Get cluster level performance right at your desk with NVIDIA® Tesla® GPU Accelerators, based on the NVIDIA Kepler™ Architecture.
With the introduction of Tesla K40 GPU Accelerators, you can run big scientific models on its 12GB of GPU accelerator memory, capable of processing 2x larger datasets. It also outperforms CPUs by up to 10x with its GPU Boost feature, converting power headroom into user-controlled performance boost. Solve complex computational problems quickly and efficiently with the Tesla K40 GPU for workstations, right from your desk.
TESLA GPUS FOR WORKSTATIONS
|Feature||Tesla K40||Tesla K20|
|Peak double precision floating point performance||1.43 Tflops||1.17 Tflops|
|Peak single precision floating point performance||4.29 Tflops||3.52 Tflops|
|Memory bandwidth (ECC off)||288 GB/sec||208 GB/sec|
|Memory size (GDDR5)||12 GB||5 GB|
DRIVERS AND SUPPORTED OPERATING SYSTEMS
Drivers are available at NVIDIA Driver Downloads.
Tesla workstation products are supported under:
|TECHNICAL SPECIFICATIONS||TESLA K40||TESLA K20X||TESLA K20|
|Peak double-precision floating point
|1.43 Tflops||1.31 Tflops||1.17 Tflops|
|Peak single-precision floating point
|4.29 Tflops||3.95 Tflops||3.52 Tflops|
|Number of GPUs||1 x GK110B||1 x GK110|
|Number of CUDA cores||2,880||2,688||2,496|
|Memory size per board (GDDR5)||12 GB||6 GB||5 GB|
|Memory bandwidth for board (ECC off)2||288 Gbytes/sec||250 Gbytes/sec||208 Gbytes/sec|
|Architecture features||SMX, Dynamic Parallelism, Hyper-Q|
|System||Servers and workstations||Servers||Servers and workstations|
* Please check with your server vendor for the CPU power connector usage before turn on the server to avoid potential damage.