Hacker News Reader

Comment by randerson | Hacker News Reader

I've developed a new fear of my 2025 desktop PC being damaged by a power surge or something, because it would cost at least $2K more to replace than I paid for it, assuming I can even find parts now. Compared to the rest of my adult life when I used to secretly pray for something to fail so I would have a reason to upgrade.

View on HN

Living in developing countries taught me to never plugin expensive computers without a surge protector UPS.

Honestly even in "developed countries" it's not worth blindly trusting that the power in your house/building is clean. It's cheap and easy enough to just put any expensive hardware on a UPS rather than speculating what's going on behind the walls.

Doesn't even sound like a developed country to me. Is that the US or something?

I work on embedded systems. I can often see whether my A/C or other appliances are running on my oscilloscope signals. They often affect the output of USB power supplies.

Eh, if these surges are rare enough, then you are statistical better off just risking your 'expensive' hardware to a one in a trillion possibility rather than spending money on gear you don't need.

Do you live in a bunker to protect against artillery shells?

Do you still need a UPS if you have one of those household (Powerwall style) battery packs? Also Apple switched mode power supplies are pretty well built.

But then again there's horror stories like

https://www.reddit.com/r/applehelp/comments/1maegvb/i_burned...

Yes. The power walls are like cheap UPS topology. You could still get whacked with a transient from the grid before the ATS decides to island the house.

My understanding is that home batteries are not UPSes, they don't go through the battery. They have a switch between power company, solar, or battery. I think that means would be exposed to surge from power company.

You can install a whole house surge protector. Those go in the panel and would protect from different sources.

Depends on how they are configured, I think in some regions (where power outages are very rare), they are wired to sync up with external network, and without external network they shut down as well.

Lightning can mess you up in every country lol. Had to replace a PSU because of that, thankfully it was just that and minor damage to GPU.

Lightning damage is mostly an issue if the last-mile power lines are above ground. In my experience, power surges in urban areas with a decent grid are so rare that people generally don't bother protecting their devices.

I have lived in the DC metro area inside the beltway or in Sillicon Valley my entire adult life and have only had above ground power wiring. Despite tree ordnances and wind storms and a grid so aged if we see lightning we lose power.

I've heard that before, that the US apparently loves above ground power lines. In NL it's only the long distance ones that are above ground. Even in most rural areas, I think everything is below ground.

Yes, we love them on account of our country having approximately 230 times the surface area and the Netherlands having approximately 13x the population density. We not only have vastly more line to run, but also many, many fewer people per square mile to absorb the costs. Underground line is expensive.

It turns out that the fires caused by above ground power lines are also quite expensive, at least in certain areas.

https://www.firevictimtrust.com/

That explains rural areas but not urban areas. We've got above-ground in rural areas but pretty much all urban stuff is underground. We get maybe one power cut a year, usually for scheduled maintenance work, and no problems with surges and whatnot.

Yes, where the population density supports it the US has tons of underground.

Grandparent was talking about the DC metro area and Silicon Valley. We're not exactly talking about Montana here.

Expensive to install, less expensive to maintain.

> Lightning damage is mostly an issue if the last-mile power lines are above ground.

So 99.99999% of the world.

But not where 99.99999% of the population lives.

Where I live it's not an issue.

Lightning in the first place its much much less common in some countries. Add to that under ground residential power lines and adequate lightning rods and it's not something most people here have to worry about.

Commercial uses layered surge protectors (Type I, II, and III), which is also recommended for other users but rarely followed.

In surge prone areas, at a minimum I would have good quality whole-house surge protector (eg Siemens 140 or Eaton 108), and a good quality surge protector strip for any computer/TV/phone charger.

I also put surge protectors in front of expensive white goods like the fridge, washer/dryer, dishwasher, and garage door opener. Besides being costly to replace these can contain "sparky" motors and this provides protection in the other direction too. Over time smaller surges can degrade the main surge protector for your computer.

Nothing (reasonable) can protect against direct lightning strikes, but for anything less it should provide decent protection.

> Nothing (reasonable) can protect against direct lightning strikes

Belkin make a number of surge protectors which offer a connected equipment warranty in the UK. Admittedly: financial protection, not data protection, but I felt it was worthwhile for the peace of mind.

https://www.belkin.com/id/p/6-outlet-surge-protection-strip-...

>Admittedly: financial protection, not data protection

You should have data backups regardless, because there are plenty of ways to lose data that don't involve power surges.

Have they ever paid out on one of those, or is it like CAs who offer liability protection for their certificates carefully set up in such a way that they never have to pay out.

