Nvidia Stops Distribution Of GeForce GPU’s To HPC Clients
Nvidia intends to boost the sales of their new Quadro and Tesla series graphics cards with a rather “inconvenient” way. The famous GPU company intends to make all HCP clients buy their new graphics cards instead of getting the older GeForce ones to save costs. While this makes sense since Nvidia wants their new GPUs to hit the market as a success, this decision seems a bit harsh, more if you think about what Nvidia intends to do with any client that decides not to cooperate.
According to DigiTimes, Nvidia will “force all distribution partners to ship only those two graphics cards instead of the GeForce ones to server and high-performance computer (HPC) clients”. The company wants to make their Tesla series the best option for HPC datacenters. This can also be seen in their Accelerated Computing page:
NVIDIA Tesla platform leverages massively parallel graphics processors to provide significantly higher throughput for compute intensive workloads and resulting in dramatic savings in datacenter cost. With over 450 applications optimized in a broad range of industries, including 10 out of 10 top, every HPC datacenter will benefit from Tesla Accelerated computing platform.
All GPU distributors that won’t stick with this requirement and go against Nvidia’s rule will get reduced graphics card supplies in the future. As Digitimes states: HPC and server players are concerned that the new move could lead to increased risks as equipping Quadro and Tesla series graphics cards will significantly raise the end-products’ prices. Nvidia is always prioritizing the shipment of GPUs to HPC players so if they would still use the older GeForce graphics cards, the company would lose a huge sales target group.
Clients that own both server and graphics card businesses such as Asustek Computer, Micro-Star International (MSI) and Gigabyte Technology are also expected to be strictly monitored by Nvidia so as to ensure that the older GeForce graphics cards will stop being shipped.