Tesla K20X

GPU Code: GK110

6 GB

GDDR5

732 MHz

Boost Clock

235 W

TDP

28 nm

Process
Share on Social Media
Share Tesla K20X specifications with others

Technical Specifications

Core Specifications
Architecture: Kepler
Process Size: 28 nm
Base Clock: 732 MHz
Boost Clock: 732 MHz
Shader Units: 2688
TMUs: 224
ROPs: 48
Memory & Performance
Memory Size: 6 GB
Memory Type: GDDR5
Memory Bandwidth: 249.6 GB/s
Bus Interface: PCIe 3.0 x16
TDP: 235 W
Suggested PSU: 550 W
Cache & Die Information
L1 Cache: 16 KB
L2 Cache: 1.5 MB
Die Size: 561 mm²
Transistors: 7.1 billion
Transistor Density: 12.6 million/mm²

API Support

DirectX
11
OpenGL
4.6
Vulkan
1.2
Cuda Compute
3.5 sm_35

Physical Specifications

Board Length: 267 mm
Power Connectors: 1x 6-pin + 1x 8-pin

Quick Facts

Release Date: 2012-11-12
Manufacturer: NVIDIA
Generation: Tesla Kepler(Kxx)

Performance

Pixel Fill Rate 41.0 GPixel/s
Texture Fill Rate 164.0 GTexel/s
FP32 Performance 3.9 TFLOPS

CUDA Toolkit Compatibility

Cuda Compute
3.5 sm_35
Architecture

Kepler

Supported CUDA Toolkits
5.0 5.5 6.0 6.5 7.0 7.5 8.0 9.0 9.1 9.2
Note: Support dropped in CUDA 10.0

Kepler architecture (GK110/GK208) - Support dropped in CUDA 10.0