>In surge prone areas

What areas are surge prone?

Areas with lots of thunderstorms. Also more rural areas with long power lines with few taps off for customers — the long runs are both exposed to many nearby strikes and accept induction well, and the few customers are fewer power sinks to dissipate the spike. So, you're more likely to get hit, and hit harder.

Sounds like if you're in an urban area with buried lines, you don't have to worry?

I would recommend a circuit surge protector in urban areas.

Lightning getting through some structure and hitting the electric lines happens. Even when they are buried. It's less of a problem when the ground absorbs a lot of the power before it even get into copper, but it's even less of a problem if there's some cheap device that will burn and protect you from it.

In urban areas you probably can just have the whole-house surge protector and skip the rest, since that protects all costly electronics not just a single device. With just a surge strip on the PC I'd say you're a tad under-protected, yeah.

Incidentally whole-house surge protection is now required by code in new houses. Existing buildings aren't required to upgrade, but by my reasoning what's good for the goose is good for the gander.

No, just worry less about big lightning, but urban areas have many more sources of different types of spikes and line noise

Open aerial wiring can shortcircuit two phases, bringing a low impedance surge that can damage most electric and electronic equipment.

The California bay area, at least all the sides of it I’ve lived on. We currently have a whole house battery, whole house surge protector, a second surge protector, and a UPS between the router/nas/etc and PG&E.

It’s not good enough. At least the power stays on once the grid stops bouncing (or once I manage to log into the rebooting battery gateway computer to have it flip the “off grid” breaker, or go outside and flip the manual one by the meter).

Color me skeptical. I've lived in several different Bay Area cities for decades. There are usually a couple power outages per year but I've never experienced a surge strong enough to cause equipment damage.

Huh, interesting. I've lived in the bay area for 22 years (first 6 of them in various places from SJ up the peninsula, remainder in SF), and I've never experienced damage to anything due to a power surge.

Not saying you're lying, but I do wonder if your experience is typical.

I've lived all over the peninsula, your experience is not normal.

I had far more power outages during my late teenage years in suburban Dallas than I've ever had in the bay. That was due to a bad transformer in the neighborhood which took years to replace, but once it was replaced everything was perfect. The moral of the story being: if your power is bad, it's probably because some piece of mains infrastructure near you is broken.

I had a string of annoying outages in 2023-2024, but it was all due to main service upgrades on my street, can't really complain about that.

Are you in an area with a bad electrical grid or something? In 40+ years I've never had a single device get fried from a surge/storm. My "surge protector" power strips are from the 90s and probably don't even work.

Not too bad, just rural. We used to lose stuff every 10 years or so.

One day The Big One came along and fried nearly everything. "Once burned, twice shy."

Hopefully someone can learn from my mistake and not have to do it post-mortem.

This. Same timeframe and I've lived through both lots of lightning storms and in areas with lots of power failures. Some of them intermittent and essentially caused by transformers blowing up. Like earlier this winter, we had multiple storms where you'd hear a transformer blow up, in many cases even seeing the sky light up as well from it, power going out, couple seconds, power coming back, next transformer blowing out, rinse, repeat.

On the other hand I've read about plenty of stories of the "cheap" UPSs you'd usually buy as a consumer (not to name any brands coz I've never had any) actually causing such issues in the first place. Without any actual surges from the grid.

That said, being totally not superstitious (for real, but someone's gonna "kill me" if they find out I wrote this and something dies from a surge...), now I guess I need to knock on wood like seventeen times ...

I do use surge protectors when we're on generator power temporarily.

The things people often call "transformers blowing up" are usually not transformers blowing up.

Instead, it's usually just overhead wires that are too close or literally touching, often from influences like wind and ice. The electricity arcs between the wires, creating bright blue-white flashes that can be seen from far away, sometimes with instantaneous heat that makes hunks of metal wire evaporate explosively. It can be violent and loud, and repetitious as different parts of even a single run fail.

Transformers can certainly blow up, but that's less common. They're (generally) filled with oil for cooling purposes, and they're massive things that tend to take time to get hot. A failed transformer can produce arcing and blue-white light, but if things are that hot then the oil is also ready to burn.

And when the oil burns it isn't blue-white -- it burns with about the same yellow-orange color we saw the last time we accidentally flambéed dinner on the kitchen stove, or a Hollywood fireball.

A bright flash without a fire is probably not a transformer.

Here's a video of a transformer actually-exploding (note the prominent fireball): https://www.youtube.com/watch?v=oFkfd31Wpng

And here's a video of what someone describes as a transformer exploding, even though there are no transformers in the video: https://www.youtube.com/watch?v=rHVh0KwG_0k

Haha, I hear you. But yes, it really is transformers blowing up sometimes. Sometimes it really is just branches blowing up the line, sure.

