HPC&GPU
HPC (High performance computing) and cloud computing center in the present structure are in desperate need to reach the maximal performance, while by the application of virtualization on high density servers can better cater to all demands only with one single machine. Furthermore, in the fields such as life sciences, engineering analysis, molecular dynamics, medical diagnosis, electronic design automation, financial analysis, fluid mechanics, geological sciences, and graphic processing, etc., greater demands on floating-point operation are submitted; applications such as streaming media processing, computer aid design, render farm and so on, are all demanding more floating-point operation, greater memory capacity and storage, where GPU accelerated computing is emerging.
P90G6
A Breakthrough Platform for AI and HPC Applications
- Modular architecture to provide customers with the best performance platform
- Support 4th and 5th Gen Intel® Xeon® Scalable Processor up to 385W
- Support up to 10 active or passive GPU Cards
- DDR5 and PCIe Gen5 ready platform
- Best GPU Communication with NVIDI® NVLink™ Bridge
- Inventec-designed switchboard provides maximum bandwidth for GPU direct RDMA
- 3200W/ 2700W/ 2400W CRPS 2+2 / 3+1 Redundancy
- Offer powerful thermal solutions, CPU up to 400W, GPU up to 450W