UK.gov's nuclear strategy is 'slow, inefficient, and costly' dlvr.it/TMRN7R

Still need to do some cable management and it remains to be seen whether that PSU will be adequate, but it posted. 24 Zen4 cores 128GB DDR5 5600 RDIMM 2x RTX 3090 TI 24GB w/ NVL And don’t worry those GPUs are power limited to 250W. The PSU shouldn’t explode.

Meanwhile, if you haven't figured it out already GPT-5 is a huge improvement, just mostly for OpenAI's bottom line. This thing is a big cost cutting measure and not necessarily in a bad way: My latest for The Register theregister.com/2025/08/13/gpt…

Got GPT-OSS 120B running tonight in Llama.cpp. ~60Tok/s with minimal expert offload to DRAM. Had to compile a pull request to make it play nice to Open WebUI, but I have tool calling now. Need to figure out a system prompt cause my god this thing loves tables.

Need one of these for review. I want to type -tp 8 into vLLM.

Gentle reminder that GPUs aren't just for AI. In the right hands they can advance life saving science. Boffins at LLNL used the No. 1 ranked supercomputer, El Capitan, to build a real-time tsunami forecast system. My latest for @theregister theregister.com/2025/08/13/tsu…

Main thing confusing me: why give up an optimized training stack built around Nvidia to take a chance on Huawei? Like other than national pride. The 910C doesn't support FP8, so back to BF16. Were they using them for RL? Latest for The Register theregister.com/2025/08/14/dod…

That's my arm if you're wondering... Not sure why I felt the need to share that, but now you know

OpenAI's decision to use MXFP4 datatypes as the standard for gpt-oss is a big deal and sets the tone for the rest of the industry. Is it perfect? No, but it's a huge improvement over FP4 or INT4. Find my deep dive @theregister theregister.com/2025/08/10/ope…

Compiling code an Raspberry Pi 5 is painfully slow

Sneak peek at something I've got cooking.

Called it! GPT-5 was a cost cutting measure. "We have better models, and we just can't offer them because we don't have the capacity. We have other kinds of new products and services we'd love to offer," Altman said. ICYMI: theregister.com/2025/08/13/gpt…

Cisco, Arista, Broadcom and every other Ethernet equipment vendor is jazzed about AI and its not hard to see why when when ever GPU is an excuse to sell 3-5 of the fastest switch ports money can buy: My latest for @theregister theregister.com/2025/08/15/eth…

Last I checked, Uncle Sam hasn't yet resorted to using human batteries to power major industry or AI development. But sure, compare the US government to the Matrix: theregister.com/2025/08/18/chi…

AMD Radeon Inference perf in Llama.cpp is an interesting conundrum. The Vulkan backend offers higher token gen but ROCm is vastly superior in prompt processing. Which do you opt for?

The way the report describes the B30A it sounds a lot more like a B300 NVL than a H20 replacement... My latest for The Register theregister.com/2025/08/19/nvi…

So wait. Is Arm really getting into silicon? Seems like a sure fire way to piss off your customers. Maybe chiplets would be okay? Most Arm customers take off-the-shelf cores anyway. My latest for The Register theregister.com/2025/08/19/arm…

In my latest hands on for @theregister I break down everything you need to know to run large language models in the privacy of our home using Llama.cpp. theregister.com/2025/08/24/lla… #AI #PC #LLM #HomeLab

After three years of 24/7 service in my homelab my R9 3900X met its end on Sunday. During routine thermal paste change the cooler ripped it from the socket bending several pins. Alas even when bent back into position it refused to post. RIP my friend #AMD #CPU #PC #Homelab

GPT-5 is the completely broken. I want o4-mini back.