religious thing. Both sides, anti nuclear who seem to ignore the great benefits it can bring and the pro nuclear who can't seem to see how it sucks (not because it's dangerous or not green, but because it's expensive and doesn't work half of the time because of maintenance or heat - oh and also because we cannot build new reactor in reasonable time and budget anymore )

It still produces plenty of energy, and that much more would you have to create mostly from coal otherwise.

With nuclear being 6.3 percent and the rest of this half of the circle being coal and a bit of gas.

It's not 100% renewable in the Mix. Thats not what he said. It just means, that he picked a supplier, which exclusively buys renewable energy on the energy market. Obviously it's still the same energy mix, which arrives on your house. But your money won't support Coal Power Plants.

My rule was something like 2Eur /10W/Month(~2A 5V Raspi) which comes down to 2,4Eur/W/Y.

Just for my curiosity though, Where did your 1Eur come from? The prices were >20ct for 10 years now. Or do you calculate without taxes and stuff?

I suppose it was a "left over" from my time in France where the price was just above 10ct/kWh.

Electricity is a significant opex for many products including Agriculture. For example, I've seen Farmers cost as high as 50% for power, generally to move or pump water

There's a reason used servers are so cheap on ebay: if you run them 24/7 their electricity cost makes them unattractive compared to new hardware (often within a year).

Highly disagree.

Said in another comment, my server averages around 100-200W. It's equivalent to running a gaming PC.

The reason there's a lot of cheap servers on eBay is that their support contracts expired and are absolutely ridiculously expensive to maintain for a big company. It's easier to get a new server (which will likely have performance and efficiency upgrades as well) than to continue maintaining a server that no longer has a support contract.

The reason there's not many servers in people's hands is size and noise. Servers aren't meant to go in a house; most people don't have server racks in their home. And by default, the fans are jet engines, even behind closed doors.

You probably live in the USA, because 200w 24/7 would cost me about 600€/year which is definitely significant.

one of the upsides to living in texas, lots of cheap wind power lol. even with more expensive electricity i can still run old hardware for cheap. means less e waste also.

https://app.electricitymaps.com/zone/US-TEX-ERCO

Your cheap TX electricity is, unsurprisingly, mostly gas. And if you look at data from 12h ago it was more coal than wind.

Yearly averages are probably the most sensible to look at. Wind is at 20% which is still nothing to sneeze at:

> According to ERCOT, nearly half of Texas’ electricity was generated by natural gas-fired power plants in 2019. Coal-fired plants and wind power each generated about 20 percent, while the state’s two nuclear power plants — the South Texas Project near Bay City and Comanche Peak near Glen Rose — supplied a total of 11 percent. Solar, hydroelectric and biomass resources provided most of the remainder

https://comptroller.texas.gov/economy/fiscal-notes/2020/augu...

I live in the USA - California, and 150W 24/x would cost about $550/yr. In fact it would cost me a lot more, as my family shifted to time-of-use billing to save money (yeah, it means we don't do certain things between 4-8pm, but with some adjustment it's made a difference in our bills).

A fan replacement and activating quiet mode on the server (usually a bios option or sometimes a hardware tweak) helps with the noise a ton at the cost of slightly less thermal ceiling.

That's why I said "by default". I have done the IPMI toggles needed to lower the fans but it's still not quiet by any stretch of the imagination.

They are still a great deal for burstable workloads or to use as a workstation - only run it when needed so no need to run 24/7.

If you need really cheap (and local!) compute for short periods of time, picking up one of those servers can still work out cheaper than renting from a cloud provider.

This is why for my home server I didn't get second hand servers but built the following in a 2U case that fits in my network closet:

Asrock Rack c246 WSI (Mini ITX workstation board) 32GB DDR4 ECC Udimms (16GB x2) Core i3 9100T (25w, low power 4 core part, supports ECC, and quicksync) 6 IronWolf 4TB NAS drives

The 6 drives are in RaidZ2 so give 16TB usable space, and at full pelt the max power consumption is 60W for everything.

My R730 averages at around 100-200W with two 12-core CPUs, 128GB RAM and 8 12-14TB drives.

Servers are more efficient than people give them credit for. It's close to a gaming PC in electricity usage. Certainly not 60W (or what people can get with NUCs and RPi clusters) but for the power I get, it's very much worth it.

I'd guess the issue isn't so much peak consumption but idle.

If we average that out to 150W, that'd be 1250-1300Kwh/yr. Current Bay Area electric prices means that'd cost in excess of $500/yr to run. I'd say that servers are not more efficient than people are giving them credit for...

And at Seattle prices, that's $150, so I don't exactly see what the purpose of your anecdote is other than to mention how pricy your electricity is.

Edit: But even at $500, that's pretty much the price of a low-spec VPS or VM per year. So for a fairly low initial price to buy the server, you're getting far more performance for the same price you'd pay someone else to use theirs.

Servers are more efficient when fully utilized. A commercial grade server in a homelab is more likely to idle all the time, making it very energy wasteful. Servers need to be right-sized, instead of "more is better".

What's it like at idle though?

Since it's a home server, when it's not receiving backups from computers in the house, or streaming media via Plex it's sitting there at < 15w, which is low enough not to worry about.

My favourite 'home server' board for a while was J3160DC-ITX, which runs at 6W TDP and runs off a 12V brick.

Well the electricity prices (according to the article) rose 55% in Germany and 83% in Finland, while the server prices are just going up by 10%, so electricity is just a part of it.

Don't use a "server" server, unless you really need that. Use a desktop from a few years ago, from some oem who advertises power efficiency. Stuff it full of hdds and ram.

I've made very good experiences with Fujitsu. The power consumption figures they publish in their energy consumption white papers has been matching my measurements very well, so you can check ebay for used machines that match your performance/price expectations and then google eg "p920 energy white paper" and check the pdf. (Note, some models are available with different PSUs, ie 90+ and 85+.)

My "home server" mostly idles at 14W with one hdd running, one in standby and one ssd.

This is the way. You don't need high density at home (usually) so don't pay the price for it. Get a mobo with ECC ram that fits an ATX case if you want server grade stuff. If you don't use stupid tech like VMware it won't eat your resources either (not that VMware is bad in a datacenter, but it is quite resource hungry in comparison to virt-manager or Proxmox).

I mean unless databases or Java are your favorite things in the world.

Configure power saving as well, that way your CPU really doesn't consume a lot.

> mobo with ECC ram that fits an ATX case

yes, and careful choice of graphics card; GPU computing is special purpose and can be costly in an urban area

Indeed, I don't know how dense the megascalers build, it looks less dense than what I see at work compared to the few videos that are out there.

Same conclusion here. Initially I used an old notebook with USB cases for full-size HDDs, nowadays it'n an older desktop PC in a midi tower, with some extra memory and HDDs. Idle power is <20W, which is most of the time. I really wanted a proper server for the cool factor, but it's just so expensive to maintain.

