Nvidia unveils Rubin HPC GPU with over 50,000 CUDA cores
At Supercomputing 2026, Nvidia unveiled its Rubin HPC GPU, featuring more than 50,000 CUDA cores, up from 28,672 in the current Rubin GPU. The chip, which implies a four-die layout, targets agentic workflows and scientific simulations.
Coverage timeline
Techmeme
Jeffrey Kampman / Tom's Hardware : Nvidia unveils its Rubin HPC GPU with “more than 50,000” CUDA cores, up from 28,672 in the current Rubin GPU, for agentic workflows and scientific simulations — Rubin HPC puts “over 50,000” CUDA cores on a chip, implies four-die layout — The Supercomputing 2026 conference is under way …

Related stories
Nvidia to invest $3B in Meta's equity fundraise as firms expand Muse partnership
Nvidia and Meta are expanding their Muse partnership, with Meta deploying more Vera Rubin and Vera CPUs. Nvidia will invest $3 billion in Meta's planned equity fundraise, which seeks up to $95 billion via stock offerings.
1 reports · 1 sources
ByteDance, Alibaba, Tencent order 3M+ Nvidia RTX PRO 5500 GPUs worth $35B+
According to sources cited by Reuters, ByteDance, Alibaba, and Tencent have ordered more than 3 million of Nvidia's new RTX PRO 5500 GPUs, with a combined sticker price exceeding $35 billion. The rush aims to secure less powerful, export-compliant chips amid ongoing restrictions.
1 reports · 1 sources
Google unveils TPU 8tx with AMD CPUs and TPU 8i Max for long-context tasks, launching 2027
Google announced two new specialized TPUs: the TPU 8tx, which will be paired with AMD CPUs, and the TPU 8i Max, a dual-chip package designed for long-context window tasks. Both are scheduled to launch in 2027, broadening Google's AI hardware portfolio to include confidential model training and higher memory capacity.
2 reports · 2 sources
Open-source projects unlock hidden Tensor Core throughput on GeForce GPUs, doubling RTX 4090 performance
Open-source developers released projects on GitHub that unlock higher Tensor Core throughput on NVIDIA GeForce GPUs by modifying boot-time configuration, without driver swaps or BIOS flashes. Tests reportedly doubled the FP16/FP32 and BF16/FP32 throughput of an RTX 4090 from about 165 TFLOPS to over 330 TFLOPS, with gains also observed on RTX 3090 Ti, RTX 4070 Ti SUPER, and RTX 2080 Ti. The projects target RTX 20, 30, and 40 series, with some experiments done on Windows 11.
1 reports · 1 sources

News updates every hour. Subscribe in your RSS reader to get new stories automatically.
Subscribe