NVIDIA P100 PCIe 16GB vs NVIDIA P100 SXM2 16GB
The P100 SXM2 16GB delivers 13% more FP16 throughput than the P100 PCIe 16GB (21 vs 19 TFLOPS dense).
NVIDIA P100 PCIe 16GB
Full specs →NVIDIA P100 SXM2 16GB
Full specs →Specifications
Performance (TFLOPS)
FLOPS by Precision
What actually differs
Both GPUs launched in 2016: the P100 PCIe 16GB on Pascal and the P100 SXM2 16GB on Pascal.
Dense throughput for the P100 PCIe 16GB against the P100 SXM2 16GB: FP64 4.7 vs 5.3 TFLOPS, FP32 9.3 vs 11 TFLOPS, FP16 19 vs 21 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.
Memory is 16 GB against 16 GB, fed at 732 GB/s versus 732 GB/s. More memory per GPU means larger models fit before you have to shard across cards, which often matters more than raw TFLOPS for inference.
Power budgets are 250 W for the P100 PCIe 16GB and 300 W for the P100 SXM2 16GB. At FP32 that works out to 0.04 against 0.04 TFLOPS per watt.
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
Is the P100 PCIe 16GB faster than the P100 SXM2 16GB?
At FP16 precision the P100 SXM2 16GB reaches 21 TFLOPS dense against 19 TFLOPS for the P100 PCIe 16GB. The performance table on this page lists every published precision for both GPUs.
Which has more memory, the P100 PCIe 16GB or the P100 SXM2 16GB?
Both GPUs carry 16 GB of VRAM. Memory bandwidth is 732 GB/s for the P100 PCIe 16GB and 732 GB/s for the P100 SXM2 16GB.
How much power do the P100 PCIe 16GB and the P100 SXM2 16GB draw?
The P100 PCIe 16GB is rated at 250 W TDP and the P100 SXM2 16GB at 300 W. On FP32 throughput per watt, the P100 PCIe 16GB is the more efficient part.
Get Comparison Updates
New GPUs added weekly. Be the first to see how they compare.