Were there significant downsides to the laptop solution?

I have a similar setup with a NUC mounting storage from a small ARM-powered NAS, but I was thinking of changing this to an old ThinkPad X230 with just internal storage and offsite backup.

It was a really old beater, and the problems, I think, stemmed from there - think of electrical problems, because of old USB ports and such. Otherwise it was chugging along like a champ. What I like in my current setup more, is that I could upgrade it as my needs increased - in the desktop, I could fit 5 HDDs and 20G RAM, impossible with my old laptop.

Take a look at the Tiny Mini Micro[1] project from Serve the Home for some very enthusiastic reviews of small 1 liter-sized desktop computers from Lenovo, Dell, and HP. I use a tiny Lenovo that's been rock solid, and uses almost no power.

[1] https://www.servethehome.com/introducing-project-tinyminimic...

As a rule of thumb, 1W costs about $1 per year.

I've been wondering over the past few months if we can build a "public" cloud. I have some old machines gathering dust, I'm sure others do too. Is it feasible to build an ad-hoc network of such computers where the compute power is rented out. For ex: I can connect my old lenovo i5 laptop to this network for 6hrs a day and charge 1.2x my electricy cost. Sort of like on demand EC2 instances, but backed by torrent like tech. It'd make computers pay for themselves over time

In order to keep my eletricity bill for a running system in my apartment to then minimum I researched the specs & prices from the "The Power Consumption Database"[1] and found an intel nuc running on 5watts idle. Thats the same class of power consumption of a raspi 4 8GB, but easier to maintain. Now I'm waiting for my a Coral TPU USB-Accelarator to keep the peak consumption down for some recognition tasks in homeassistant.

[1]http://www.tpcdb.com/list.php?type=13

While the pricing is interesting, I think the fact it is driven by the input costs of electricity is even more telling.

The fact that electricity costs is the dominating (edit, correction: marginal) cost of cloud computing basically means that:

1) We're up against physical barriers to lowering the cost (and ecological impact) of our clouds.

2) Optimizing for less compute will pay multi-fold dividends over the next few years where we can expect energy prices to continue rising.

3) Building more power-efficient chips is a big deal, and I'm glad to see more general ARM deployment, given their efficiency gains.

To be fair, Hetzner is notorious for pushing down prices for everything as far down as they can by any means necessary. So this most likely makes their price point more sensitive than many other providers to any cost increase. Here is a great walk-through of one of their data centres [1] and you will see how streamlined everything is down to the hardware components. Nearly nothing goes to waste.

[1]: https://yewtu.be/watch?v=5eo8nz_niiM

I mean if your competition is Amazon I image you'd have to be like that.

Their competition charges XXX% markups on everything but EC2...

EC2, particularly traffic, has massive markups too.

EC2 is definitely at a premium. If you are pricing CPU and RAM, the big three cloud providers are near the top of the pricing continuum. Hetzner is usually at or near the bottom end of the pricing continuum.

If you are including the value-add of other cloud features, then it is not a like-for-like comparison.

Amazon's margins are astronomical. Hetzner doesn't need to give up any where near as much consumer surplus as they do in order to compete on price, but having a solid offering at very low price point is how they have chosen to differentiate themselves from other providers on the VPS to cloud spectrum. Arguably they are competing more against Linode, OVH, digital ocean etc..

Massive respect to them for linking the change to electricity prices rather than just "inflation". They know that by doing this there will be pressure to reverse the increase once electricity prices normalise.

For sure, and I agree, but that's underscoring my point.

The marginal cost of cloud computing isn't capex, it's opex. And of that opex, it's mostly electricity. We aren't paying for the hardware so much as we're paying for the cost of keeping it turned on and cool.

Why "notorious" though? Isn't this a good thing for them and for customers as well?

I would expect "famous", since this seems lika a feat/achievement to me.

"Notorious" isn't always negative—the connotations of the word depend on context (and probably regional use?) in sometimes confusing ways.

Indeed! It's a really interesting usage. Maybe it developed in a similar way as "badass" to convey a sense of (positive, approved) vaguely dangerous wildness and coolness.

It kind of makes sense here -- Hetzner's competitors probably hate this "notorious" behavior of being so good at ruthlessly[1] slashing prices

[1] See, another example of this sort of thing. Cool!

Interesting, didn't know that. Thanks for elaborating!

You still don't! I was speculating :)

Those consumer PCs don't look too energy-efficient, though.

Wow, that's really cool. Love the old midi-towers.

Hetzner has a data center in Finland as well. While the electricity price in Finland is going up, it is not going up as much as in Germany due to availability of nuclear power. If you want to use Hetzner services I recommend picking a location in Finland.

How is the peering to Finland nowadays? Last I heard it was all still going through their German centres.

From Telekom Germany:

Telekom -> Level3 (*.edge1.helsinki1.level3.net et al.) -> Hetzner Helsinki (core*.hel1.hetzner.com)

> Optimizing for less compute will pay multi-fold dividends over the next few years where we can expect energy prices to continue rising.

I suspect this is already happening in places like Google and paying significant dividends given the scale of code replication.

Or people will separate out heavy tasks that aren’t time critical to run them when power is abundant (when there’s wind or sun) and not when it’s scarce.

People do.

https://www.computerweekly.com/news/252481957/Google-creates...

Some of the hyperscalers area already experimenting with this - https://blog.google/inside-google/infrastructure/data-center...

ARM on desktop => ARM for development => ARM on servers => ARM on cloud.

Apple's multi-billion contracts with AWS are ending this or next year, Apple has their own power efficient systems and will compete with AWS, as they have to maintain/increase growth.

> Apple has their own power efficient systems and will compete with AWS

Are you saying Apple is going to roll out an AWS competitor to the public? I find that highly unlikely.

What other verticals does it have to seek out next, if growth is demanded by shareholders?

There's all kinds of consumer-oriented things they can expand into, like they're currently doing with sports on ATV+. I don't see how competing with AWS could ever work out for them, absent some anti-competitive stuff like allowing additional functionality for apps that use AppleCloud for their backend.

Apple shareholders aren’t demanding growth - they’re doing fine and given Apple’s history it’d make more sense to expand their consumer footprint, especially internationally, since that’s an area where Apple has considerable strength.

Cloud computing is also a hard field to move into: you need a significant amount of capital up front but customers will be slow to migrate. Google is struggling, I’m not sure what pitch Apple would come up with which is obviously better.

But you’ll only be allowed to deploy code that’s been reviewed by Apple.

That would be an interesting restriction/value-add. So many things block cloud IPs due to the heavy abuse they get from them. Apple could offer high deliverability IPs and block clearing as services.

This does seem to be the normal course for services companies, and that's where Apple is headed.

I dunno. I could see it... As long as you access the "iCloud Web Service" from an iDevice or a Mac there will be loads of useful and functional tools and features for you to take advantage of. It'll be 4x more expensive than AWS but the integration with Apple's ecosystem will be fantastic!

