26 related articles

Open-source GPU kernel library fast_trimul optimizes triangle multiplicative update operations in AlphaFold3 family models, achieving 4.5-6.8x speedup with 2.2-2.4x memory reduction for short sequences.

Is GPU parallel simulation the only choice for robot reinforcement learning? UniLabSim argues CPU simulation remains competitive. We analyze the hidden costs of GPU simulation, CPU flexibility advantages, and the tech and business logic behind this compute debate.

Explore how contract-grade verifiers validate LLM-generated GPU kernel correctness, addressing trust issues like race conditions and out-of-bounds access in AI code generation.

Fixed the random seed but GPU training results still differ? This article explains floating-point non-associativity, non-deterministic CUDA ops, and provides a complete PyTorch deterministic training configuration guide.

Basalt is an open-source Minecraft launcher built with Rust and Tauri, featuring artwork-led design, multi-instance management, CurseForge/Modrinth support, smart crash diagnostics, and zero telemetry.

Analysis of how a single NVIDIA B200 GPU surpasses Groq LPU and approaches Cerebras performance through software optimization alone, covering CUDA kernels, TensorRT-LLM, and FP8 quantization.

Just 3 days after MiniMax H3's release, the community delivers a Turbo LoRA that generates quality video in only 10 sampling steps, supporting both I2V and FLF2V modes.

nvidia-smi showing 100% GPU utilization doesn't mean optimal training efficiency. Learn about DCGM, PyTorch Profiler, and MFU metrics for diagnosing real GPU training bottlenecks.

OpenAI's internal model GPT-5.6 reportedly autonomously rewrites production kernels, achieving ~20% service cost reduction. Deep analysis of this AI recursive self-optimization event's technical plausibility and industry impact.

OpenAI's internal model GPT-5.6 reportedly autonomously rewrote production compute kernels, achieving ~20% cost reduction. Deep analysis of this AI recursive self-optimization event's technical plausibility, industry impact, and key questions.

Moonshot AI open-sources FlashKDA, providing high-performance CUDA kernels for Kimi Delta Attention. Explore its technical principles, performance gains, and value for long-context training and inference.

Moonshot AI open-sources FlashKDA, providing high-performance CUDA kernels for Kimi Delta Attention. Learn about its technical principles, performance gains, and value for long-context training and inference acceleration.

Alibaba open-sources a 2.4 trillion parameter Qwen model and launches the Qwen Token Plan. Chinese models surge, Kimi K3 tops global rankings, and China's AI is reshaping the global competitive landscape.

Researchers propose the Deterministic Attention-Transformer, measuring just 0.63 J/token on NVIDIA H100 GPUs. Explore the tech behind it and its green AI implications.
CUDA Kernel Fusion: A Practical Guide …
Learn how CUDA Kernel Fusion merges multiple GPU kernels to reduce global memory traffic and launch overhead, with real-world examples from AI inference and deep learning.

A self-learner completed a full progression from math foundations and core ML to deep learning in 6 months—hand-writing a Transformer and implementing gradient boosting from scratch. This article breaks down the highlights and blind spots of this real roadmap.

DeepSeek is entering AI chip development, targeting compute autonomy. This article analyzes its motivations, software-hardware synergy, chip R&D challenges, and impact on China's AI vertical integration.

SGLang's team converted expert knowledge into agent skills, achieving 71.4% throughput gains, TTFT reduced from 456ms to 168ms. A deep dive into agent-assisted kernel optimization methodology.

ComfyUI-INT4-Fast brings W4A4 quantized inference to ComfyUI. RTX 3060 (6GB VRAM) generates 1024×1024 images in 17s. Per-layer mixed-precision routing balances speed and quality for Flux models.

In-depth analysis of GPT-5.6 Ultra's sub-agent collaborative reasoning, the global rise of Chinese AI models, world-model evaluation gaps, and AI's real-world deployment challenges and bubble warnings.