Nvidia A100
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 8
- Tensor Core FP32 methods speed up MLIP molecular dynamics
NVIDIA Technical Blog reports that TF32x3 and BF16x9 Tensor Core methods accelerated PET machine learning interatomic potential workloads without retraining or changing model weights or architecture. On an A100, TF32x3 sped up major GEMMs by up to 1.9x and complete molecular dynamics steps by up to 1.5x, while BF16x9 on an NVIDIA B300 reached 2.5x and 1.7x speedups, respectively. The tests also examined force accuracy, energy conservation, structural relaxation, and numerical behavior close to FP32.
NVIDIA Technical Blog · 🔥 0 - Core 4 Solutions Lists 300-Rack A100 AI Supercomputer on eBay
Core 4 Solutions listed a 300-rack AI supercomputer on eBay for $59.7 million, according to Tom’s Hardware. The advertised cluster includes 9,600 Nvidia A100 accelerators and 4,200 AMD EPYC Rome 64-core processors, with $50 overnight delivery, free FedEx Ground shipping, or free local pickup. The reports do not confirm that the equipment is operational or that a sale has been completed.
Tom’s Hardware · 🔥 0 - FluxVLA Engine V2 expands the embodied AI development stack
LimX Dynamics updated the open-source FluxVLA Engine V2 technical report, adding online reinforcement-learning post-training, broader simulation evaluations, inference optimization, and deployment testing across ALOHA, Rebot, dual-arm Franka, and LimX Oli. The project reports that SmolVLA improved by 10.0 percentage points on LIBERO-10 and 15.4 points on ten RoboTwin Clean tasks after task-specific PPO post-training, while optimized inference paths across eight model-device combinations ran 2.31 to 9.64 times faster. The stack also supports multiple VLA architectures, including OpenVLA, π0.5, DiT4DiT, GR00T, and API-only policies, through shared data, training, evaluation, inference, and deployment workflows.
雷峰网 AI · 🔥 0 - AI Compute Rental Market Shows a Sharp Supply Mismatch
A TMTPost analysis reports that H100 median rental prices fell from more than $7 per hour in early 2024 to about $3.4 by mid-2026, while some lower-end compute-node rents dropped from 30,000 yuan to below 6,000 yuan a month. It cites industry estimates that China’s intelligent-computing centers average about 30% utilization, versus continued shortages of newer Blackwell B200 cards priced above $8 per hour. The report attributes the divergence to older GPUs losing rental demand as major customers build their own clusters and model API prices fall, while supply remains uneven across locations and generations.
钛媒体 · 🔥 0 - GHARP brings real-time Gaussian head animation to mobile
GHARP is a method for animating 3D human heads in real time from a few images of a subject and a driving expression signal. Its authors say it separates an offline identity-reconstruction stage from a lightweight runtime animation stage optimized for mobile devices. The runtime predicts expression-dependent residuals in the semantically structured latent space of a pretrained reconstruction model. The authors also report that ambiguous body appearance can cause blur and temporal flicker, which they address with a body alignment network. They claim state-of-the-art quality on Ava-256, execution up to 13 times faster on an A100 GPU, and eight times fewer Gaussians; the supplied evidence does not independently verify these results.
Apple Machine Learning Research · 🔥 0 - AI Demand Raises NVIDIA GPU Rental Prices
Wccftech, citing Ornn Data statistics reported by a16z, says NVIDIA GPU rental prices increased from January 1, 2026. The H200 recorded the largest rise at 143%, followed by the B200 at 111%, RTX PRO 6000 at 73%, H100 at 54%, and A100 at 9%. The report also says the Tesla V100 and other older NVIDIA GPUs have gained popularity during the AI boom.
Wccftech · 🔥 0 - Microsoft open-sources TRELLIS.2, a 4B image-to-3D model with PBR output
Microsoft has published TRELLIS.2 on GitHub, a 4-billion-parameter image-to-3D generative model built on a "field-free" sparse voxel representation called O-Voxel, which the repo says handles open surfaces, non-manifold geometry and internal enclosed structures and predicts PBR surface attributes such as base color, roughness, metallic and opacity. The project reports generation times on an NVIDIA H100 of about 3 seconds at 512³ (2s shape + 1s material), roughly 17 seconds at 1024³ and about 60 seconds at 1536³, plus under 10 seconds on a single CPU to convert a textured mesh to O-Voxel and under 100ms on CUDA for the reverse. Code, data tooling and training scripts are provided and the TRELLIS.2-4B weights are on Hugging Face, but the code is Linux-only and needs an NVIDIA GPU with at least 24GB of memory (verified on A100 and H100, CUDA 12.4).
GitHub Trending · Python · 🔥 0 - SparseDecoding: decoding-aware pruning for faster LLM inference
The paper introduces SparseDecoding, a training-free network pruning framework aimed at the decoding stage of LLM inference. It builds calibration matrices from layer-wise activations collected during dense-model autoregressive generation (excluding prefill) so the pruning objective matches decoding activations, and pairs this with an N:M sparse matrix-vector kernel using bitmask indexing and fixed-step traversal. The authors report that on Llama-3.1-8B, Llama-3.3-70B and Qwen3-14B/32B it outperforms fixed-text calibration on long-form generation benchmarks, with up to 1.48x end-to-end wall-clock decoding speedup on A100 GPUs.
Hugging Face · Papers · 🔥 0
Experience and discussion from the community
Share my Nvidia A100 experienceAsk about Nvidia A100
Nobody has shared their experience with Nvidia A100 yet.