If you try to access it from a PC or an Android device though you'll have your bandwidth throttled and all your images will be reduced and compressed like it's 1999.

I wouldn't be too surprised. They already have or are rolling out cloud development environments with XCode cloud.

They don't have anywhere close to the capacity to produce the necessary amount of silicon

They used aws, gcp, azure.

The aws contract will end this or next year, I don’t recall.

They’re offering Xcode cloud build.

It’s only logical for them to also grab a piece of market share of a highly profitable / high margin market.

They’re doing the same with EV, AR, etc.

I am long apple

> The fact that electricity costs is the dominating cost of cloud computing

In the article, they say that electricity has gone from .29 to .45 euros, a 55% increase. If electricity was the dominating cost, seems like they would have raised the price way more than 10%.

The article didn't seem to say how much they're raising prices, only that they're raising prices due to electricity.

Where are you pulling seeing the price change %? I may need to edit that to marginal, if it's only 10%. :)

The article gives the prices in a table in you scroll down a bit :

electricity/air conditioning rate old price* new price*\

Germany € 0.29/kWh € 0.45/kWh

Finland € 0.12/kWh € 0.22/kWh

That's the price for electricity/air conditioning (presumably for Hetzner), not how much their products will rise in price.

It is the electricity/air rate for their colocation service

> we are forced to increase the prices of many of our products by approximately 10%

First paragraph: "we are forced to increase the prices of many of our products by approximately 10%"

Missed that. Thanks!

Assuming that the 10% price raise is equal to the increase of the electricity cost, that implies that after the increase the cost of the electricity represents slightly more than 25% of the price (which must also include the profit, besides the other costs).

So the electricity is not really dominating the costs, but it might still be the largest component of the cost.

There is an estimate that Amazon has a 61% profit margin for cloud services. Hetzner must have a lower profit margin, but nonetheless the electricity could be as much as half of the costs.

4) Cloud computing will be outsourced to low energy cost countries.

I wonder if latency allows this

It depends on your requirements. Serving user-facing content is likely to be a task best suited to data centers geographically close to the user base. There are plenty of non-user-facing tasks that could be offloaded to another location.

As an example, the product I am developing spends most of its time calling third party APIs to do bulk data extraction and processing. That can happen anywhere. After data processing and some batch ETL work, we surface that to our end users. The output of that ETL needs to go somewhat close to our users, but everything else could easily happen network-seconds away.

With cloud egress cost though, moving the data to a low-cost region could be expensive itself. I wonder when it would pay off.

I am not shipping more than 20TB/host/month: https://www.hetzner.com/cloud

Also not shipping more than unlimited/host/month: https://www.ovhcloud.com/en/vps/

Start with a cloud provider that is better in terms of cost profile, and then move only the stuff that needs wider distribution to a higher cost geo-distributed provider.

4) Cloud computing and the electricity used is impacting housing provision:

https://www.theregister.com/2022/08/01/london_electricity_su...

How many data centers already use waste heat for district heating?

gonna go out on a limb and say 0... made me look though and apparently it is starting to happen.

In March (2022), Fortum and Microsoft announced our joint plan for a ground-breaking data centre region in the Helsinki, Finland metropolitan area. The data centres will use 100% clean electricity and the waste heat they generate will be used to heat homes, as well as business and public premises in the area. I must say I am immensely proud of this project – but I believe it’s only the beginning.

I wouldn't talk about "clouds" as being limited by energy prices, as though they're uniform. Hetzner is much, much cheaper than AWS/Azure/Google. There's a huge amount of fat/gross margin that could be cut from the pricing of the big ones.

It's been like that for a long time (not only for cloud computing, but any big IT infra). Power is everything. To a point, that you measure datacenters in megawatts, when you bring new HW to datacenters you most often talk about improvement in compute per watt, not absolute compute.

It's not only because of the cost of power, but also because power is something that's hard to increase in an existing datacenter - you can put in new servers easily, but putting new power lines and upgrading all the infrastructure is usually just too expensive.

Server pricing, potato pricing, I suppose every product has some energy component baked into the price. Probably more than I might even guess…

High electricity price, and high energy cost in general, is probably the biggest regressive tax that humans face.

Not just in direct energy and gas consumption, but in the baseline increases that flow through to all the essential goods that households purchase.

New energy generation technologies must not only be cleaner, but also more efficient. It is so crucially important to raising standards of living to ensure a reliable supply of cheap energy.

Love Hetzner great to see a hosting company thats being upfront and transparent about its prices.

So tired of hosting companies doing dodgy stuff with their pricing like giving you the first month 99% off but wording it like you will pay that month to month.

AWS has never raised prices on me.

Don't worry they will. At least in Germany :-)

Electricity pricing is insane and will only get worse. I doubt AWS will operate a whole region at a loss.

Amazon invests a lot in its own wind and solar parks. Because of that, energy prices on the open market shouldn't affect them too much.

Interesting, but looking at their map they will go live in Germany in 2025. Which is way too late to ride this out.

Continental Europe has a synchronous power grid, so Amazon can also use electricity produced outside of Germany for its German data centers.

Transit lines even in Germany are already at capacity, and the French are buying up whatever capacity they can because of their nuclear foolishness.

Buuut... Just because you made a good business move and hedged one of your raw materials, doesn't mean your competitors did. If your competitors didn't, then you might as well rise prices and just take the extra as profit.

Or... look at it another way, you bought some nice energy futures contracts when they were cheap. You might earn more selling those energy futures contracts now they're very valuable than you make putting that energy into servers and selling the compute.

AWS EC2 is like 6x the price of Hetzner cloud (4x with reserved instances). I imagine AWS has a lot of room to absorb price swings, if they want to.

Dunno about you, but that tells me something about their profit margins.

They had boutique prices (think 10x Hetzner price) to begin with. And cost control is notoriously difficult.

because everything except the loss leader (EC2) has triple+ digit markup

What is "loss leader"? Do you mean they make a loss on EC2, or that it's the bottom entry if you'd sort by profit? I'd assume the former but making a loss at AWS' price level would be extremely impressive.

Maybe at least increasing energy costs will force people and companies to ditch Electron-based "programs" (or Java/NodeJS in case of servers). Native is much more energy efficient by definition.

Native may also mean the software may not exist, because it’s too expensive, which is more sustainable.

I don’t know about the energy usage of NodeJS but the jvm can be pretty efficient. Switching implementations or tuning an implementation is easier than rewriting an app to native c++ https://dl.acm.org/doi/abs/10.1145/3475716.3475774?sid=SCITR...

first I’ve heard of java not being energy efficient

No. Most people do end user computing (where the Electron-based apps run) on laptops, which, for battery size/weight/runtime purposes, are already power constrained.

The efforts are going into making browser engines and processors more efficient.

> Native is much more energy efficient by definition.

