PRIMERGY NVIDIA® Tesla® K20 and K20X GPGPU
The NVIDIA® Tesla® K20 and K20X companion processor is built for GPU computing. It features up to 2688 application-acceleration cores per board, dramatically increasing performance compared to a traditional workstation. By adding a Tesla companion processor, engineers, designers, and content creation professionals can add over one Teraflop of computing potential to their workstation.
Features & Benefits
|Features ||Benefits |
- K20 = 2496 Cores; K20X=2688 Cores
- Tesla K20 delivers up to 3.52 / 1.17 teraflops, the Tesla K20X delivers up to 3.95 / 1.31 teraflops of singel / double-precision peak performance
- Meets a critical requirement for computing accuracy and reliability in datacenters and supercomputing centers. Offers protection of data in memory to enhance data integrity and reliability for applications. Register files, L1/L2 caches, shared memory, and external DRAM all are ECC protected.
- up to 5 GB / 6 GB of GDDR5 MEMORY per GPU
- Maximizes performance and reduces data transfers by keeping larger data sets in local memory that is attached directly to the GPU.
- Maximizes the throughput by faster context switching that is 10X faster than previous architecture, concurent kernel exectuion and imporved thread block scheduling.
|Max. number per system unit||2x in PRIMERGY RX350 S7 / S8
2x in PRIMERGY TX300 S7 / S8
1x in PRIMERGY CX270 S2
|Graphics memory speed||2.6 GHz
|Graphics memory bandwidth||208 GB/sec
|Power consumption||225 W / 235 W