Hacker News: trsohmers

New comment by trsohmers in "John Ternus to become Apple CEO"

trsohmers — Mon, 20 Apr 2026 20:53:50 +0000

Software people, in my very direct experience, are terrible at hardware... While in jest, I do think most software engineer's understanding of hardware abstractions is pretty poor and does disservice to the hardware they run on.

I know between Moore's Law and Gate's Law which one I would prefer to be the industry standard... https://en.wikipedia.org/wiki/Andy_and_Bill%27s_law

New comment by trsohmers in "Running Tesla Model 3's computer on my desk using parts from crashed cars"

trsohmers — Wed, 25 Mar 2026 22:22:53 +0000

It actually stands for "lizard brain"... it is (or at least was) an Infineon Aurix control and monitoring microcontroller, they may have changed to a newer one.

New comment by trsohmers in "Positron's $230M Funding Led by Financial Trading Firms"

trsohmers — Wed, 04 Feb 2026 14:09:17 +0000

Feel free to ask me any questions!

Website: https://positron.ai

Positron's $230M Funding Led by Financial Trading Firms

trsohmers — Wed, 04 Feb 2026 14:07:55 +0000

Article URL: https://www.eetimes.com/positron-230-million-funding-led-by-financial-trading-firms/

Comments URL: https://news.ycombinator.com/item?id=46885964

Points: 1

# Comments: 1

The New Chips Designed to Solve AI's Energy Problem

trsohmers — Mon, 28 Jul 2025 17:36:24 +0000

Article URL: https://www.wsj.com/tech/ai/the-new-chips-designed-to-solve-ais-energy-problem-1ba9cac1

Comments URL: https://news.ycombinator.com/item?id=44713225

Points: 2

# Comments: 0

New comment by trsohmers in "Show HN: Buckaroo – Data table UI for Notebooks"

trsohmers — Sun, 18 May 2025 21:46:14 +0000

Only with the oscillation overthruster flag enabled.

New comment by trsohmers in "Charlie Javice convicted of defrauding JPMorgan in $175M startup sale"

trsohmers — Tue, 01 Apr 2025 04:32:46 +0000

I was put on it in 2015 after an acquaintance of mine that was previously on the list recommended me… I only heard from Forbes a few days before the list came out, they asked me for a photo and asked if I approved the 2 sentence blurb they prepared, and that was it. For years afterwards they would try to get me to come to their events, but I never had any interest, and I assume that was how they made money… but I never paid anything to be on the list or had real interest in being on it, and I don’t think it led to anything other than my technically illiterate parents thinking that it was impressive.

New comment by trsohmers in "Llama 3.1 405B now runs at 969 tokens/s on Cerebras Inference"

trsohmers — Tue, 19 Nov 2024 06:21:42 +0000

Based on their S1 filing and public statements, the average cost per WSE system for their (~90% of their total revenue) largest customer is ~$1.36M, and I’ve heard “retail” pricing of $2.5M per system. They are also 15U and due to power and additional support equipment take up an entire rack.

The other thing people don’t seem to be getting in this thread that just to hold the weights for 405B at FP16 requires 19 of their systems since it is SRAM only… rounding up to 20 to account for program code + KV cache for the user context would mean 20 systems/racks, so well over $20M. The full rack (including support equipment) also consumes 23kW, so we are talking nearly half a megawatt and ~$30M for them to be getting this performance on Llama 405B

New comment by trsohmers in "Meta's open AI hardware vision"

trsohmers — Tue, 15 Oct 2024 19:56:48 +0000

Do you think that the 16k GPUs get used once and then are thrown away? Llama 405B was trained over 56 days on the 16k GPUs; if I round that up to 60 days and assume the current mainstream hourly rate of $2/H100/hour from the Neoclouds (which are obviously making margin), that comes out to a total cost of ~$47M. Obviously Meta is training a lot of models using their GPU equipment, and would expect it to be in service for at least 3 years, and their cost is obviously less than what the public pricing on clouds is.

New comment by trsohmers in "Visit Bletchley Park"

trsohmers — Fri, 30 Aug 2024 01:48:12 +0000

+1 this commenter. I just visited the UK for the first time at the beginning of this month and had a fantastic ~3 hours at Bletchley Park, but felt I had to cram TNMOC and the amazing Colossus live demonstration (where I asked a million questions) and everything else in the museum in the 90 minutes I was there. If I assume other HN readers are like me, I would dedicate at least 2.5-3 hours for TNMOC to actually get a chance to actually see and play around with their extensive collection of vintage machines.