"much more" is not true. It may be much more efficient in some cases, but with modern runtimes for interpreted code for many operations it is very close to native performance.

I have a bunch of servers in an apartment I use as a secondary office space. Monthly electricity cost increased by 230%. It’s like that since January. Hetzner’s 10% seems too low, although the increase is calculated on the total price of the server, so I guess it could be sufficient. I won’t be surprised if they further increase their prices in the near future.

I currently pay less per month to Hetzner for a rented dedicated server (from the second hand "server auction" section) than I would pay in electricity alone if I ran the thing in my home. Yeah, electricity is quite expensive now in Germany and Europe. ^^

I think Germany has the most expensive electricity in the world, and AFAIK that was 2021, so its even worse in 2022.

The result of shutting down nuclear plants. ¯\_(ツ)_/¯

While I'm a fan of nuclear power myself, I don't think this is true. Nuclear is only good for base loads, current high prices are mostly driven by peak demand.

Wouldn't it be the case that the nuclear portion would be price stable, so any fluctuations in say the last 30% of power to reach peak power would be less severe in the final bill?

No, the most expensive power plant defines the price.

In France, yes.

Well the green energy proponents (which we are now learning were on the payroll of Gazprom lobbyists) launched a FUD campaign against nuclear plants which were completely safe and working fine post-Fukushima.

This chess move by Putin is paying dividends because Germany is shifting towards populism/right wing as is the trend in Western Europe (can you really blame Parisians?) and realizing it can't replace Russian gas without incurring serious internal turmoil and economic ruin.

Now we are seeing this impact everything. Even hetzner isn't safe.

  Germany € 0.29/kWh € 0.45/kWh
  Finland € 0.12/kWh € 0.22/kWh

Damn, why is electricity in Germany over twice as expensive as in Finland?

I doubt Germans have twice the purchasing power as Finns.

Politics

Finland has oil. Germany imports oil from Russia.

You're thinking of Norway. Finland does not have any oil reserves at all.

You're right

Only 0.2% of the electricity is generated from oil in Finland. But Finland produces a good quarter of its electricity from nuclear power. Maybe that makes a difference?

It’s hard to export electricity from Finland due to it’s geographic location and it’s main connections are with Sweden where electricity is also much cheaper than in continental Europe.

And the national grids of Sweden and Norway can't really keep up with the current load. The price of electricity up north is dirt cheap due to lots of hydro but the grid just can't get it to where the people are (south). So you get crazy 100x price differences between the two ends of the countries.

This is also why both Sweden and Norway want to invest into new transfer lines to northern Finland as the national grid of Finland is actually in good enough shape that any electricity you get to any part of it can actually be sold to a customer anywhere in the country. This is why Finland is a single price zone when both Sweden and Norway are split into multiple zones (4 and 5)

Gas and oil use for electricity is very low. Electricity is over 85% renewable and nuclear with most of the rest coming as industrial or district heating by products

District heating is quite often coal or oil for now but quickly changing to non carbon forms.

Here is a real time view of the electricity market in Finland https://www.fingrid.fi/en/electricity-market/power-system/

Germany on the other hand bet heavily into gas (and mostly as pipelines to Russia without building any LNG terminals of their own for redundancy) while trying to change over to renewables. Now gas is very expensive and the change over is not finished yet so they have to burn expensive gas and keep their old coal plants running (very expensive in terms of co2 credits)

edit: Electricity is really expensive now in Europe in general due to sanctions on Russia (and their retaliation cuts). Hopefully the new Olkiluoto 3 reactor goes into full production mode before winter to help a bit (it is running at partial power during testing at the moment)

Germany already had the highest electricity prices in the world even when gas was cheap. Green energy has become an ersatz-religion here.

Green energy is in the costs of 8 cents/Wh. Coal is more expansive as gas always was. Only nuclear is calculated with 2 cents/Wh ignoring the costs of the nuclear waste as this paid for by taxes in addition.

Are you missing a 'k' or 'M' on those units? 8 cents per watt-hour is obscenely expensive. A hibernating Mac m1 would cost over a dollar an hour. Just to hibernate.

Thanks for correcting.

Also ignoring the cost of cooling water they take for free from rivers and habitats. As half of the German water consumption are used for cooling gas, coil and atom facilities. As stated in this article.

"In Deutschland zum Beispiel geht etwa die Hälfte des gesamten Wasserverbrauchs für das Kühlen der Atom-, Kohle- und Gaskraftwerke drauf [...]"

https://www.heise.de/tp/features/Krisenstab-eingesetzt-Atomd...

Pretty sure that's included.

Pretty sure? Like it is gut feeling? If you can read German, here you can find an article from the Linke party. There are several of links to papers from the German government to the cost of that mess. I don't want to offend you, but your comment has really no value and also is wrong.

https://www.hubertus-zdebel.de/atommuell-kommt-teuer-offizie...

Nope, if you want to make a true comparison, then nuclear is in the costs of about 40 Cents/kWh instead of the postulated 2 Cents/kWh.

Because as we do not have a final deposit for nuclear waste there are no costs for it. There are only models with suggested costs. But all of this is not priced into it.

A German paper produced by the independent "Wissenschaftlicher Dienst des Bundestages": https://www.bundestag.de/resource/blob/887090/1867659c1d4edc...

There is no way storing the spent fuel would cost so much that it would generate 38c/kWh of costs.

For example the total lifetime costs of Onkalo (Finlands spent nuclear fuel storage site) is estimated to be 3.5 billion.

Lets take the new OL3 reactor as a example. It produces at 1600MW which comes to around 14 million MWh in a year so 14 billion kWh per of electricity produced yearly (slightly less in reality with downtimes for refueling and maintenance). That would mean that if storing the spent fuel would actually cost 38c/kWh just OL3 would generate ~5.3 billion euros worth of fuel storage costs per year which is more then Onkalo lifetime total costs. And OL3 has a planned operational life of 60 years.

And Onkalo is supposed to store all of Finlands already produced spent nuclear fuel and the fuel produced in the next 100 years.

Even if you add dismantling costs of the plants at the end of their lifetime there is no way to get to such crazy numbers even if it costs 2 or 3 billion euros per reactor to rid of. Lets say with 60 year lifetime of OL3 you get around 840 billion kWh produced and decomissions costs 5 billion it would be ~0,6c/kWh of cost. Yes it costs but not tens of cents. For the older plants that produce less it would be around 1c/kWh.

In Finland they started to collect money into a fund for decomissioning and spent fuel storage in the 80s and it currently has a bit over 20 billion euros in assets (the fund invests the money on markets so it fluctuates obviously). So basically they try to do their best to estimate how much money they need to save and collect it during the lifetime. For now it looks like the amount has been roughly correct. And still the for profit companies keep investing into new nuclear plants with this cost already in the price so it doesn't seem to be too much.

