--- title: "#gpu articles | Hashnode" description: "Explore 544 articles tagged #gpu on Hashnode. Tutorials, guides, and how-tos from the community." url: "https://hashnode.com/tag/gpu" --- Tag # #gpu 544 posts · 18 followers [reshmi aravind](https://hashnode.com/@raravind007) · [12h ago](https://hashnode.com/posts/from-gpu-pids-to-model-ids-which-vllm-model-is-using-my-gpu/6af45b9e6c263bd581ab7c77) ### [From GPU PIDs to Model IDs: Which vLLM Model Is Using My GPU?](https://practical-ml-systems.hashnode.dev/from-gpu-pids-to-model-ids-which-vllm-model-is-using-my-gpu?utm_source=hashnode&utm_medium=feed) When running LLM inference workloads, tools like nvidia-smi and nvtop help identify GPU utilization, memory consumption, and the processes using those resources. But they don't directly answer another 11 min read · [practical-ml-systems.hashnode.dev](https://practical-ml-systems.hashnode.dev/from-gpu-pids-to-model-ids-which-vllm-model-is-using-my-gpu?utm_source=hashnode&utm_medium=feed) [Shubham Tatvamasi](https://hashnode.com/@ShubhamTatvamasi) · [2d ago](https://hashnode.com/posts/vllm-deep-dive-how-pagedattention-continuous-batching-and-the-v1-engine-serve-llms-in-production/6af1a0c4a32bf9f449f76bd3) ### [vLLM Deep Dive: How PagedAttention, Continuous Batching and the V1 Engine Serve LLMs in Production](https://blog.shubhamtatvamasi.com/vllm-deep-dive-how-pagedattention-continuous-batching-and-the-v1-engine-serve-llms-in-production?utm_source=hashnode&utm_medium=feed) Every layer above it routes, scales and governs requests. vLLM is the layer that turns GPU memory into tokens. 52 min read · [blog.shubhamtatvamasi.com](https://blog.shubhamtatvamasi.com/vllm-deep-dive-how-pagedattention-continuous-batching-and-the-v1-engine-serve-llms-in-production?utm_source=hashnode&utm_medium=feed) [3](https://hashnode.com/posts/vllm-deep-dive-how-pagedattention-continuous-batching-and-the-v1-engine-serve-llms-in-production/6af1a0c4a32bf9f449f76bd3) [Viswaretas Kotra](https://hashnode.com/@viswa-kotra) · [1d ago](https://hashnode.com/posts/keeping-training-state-when-the-gpu-node-changes/6af36ff0293fffa36e514a86) ### [Keeping training state when the GPU node changes](https://nodus-compute.hashnode.dev/keeping-training-state-when-the-gpu-node-changes?utm_source=hashnode&utm_medium=feed) A controlled Nodus node replacement preserved a checkpoint and advanced training from step 8 to 9. Evidence, timing boundaries and limitations. 6 min read · [nodus-compute.hashnode.dev](https://nodus-compute.hashnode.dev/keeping-training-state-when-the-gpu-node-changes?utm_source=hashnode&utm_medium=feed) [netproxy](https://hashnode.com/@argumentmoney8117) · [1d ago](https://hashnode.com/posts/google-s-new-chips-split-ai-in-two-directions-private-and-enormous/6af29301bcd0b634103a91fb) ### [Google's New Chips Split AI in Two Directions: Private and Enormous](https://pharaoh-pig.hashnode.dev/google-s-new-chips-split-ai-in-two-directions-private-and-enormous?utm_source=hashnode&utm_medium=feed) On Monday, Google added two systems to its eighth-generation TPU line: TPU 8tx and TPU 8i Max. With the TPU 8t and 8i unveiled back in April, that makes four TPU systems announced in a single year — a 3 min read · [pharaoh-pig.hashnode.dev](https://pharaoh-pig.hashnode.dev/google-s-new-chips-split-ai-in-two-directions-private-and-enormous?utm_source=hashnode&utm_medium=feed) [sheephess9527](https://hashnode.com/@sheephess9527) · [1d ago](https://hashnode.com/posts/where-does-a-70b-model-actually-fit-dual-rtx-3090-vs-dgx-spark-64gb-vs-128gb/6af28648870d07ceeb681e1f) ### [Where does a 70B model actually fit? Dual RTX 3090 vs DGX Spark 64GB vs 128GB](https://sheephess9527.hashnode.dev/where-does-a-70b-model-actually-fit-dual-rtx-3090-vs-dgx-spark-64gb-vs-128gb?utm_source=hashnode&utm_medium=feed) 70B dense models are the point where a single consumer GPU stops being enough. Here is what