NVIDIA Rubin CPX vs NVIDIA Vera Rubin Superchip
Side-by-side specifications, performance, and rental pricing for datacenter AI workloads.
Rubin CPX: Rubin, 2026 Vera Rubin: Rubin, 2026
Memory
128 vs 576 GB
448 GB more for the Vera Rubin
NVIDIA Rubin CPX
Full specs →Architecture
Rubin
Year
2026
Memory
128 GB
NVIDIA Vera Rubin Superchip
Full specs →Architecture
Rubin
Year
2026
Memory
576 GB
Specifications
Rubin CPX Vera Rubin
Architecture
RubinRubin
Launch Year
20262026
Form Factor
—Superchip
Memory
128 GB576 GB
Memory Bandwidth
—44.0 TB/s
TDP
——
Process Node
——
Spec Rubin CPX Vera Rubin
Architecture RubinRubin
Launch Year 20262026
Form Factor —Superchip
Memory 128 GB576 GB
Memory Bandwidth —44.0 TB/s
TDP ——
Process Node ——
Performance (TFLOPS)
Rubin CPX Vera Rubin
FP64
No verified data available67 TFLOPS
FP32
No verified data available260 TFLOPS
TF32
No verified data available4,000 TFLOPS
sparse not published
BF16
No verified data available8,000 TFLOPS
sparse not published
FP16
No verified data available8,000 TFLOPS
sparse not published
FP8
No verified data available35,000 TFLOPS
sparse not published
FP6
No verified data available35,000 TFLOPS
sparse not published
FP4
No verified data available No verified data available
INT8
No verified data available500 TOPS
sparse not published
Tensor Core
FP64 (TC)
No verified data available400 TFLOPS
Precision Rubin CPX Vera Rubin
FP64 No verified data available67 TFLOPS
FP32 No verified data available260 TFLOPS
TF32 No verified data available4,000 TFLOPS
sparse not published
BF16 No verified data available8,000 TFLOPS
sparse not published
FP16 No verified data available8,000 TFLOPS
sparse not published
FP8 No verified data available35,000 TFLOPS
sparse not published
FP6 No verified data available35,000 TFLOPS
sparse not published
FP4 No verified data available No verified data available
INT8 No verified data available500 TOPS
sparse not published
Tensor Core
FP64 (TC) No verified data available400 TFLOPS
FLOPS by Precision
What actually differs
Both GPUs launched in 2026: the Rubin CPX on Rubin and the Vera Rubin on Rubin.
Memory is 128 GB against 576 GB. More memory per GPU means larger models fit before you have to shard across cards, which often matters more than raw TFLOPS for inference.
For LLM training and inference, weight the FP8 and FP16 rows and memory capacity most heavily. For scientific and HPC workloads, the FP64 row is the one to read.
Frequently asked questions
Which has more memory, the Rubin CPX or the Vera Rubin?
The Vera Rubin carries 576 GB of memory versus 128 GB for the Rubin CPX.
Get Comparison Updates
New GPUs added weekly. Be the first to see how they compare.