That is only right if we completely ignore the cost of the final disposal and such. As always in this calculations they forget to add the cost for the 15.7 billion years of nuclear waste maintenance. If we talk about the cost for the normal consumers (tax payers) this must be included. Don't you think? But if we do the profit talk, you're totally right. What a fantastic business model this is, isn't it?

Nuclear waste doesn’t last 15.7 billion years. It’s radioactive for hundreds to thousands of years. I expect this is why the waste site would cost $3.5B to store forever. Waste isn’t terribly heavy or large so the storage is largely fixed to dispose of many, many reactors.

And there’s a bit of unused space in Finland :)

After a few years the fuel is cool enough to passively cool. At that point you just bury it underground.

The facility (Onkalo) does not need any maintenance once it has been filled up and closed as you just fill the whole thing making access to it very very hard as it is 520m under ground so a couple kilometers of tunnel to dig through to get to it. (clay sediment at the bottom where the vessels are and concrete for the pathway up)

As I said the 3.5 billion is the projected total lifetime costs of the permanent storage site Onkalo. So that is the total cost (to the best of our ability to estimate for it)

What, the nuclear waste handling, long term storage and disposal, as well as disposal of old nuclear plants? It is not really included in any of the price calculations I've seen around Germany, when it comes to what either the plant operators put out (in their business reports), the nuclear lobby is publishing or the politicians are quoting.

One of the most used metrics is the so called Levelized Cost of Electricity (LCOE). The LCOE includes investment, maintenance and operation, fuel and "carbon" costs, as well as waste costs. However, the waste costs are only calculated for the years of operation, so it's not really a great number to look at in terms of actual total cost.

Regardless, the LCOE the German federal government uses is 4-8 cents/kWh for onshore wind, 2-6 for photovoltaic, 14-19 for nuclear.

The (former) nuclear plant operators in Germany are required to pay into a fund that is meant to be used for covering the costs of decommissioning and disposal of the plants, as well as long term storage (except not everything is covered[0]). This is not included in the LCOE. Anyway, operators are required to pay a total of 47 billion EUR into that fund[1].

The old East German plants that got shut down right after the German reunification and are in the process of slowly being deconstructed and disposed of[2] are now estimated to cost 3-7 billion EUR a piece for full disposal (but not storage). If that's any indication, even when considering "lessons learned" will lower costs for future decommissioning and deconstruction, those 47 billion EUR is very unlikely to cover the eventual costs of deconstructing those 20-odd decommissioned German nuclear facilities and store spent fuel and contaminated materials for the first few decades. Some economists from the DIW (the private German Institute of Economic Research, not exactly tree huggers) calculated the eventual cost until 2100 will be closer to 169 billion EUR in 2019s money.

There are some eco lobbying organizations - obviously not without their own bias - that make claims such as "the true cost of nuclear is 42.2 cents/kWh while wind is 8.1 cents/kWH"[3]. I have previously looked through their numbers and argument in more detail, and it sounded mostly very sound to me, however I am a layperson.

Regardless, the fact is: we do not actually know yet the total cost of nuclear power.

[0] E.g. German tax payers have spent 1.5 billion EUR just on exploring potential long term storage sites so far. Just exploring.

[1] This is still litigated a bit, as the German legislature only really had the great idea to make a law about that in 2016. That's another reason nobody historically included it or now includes it in the LCOE.

[2] E.g. the old Greifswald plant. When it operated, it employed 10,000 people. Now 32 years after it officially ceased operation in 1990, it still employs around 1,000 people. Full deconstruction and disposal is now expected to happen by 2028 and will eventually have accumulated costs of a total of 6.5 billion EUR according to the company in charge of the disposal. Not including long term storage costs. Since the operator of those things used to the East Germany, it's now the Federal Republic of Germany's full responsibility to pay for that.

[3] https://www.bund-sh.de/energie/atomkraft/hintergrund/die-wah...

The problem is called Merit-Price-Fixation and is the method used all over Europe to "define" the costs for energy. It is also called Windfall earnings.

I heard that in Austria it is under discussion that this method is not well suited for price fixation. I also heard the first tones here in Germany.

Finland has a lot of nuclear and hydro, which are stable sources. Hydro can get dirt cheap since production is basically constant even if demand is low.

Hydro may be stable, but prices are soaring in Norway as well, despite being mostly hydro. The entire energy market in Europe is having a tough time.

Germany shut down all their nuclear plants about 10 years ago, opting to rely on natural gas from Russia.

Finland invested in nuclear and is bringing their latest plant online this fall.

It doesn't take a genius to see which one would be hit harder by sanctions on Russian gas.

It didn't take a genius even 10 years ago. That decision to get out of nuclear power, based only on emotions, was a grave mistake.

> Germany shut down all their nuclear plants about 10 years ago...

"Nuclear power in Germany accounted for 13.3% of German electricity supply in 2021, generated by six power plants, of which three were switched off at the end of 2021, the other three due to cease operation at the end of 2022 according to the complete nuclear phase-out plan of 2011." https://en.wikipedia.org/wiki/Nuclear_power_in_Germany

And spot market prices last December when all six plants were running at full power were as high as they are now. This has nothing to do with nuclear energy. It's demand pushing Germany to run more expensive power plants, which in turn dictate the price of electricity for all power plants, so pointing at the electricity mix of Germany is a waste of time.

> Finland invested in nuclear and is bringing their latest plant online this fall.

How exactly has a not yet operating reactor that is vastly over budget and time any influence on the current price though?

1. The Hetzner article is about future prices

2. Supply and demand

It effects the futures which effects any fixed price contract they might want to try and negotiate. Going for spot price is very risky at the moment with what winter is looking like to be.

Also there are the 4 existing reactors that are running just fine at the moment. (at this moment production is over 25% nuclear)

Not that I know of. We still have 3 nuclear stations in production. One of them is Isar 2 next to Munich. They will stay in operation until the end of this year.

That said, there are debates right now to prolong their operation until another few years. But that has other constraints.

The instantaneous market price that everyone pays is the marginal cost of the most expensive power source currently online.

For nearly every country worldwide, that is gas turbines most of the time. Even if 90% of power generation is coming from solar, it's the 10% gas that sets the price (since the marginal cost of installed solar is near zero).

Gas prices in Europe vary by region, but are pretty much determined by gas transport capacities to Germany. Ie. Germany has the highest prices because of government mandates to fill storage. Any place with good gas transport links also has high prices. Places with few/no pipelines to Germany and an independent supply have low prices.

And that is why Germany is twice Finland (Finland gets gas from Norway, and neither has big pipelines to Germany).

> (Finland gets gas from Norway, and neither has big pipelines to Germany).

Not really. Finland used to get gas from Russia but now it is mostly from a pipeline through Estonia to Latvia where it arrives in ships from wherever happens to be the cheapest (we are in process of building the first LNG terminal in Finland). Could be Norway but can also be USA, middle east, etc

Finland just doesn't use much gas for electricity. What we use mostly goes into industrial processes. We do have a couple gas peaker plants but those are only used in the winter if needed. Actual load balancing happens through hydro (and imports from Norway/Sweden but that is also mostly hydro)

There is no LNG terminal in Latvia the closest one is in Lithuania. Also AFAIK it doesn't even have enough capacity for the three baltic states let alone Finland. So chances are you most of the gas your buying through Estonia is Russian, just sold be some reseller.

