HPE Apollo 6500 - Enterprise platform for accelerated computing
The HPE Apollo 6500 was created as the industry-leading enterprise for accelerated computing. This next-generation high performance computing (HPC) system is purpose built for deep learning that delivers a 3X faster model training than previous generations.2 The platform supports up to 8 industry-leading GPUs, fast GPU interconnect, high bandwidth fabric, and a configurable GPU topology to match customer workloads.
The HPE Apollo 6500 Gen10 System is ideal for a wide range of workloads, including artificial intelligence (AI), deep learning, and complex High Performance Computing (HPC) workloads of simulation and modeling. Deep learning use cases include the following: identify vehicles, pedestrians, and landmarks such as in autonomous vehicles; monitor oil field drilling rigs to prevent disasters; pattern matching for fraud detection; speech recognition and translation; and pharmaceutical design. HPC use cases could include: seismic simulation and modeling (reverse time migration), financial risk analyses, electronic design automation (EDA), etc.
Fast three – HPE Apollo 6500 Gen10 advantages
- Better performance and economics with up to 8 GPUs that communicate in a single node powered by only 2 processors.
- Easier server maintenance, with a 4U design that allows space to freely access components in addition to tool-less simple slide-out modules, easy slide-out switching between PCIe or NVLink GPU modules, and easily configurable topologies executed within the BIOS. Also, the HPE Apollo 6500 Gen10 offers extensive server and infrastructure management capabilities with industry-unique HPE iLO and HPE OneView.
- Thermal optimization with a design that provides far better cooling capabilities and proven ability to perform well with high bin CPUs up to 165 W and 3.2 GHz, and well-grounded support for higher bin processors in the next refresh.