A branch hitting a wire, happenes all the time here too. Lots of trees in this community. The video of a transformer you shared: that's not the transformer I'm talking about. That's at a transformer station.

I'm talking transformer on a street pole. The kind that hangs right across the street from me. This kind: https://www.youtube.com/shorts/y3E7avUvj6I

See it's the kind in your second video. It's a transformer. You just chose a narrower definition I suppose. It's a https://en.wikipedia.org/wiki/Distribution_transformer ;)

And yes, I know it's transformers and not just wires (but also wires do happen definitely) coz I do walk the neighborhood regularly and I can tell when a transformer is new vs. old up there. Ours is old. The ones a few streets over sometimes are very new and I see the Hydro trucks go by the next day(s) to make them new ;)

Again, like seventeen times knock on wood but the ones next to us have not actually blown up. But three streets over, seen the new ones. Literally last weekend, we had an ice storm come through and while no blowouts we could see or hear, the outage map showed plenty of failure.

Residental-scale transformers can and do explode. Shorts happen not-infrequently with freezing rain and ice storms especially causing issues - the internal oil gets displaced by the water, and the dirty water causes an internal short. It wipes out power to a few blocks here when it happens, but we get an outage due to it every year or two.

They can. They do.

But when the wind is whipping along on a warm day and there are bright flashes and audible bangs, that's (usually!) not signs of transformers blowing up... even though the popular vernacular often erroneously describes it that way.

It happens. The power company was very unhappy with my boss for destroying one of their transformers. The thing is while circuit breakers react very quickly to extreme overcurrent situations (shorts) they're much slower to react to loads which are only a bit over the limit, and if short enough won't react at all. Very common with heavy motors.

And that's exactly what the problem was--we had a whole bunch of really heavy motors. Getting ready to start for the day you flip on the switches and the big machines start to spin. The transformer on the pole was rated higher than the main breaker for the plant--but the transformer apparently was more sensitive to the temporary loads. Once the problem was identified it was resolved by staging it, instead of flipping them all on they were flipped on over 5 minutes.

It's not just cheap UPSes, it's cheap surge protectors as well. They exist because the vendor can throw in a MOV costing a few cents and increase the price of the power strip by 50%, not because they're any good. MOVs are sacrificial components which have either degraded to uselessness by the time they're actually needed or, if they're still working, can explode or catch fire from the energy dissipated. Even if they don't, all they're doing is converting an x-kV spike on active into an around-x-kV spike on neutral or ground. If you want to do it properly, use a series tracking filter, not a "surge protector".

No offense, but can you tell me how my 4.5 kW generator is gonna generate that kind of power surge?

One scenario: there's a short circuit somewhere, say rats chewing through insulation. This can cause a very high current through the short. A non-inverter 4500 watt 120 volt generator might have 0.2 ohms coil resistance, so the short circuit current can hit 170 volts / 0.2 ohm = 850 amps. When the shorted branch's circuit breaker trips, the inductance in the generating windings wants to keep that 850 amps flowing for at least a few microseconds, and it gets distributed across everything else that's still connected. Depending on what else is connected (hopefully including some surge protectors) the peak voltage can get into many kilovolts.

The circuit is something like this:

  voltage source -- parasitic inductor --+- circuit breaker -- short
                                         |
                                         +- circuit breaker -- your PC

More generally, for the previous poster, look at what happens when a magnetic field collapses suddenly, you can get kilovolt spikes. There's probably a ton of YouTube videos demonstrating this in various ways, it sounds like the sort of thing that Electroboom would do. Normally this is handled via snubber circuits which dissipate the energy before it can do anything, but in exceptional cases it could end up going where it shouldn't.

Definitely use quality surge protectors on expensive equipment connected to generators.

PSA: UPSes and GFCI/GFI extension cords won't work properly when connected to a stand-alone generator with a bonded neutral. I've tried using enterprise UPSes on such generators, but they absolutely won't work. In such scenarios, debond the generator's ground from neutral, apply a very large warning label to it being debonded, and drive a massive ground rod electrode into the ground as close to the generator as possible and ground the neutral there. This does work and is much safer because there's a stable voltage reference source. It's more of a hassle but can be necessary for some off grid and temporary scenarios.