Ok I got the country wrong but it is being bought from that terminal.

https://www.helen.fi/en/news/2022/helen-to-break-away-from-r... with similar moves by the other gas buyers in Finland.

Only pipeline connection from Russia to the Baltics is this pipeline (Balticconnector) as it connects to the Finnish gas grid that has its own pipeline to Russia and through the national grids and the pipeline between Lithuania and Poland that I guess allows access to the Russian pipeline in Poland.

So no clue where Estonia would get that Russian gas from.

The capacity of the LNG terminal in Klaipeda is 23,9 TWh per year, Lithuania alone consumed around 24 TWh per year in before this year. While consumption is probably going to be lower this years due to the massive spike in price there still is no way it can provide enough gas for Lithuania, Latvia, Estonia and Finland.

>Only pipeline connection from Russia to the Baltics is this pipeline

No. Latvia has a direct connection to Russia at Korneti which is where the majority of natural gas consumed in the Baltic states is imported at. So even if Finland buys any gas from the terminal in Lithuania that only means that more Russian gas is consumed in the rest of the Baltics (so effectively pretty to close to being the same as just buying it directly from Russia)

Not really. That’s the theoretical microeconomics approach if all kWh were “rebid” instantaneously. In fact, wholesale distributors can dampen the cost by diversifying in time (with contracts), location (by distribution) and cost (different plants).

The primary reason is: Winter is coming...

But there is more:

- They wanted to be super green so lets go with sun and wind

- But, hey sun and wind are not reliable (not talking about some problems that the grid is not very good prepared to handle the peaks), especially not in winter

- Let's use gas to cover that, Russia is so friendly country, and its all about business, what kind of maniac would stop natural gas, when we pay billions for it

- Hmm, well there is 'a country' that is using natural gas to do politics, but hey, they got Crimea now, what would they want more, when we are still paying that billions to them. Let's build another gas pipe, and pretend nothing can go wrong, while ignoring that negativistic east-european countries preparing for worst case scenario

- And, what about shutting down that completely functional nuclear power plants ? We want to be green, lets do that.

- Oops, our lovely natural gas supplier attacked another country

- Oops, our natural gas power plants are somehow costly to run now, lets burn some coal

- Oops, coal is also getting a bit expensive

- Oops, our not very friendly natural gas supplier, occasionally stops natural gas now

- Ooops, our natural gas reservoirs will be not full before winter

- Oooops, winter is coming...

> They wanted to be super green so lets go with sun and wind

This is a complete mischaracterization of why Germany invested so much into renewables. Like... 20 years ago? Germany saw the writing on the wall when it comes to how they were getting their energy which at the time was mostly coming from coal. They didn't want to go Nuclear for whatever reason so that left renewables and natural gas which wasn't preferred but it was still significantly better.

They didn't do it just to "be super green". They did it because coal is all around bad and wasn't the future that Germany wanted--and they were right. They absolutely made the right decision back then and investments by the German government are a big reason why solar is the cheapest form of energy right now (thanks Germany!).

If anything they didn't invest enough in renewables. They obviously didn't expect the effects of global warming to come this quickly. A big reason why electricity is so expensive is because the rivers in France that supply water to their nuclear reactors dried up. I mean, how do you predict something like that happening even 10 years ago? Even if that was a possibility included in some climate change report how seriously would politicians take such a prediction?

I would say the big mistake was to only use pipelines to Russia for gas. They should have built LNG terminals as redundancy. And maybe push for another pipeline to Algeria either through France and Spain or Austria/Switzerland and Italy.

They basically left the door open to abusive trade partners by not building any redundancy on such a critical resource. And it is not just electricity Germany needs gas for as it is also a very important resource for lots of industrial processes (and used for heating a lot but that should be replaced by heat air pumps)

Obviously such redundancy would have pushed up the price of gas in the short term at least. Easy to say this stuff in hindsight.

https://www.gem.wiki/Midi-Catalonia_Pipeline

I’m just going to point out that Trump said this exact thing and was ridiculed for it: https://www.washingtonpost.com/world/2018/09/25/trump-accuse...

I'm quite sick of that "reverse ad-hominen" attack on politics of discrediting ideas just because somebody stupid agrees with them.

I don't say that that being super green is bad, the problem was to rely almost solely on natural gas (from those we don't name) to stabilize the grid. Well, and now using that bad coal to fix that... I don't know, that also doesn't seem to be very wise, although I know, its only the consequence of the first mistake, but it completely negates what they tried to achieve.

Every single country relies on either natural gas or hydro power to stabilize the grid. What alternative did you expect?

If there is a problem, it's either on the share of the generation that they decided to let for those "stabilizing" sources (it is much larger than the short term instabilities on their grid) or on the lack of diversity of their supply.

This picture shows you this crazyness: https://lampopumput.info/foorumi/attachments/1660195161813-p...

In north Norway the electricity costs: 0.75 c/kWh In south Norway the electricity costs: 307.00 c/kWh In Latvia it costs: 1,240.00 c/kWh

It is like that every day everywhere in Europe. In Finland the transfer line to Estonia means a lot to our local prices. If we cannot get that line fully utilised (in normal non-windy day, and I bet producers want it to stay around 99%), it means that prices are fixed to Estonia (which you can see in that picture). If we get the line fully loaded, Finland divides to its own market and prices drop considerably.

In Norway and Sweden you can see multiple prices inside a country. It is because their internal lines are not sturdy enough to transfer enough electricity to south (in nordics it is common that electricity is made at north (e.g. a lot of rivers), but needed at south). The random energy (wind) has made this quite crazy. On windy days, the electricity price can even be negative. But wind energy is not reliable at all. But nobody builds or wants to run anything else as wind is easily the cheapest on windy days, and it is not that easy to always start and stop other facilities, which operate with losses on windy days. In general Germany jumped to a dwell, and everyone followed, and now we are where we are.

> In Latvia it costs: 1,240.00 c/kWh

Note that this is particular price point today (11.08.2022) at 12 - 13. Record was some other day for 2100 c/kWh.

However at particular moment, you can get 27 c/kWh fixed price contract. (last year it was 6 c/kWh)

It's not that the energy is made in in the northern parts of Norway. Most energy is from the south west of Norway, but the southern parts of Norway has connections to other parts of Europe and is affected by the energy crisis in Europe.

But the part about capacity internally is true, so the average price in northern Norway today is 0.00061 EUR/kWh(yes, you read that correctly) while it's 0.35 EUR/kWh in the south. That's 574 times as expensive!

Norway is making a ton of money on energy export now.