An "Observatory" for a Shy Super AI?

trsohmers — Fri, 09 Aug 2024 15:49:38 +0000

Article URL: https://substack.com/@robreid/p-147352947

Comments URL: https://news.ycombinator.com/item?id=41202970

Points: 3

# Comments: 0

New comment by trsohmers in "Codestral Mamba"

trsohmers — Tue, 16 Jul 2024 17:06:23 +0000

They meant that there is no support for Codestral Mamba for llama.cpp yet.

New comment by trsohmers in "Rex Computing"

trsohmers — Thu, 30 May 2024 02:27:40 +0000

We had a basic LLVM backend that supported a slightly modified clang frontend and a basic ABI. We tried to make it drastically easier for both the programmer and compiler to handle memory by having all memory (code+data) be part of a global flat address space across the chip, with guarantees being made to the compiler by the NoC on the latency of all memory accesses across one or multiple chips. We tested this with very small programs that could fit in the local memory of up to two chips (128KB of memory), but in theory it could have scaled up to the 64 bit address space limit. Compilation time for programs was long, but fully automated, specifically to improve upon problems faced by Cell and other scratchpad memory architectures… some of our original funding in 2015 from DARPA was actually for automated scratchpad memory management techniques on Texas Instruments DSPs and Cell (our paper: https://dl.acm.org/doi/pdf/10.1145/2818950.2818966)

This was all designed a decade ago, and REX has been in effectively hibernation since the end of 2017 after successfully taping out our 16 core test chip back in 2016, but being unable to raise additional funding to continue. I have continued to work on architectures that have leveraged scratchpad memories in different ways, including on cryptocurrency and machine learning ASICs, including at my current startup, Positron AI (https://positron.ai)

New comment by trsohmers in "Rex Computing"

trsohmers — Thu, 30 May 2024 01:06:08 +0000

Founder of REX Computing here; I highly recommend checking out my interview on the Microarch Club podcast linked elsewhere on the thread; will also answer questions on this thread if anyone has them.

New comment by trsohmers in "Putting a Wet Towel on a Tesla Supercharger Handle Gets Faster Charging Speeds"

trsohmers — Thu, 09 May 2024 22:41:59 +0000

This is a lesson that like all good Hitchhikers, you should always carry a towel.

New comment by trsohmers in "Building Meta's GenAI infrastructure"

trsohmers — Tue, 12 Mar 2024 18:22:46 +0000

Significantly more than that; MFN pricing for NVIDIA DGX H100 (which has been getting priority supply allocation, so many have been suckered into buying them in order to get fast delivery) is ~$309k, while a basically equivalent HGX H100 system is ~$250k, coming to a price per GPU at the full server level being ~$31.5k. With Meta’s custom OCP systems integrating the SXM baseboards from NVIDIA, my guess is that their cost per GPU would be in the ~$23-$25k range.

New comment by trsohmers in "Who uses Google TPUs for inference in production?"

trsohmers — Mon, 11 Mar 2024 21:57:37 +0000

The quote from the linked press release is that they do training on TPUv4, while inference is running on GPUs. I have also heard this separately from people associated with Midjourney recently, and that they solely do training on TPUs.

New comment by trsohmers in "Dune: Part Two Is the Best Sci-Fi Film of the Decade"

trsohmers — Thu, 29 Feb 2024 19:20:59 +0000

I’m right on the millenial/gen Z divide and an inner selfish purpose for me working on AI/ML is just to enable a creation of Jodorowsky’s 10 hour version of Dune with soundtrack by Pink Floyd.

New comment by trsohmers in "Groq runs Mixtral 8x7B-32k with 500 T/s"

trsohmers — Tue, 20 Feb 2024 06:31:11 +0000

Long story, but technically REX is still around but has not been able to continue to develop due to lack of funding and my cofounder and I needing to pay bills. We produced initial test silicon, but due to us having very little money after silicon bringup, most of our conversations turned to acquihire discussions.

There should be a podcast release (https://microarch.club/) in the near future that covers REX's history and a lot of lessons learned.

New comment by trsohmers in "Groq runs Mixtral 8x7B-32k with 500 T/s"

trsohmers — Mon, 19 Feb 2024 19:02:35 +0000

I thought that was clear through my profile, but yes, Positron AI is focused on providing the best performance per dollar while providing the best quality of service and capabilities rather than just focusing on a single metric of speed.

A guarantee to match the cheapest per token prices is sure a great way to lose a race to the bottom, but I do wish Groq (and everyone else trying to compete against NVIDIA) the greatest luck and success. I really do think that the great single batch/user performance by Groq is a great demo, but is not the best solution for a wide variety of applications, but I hope it can find its niche.