NVIDIA doubled Blackwell performance through continuous software optimization, refining kernels, compiler paths, and inference runtimes so the same hardware delivers significantly more useful AI throughput over time. Initial gpt-oss-120b performance on an NVIDIA DGX Blackwell B200 system with the NVIDIA TensorRT LLM library was market-leading, but NVIDIA’s teams and the community have significantly optimized TensorRT LLM for open-source large language models. The TensorRT LLM v1.0 release is a major breakthrough in making large AI models faster and more responsive for everyone. Through advance
Telecommunications ArchivesBlackwell’s leadership comes from extreme hardware-software codesign. It’s a full-stack architecture built for speed, efficiency and scale: The Blackwell architecture features include: NVFP4 low-precision format for efficiency without loss of accuracy Fifth-generation NVIDIA NVLink that connects 72 Blackwell GPUs to act as one giant GPU NVLink Switch, which enables high concurrency through advanced tensor, expert and data parallel attention algorithms Annual hardware cadence plus continuous software optimization — NVIDIA has more than doubled Blackwell performance since launch using software
Telecommunications Archives…Paired with the NVIDIA Nemotron open model family, enterprises can run high-performance AI workloads where their data resides, whether in on-premises, hybrid or sovereign environments, without sacrificing performance or governance…
…optimizing platform power. Eighty-eight NVIDIA custom Olympus cores, 1.2TB/s of memory bandwidth and a high-speed, on-chip fabric results in a CPU platform that combines core performance and…
…This enables operators to optimize performance and efficiency before a rack is installed in the real world. Compute Is Data: Real-World Data Is No Longer the Moat Real-world data used…
…explore thousands of design variants overnight, producing higher-performing components with lower energy consumption. NVIDIA is using Flexcompute technology for the design and optimization of advanced optical and photonic devices. https://blogs…
…and Google Cloud have collaborated for more than a decade, co‑engineering a full‑stack AI platform that spans every technology layer — from performance‑optimized libraries and frameworks to enterprise‑grade cloud...
…The 3x scaling in GPU numbers delivered a 2.8x scaling in performance, a 93% efficiency rate thanks in part to software optimizations. Efficient scaling is a key requirement in generative AI…
…Jetson Thor delivers optimized performance across Qwen models like the Qwen 3.5-35B-A3B model, which reasons at 35 tokens per second, making real-time interactivity possible. Any developer can fine…
…By combining BlueField’s acceleration capabilities with DOCA Argus’ proactive threat detection, enterprises can secure AI factories without compromising performance or efficiency. Cisco is collaborating with NVIDIA to deliver a Secure AI…
…worked together for over a decade to advance KLA’s physics-based AI with optimized high-performance computing solutions that tap into GPUs and the CUDA ecosystem. The value of process control…
…Learn more about how to calculate lowest cost per token and download the NVIDIA guide on Cost-Latency-Performance Optimization for AI Factories . Start building AI factories on NVIDIA’s full-stack…