Stefan Mileschin

GPU optimised cloud

Before Amazon released AWS, Linode championed the concept of cloud computing and it is still used by developers and experienced users because it is slightly cheaper.

Now Linode is upping its game by bringing Nvidia Quadro RTX 6000 GPU cards to its cloud. Linode today launched new GPU-optimized cloud computing instances tailored specifically for developers and businesses requiring massive parallel computational power.

The new instances are built on CUDA, Tensor, and Real-Time Ray Tracing processing cores. Linode said it was one of the first cloud providers to deploy Nvidia latest GPU architecture.

Linode Research and Development Engineer Alejandro Peña claimed Linode’s GPU instances deliver a new level of accelerated computing power coupled with the simplicity, affordability and award-winning support that has become a hallmark of Linode’s position in the industry.

These are not virtual GPUs. Users get access to dedicated GPUs and CPU to run their workloads. Linode teams have optimised the Linode running GPUs to ensure that there is no bandwidth bottleneck between CPU and GPU.

Users can get dedicated GPU Linode for $1,000 per month. The card alone costs over $5,000, not including the cost of CPU memory and other components, plus the electric bill.

https://fudzilla.com/news/memory-and...me-with-nvidia