SiPearl team up with Nvidia
SiPearl says it's teaming up with Nvidia in a collaboration to provide a joint offering combining SiPearl HPC microprocessors with Nvidia's accelerated computing and networking portfolio.
SiPearl and Nvidia are aiming to combine both companies’ portfolio of hardware and software solutions. The companies will develop a proxy platform for porting activities and SVE workload analysis combining the strengths of SiPearl CPU (such as HBM memory) and Nvidia GPU (including massive parallelism and throughput). The collaboration will include joint efforts with third-party European research institutions on elements such as SoC and NoC simulation capabilities in open-source and research-oriented modelling tools, a press release reads.
On the software side, the companies aims to accelerate the development and growth of the Arm HPC ecosystem working closely with several European Centres of Excellence. On the hardware side, SiPearl and Nvidia will work together to make sure the Nvidia accelerated computing and networking portfolio continue to work seamlessly with SiPearl ‘Rhea’ and future CPUs.
“We are excited about this collaboration between SiPearl and NVIDIA. It holds great potential for the European supercomputing space, the ARM ecosystem and accelerated computing,” says Dr. Thomas Lippert, Head of Jülich Supercomputing Centre in the press release.
“HPC research has the potential to transform society with beneficial discoveries,” says Marc Hamilton, vice president of Solutions Architecture and Engineering at Nvidia. “Nvidia is collaborating with SiPearl, a leading innovator using Arm based CPUs to integrate Nvidia accelerated computing and equip HPC researchers with the extreme performance demanded by next-generation applications to fuel these discoveries.”