Earlier this year nVidia and Fujitsu announced that the CX400 S1 would offer high performance GPU-accelerated computing. The CX400 is already utilised for high density, compact form factor applications with half-wide, dual socket server nodes. Combining the CPU multicore power with the parallel processing cores of the GPU has given the HPC industry a new computing model.
This shift of complex computing tasks to the Cuda enabled GPUs allowed program logical functions to remain on the CPU, giving drastically increased performance levels than with sole CPU systems - while offering a dramatic decrease in power consumption.
The CX400 S1 is available now through our sales team, for more information see our CX400 product page, or talk to our sales team on +44(0)20 7352 7007.