GPUWorkstationReleased Feb 6, 2017

NVIDIA Quadro GP100

16 GB HBM2 · 4096-bit · 3,584 cores

Workstation card built on NVIDIA's GP100 compute GPU with 16 GB of HBM2 and fast double-precision and FP16 performance, aimed at simulation, CAE and deep learning development. Two boards can be linked with NVLink to access 32 GB of combined memory in supporting applications.

General

Manufacturer: NVIDIA
Model: Quadro GP100
Segment: Workstation
GPU chip: GP100
Architecture: Pascal
Release date: Feb 6, 2017
Price: Search on Amazon

Memory

Memory size: 16 GB
Memory type: HBM2
Bus width: 4096-bit
Bandwidth: 717 GB/s
Memory clock: —

Clocks & silicon

GPU clock: 1,304 MHz
Boost clock: 1,442 MHz
Shader units: 3,584 CUDA Cores
Tensor / AI cores: —
RT cores: —
Transistors: 15.3 billion
Technology (nm): 16 nm
TSMC 16 nm FinFET
TDP / TBP: 235 W

Interface & APIs

Bus interface: PCIe 3.0 x16
SLI / multi-GPU: NVLink (2 boards)
DirectX: 12 (12_1)
OpenGL: 4.6
Vulkan: 1.3
Other APIs: OpenCL 3.0, CUDA 6.0

AI support

Precisions: FP8FP16BF16INT8INT4TF32FP64
AI accelerator: None (no Tensor Cores / RT Cores)
Peak AI TOPS: —
Notes: No Tensor Cores and no RT Cores. GP100 (CUDA compute capability 6.0) runs FP16 at twice the FP32 rate and FP64 at half the FP32 rate, but lacks the DP4A INT8 dot-product instructions of other Pascal GPUs. No BF16, TF32 or FP8.

All NVIDIA GPUs →

Workstation gpus from NVIDIA announced around the same time as the NVIDIA Quadro GP100. Add any of them to your compare list to see the specs side by side.

gpu