There is also an energy consumption tax in Germany:

https://de.wikipedia.org/wiki/Stromsteuergesetz_(Deutschland...

I don't know if Finland has something equivalent or not.

0.22€/kWh is actually pretty reasonable looking at what current long term contracts are offered in Finland.

Interesting that the colo electricity cost goes up that much, that’s a lot more expensive than what I will pay as a normal consumer, I’d have thought they’d get better discounts.

I'd guess that they used to have very good discounted contracts but are not able to renew them anymore.

The cost they charge also includes the cooling, so you could see it as for example 30c/kWh for power, 15c/kWh for cooling to compare it to a consumer contract. Maybe they have published their ratio of that split "on average" before.

There's a more than probable markup for redundancy and management.

Anyway, I believe industrial power billing is quite a trip (at least in Spain), and probably have a miriad of different prices depending on the hour of the day, and different guarantees of delivery compared to a regular user.

In many places if you as a business are willing to take intermittent power cuts (with varying warning times) you can get significant discounts.

I believe in EU household electricity is largely subsidized and the industry always paid more.

In Finland it is the other way around. Household pays the full price with industry paying the same but get huge tax breaks.

Basically residential and companies pay the same but are taxed very differently in a form that ends up with companies paying way less electricity taxes.

It’s the other way around, at least in Germany. Households pay more than industry for electricity. https://www.globalpetrolprices.com/Germany/electricity_price...

Well. I have to admit you're right. But it seems to not be universally true.

In Poland I have found that price for companies can be 3x higher based on at least one source. C11 - the tariff for the companies seem to cost 1.6 PLN / kwH [1] G11 - the same tarrif for households seem to cost more like 0.4-0.5PLN / kwH [2]

[1] https://www.energa.pl/mala-firma/umowy/prad-dla-firm.html [2] https://www.energa.pl/dom/umowy/taryfa.html

I'd guess this is because of price regulation, not subsidy.

What's the difference?

No money transfer from the state to the energy company.

Nope. Wrong. The economy of scale works also for electricity. We don't have subsidized electricity. The other way around we had additional premiums to pay for green energy. That got yanked and is now paid by the general taxes.

In colos, the cost of electricity typically also pays for air conditioning since it's a perfect linear relationship - more power goes in, more AC needed to get the heat out.

Plus some overhead costs (but with Hetzner, I'd assume they're minimal - they have razor thin margins).

Colo costs are just insane compared to consumer prices.

At home: 12 or 42 €/month (details¹), depending on whether I want gigabit upload or not. Advantages:

- fast LAN speeds,

- don't need to go anywhere to do hardware maintenance/replacement/additions,

- low cost,

- partially solar powered if you have solar panels,

- don't need to make things fit in a rack (can be a raspberry pi, an old laptop, or whatever you like),

- likely you have more space available, and you get to share that gigabit uplink if you pay for it for your server.

Colocation: €128 / month (120 base price, 8 electricity in germany; if you want to fly to Finland for accessing your hardware, then electricity is half the price). Advantages:

- redundant power supply,

- maybe noise if you don't have a separate room for this stuff,

- you can do this also if your home has poor internet.

¹ I pay €50/month for my internet connection regardless (I didn't buy a bigger bundle for the server), the server uses ~35W average which makes ~26 kWh/month which makes ~9 €/month. Though, I'd pay the base connection fee for electricity anyway so the added cost is even a bit lower. Rent is about €6 per m², but I have my stuff in between storage (above and below) so the column above those 1.5 m² is basically cut in three, coming out to about 3 €/month added cost. The added cost of hosting a server at home is about 12 euros per month. If I wanted to upgrade to gigabit symmetric internet, that would be an added 30 €/month (total 42 €/month). At 35W it doesn't make much of a difference for cooling, but there is no AC installed in my home anyway.

Does your ISP allow you to use their service to host a server? Most expressly prohibit it in their terms and conditions.

I remember in 2005 or so, there were people that got letters from their ISP to stop hosting things, and blocking port 25 was common until about 5-10 years after that but then net neutrality became more of a topic. Haven't heard of any such trouble since then.

There are also ISPs that expressly encourage it. Formerly I was with XS4ALL who encouraged it when I asked them on the phone (long time ago), nowadays their spiritual successor is Freedom Internet. I think the name says enough :)

That is pretty uncommon in Germany (possibly illegal?), here I can host whatever I want.

> At home: 12 or 42 €/month (details¹), depending on whether I want gigabit upload or not.

For me it's either 40€ a month if I want Gigabit down or move to a different city in case I want Gigabit up. Which is actually pretty good for current German standards; most can only dream of Gigabit down. Also, your home link is meager compared to a Gigabit link with proper redundant links to IXPs.

If you have a big home with solar power, appropriate cooling, a good connection and space to stuff servers away then yes, you can outprice collocation - but that's excluding a lot of costs you had before; if I was already running a server farm, collocating my servers would be even cheaper ;) If, however, you live in a flat in a city or in a rural area with bad connection [0], these prices are very competitive.

[0] So, most of Germany.

> At home: 12 or 42 €/month (details¹), depending on whether I want gigabit upload or not.

Your consumer gigabit connection is not comparable to a DC gigabit connection. Your ISP is over subscribing 20x and hoping people don’t use anywhere near their capacity. To make it fair you should compare it to a business plan with a proper SLA.

> business plan with a proper SLA

If I am that serious, then I'd also want the redundant power and would have a budget for it. For business, 130€/month is going to be a fraction of the cost of your first employee. The scenario posed was "as a normal consumer".

Their cost is 1/3rd more just for running the AC (your electricity use generates heat that has to get out). And savings from discounts probably get eaten up by the costs of redundancy, diesel backups etc.

Just write efficient software, then you won’t need that many servers.

While this is true to a degree, development costs also come into play here. Let's simplify the problem a lot for a moment. Suppose the only thing you care about is cost and the only things you can control are the hardware you're running on and the code it runs. You can optimize your code to use less hardware, which would lower operational costs. But you also have to pay developers to optimize the code. In some cases, it may be more cost effective to run less efficient code rather than pay for the development time to optimize.

I do sometimes wonder about the global cost to run Python and Ruby versus equivalent but literally-more-efficient programs that otherwise do the same thing in e.g. Go or Nim.

You can pay lots of mediocre engineers mediocre salaries to ship tons of inefficient mediocre code. Or you can pay a few rockstar engineers high salaries to ship small and efficient software. Basically the WhatsApp model.

Well, let's play that game. 0.22€/kWh. So a 1kW server costs almost €2000/year. If you write Python vs. Go, the performance increase can be 10 fold, so you'd save €1800/yr. So when you're running 55 Python servers, you could earn 1 dev by rewriting (assuming €100k/yr, all in).

Not to upset anyone: there are a lot of assumptions. It's just ballpark numbers. There's a large difference between Python v. Go for tensorflow and as a Django backend.