> In such scenarios, debond the generator's ground from neutral

eeeeep. Please for the love of all that is holy, CONTACT AN ELECTRICIAN before messing around with that - or before creating a ground bond where none should be (i.e. TT grid [1]). You may end up endangering yourself if you do not exactly know what you are doing - in the case of TT, you get ground potential difference current from other parts of the grid flowing to ground via your generator's bond. Best case you're getting problems with electrochemical corrosion (including in your foundation), worst case enough current flows to turn your bond wire into a thermal fuse.

Also, take great care if your grounding is provided via municipal water service, or if your original grounding rod has dried out to the point it's ineffective.

Let me repeat: LET ELECTRICIANS DEAL WITH GROUNDING AND SURGE PROTECTION. Floating grounds and improper ground connections CAN BE LETHAL OR POSE A SERIOUS FIRE RISK.

AND YES THAT INCLUDES "ISLAND" SCENARIOS OR EMERGENCY POWER INPUTS (e.g. via CEE plugs and transfer switches).

[1] https://de.wikipedia.org/wiki/TT-System

I'm not sure I'd leave something like this to an electrician. Or if so at least make that electrician be experienced in this field. I think you'd want an electrical engineer to be involved with the plan to some degree.

Electrical engineers don’t know code requirements and wiring guidelines for household electrical wiring. They’re absolutely not the correct default. Electricians with specialization in generator setups, sure, but an electrician engineer on average is likely going to be more uninformed on code requirements than an electrician.

Electrical engineers know the theory but lack the practical knowledge which grid form is used at your specific address (yes, here in Germany we have a few towns where one half side of a street runs TT and the other one is already migrated to TN-C or TN-C-S).

An electrician specializing in lightning protection, uninterruptible power installation or in radio installations can sort out all of that far better than an engineer can.

That extra unbonded ground rod is the worst thing you can possibly do to make your generator vulnerable to lightning strikes.

That's an extreme edge-case and a strawman. Anyone operating temporary equipment on a generator during a severe storm will obviously unplug sensitive stuff to not take unnecessary chances regardless of safety precautions already in place.

Ground rods are required in certain situations according to the NEC.

Ground rods are for lightning protection, transient surges (over voltage), and induced surges; not for short protection, ground faults, or making ordinary extension cord use of bonded generators "safer".

Typically, they're required whenever it's a system that powers a building on its own, i.e., off-grid setup or with a floating neutral generator connected via a switched neutral transfer switch.

You can unplug everything and open all the switches, but a nearby lightning strike will still fry your generator through that unbounded ground rod. Lightning ground potential is very eager to take the shortcut to your other ground rods through a few millimeters of insulation and open switches on the path through your generator and house wiring, when the alternative might be tens of meters of dirt :)

I don't care what the NEC doesn't say, NFPA 780 says you have to bond all ground rods.

GFCI works correctly either way. Their operating mode doesn't care at all about ground: Whether bonded, not bonded, or not even present (look, ma! only two wires!), they still perform the same way.

They respond to an imbalance in current flow betwixt line and neutral. What goes out must return; if it doesn't, then switch off.

Ground is not part of the equation at all.

If your close neighbours have surge protectors then you benefit little from installing your own.

Another perspective: we should install whole house surge protectors if we can afford them, not only for ourselves, but to help our neighbors - even if in reality the help is minimal and they need their own as well. In the best case scenario, if everybody in a neighborhood has them, each individual house will be more resistant to surges than if they were the only house with one (five houses with surge protectors nearby is a lot better than one) - everybody wins.

You might as well phrase that as "If your close neighbours have gotten vaccines then you benefit little from getting your own."

We live in a society. Everybody chips in. And each surge protector adds to the robustness of the grid.

Yes, that puts it down perfectly. That’s why some don’t ever see the benefit of installing their surge protector whereas others install one way too small for their situation and find them useless anyway.

Eh. Most nice power strips are also surge protectors.

Why? If the voltage spikes on the grid (that's what I understand a power surge to mean), wouldn't even more of it end up in your house (that is: the grid voltage spike even higher) if the neighbors have equipment that doesn't let their devices consume some of that energy?

Edit: wait, maybe I figured it out: those devices must be consuming the excess rather than blocking it. Is that it?

Yes, energy dissipates. Although one still needs to look out for the distance from the neighbours as your ground can be different from your neighbours ground potential.

When I lived in Costa Rica, I lost three surge protectors in a year to power surges. During one such power surge, I didn't notice that the red light indicating surge protection was already out, and a power surge fried my (knockoff) Macbook power adapter, leaving me without a way to work for a day.

There's no real indicator on those things.

It's a 10 cent component that would require a $10k machine (this is old data) to do a non-destructive test.

