NVIDIA has just announced a new GPU platform called the Tesla P100. However this new GPU is not intended for gamers but rather for hyperscale datacentre applications.
The Tesla P100 platform will include the company's Pascal GPU architecture along with the newest memory, semiconductor process and packaging technology to allow for the best graphical and computing experience from such a densely packed card.
NVIDIA will be manufacturing the chip using 16nm FinFET manufacturing technology. The Tesla P100's chip will have over 15 billion transistors and will include 16GB of second generation die stacked High Bandwidth Memory (HBM2). NVIDIA has made it easy to link multiple P100 GPUs using its NVLink technology.
The company also unveiled the DGX1 which is a full server appliance platform that combines the power of eight Tesla P100 GPUs with two Intel Xeon Processors in 3U chassis that can easily be mounted in a server rack. The system will require 3500W of power and will cost $129,000 but with the high price tag comes 170 teraflops of performance.
The DGX1 will offer developers and researchers the necessary power to achieve new milestones in the deep learning projects that will pave the way for the future of AI applications. The system will be available directly from NVIDIA in May.
NVIDIA is widely know as a provider of GPUs for gaming PCs but the announcement of the Tesla P100, the DGX1 and its work in self-driving cars could have a real impact in reshaping what the company is known for.