-
Drivers
-
Products
-
Processors
-
Technologies
-
NVIDIA GRID
-
3D Vision
-
Platforms
-
-
Communities
-
Support
-
Buy
-
About NVIDIA
NVIDIA® Tesla® GPUs are built for graphics and visual processing. They deliver extreme co-processing power for the NVIDIA Maximus™ platform to accelerate the most demanding parallel-computing tasks. By adding a Tesla GPU to your workstation, you can take advantage of the newest simulation and rendering tools within industry-leading applications and see results in as little as half the time. You can also accelerate some of the most complex tools exponentially faster than by adding a second CPU. It's an unbeatable solution for getting more done in less time.
Need a personal super computer? Tesla Workstations for High Performance Computing
Tesla K20 GPU |
||
| NVIDIA CUDA Cores | 2496 | 448 |
| GPU Memory | 5GB GDDR5 | 6GB GDDR5 Memory |
| Memory bandwidth (ECC off)1 | 208 GB/sec | 144 GB/sec |
| Peak double precision floating point performance (board) | 1.17 Tflops | 515 Gflops |
| Peak single precision floating point performance (board) | 3.52 Tflops | 1030 Gflops |
| Applications | ||
| Availability | Now | Now |
Tesla 20-Series |
Tesla 10-Series |
| Tesla C2050/Tesla C2070 (PDF 227 KB) |
Tesla C1060 (PDF 162 KB) |
1 - With ECC on, 12.5% of the GPU memory is used for ECC bits. So, for example, 6 GB total memory yields 5.25 GB of user available memory with ECC on.