Not only should you get rid of them, but also they are a fire hazard.

Also, do not accidentally plug surge protectors into each other, metal oxide varistors can star fires _without_ meaningful surge conditions when you do so.

I prefer to buy products without MOVs entirely due to the risk, with the exception of one, Tripp Lite Isobars; but I prefer to use series mode protectors such as Brickwall or SurgeX.

> Not only should you get rid of them, but also they are a fire hazard.

Are they not a fire hazard even when new? MOVs do tend to degrade with use (especially after they've gone conductive to snuff one or more surges). But AFAICT we can't really know, without potentially-destructive testing, whether a given MOV is in good shape -- whether installed last week, last year, or 30 years ago.

> Also, do not accidentally plug surge protectors into each other, metal oxide varistors can star fires _without_ meaningful surge conditions when you do so.

What is the mechanism that increases risk for MOV-sourced fires in this arrangement?

I've also noticed that many of the power supplies I've taken apart (for very pedestrian consumer goods) have internal MOVs on their line input. Whatever the mechanism is that increases risk, isn't using one external surge protector already doing that in these instances?

> I prefer to buy products without MOVs entirely due to the risk, with the exception of one, Tripp Lite Isobars; but I prefer to use series mode protectors such as Brickwall or SurgeX.

I prefer to avoid MOVs, too. Broadly-speaking, diodes seem like a better way to do it. (Transtector is another reputable brand that uses diodes.)

---

That all said, I've noticed over the years that problems with dead (presumed-to-be-hit-by-a-power-surge) electronics tend to follow particular structures. And the reason for this seems related to grounding more than it is anything else.

So when I find someone (a friend, a client, maybe someone online that I'm trying to help) complaining about repeated damage, I often ask about grounding. Almost always, it turns out that they've got multiple grounding points for the electronics: The electric service has one ground rod, and the telephone/cable feet/satellite/whatever is connected to some other ground.

This might be a dedicated rod, maybe a metal pipe; whatever it is, it is distinct from the main service ground. It happens all the time. (It is worth noting that the NEC prohibits this kind of configuration unless extraordinary effort is put forth. See 800.100(d), for example.)

The way that MOVs -- and avalanche diodes alike -- behave combines with the fact that the earth is an imperfect conductor, such that having multiple ground points promotes dynamic ground loops that can provide quite large potential -through- the electronics that we seek to protect.

The problem appears suddenly, and repetitiously. Everything is fine, and then ZANG: The cable modem gets smoked along with the router it is connected to. So the modem goes back to Spectrum or wherever to get swapped, and the router gets replaced again, until the next time: ZANG.

TV connected to satellite receiver, with coax incorrectly grounded? ZANG. Over and over again.

I'd see it all the time when I was a kid back in the BBS days: The phone line was grounded improperly, and computer was the only thing that connected to both electricity and the telephone line. Some folks would go through several modems over the course of a summer, which was very expensive -- while most people had no problems at all. Next-door neighbors would have completely different failure rates.

Structures with correct grounding tend to do very well at avoiding these issues, and I've fixed these conditions in subsequent years more times than I can count.

(A coworker installed a phone system at a business once, wherein he made extensive use of Ditek surge suppressors -- on the incoming POTS lines, and on the power inputs. It blew up one day. So he called Ditek to try to get at least the cost of the phone system hardware covered. They asked him to draw up a map of how the building was grounded and send that over, so that's exactly what he did. When they saw his map, they very quickly identified a ground loop and denied the claim.)

"What is the mechanism that increases risk for MOV-sourced fires in this arrangement?"

I wondered the same thing, and failed to find a satisfying explanation.

I can find plenty of reports of MOV fires, especially in situations where there's a persistent over-voltage, e.g. a 120 V site actually having closer to 240 V due to a floating neutral. But I don't see how chained MOVs make that worse in general. This blog post has some nice photos:

https://www.electrical-forensics.com/SurgeSuppressors/SurgeS...

No clue about the actual reliability of this[1] article but the mechanism mentioned (new pathways due to changes in crystalline structure due to uneven heating) sounds possible.

1. https://incompliancemag.com/how-and-why-varistor-failure-occ...

Reread your wondering and now conclude its about chained situations which this also does not answer.

Heavy industry can also cause these kinds of power surges to happen.

Last year an aluminum smelter in Iceland had a transformer blow which caused a big power surge on parts of the very well developed national power grid. The surge caused damage to electronics in some households and companies near to the smelter.

I lost an audio mixer to a bad surge last year. I don't know whether it was additional load or just really bad fluctuations that damaged the device. Nothing else bit the dust, but the, digital board in this mixer got bricked.