the memory math says for the three setups people usually compare. The baseline: Llama 3.3 70B at Q4\_K\_M Abou 3 min read · [sheephess9527.hashnode.dev](https://sheephess9527.hashnode.dev/where-does-a-70b-model-actually-fit-dual-rtx-3090-vs-dgx-spark-64gb-vs-128gb?utm_source=hashnode&utm_medium=feed) [CloudGPU](https://hashnode.com/@cloudgpu) · [1d ago](https://hashnode.com/posts/make-a-failed-gpu-video-trial-explain-the-next-decision/6af2b6360e83b734bb0fb7de) ### [Make a failed GPU video trial explain the next decision](https://cloudgpu.hashnode.dev/make-a-failed-gpu-video-trial-explain-the-next-decision?utm_source=hashnode&utm_medium=feed) I work on CloudGPU. A failed AI video trial can leave two very different things behind: a usable explanation, or another file that nobody can interpret. The difference is usually decided before the ne 3 min read · [cloudgpu.hashnode.dev](https://cloudgpu.hashnode.dev/make-a-failed-gpu-video-trial-explain-the-next-decision?utm_source=hashnode&utm_medium=feed) [Christopher Smith](https://hashnode.com/@et-ducky) · [2d ago](https://hashnode.com/posts/xid-79-gpu-has-fallen-off-the-bus-on-linux-and-how-to-find-the-cause/6af23db875fecf45c41165d0) ### [Xid 79 GPU Has Fallen Off the Bus on Linux and How to Find the Cause](https://etducky.hashnode.dev/xid-79-gpu-has-fallen-off-the-bus-on-linux-and-how-to-find-the-cause?utm_source=hashnode&utm_medium=feed) This post first appeared on the ET Ducky blog. I build ET Ducky, an RMM that reads ETW on Windows and eBPF on Linux. Xid 79 is an error code the NVIDIA driver writes to the Linux kernel log. NVIDIA's 5 min read · [etducky.hashnode.dev](https://etducky.hashnode.dev/xid-79-gpu-has-fallen-off-the-bus-on-linux-and-how-to-find-the-cause?utm_source=hashnode&utm_medium=feed) [Viswaretas Kotra](https://hashnode.com/@viswa-kotra) · [2d ago](https://hashnode.com/posts/six-wrong-normalization-formulas-passed-a-correctness-check/6af1d05036a54cca5df5fe55) ### [Six wrong normalization formulas passed a correctness check](https://nodus-compute.hashnode.dev/six-wrong-normalization-formulas-passed-a-correctness-check?utm_source=hashnode&utm_medium=feed) Six incorrect normalization formulas passed with identity inputs and failed with varied parameters in separate synthetic CPU and T4 CUDA runs. 9 min read · [nodus-compute.hashnode.dev](https://nodus-compute.hashnode.dev/six-wrong-normalization-formulas-passed-a-correctness-check?utm_source=hashnode&utm_medium=feed) [Sebinraj](https://hashnode.com/@sebinrajs) · [3d ago](https://hashnode.com/posts/the-work-that-never-ran-how-a-kernel-workqueue-bug-froze-a-gpu-and-the-test-that-catches-it/6af0b093ddd20b974b443491) ### [The Work That Never Ran: How a Kernel Workqueue Bug Froze a GPU, and the Test That Catches It](https://sebinrajs.hashnode.dev/the-work-that-never-ran-how-a-kernel-workqueue-bug-froze-a-gpu-and-the-test-that-catches-it?utm_source=hashnode&utm_medium=feed) How one missing list insertion left GPU fences unsignaled forever, and how a stress test turns a hard-reboot hang into a clean failure The scariest bugs are not the ones that crash. They are the ones 17 min read · [sebinrajs.hashnode.dev](https://sebinrajs.hashnode.dev/the-work-that-never-ran-how-a-kernel-workqueue-bug-froze-a-gpu-and-the-test-that-catches-it?utm_source=hashnode&utm_medium=feed) [Viswaretas Kotra](https://hashnode.com/@viswa-kotra) · [2d ago](https://hashnode.com/posts/the-reward-is-part-of-the-dataset/6af192e0befea3fedc931883) ### [The reward is part of the dataset](https://nodus-compute.hashnode.dev/the-reward-is-part-of-the-dataset?utm_source=hashnode&utm_medium=feed) Six wrong normalization formulas passed a CPU checker with identity inputs. What this means for kernel RL feedback and held-out evaluation. 3 min read · [nodus-compute.hashnode.dev](https://nodus-compute.hashnode.dev/the-reward-is-part-of-the-dataset?utm_source=hashnode&utm_medium=feed)