Install
GPUs & Accelerators
AI/graphics processors, memory bandwidth, and massive‑scale training chips.
- 25 Tracked terms
- Last 30 days Feed window
What this topic collects on
An article joins this feed when it matches these terms. Each one is also a search of its own.
Related topics
Latest in GPUs & Accelerators
Training my first language model on a GTX 1650
1+ week, 14+ hour ago (673+ words) A 4 GB GPU, a small corpus and a lot of things that went wrong in instructive ways. The story around AI.001, with a few examples of what the model actually produced. Using ChatGPT, or pulling a model through Ollama, makes…...
I Trained a Language Model From Scratch on a 4 GB GPU
34+ min ago (526+ words) I trained a 22-million-parameter language model from scratch on a laptop with 4 GB of VRAM. It learned to imitate technical documentation, got stuck repeating fragments of language, and never produced a working Fibonacci function. Near the end, CUDA crashed. The…...
I Put Three AI Services on One 24 GB GPU. It Used 17.8. Here's What Fits.
1+ hour, 18+ min ago (884+ words) The Local LLM Lab — Part 9 One controlled experiment. One number. One verdict. The "how much VRAM do I need" advice on the internet comes in one flavour: a spec sheet. "27B in FP8 needs 28 GB. 70B in Q4 needs 42 GB. Buy a 48 GB card....
My GPU Was Busy. I Finally Found Out Who Was Using It.
1+ hour, 23+ min ago (645+ words) The Local LLM Lab — Part 10 One controlled experiment. One number. One verdict. Every homelab has a moment like this: the GPU is at 100%, the fan is screaming, and you have no idea which of the twenty services on the box…...
GPU Utilization: The Metric That Matters More Than GPU Count
1+ hour, 31+ min ago (1253+ words) For years, GPU capacity has been discussed in terms of one number: how many GPUs are available? A data center has 1,000 GPUs.A cloud provider has 10,000 GPUs.An AI factory is planning 100,000 GPUs. These numbers are useful for describing infrastructure…...
Samsung starts mass-producing the Exynos 2700, and it could finally challenge Snapdragon
1+ week, 1+ day ago (273+ words) Samsung is reportedly stepping up its Exynos push with the Exynos 2700, which has now entered mass production. Built on Samsung’s second-generation 2nm process, the chip is expected to power several Galaxy S27 models. Internal testing has reportedly shown promising results against Qualcomm’s…...
AI waste disposal app 'Gajyeoga' to launch pilot service in Seoul district
3+ hour, 11+ min ago (271+ words) South Korea's Ministry of Science and ICT and the Ministry of Climate, Energy and Environment said they will begin a pilot run of the AI-based waste disposal support app 'Gajyeoga' in Seoul's Eunpyeong district from Oct. 13. The app identifies waste…...
Surged Over 7x From Its Low, How Does L1 Blockchain Pearl Price GPU Compute?
7+ hour, 48+ min ago (1421+ words) The Bloomberg terminal displays GPU price indices from Silicon Data, Ornn, and Compute Desk. All three track H100 rental prices, yet they cannot agree on the actual value of a single H100. Most GPU compute flows through private bilateral contracts at discounted…...
SCSKiller Targets the Stutter Your GPU Can’t Outrun
6+ hour, 10+ min ago (759+ words) SCSKiller moves shader compilation before play. Its striking PC results target one kind of stutter—and the new update is tackling compatibility snags. SCSKiller Targets the Stutter Your GPU Can’t Outrun SCSKiller is attracting attention for a promise PC players…...
What I Learned Benchmarking the Same LLM on My Phone's NPU, GPU, and CPU
3+ hour, 40+ min ago (861+ words) Most discussions about NPU inference are either vendor press releases or synthetic benchmarks on models nobody ships. I wanted something different: real numbers, from a real app, doing real work, on a phone you can actually buy. This post breaks…...