NVIDIA Labs built an AI agent skill that translated all 24 public TileGym operators from cuTile Python and Triton-TileIR into cuTile Rust, reaching 99.5% of cuTile Python performance on average across…
NVIDIA's Nemotron 3.5 Lightning 30B-A3B model demonstrates how a Mixture-of-Experts (MoE) architecture can activate only 3B of its 30B parameters per token by routing each token through a selected sub…
NVIDIA detailed NVLink 6, the scale-up networking fabric in its Vera Rubin NVL72 rack-scale compute engine, which connects 72 Rubin GPUs into a single scale-up domain and uses a multi-layer resiliency…
NVIDIA detailed power-efficiency technologies in its Groq 3 LPX rack, a low-latency accelerator added to the NVIDIA Vera Rubin platform for the highest-interactivity inference tiers. The Groq 3 LPX de…
NVIDIA reported that combining its Transformer Engine library with the JAX Python library raised DeepSeek-V3 MoE training throughput on NVIDIA GB200 GPUs from an unoptimized baseline of 103 TFLOPS/GPU…
NVIDIA's NIM 2.0.12 optimized serving stack delivered up to 2.5x higher output-token throughput than the open-source baseline when serving the Nemotron 3 Ultra model on four B200 GPUs, reaching 1,997 …
NVIDIA's BioNeMo Inference Runtime (BioIR) was used in the expansion of the AlphaFold Database (AFDB), accelerating protein-complex structure generation across 4,777 proteomes — about 31 million candi…
NVIDIA's supply chain operations team worked with Palantir to build a Digital Supply Chain Intelligence command center in Palantir Foundry, a unified data layer that connects materials, manufacturing …
NVIDIA Dynamo, an open-source inference framework, supports encode-prefill-decode (EPD) disaggregation to accelerate multimodal model serving, achieving up to 5x faster time to first token (TTFT) and …
NVIDIA released CUDA Toolkit 13.4, adding Windows on Arm support, early developer preview for the NVIDIA Rubin GPU architecture, and enhanced GPU management capabilities including Multi-Process Servic…
NVIDIA announced CUDA Rust, a native GPU programming toolchain for Rust, with two tracks: SIMT (cuda-oxide) and Tile, to mature through 2027. The SIMT track, cuda-oxide, is a custom rustc codegen back…
NVIDIA introduced a memory-driven Chief of Staff agent built with NVIDIA NemoClaw that maintains a human-readable knowledge layer called the self model to track people, projects, priorities, and worki…
NVIDIA announced that new compact open models, including Nemotron 3.5 Lightning and Qwen3.8-27B, can now run reasoning and agentic AI workloads on Jetson edge devices, eliminating the need for data ce…
A new guide explains how to carry user identity across federated Kubernetes and AI platforms, addressing the challenge of maintaining consistent authentication and authorization as users move between …
NVIDIA has launched the beta of its Personal AI Router (PAIR), a virtual inference router that distributes AI inference requests across multiple systems on a local network to relieve bottlenecks in mu…
NVIDIA published a step-by-step tutorial demonstrating how to debug, benchmark, and optimize CUDA code using modern tools such as Compute Sanitizer, Nsight Systems, NVTX, CUB, and CCCL, applied to an …
NVIDIA published the third post in its AI model co-design series, detailing how speculative decoding accelerates large language model inference by predicting multiple tokens per iteration with a small…
NVIDIA and CrowdStrike evaluated an agentic attack-defense system in an isolated environment, reporting that CrowdStrike's Blue Solano defensive model is 13% more accurate than the leading proprietary…
A practical framework for sizing GPU resources for AI inference workloads and optimizing total cost of ownership (TCO) is outlined, emphasizing use case, token patterns, latency targets, concurrency, …
NVIDIA and Anthropic have integrated the NVIDIA BioNeMo Agent Toolkit into Claude Science, enabling AI agents to run protein structure prediction with multiple-sequence alignment using NVIDIA NIM micr…