AI hardware
NVIDIA Vera Rubin Superchip
FP16 throughput
4×10¹⁵ FLOP/s
Memory
268 GB
NVIDIA Vera Rubin Superchip is an AI accelerator designed by NVIDIA (GPU).
It delivers 4×10¹⁵ FLOP/s of tensor FP16/BF16 throughput, 1.8×10¹⁶ FLOP/s at FP8. It carries 268 GB of memory at 20 TB/s of bandwidth.
Full record
- Manufacturer
- NVIDIA
- Type
- GPU
- Tensor FP16/BF16
- 4×10¹⁵ FLOP/s
- FP8
- 1.8×10¹⁶ FLOP/s
- INT8
- 2.5×10¹⁴ OP/s
- Memory
- 268 GB
- Memory bandwidth
- 20 TB
More from NVIDIA
SourceEpoch AI, 'Machine Learning Hardware'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/machine-learning-hardware. Licensed under CC BY 4.0.
← All ai hardware