How did you know it was a power surge? Not doubting your comment, just interested in knowing if this ever happened to me

Just saw this days later - I'm not sure if it was a surge specifically, but it got bricked after a really wacky power swing. So maybe not a surge, exactly.

EE living in a rural location here: transient related failures do happen in my experience. Rare but they happen. And I've known of people who had everything in their house fried. For me it's just been a couple of Ethernet ports. Power strips don't provide much protection fwiw. Always worthwhile checking that your electrical service is properly grounded.

I had exactly one device fried in my life. This was when I lived in FL in 1999. Lightning took out my sweet new 56k modem. My PC was fine.

You don’t need a storm. A fault like an open neutral can very easily fry things.

I live in Oakland, and the Easy Bay has had its share of random outages without incident. The last outage was just a split-second blip, but it broke a $1k computer monitor.

Not completely correct, nuanced, or comprehensive.

Direct lighting strikes cannot be defended against without extreme costs. This type of risk is generally extremely unlikely except for certain niche use-cases like equipment or facilities on tall peaks.

Transients from lightning (E2) nearby and distant nuclear detonations can be defended against, and often require additional protection of telco and internet entry points. Whole house type 1 SPD devices exist for residential applications. This is much more likely than direct lightning strikes, especially in certain areas and can be defended against for reasonable cost. The main issue of lacking it is the unseen, cumulative degradation of semiconductor components that lead to instantaneous or eventual failure, especially in high value devices like electrically-communicated motors in HVAC systems. There is no reasonable expectation of defense against a direct lightning strike even with type 1 SPD, and there are different types of lightning with vastly different amounts of energy. A positive strike direct hit will totally fry anything and everything.

What generally isn't defended against at all in any infrastructure or system except some military equipment is H/NEMP E1 (short duration impulses) or E3 (E3a or E3b; long duration surges larger than lightning) such as from unusual space weather events or nuclear blasts.

> Nothing (reasonable) can protect against direct lightning strikes

This is one reason why you bury power cables.

Burying power cables doesn’t help against direct lightning strikes. If lightning hits your house, underground power cables will do nothing.

Living in California taught me this

I grew up in San Jose CA in the 80's and 90'd on a street with a perpetually bad transformer. We had UPSes on every computer and proper surge protectors on everything of value.

Over the last two years I bought 2 4TB SSDs, 64GB DDR5 ECC UDIMM and 4 14TB HDDs.

I couldn't justify buying any of them today.

I bought a PC in early 2021 IIRC. It was good for the time and a good deal for a high end PC. IIRC it was $2800 and had a 6900 XTX. Last year I accidentally killed it. The CPU temps were higher than I'd like (~85C). the thermal grease can become hard and ineffective over time so I figured I'd replace it. Instead, it had become like cement and by twisting the AIO off, I snapped the socket on the motherboard.

This was an expensive mistake as I both looked into buying a replacement motherboard and CPU but that quickly gets to the price of a new PC. Paying someone to rebuild my PC is expensive and I'm beyond the age of wanting to fully remove a motherboard and effectively rebuild my entire PC myself. So I didn't know what to do with it.

Anyway, I ended up buying various alternatives like a NUC with 32GB of RAM, a laptop (with a 4080) and a Mac Mini. But I also ended up buying a new 9800X3D PC with a 5070Ti. Like I said, it was an expensive mistake.

But I decided for no particular reason to upgrade the (already good) 32GB of DDR5-6000 to 64GB with a $200 kit of DDR5-6000. This was in July I think. I also upgraded my laptop to 64GB for no readily apparent reason.

I recently checked and that $200 64GB kit now costs $950. SSDs are through the roof too but through complete accident I'm surrounded by about 5 PCs and a bunch of spare RAM. I don't see myself upgrading anytime soon.

I will say that there are some good deals (relative to current pricing) for combos including CPU, motherboard and memory or even some pretty good prebuilts.

Silver lining: literally all Macs are a total steal right now.

Good Mac Pro models are still spendy, but the M3/M4 laptops are great if your software use-cases are met. =3

I’m still doing great even with an M2!

Still loving my M1 to be honest!

It’s super annoying. I’m salivating over all the new announcements but my M1 16GB 1TB will likely last another 5 years.

Same here. I've never had a machine feel so great for so long before!

The SSD don't last forever, after about 3 to 4 years of daily use the drive/system should be replaced. At >5 years, one could hit retention issues and corruption losses.

Good excuse to upgrade though, as a $1500 recovery bill would not be cool. Best regards =3

Never used a computer less than 8yrs and never had an ssd have an issue in that time.

