auspiv

128GB DDR5 module - https://store.supermicro.com/us_en/supermicro-hynix-128gb-28...

It is of course RDIMM, but you didn't specify what memory type you were looking at.

On a somewhat prominent /fit/ness forum, they consider the pinnacle of lifting to be 4/3/2/1 plates for the big 4 lifts. The bar is 45 lb. Each plate is 45 lb. This equates to

405 lbs for deadlift

315 lbs for squat

225 lbs for bench press

135 lbs for overhead press

If you can do these weights for these exercises, you are in the top 1% for strength.

22600 oil/gas wells drilled in 2022 in the US [1]. Fervo is proposing to do ~20 MW per injector/producer pair [2][3].

that's 11000 potential pairs per year x 10 years = 110000 pairs = 2200000 MW = 2200 GW. total US usage is 4 trillion kWh for 2022 [4], which is 456 GW averaged over the whole year.

so in the 10 years that it takes to build 85% of nuclear reactors, the oil/gas industry could drill enough geothermal wells to satisfy 4x the electricity requirements of the entire US.

[1] https://www.statista.com/statistics/965763/number-of-oil-and...

[2] https://fervoenergy.com/fervo-energy-breaks-ground-on-the-wo...

[3] presentation in Denver May 2024 by fervo completion engineer said 40 total wells

[4] https://www.eia.gov/energyexplained/electricity/use-of-elect...

UUID7 basic information/examples - https://uuid7.com

"Synthetic oil is a lubricant consisting of chemical compounds that are artificially modified or synthesised. Synthetic lubricants can be manufactured using chemically modified petroleum components rather than whole crude oil, but can also be synthesized from other raw materials. The base material, however, is still overwhelmingly crude oil that is distilled and then modified physically and chemically. The actual synthesis process and composition of additives is generally a commercial trade secret and will vary among producers."

https://en.wikipedia.org/wiki/Synthetic_oil

Fervo Energy is making great progress in this area. They've drilled a injector/producer pair, frac'd between them (exactly like oil/gas), and pump water at ~60 L/s. Whatever goes into the injector comes out the producer a few hours later much hotter. Their proof of concept produces 3 MW, and uses ~1MW to power the injection pumps. They are doing a full scale plant in Utah now, and expect ~8 MW net for each injector/producer pair.

https://eartharxiv.org/repository/view/5704/

https://fervoenergy.com/fervo-energy-breaks-ground-on-the-wo...

"The deepest man-made hole, which extends 12,262 meters below the surface of Siberia, took nearly 20 years to drill. As the shaft went deeper, progress declined to less than a meter per hour—a rate that finally decreased to zero as the work was abandoned in 1992. That attempt and similar projects have made it clear that conventional drills are no match for the high temperatures and pressures deep in the Earth’s crust."

This is true. Neither the article nor the CEO of the microwave drill company say why.

At that depth, rock isn't a solid. It behaves plastically. The traditional tri-cone bit used could make progress, but it kind of just started "massaging" the rock. The bit (as all bits do) wore out. They pulled the drill string out of the hole to put a new bit on. The borehole would close back in during the multiple days it took to pull 40kft of drill string out, change the bit, and put it back in. Progress was not possible.

Unless the microwave drill includes some enormous cooling system (that works 40kft down hole even when the drillstring is removed!), they will face the same issues.

Also, separately, I saw pictures of the resulting lab-drilled hole on LinkedIn the other day. The hole shows a high rugosity (qualitative description of the roughness of a borehole wall). Similar photo here - https://spectrum.ieee.org/media-library/image-of-a-rock-surf.... That's a ugly hole, and would be very difficult to run casing into it.

Further - traditional well drillers' #1 focus is controlling pressure downhole (typically done by varying the density of the drilling mud). If the pressure becomes too great, a blowout can happen, which is bad news for everyone involved (see the BP Deepwater Horizon incident). For geothermal wells, they presumably will try to avoid hydrocarbons. However, rock far above water boiling point can cause a BLEVE (boiling liquid expanding vapor explosion), which is also undesirable. Super curious to see how they intend to control bottomhole pressure with statements like this - "Instead of pumping fluid and turning a drill, we’ll be burning and vaporizing rock and extracting gas, which is much easier to pump than mud."

Considering it takes 8x A100 GPUs (80GB VRAM) to train GPT-2, I think it'll take far more than a single 4090.

Intel has been doing "tick-tock" for almost 20 years - https://en.wikipedia.org/wiki/Tick%E2%80%93tock_model

Many US states also have their own (free) CORS networks, primarily from transportation departments. NOAA helps keep track of them all. Not many of these are realtime unfortunately.

https://geodesy.noaa.gov/CORS/

US overview:

https://www.arcgis.com/apps/webappviewer/index.html?id=19038...

The customer isn't exactly small either - "UniSuper is an Australian superannuation fund that provides superannuation services to employees of Australia's higher education and research sector. The fund has over 620,000 members and $120 billion in assets (funds under management and total member accounts at 7 July 2021)."[1]

https://en.wikipedia.org/wiki/UniSuper

I see, yes controlling at 2 Hz will not be easy no matter how you slice it. What's preventing you from doing more? At one point I experimented with mine. I think below 5 Hz things got choppy. Greater than 10 didn't do a whole lot.

For reference, Arduplane (RC plane OS/autopilot/etc) uses 50 Hz for defaults - https://ardupilot.org/plane/docs/parameters.html#sched-loop-..., and quite a bit higher for copter (at least 400 Hz).

