NVIDIA's Tesla P100 GPU Gets a PCIe Version

The NVIDIA Tesla P100 GPU, capable of replacing dozens of traditional CPU nodes to speed upwards the processing of complex algorithms, will now be offered in a PCIe course factor, the visitor announced today.

PCIe makes it easier for the NVIDIA Telsa P100 to be installed in the server banks that can benefit most from its power to accelerate computing tasks, especially when it comes to Motorcar Learning algorithms that ability artificial intelligence applications. When NVIDIA originally appear the P100 in April, CEO Jen-Hsun Huang described his visitor as 'all-in' when it comes to artificial intelligence and virtual reality.

The NVIDIA Tesla P100 is based on the Pascal GPU architecture. With support for PCIe Gen three, which offers a bandwidth of 32GB per 2d, each next-gen graphics processor provides the throughput of more than 32 CPU-based nodes.

Related Manufactures

Nvidia Unveils New Supercomputers and AI Algorithms

Its 18.7 Teraflops of half-precision performance, however, means that the PCIe version will have a slight functioning hit compared to the original NVIDIA Tesla P100 and its 21 Teraflops offering. That difference won't exist much of an upshot for massive server banks just it could be noticeable for smaller installations, such every bit democratic auto testing.

Non-server farm customers for whom that functioning driblet presents an effect will likely consider the NVIDIA in-house super-computer offering, the NVIDIA Tesla P100-powered DGX-1. Using eight of these overpowered cores, it delivers 170 Teraflops of half-precision top performance—the equivalent of 250 CPU-based servers.

The showtime PCIe-based P100s volition exist available in servers from Dell, HP, IBM, and other companies starting time in fall 2022. I of its beginning applications volition be in Europe's fastest supercomputer, the Piz Daint system at the Swiss National Super-Computing Centre in Lugano.