Switching languages is highly impractical for most teams. It is possible in some scenarios code written by an experienced Go developer is 10x faster than code written by an experienced Python developer. However, I think it's unlikely that a team writing Python is going to be able to write Go code which is as efficient as an expert Go developer. Even if they could, I think the productivity (in terms of being able to quickly implement features) of writing Python is significantly higher than Go. If only for the massive amounts of existing packages available, not including the language itself. I don't think this loss would be made up by a single additional developer.

I'm certainly not saying that's not possible to have a highly productive team of Go developers. But I think expecting a team of Python developers to become Go experts is not likely to work well. (Dropbox potentially being a notable exception, although I don't know how they navigated this change organizationally.)

Every algorithm I write ends up being O(2^n)

That could just be because your inputs are very small

Same vibes as "I would simply write bug-free software"

If you're operating at a small scale of a few servers (like most people using Hetzner), re-engineering the code that you already have to be more efficient will cost considerably more than paying your server bills for a year.

Really annoying that they're not publishing the new prices now, we have to wait until 1st September.

yes, that's frustrating, but they would have to list both current and future prices or make some a list of changes instead of just updating the actual price when they go current

At least they're not changing pricing for already provisioned products until 31st December.

They are publishing the prices now, just not in the main pricing page until the prices are current:

  The new prices for the colocation electricity/air conditioning rates are:
  electricity/air conditioning rate  old price\*  new price\*
  Germany  € 0.29/kWh  € 0.45/kWh
  Finland  € 0.12/kWh  € 0.22/kWh
  
  \*These prices do not include VAT.

But those are only the rates for collocation. The other server products (rented dedicated servers, VPS, etc) will become more expensive, too. Those prices are not yet announced, are they? (except the rough estimate of 10%)

Ah, it's true it isn't super clear... So "around 10%" and:

  The products that do not directly use electricity are not affected: IPs, domains, SSL certificates, racks, setup fees, and a few other examples.

Will this affect their US cloud (ASH)? I didn't see it listed but they mention all locations.

Please see: https://docs.hetzner.com/general/others/price-adjustment/#wh... "in all locations" = Yes, the ASH location will also be affected. --Katie, Hetzner

Guess this is something that has been going on for much longer already.

Leaseweb has been raising their prices because of higher electricity costs on February 1st, and recently DigitalOcean raised their price by even 20%.

If I am running Hetzner, I will be looking at an aggressive North American expansion. The energy cost gap between US and Europe are going to get bigger.

> The energy cost gap between US and Europe are going to get bigger.

Only briefly. The fracking boom is basically over:

https://www.technologyreview.com/2021/07/01/1027822/fracking...

We're presently running a good portion of our heating and electricity via the cheap gas that the fracking boom unlocked but that's going to taper off pretty quickly over the next decade. Because fracking was never a long-term solution.

Here's what's probably going to happen: Places like Europe will be investing heavily in renewable energy and locking in long-term contracts to get things like solar panels and wind turbine blades. Meanwhile the US will continue "business as usual" thanks to the largely for-profit power sector that only looks at next quarter. Then as the cheap gas from fracking runs out the US will be screwed and have the same problem that Europe is currently having.

This is a flawed analysis.

1. Permian basin itself will provide fuel for at least 300 years and right now if you have skills, you can get well paid job in Texas.

2. Texas has so much Wind and Solar footprint already it will make most European countries headspin, and we are not even adding Iowa! So, US has massive set up already and the best part is, unlike Germany, US has windmills where Wind is.

Fracking changed the market but the Western Hemisphere is not dependent on it. There is enough oil and gas in the North and South American continents to last hundreds of years by all projections.

What’s likely to happen is the global market will split into three regional markets: the Americas, Europe and Africa, and Russia and Asia. The US will recalibrate for more local feedstocks and very much continue business as usual. The other markets will realign producing regions and consuming regions along political lines and eventually get back to business as usual. The difference will be the prices in each region, which will go back to pre-globalization independent activity.

Other than that, yes, renewables and other sources will see heavy investment, but they depend on raw materials that come primarily from unstable (or newly unstable) political regions. I’d say it very much remains to be seen how this affects prices between renewables and petroleum.

What’s clear is that we are all heading into an “all of the above” energy policy regime that will no longer enjoy the perks and stability afforded by the global market.

Worth noting that most of Hetz infrastructure is deployed in EU, where cost of energy is hugely affected by events in Ukraine and Russia.

Writing this on a new account because my main has been shadowbanned

Why don’t they switch to an alternative cooling method like ground cooling?

aaaand thank goodness i'm on an american server host or else i'd probably be getting the same rn.

I'd have thought that these cloud providers would be pretty well insulated from the retail energy prices by their renewable energy generation.

That is not much use if the future contract prices at the moment are very high. 24 month consumer deals seems to be around 25 to 30 cents/kWh currently or even more, 12 month even more. And some sellers have stopped selling these.

Overall there is nothing much to do as unless you own the production there just isn't insulating contracts available.

Worth it, especially because of the reasons the prices of electricity went up.

Worth how? Sanctions achieved nothing on the military front. Is it more of a scaremongering thing? 'If we didn't do it, things would have been much worse'?

I am not talking about strategy but my own feeling that I don't mind to pay the increase. My wife family was wiped out in the holocaust. I don't think that sitting and doing nothing is a better option. And I do really not wanting to start an argument about it. As I said, it is a feeling and I am not a general and I have no knowledge of what is best.

"We have to do something. This is something. Therefore, we have to do this."

Never mind that it doesn't have any positive effect, and a lot of negative effects.

I am discussing the idea in general. That's perfectly fine if you personally don't agree. Although I do not agree that there are only two options: 1) Do this, 2) Do nothing. It is always possible to.. 3) Do something else.

German here. The causes of Germany's sky high electricity prices are not the Ukraine war.

A large part of the current spot price hike is the French bet on going all-in on nuclear: first, half the reactors is offline because they are simply broken down - they bet on Flamanville to replace parts of the extremely aged fleet which infamously failed. Then, the drought led to rivers drying up which means the reactors that are operational simply cannot get cooled, which led to a further reduction in capacity right when AC demands a lot of power, which in turn led to the French buying a shitload of capacity from Germany, and our spare capacity is mostly gas peaker plants since we phased out many coal stinkers over the last year.

The second part is regulatory... the fees that have to be paid for by everyone other than heavy industry for modernizing the grid and build out renewables are a large part, as well as the taxes that add on top of that.

And the last part is that states like Bavaria did their very best to impede the build-out of renewable energies (especially wind farms) and, worse, transit lines that could ship cheap North Sea wind energy to Bavaria - that led to an oversupply in the North where wind and solar have had to tune out because the energy could not be transferred to where it's needed.

If that is the case then the whole point is mute. However, what about Finland which was also specified?

Putin's starting to get on my nerves.

Russophobia is a costly endeavour!

Reply on news.ycombinator.com