This advice brought to you by the "change your oil every three months" crowd.

I removed my anecdote and flash wear explanation, because of cranky folks like yourself.

The corrosion inhibitors in petrol engine oil get fully depleted within about a year with most brands. One may certainly sell the machine before you see acidified lubricant related problems, but the motor will not reach its full operational lifespan ( https://en.wikipedia.org/wiki/Bathtub_curve .)

I do agree that anyone with a CVT style transmission likely won't have to worry, as that entire section will probably need replaced before you see significant hydrodynamic bearing damage.

"Buy cheap, buy twice" as they say...

ymmv, best of luck. =3

As a side note, I absolutely cannot imagine being upset of having a machine lasting long.

Sure it's nice the shiny new thing but has capitalism infiltrated people's mind that much? All my previous laptops died on me several times and became frankestein's monsters before I let them rest for a final time (to be often repurposed to other family/friends' machines).

With intermittent use one may get a lot more life out of the SSD than other users, but eventually flash will run out of spare-sectors and start to fail.

Most M1 systems I saw use on-board BGA110 NAND flash, and thus maintenance/upgrades on the SSD are difficult. Most users don't have a hot air rework station or x-ray inspection machines to do this modification correctly.

The MTBF statistics can move around a lot depending on the use-case, but eventually people will run out of luck ( https://en.wikipedia.org/wiki/Mean_time_between_failures .)

Not sure why people get upset by this fact, as not all Apple hardware models were built to be disposable. =3

I mean they're still expensive, they just seem relatively good value because everything else has gotten more expensive.

Yeah, that's what @SlightlyLeftPad said.

All value is relative, so everything else getting more expensive is essentially the same as macs getting cheaper.

Just a matter of time until Apple also increases the price...

It's not the same, unless you are simultaneously also getting richer.

Other electronics have gotten more expensive, yes. But other hobbies haven't.

Anyone have a good take on how well Asahi linux keeps the power management working on mac hardware? The biggest killer feature for me of mac hardware is the battery/weight. I have found it hard to get a good laptop in the linux ecosystem mainly because of power consumption. If Asahi doesn't really impact the battery life then I would seriously consider going that route. Similar question about support for pytorch on linux/arm64 / Asahi.

I think it's improved from when I last tried it, but it still isn't great. You can get like 60% of the battery life compared to macOS.

Someone with more recent knowledge correct me on this, but I believe idling is the biggest power drain in Asahi. You will want to shutdown and/or hibernate whenever possible.

How good is Mac virtualization? Would it be doable to put an Ubuntu inside a VM and just run it full screen all the time?

Bought a used MacBook Air M2 past summer to run Asahi linux exclusively on it, the installation went hassle-free. One charge lasts 9+ hours easily, sometimes up to 12 hours. Thunderbolt, DP Alt Mode and TouchID would be nice to haves, but I'm super happy how everything runs. Thank you everyone on the Asahi team!

I think the support for linux/arm64 is already very good in general, can't answer on pytorch though. The only app I'm really missing is Signal Desktop. The virtualization to run games is a noticeable performance hit and shows occasional glitches in the Steam overlay, but all my games run smoothly.

Hey, just letting you know you can get aarch64 Signal Desktop from Terra repository!

Too bad I can’t play the games I want to play on them

Hint : GE Force Now

ssshhhhh... do not tell anyone I told you...

It's horrible. Bad quality, bad latency, can't mod the games etc. And worse you have to pay for it when you already have a more than capable computer.

i wouldnt go as far to say "its horrible".

i would never recommend it to someone who otherwise has a capable computer, of course, but it really isnt that bad. i gave it a pretty thorough test out of curiosity, and when they sponsored a few streamers i watch, it was totally fine. with the caveat that you have a decent internet connection and its probably not good for twitchy games like counter strike.

and, as far as i know, there is limited support for modding and some unsupported workarounds.

And it works on the Vision Pro via the next update.

finally!!!

Can’t wait to try that and for the f1 stuff to come out.

I used Shadow PC for a long time. Never any issues over several years. Lots of reasons in preferred it over GeForce. I can expound on that later if needed

Is the computer in question really "more than capable" if it "can't play the games [you] want to play"?

I've used geforce now on my mac before and didn't have latency issues. I wasn't using it for any competitive games where you need ultrafast twitchy response, but I did use if for plenty of FPSes and never had any issues. And I don't have super fast internet, just the basic package from Spectrum. So I wouldn't say it's bad, though admittedly it might not be the best latency achievable in the gaming world.

