The World’s Most Capable and Supported Single GPU Accelerator Board for Servers
Solve your most demanding HPC and big data challenges on the NVIDIA® Tesla® K40 GPU Accelerator, the world’s most capable and supported single GPU accelerator. Equipped with 12GB of GPU accelerator memory, it processes 2x larger datasets then prior generation products to solve the world’s most challenging computational problems – making it ideal for big data analytics and large‐scale scientific computations. It also outperforms CPUs by up to 10x and delivers additional performance with its GPUBoost feature, converting power headroom into user‐controlled performance boost.
|Peak Double Precision Floating Point Performance ||1.43 Tflops |
|Peak Singe Precision Floating Point Performance ||4.29 Tflops |
|Memory Bandwidth (ECC Off) ||288 GB/sec |
|Memory Size and Type ||12 GB GDDR5 ECC |
|CUDA Cores ||2880 |
|Thermal Solution ||Passive heatsink |
|Form Factor ||4.376” H x 10.50” L, Dual Slot |
|Warranty ||3 Years |
|PNY Part Number ||TCSCK40KITPAS |
PNY provides unsurpassed service and commitment to its NVIDIA Tesla GPU accelerator customers including a 3 year warranty, pre- and post-sales support, dedicated Field Application Engineers and direct tech support hot lines.
For authorized customers only.