Nvidia's Latest Graphics Architecture: Blackwell and the Massive GB202 Die
By
pella
Toasted golden, schmeared with insight. Top of the rack.
Summary
Nvidia's latest graphics architecture, Blackwell, features the massive GB202 die with 92.2 billion transistors and 192 Streaming Multiprocessors. The RTX PRO 6000 Blackwell is the largest configuration using GB202. It is compared to the RTX 5090 in Nvidia's lineup.
Key quotes
· 4 pulledBlackwell, their latest graphics architecture, continues that tradition.
GB202 is the largest Blackwell die. It occupies a massive 750mm2 of area, and has 92.2 billion transistors.
Nvidia’s RTX PRO 6000 Blackwell features the largest GB202 configuration to date.
It sits alongside the RTX 5090 in Nvidia’s lineup, which also uses GB202 but disables a few more SMs.
You might also wanna read
NVIDIA Blackwell Leads First Agentic AI Benchmark from Artificial Analysis
Artificial Analysis launched AgentPerf, the industry's first benchmark for agentic AI workloads. The initial results show NVIDIA's Blackwell
NVIDIA Announces RTX Spark Superchip for Laptops and Compact Desktops at Computex
NVIDIA CEO Jensen Huang announced the RTX Spark Superchip at Computex, a new processor for compact desktops and laptops built on the GB10 Su
phoronix.com·14d ago
Nvidia Unveils Vera Rubin AI Computing Platform at CES 2026
Nvidia has launched its new Vera Rubin AI computing platform at CES 2026, positioning it as a successor to the Blackwell GPU. The platform c
The Verge·5mo ago
NVIDIA confirms multi-generation RTX Spark roadmap through 2029 with Rubin and Rosa Feynman architectures
NVIDIA has confirmed a multi-generation roadmap for its RTX Spark and DGX Station product lines, positioning them under a long-term Windows
videocardz.com·14d ago
Nvidia Unveils Nemotron 3 Ultra: Top US Open AI Model Still Trails Chinese Rivals
Nvidia CEO Jensen Huang unveiled Nemotron 3 Ultra at Computex in Taipei, the company's largest open AI model with 550 billion total paramete
decrypt.co·13d ago
NVIDIA launches Nemotron 3 Ultra, a 550B open-weight AI model for long-running enterprise agents
NVIDIA has announced Nemotron 3 Ultra, a 550-billion-parameter open-weight Mixture-of-Experts language model (with 55 billion active paramet
gadgetbond.com·9d ago