Perplexity Brings Portable Computer to Windows RTX PCs for Free Local AI
Perplexity's local agent stack now runs on Windows RTX PCs, with on-device MCP servers and scheduled task automation joining the release.
// Intelligence
Gpus
A focused feed of models, agents, research and open-source releases for people building with AI.
Updated
Edge0 Streams an 8B AI Model From SSD Using Only 1 GiB
Edge0 releases an 8B sparse MoE that runs in under 1 GiB of active memory on Apple Silicon, streaming experts from SSD on demand.
Mia's AI Lab Squeezes DeepSeek V4.1 Flash's 552B Model onto Two Desktop Nodes
A community quantization squeezes DeepSeek's 552B multimodal MoE onto two desktop DGX Sparks, holding decode speeds across half a million tokens of context.
Edge0 Runs a 35B AI Model on an iPhone With Just 2.9 GiB
Edge0 streams a 35B Mixture-of-Experts model from SSD on an iPhone, holding under 3 GB of active RAM while decoding at 15 tokens per second.
Cerebras Runs Alibaba's Qwen 3.8 27B at 1,850 Tokens per Second
Alibaba's 27B dense multimodal model lands on Cerebras at roughly 1,800 tokens per second, with reasoning on by default and a 128K context on paid tiers.
PyTorch 2.14 Ships Faster Apple Silicon and 100x GPU Speedups
PyTorch 2.14 lands with a new CUTLASS-based GEMM backend, in-place fault tolerance for distributed training, and native linear algebra on Apple Silicon.
Together AI's ThunderKittens Hits 22.4 PFLOPS on NVIDIA's Rubin
Together AI ported its ThunderKittens kernel framework to NVIDIA's Vera Rubin NVL72, hitting 22.4 PFLOPS on NVFP4 GEMMs and rivaling cuBLAS.
Epoch Reveals OpenAI's Computing Power Grew 17x in Two Years
Epoch AI's new explorer estimates compute capacity for OpenAI, Google DeepMind, Anthropic, Meta, and xAI, revealing a 17x surge at OpenAI.
Cognition's Devin Breaks RSA-260 Record in Three Weeks With AI Agents
Cognition used Devin agents to build a GPU lattice siever that factored RSA-260, cutting factorization costs by roughly 10x versus prior public state of the art.
Intelligent Internet's Meta-Zenith Agent Rewrote vLLM Kernels for a 4x Speedup
An autonomous agent ran 111 trials to optimize vLLM for Qwen3.8-27B on a single RTX 5090, hitting 4.1x throughput at 65K context.
NVIDIA's CUDA Python 1.0 Makes Python a First-Class GPU Citizen
NVIDIA ships CUDA Python 1.0 with semantic versioning, a shared cuda.core foundation, and PyTorch and CuPy already building on it.
cmpunlocker Turns NVIDIA's Dead Mining GPU Into a 64GB Powerhouse
An open-source patch restores 64GB of HBM2e memory and full compute on NVIDIA's crippled CMP 170HX mining card, transforming a $200 crypto relic into a viable AI accelerator.
OpenBMB's MiniCPM5-2B Beats Larger Rivals With 128K Context on Consumer Hardware
OpenBMB's compact 2B model hits open-source SOTA in its class, beating several 4B rivals on code, math, and agent benchmarks while running locally.
H-BAC Shrinks a 327 MB Plant Disease Model Down to 6 MB
A unified pruning, quantization and distillation pipeline shrinks a Vision Transformer 54.5x while holding 95.13% accuracy on out-of-distribution chilli disease images.
FlashAttention-4's Direct-P Finally Unlocks 2.13x FP4 Speedup on NVIDIA GB200
A new attention kernel unlocks Blackwell's FP4 tensor cores for inference, hitting 2.13x BF16 forward throughput on GB200 while training stays partly in FP8.
Figure Bets $6B on NVIDIA Vera Rubin GPUs to Scale Humanoid Robots
Figure locks in a $3.5B compute commitment scaling past $6B, targeting up to 100,000 Vera Rubin GPUs to train its Helix humanoid AI.
PyTorch 2.14 Ships Faster Apple Silicon, Rebuilt Distributed Training and NVGEMM
The latest PyTorch release ships NVGEMM CUTLASS kernels for Inductor, a rebuilt nccl2 backend, native Apple Silicon linear algebra, and first-class fault tolerance.
PeasantSmith Squeezes Alibaba's 177B Qwen3.8-Flash-Next Into a 75 GB Local File
A community-built IQ2_XXS quantization squeezes Qwen3.8-Flash-Next's 177B parameters into a 75GB GGUF that runs on ~42GB of memory.
MATLOWAI Rebuilds MiniMax H3 to Run Video and Audio on 16GB Cards
A 21 GB single-file build of MiniMax-H3 renders 10-second clips with synchronized stereo audio in about 76 seconds at 4 steps.
Alibaba's Qwen3.8-Flash-Next Shrinks 125B Model to 84 GiB on One Machine
A community quant of Qwen3.8-Flash-Next shrinks the 177B MoE to 84 GiB with a per-layer mixed-precision recipe that beats standard IQ4_XS on both size and quality.
AngelSlim Shrinks Tencent's 1.5TB Hy4-Preview to 214 GiB at 6x Compression
Tencent's AngelSlim team shrinks the 770B Hy4-preview MoE from 1.5TB down to 213GB using a custom 1.31-bit quantization strategy with minimal accuracy loss.
UNSEEN Gemma 4 Removes AI Refusals With a Single Math Trick
A 26B multimodal Gemma variant with refusal directions surgically removed lands on Hugging Face in GGUF format, ready for llama.cpp.
SuperQwen3.8 Runs Uncensored on One Box at 1.77x Faster Speeds
A community fine-tune of Qwen3.8-27B strips refusals, fixes the resulting weight damage, and ships in four formats tuned for local inference.
Qwen3.8-27B Hits 417 tok/s on a Single RTX 3090 With 150k Context
A community project squeezes Qwen3.8-27B onto a 24GB gaming card with vLLM, hitting 417 tok/s batched or 82 tok/s single-user at 150k context.
No stories match these filters.