256 MB
DDR3850 MHz
Boost Clock130 W
TDP28 nm
ProcessShare on Social Media
Share GRID K100 specifications with othersTechnical Specifications
Core Specifications
Architecture:
Kepler
Process Size:
28 nm
Base Clock:
850 MHz
Boost Clock:
850 MHz
Shader Units:
192
TMUs:
16
ROPs:
16
Memory & Performance
Memory Size:
256 MB
Memory Type:
DDR3
Memory Bandwidth:
28.5 GB/s
Bus Interface:
PCIe 3.0 x16
TDP:
130 W
Suggested PSU:
300 W
Cache & Die Information
L1 Cache:
16 KB
L2 Cache:
0.2 MB
Die Size:
118 mm²
Transistors:
1.3 billion
Transistor Density:
10.8 million/mm²
Quick Facts
Performance
Pixel Fill Rate
3.4 GPixel/s
Texture Fill Rate
13.6 GTexel/s
FP32 Performance
0.3 TFLOPS
CUDA Toolkit Compatibility
Cuda Compute
3.0
sm_30
Architecture
Kepler
Supported CUDA Toolkits
5.0
5.5
6.0
6.5
7.0
7.5
8.0
Note: Support dropped in CUDA 9.0
Kepler architecture (GK104) - Support dropped in CUDA 9.0