CUDA marked its 20th anniversary with more than 11 million developers on the platform
NVIDIA Vera Rubin Sweeps MLPerf Debut at SC26 With New Scientific Libraries
Vera Rubin's MLPerf sweep establishes its performance position relative to the prior generation across both training and agentic inference workloads. The CUDA-X library releases and quantum-classical infrastructure announcements extend NVIDIA's platform into scientific computing domains beyond AI training.
When stories like this move, I write about them in an email each morning.
One free email each morning, around 8 AM US Eastern.
Check your email.
I have sent you a confirmation link. One click and you are in.
Not in your inbox? Look in spam or promotions. If it's there, mark it "Not spam" so future editions land in your inbox.
The full picture
At SC26, NVIDIA's Vera Rubin NVL72 platform swept all MLPerf Training v6.1 benchmarks in its debut and led the inaugural MLPerf Endpoints agentic inference benchmarks, delivering up to 2x more training performance per GPU than the GB300 NVL72 and more than 10x the throughput per watt on agentic inference tasks. CoreWeave reported being the first cloud provider to submit multi-rack Vera Rubin results, logging a 2.7x speedup over GB200 NVL72 on DeepSeek-V3 training. Alongside the hardware results, NVIDIA launched a wave of CUDA-X scientific computing libraries covering astrophysics, fusion energy, drug discovery, materials science, and genomics. NVIDIA and IBM announced plans for a hybrid quantum-classical platform at Fermilab pairing DGX B200 systems with IBM Quantum Nighthawk processors. NVIDIA also extended Slurm to natively schedule quantum processing units alongside GPUs, with adoption from IBM, Pasqal, Quantinuum, and others. Vera CPU racks, built around NVIDIA's first fully custom data center CPU, are expected to ship from Dell, HPE, Lenovo, and Supermicro starting in early 2027. CUDA marked its 20th anniversary at the conference, with NVIDIA noting more than 11 million developers now build on the platform.
How it developed
Sources
- Lambda Boosts AI Factory Resiliency and Efficiency With NVIDIA Photonics
- NVIDIA Expands Slurm to Bring Quantum Computing Into Supercomputing
- NVIDIA Sweeps MLPerf Training and Endpoints as Vera Rubin Debuts With Leading Performance
- NVIDIA cuPhoton Cuts Event Horizon Telescope Processing From a Month to Eight Minutes (+4 more)
- CUDA at 20: How the Platform Transformed Scientific Computing
- CoreWeave's first Vera Rubin NVL72 racks
- Google Cloud Vera Rubin GPU instances
- NVIDIA Vera Rubin and NVLink Fusion platform
- CoreWeave's Vera Rubin and Forge at FullyConnected26
BackgroundCompute: the chips, clouds and data centers behind AI
Get the daily email
When stories like this move, I write about them in an email each morning.
One free email each morning, around 8 AM US Eastern.
Check your email.
I have sent you a confirmation link. One click and you are in.
Not in your inbox? Look in spam or promotions. If it's there, mark it "Not spam" so future editions land in your inbox.