I'd be interested to know how your autopilot reacts with just the P & I portions. Everything I've read/learned about PID loops say to incorporate each term sequentially, meaning do P first, then introduce I to correct long term error, and finally introduce D to prevent overshoot. Not many use only P & D.

Austin from the referenced post here - always love to see the shoutout! What you did is pretty much what I wanted to do when I set out. But definitely not in JS haha, not a huge fan of it

Anyways, I chose to go with PID control for my control systems, which worked really well right "out of the box" with some somewhat generic values for P, I, and D.

I read/skimmed all the way through your post. It looks like your control system is more or less a P & D controller? Any reason you didn't just go with PID loops? Is this part of needing things to be stateless to work for JS (I am no JS expert, just saw a few mentions of that in the conclusion)?

The previous mixtral is 8x7B

Also used in oil/gas downhole in corrosive environments (primarily CO2/H2S). Typical tubing size is 2-7/8", which is ~10.4 lb/ft x thousands of feet. Not cheap.

https://super-metals.com/news/inconel-alloy-718-and-725-in-o...

And Sept 2022 - https://www.nytimes.com/2022/09/01/us/california-heat-wave-f...

Couldn't find a 2023 news article though

Incredible work. Crazy that this could be done by a single person. Stick it on a rotating pedestal and you've got a planar radar detector. Add some tilt and then it's not much different than aircraft/weather radar I suppose.

The cost (in $LOCAL_CURRENCY, or $570 according to another comment) isn't great but I can only imagine how many hours this took.

Given a proper budget, the sky is the limit.

Anyone know how .mil aircraft would interpret being tracked by a 6 GHz radar build by a civilian(yes I am aware that his estimated max distance is 1200m, assume he increased that by a factor of 10 with larger antennas or something)?

Yep, I've got plans. But other projects that compete for time, including raising a 1.5 and 3.5 year old. And working on a Tinder for Restaurants (https://monchmatch.com).

Plans include: Direct to, navigating from waypoint to waypoint, automated takeoff, and maybe tracking a ILS/LOC.

Is this what became Xavion or did Xavion turn into this?

The numbers coming from X-Plane are indeed super "clean" and very mathematically expected. No lag on VSI, etc

Austin from Austin's Nerdy Things here - super stoked to see my blog's traffic at 7x the daily average at 7am! Thanks to all the visitors!

I actually revisited this series a few weeks ago. I added a Flask/html frontend to control the autopilot setpoints (much easier/fluid than editing the code and restarting) using Redis as a quick intermediate "database" to store the setpoints.

I whipped up a post this morning demonstrating this to get it out there due to the volume of visitors who might find it interesting! Hopefully I find time to polish the post later.

Here it is https://austinsnerdythings.com/2024/03/24/adding-some-polish...

I'm currently messing around with a u-blox timing chip, the LEA-6T. It (and many other slightly-above-consumer grade u-blox chip) has a "survey-in" mode, where you tell it "yes, you are stationary" and it essentially averages into a fixed location. This was a $30 chip from China and it gets to 4cm accuracy in 10 min with a $10 antenna. Do that once and you have a location. If the chip reboots, you can either survey-in again, or just tell it where it is.

Highly likely the GPS module inside the Starlink device has this ability. Whether or not that functionality is exposed is up to the Starlink engineers I suppose.

Nice headline but batteries do not generate electricity

I was listening to it on audio book and after the "5000 years later" I paused it and didn't pick it back up for a long while. Slogged through the rest. Won't read/listen to it again. The ending was ok. This was my first Stephenson book. Not going to do more.

I work for a US subsidiary of a very large oil company. We are migrating from Azure to AWS for many things (it is deemed "OneCloud"). A very large number of our new EC2 instances, and even our EKS instances, were provisioned within the last 6 months as T2 instances. Some, if we were lucky, were T3. T3 was released 10 years ago. Copy + paste indeed.

These exist. I am on the fence about buying the $600 one for experimentation and repurposing it to heat/cool garage. Many have heat/cool functionality. You can put the air side either inside or outside.

https://www.amazon.com/s?k=pool+heat+pump&crid=1LWMIT8JDL7S8...

My idea is to do some ground source heat pump supplemented by solar thermal water heating to get lots of free BTUs. The issue (for the $600 one) is it can only cool the water down to 46F. Recall that cooling water means heating air (inside). But heating water goes down to I think 23F. Decisions will have to be made.

I wrote a super basic 2 axis "autopilot" with Python for X-Plane. The concepts are super simple. Getting the controller tuned properly is not.

For a surprisingly large number cases, setting a P value in the 1-3 range and a I value as P/10 works extremely well.

I now write Python for my day job. This code was written before that and now that I look at it, it is quite ugly:

https://austinsnerdythings.com/2021/10/19/coding-a-wing-leve...

A different part of my day job includes, tangentially, tuning PID control parameters for 200-500 HP electric submersible pumps in oil wells. Again, it is amazing how setting an P value in the 1-3 range and an I value P/10 works there as well.

A slightly more advanced autopilot with altitude set and hold (code is even more ugly):

https://austinsnerdythings.com/2021/12/06/coding-a-pitch-rol...

Vast majority of large, publicly traded companies (at least in the US) only do pay determination (and associated raises) once per year. Promotions can definitely happen whenever but there may be a cycle associated with that as well.

My company does week in-office, then week remote. Quite a bit of flexibility during the in-office weeks too if you have a doctor appt or whatever. I personally really like it. People get stuff done in-office, and then during the remote weeks we bust stuff out that doesn't need to be face-to-face.

We are not at all a tech company.

Equally important: how much would said geothermal cost?