Home / Hardware / NVIDIA / Rubin / R100

NVIDIA Rubin R100

Arch: RubinAnnouncedTDP 2300W
Compute Performance (FP4 Tensor)▲ 2.5x vs B200
50.0 PetaFLOPS
Dense FP4 performance — 2.5x Blackwell, the largest generational leap in NVIDIA history
R100 Rubin50.0 PF
B200 Blackwell20.0 PF
H100 Hopper2.0 PF
Memory System
288GB HBM4
8 Hi-Stacks / Next-gen 2x wider interface vs HBM3e
22.0 TB/s Bandwidth
Interconnect & I/O
3.5 TB/s NVLink 6
Bi-directional total bandwidth, Vera Rubin NVL72 support
PCIe Gen 6.0 x16
Real-World Applications
Next-Generation Reasoning Models

Rubin's 50 PFLOPS FP4 and 22 TB/s HBM4 bandwidth are purpose-built for the era of deep reasoning models. Chain-of-thought inference with million-token context windows runs natively without the memory bottlenecks that constrain Blackwell-era deployments.

Agentic AI at Scale

The Vera Rubin NVL72 platform delivers 3.6 ExaFLOPS per rack, enabling massive agentic AI systems that orchestrate hundreds of specialized models simultaneously. NVLink 6 at 3.5 TB/s per GPU ensures sub-microsecond inter-model communication.

Exascale+ Scientific Discovery

With HBM4's doubled interface width delivering 22 TB/s bandwidth, Rubin enables scientific simulations at resolutions previously impossible — from full-atom protein folding to planetary-scale climate models with AI-augmented physics.

World Simulation & Omniverse

Rubin's massive compute and memory bandwidth power real-time world simulation — physically accurate digital twins of factories, cities, and autonomous vehicle environments running AI perception, planning, and rendering simultaneously.

Full Technical Specifications
GPU ArchitectureNVIDIA Rubin
Transistor Count~336 Billion (3nm Process)
Die SizeDual-Die CoWoS-L (Reticle Limit x2)
Tensor Cores6th Gen Tensor Core Architecture
Memory Capacity288 GB HBM4
Memory Bandwidth22.0 TB/s
NVLink Bandwidth3.5 TB/s (NVLink 6)
CPU PairingVera CPU (88-Core Arm, 1.5TB LPDDR5X)
Form FactorVera Rubin Superchip / NVL72
Thermal Design Power2300W