Computer Hardware

Speed Up Data Center Graphics with the New NVIDIA Tesla V100s

November 26, 2019 • 3 min read

Data-Center_V100-Blog.png

NVIDIA’s most premium data center graphics card is now getting an upgrade. Previously, the Tesla V100 was the world’s most advanced data center graphics card, powered by the Volta architecture, coming in 16 and 32GB configurations, and offering performance of up to 100 CPUs in a single GPU.

Introducing the Tesla V100s

The brand new NVIDIA Tesla V100s is also based on the Volta microarchitecture similar to its predecessor, however packs more performance than both variants of the V100, which is attributed to the higher core clock speeds. The Tesla V100s only comes in the PCIe form factor but delivers higher clocks than the 300W Tesla V100 SMX2 (NVLink) solution. It comes with a GPU clock speed of 1601 MHz compared to 1533 MHz on the SMX2 variant and also offers higher 1.1 Gbps frequencies for the HBM2 DRAM. The combined boost to memory and graphics clocks make this Tesla variant the fastest HPC & server aimed graphics solution.

Final Thoughts

  • About a 17% increase in compute performance from the Tesla V100s
  • Tesla V100s is able to deliver a theoretical FP32 compute performance 16.4 TFLOPs, FP64 compute performance of 8.2
  • Pumps out over 1 Terabyte of memory bandwidth (1134GB/s) versus 900GB per second bandwidth of the Tesla V100
  • Comes in a 250W design

Tags

nvidia

tesla

gpu

v100

hpc

deep learning