You have to be joking, you don't own anything and enjoy your price hikes as people adopt it

What are you talking about? Geforce Now is specifically only playable with games you own.

No idea why you're down voted... I blissfully played cyberpunk 2077 for two years on GeforceNow. I still keep my membership even though I have a dedicated gaming pc now, for occasional laptop or living room pc use. It was beyond brilliant to play a hyper demanding game on a bare spec pc :-)

Mind you,I have gigabit internet. I don't know what the experience would be like on other types of internet / worldwide.

Crossover[1] is surprisingly good for this purpose if you game occasionally and don't need FPS-level responsiveness. You also need 3rd party software like LinearMouse and Mos to make a mouse usable.

[1]: https://www.codeweavers.com/crossover?srsltid=AfmBOor-7wbD-o...

The AMD395+ PCs have unified memory and since it's not tied to a garbage OS nor reasonably affected by future dram costs, it's a better choice for reasonable people, unless you're going for greater than 128GB

I bought a Mac Mini in February and maxed out the ram and storage. Now, it seems like that was a prescient move, but honestly I really only bought it for photo editing and playing the new World of Warcraft expansion (don't judge me!).

Serious question: how does WoW still appeal to players except for habit social connections to keep them locked into the game? I used to spend nights and love the game, now, even with all these expansions it feels exactly like it was in 2006 but without what happened to the gaming world in the past 20 years.

It's still fun. The social connections are also hugely important to me. One of my characters is in the same active guild that I joined in 2006. It's hard to put into words how meaningful that is to me. The game has improved, the newly re-done Silvermoon City is beautiful and richly detailed, but you are right, in many ways it's the same game as 20+ years ago, except made more casual-friendly in a lot of ways. I like it and there really isn't anything else like it out there. ...and surprising to me, if you believe Blizzard, there are around 9 million people who still play.

Alright so I 100% understand you, and now I know I'm not totally crazy.

I think because I used to play on private servers, I don't have that long-standing connection to a group, which is probably what keeps many people still there. But yeah, I'd jump on a WoW 2 but the gameplay and quest system is so outdated that just doesn't give me good vibes anymore.

I don’t know. I still fire up FF14 every couple of weeks for a few dungeon runs. No more social interactions with the various channels, I barely talk to my party even.

I think it’s just familiarity and not wanting to learn a whole new system when I’m looking to shut my brain down for a couple hours.

You pay an $1k extra just to get the model with 1TB disk. How is that a steal?

pro tip: don't

Not really a steal, just that the price differences have narrowed.

Thankfully UPSes are still cheap. Get one before Sam buys the entire yearly production of cyberpower.

It sounds like it won't affect prices that much?

> South Korean memory giant SK hynix has since said it had diversified supplies for helium and secured sufficient inventory. Meanwhile, TSMC said that it doesn’t currently anticipate a notable impact following Ras Laffan going offline, but that it’s monitoring the situation.

It also just be the typical “don’t scare the shareholders just yet” PR speak. Time will tell.

I mean you are talking about 30% of the world supply I don't think you can spin up the amount of equipment need to replace that in a few weeks.

We used those Tripp Lite LC1200 to knock down the noise floor (14dB) on remote equipment.

These line-conditioners actually perform well given the cost, but never buy used surge-arresters given the finite spike hit-count. Best of luck =3

These devices are basically autotransformers. So they reduce the noise by providing inductive filtering. But they don't really protect against strong surges by themselves.

So Tripp Lite uses a regular varistor for that, just like any other surge protector. In Europe you'd be far better off buying a voltage relay and adding it to your electrical panel, but it's not usually possible with the non-modular US electrical panels.

The simple line-conditioners were surprisingly effective, and are a fraction of the cost of lab/medical grade galvanic isolation ferroresonant transformers. =3

What’s with the ‘cock and balls’ emoticon?

welcome to the internet, sometimes people make smiley faces :)

You can find all types of people on the Internet, even some without noses :-)

Don't worry about it =3

Why are you repeating yourself every two hours?

Don't worry about it =3

https://www.youtube.com/watch?v=YDdKiQNw80c

Can never unsee that now

Don't worry about it =3

I originally interpreted <3 that way.

Must have been awkward getting a text from some people =3

It's a woffly bunny face, imo.

Don't kink shame lol =3

https://southpark.fandom.com/wiki/Harequin_Romance_Book_Publ...

I have a UPS with surge protection which I plug my computer into for this reason. Do others do the same or use something else?

I got to get me some new surge protectors and not the cheapies. maybe a small ups.

Reply on news.ycombinator.com