Combine the right mix of traditional IT, private cloud, and public cloud to achieve optimal workload placement.
Apollo 10 Systems
Now, even the smallest department can run deep learning and mixed workload HPC on industry-standard accelerated compute servers.
Entry-level HPC on a standard platform
HPE Apollo 10 series makes High Performance Computing (HPC) and deep learning affordable. Low entry costs combine with a flexible industry standard platform, custom solutions, and a choice of components. Apollo 10 series servers readily integrate into standard cluster environments.
Built on HPC expertise
We devised the Apollo 10 series based on our extensive expertise in HPC, data management, visualization, systems management, and High Performance Data Analysis (HPDA). From this foundation, you can implement optimal HPC and deep learning solutions that take advantage of responsive and flexible support along with committed account and technical resource teams.
GPU with unprecedented floating point operations
CPU with massive parallelism and vectorization
Intel® Xeon® Phi is a self-hosted many core processor with massive parallelism and vectorization for workloads like genomics. It delivers more compute per unit of energy versus comparable platforms.
Customize for required workloads
You can customize your HPE Apollo 10 server by choosing from the available CPUs, memory options, network adapters, and hard drives or solid state drives.
Factory integration and testing
Fully integrated and factory tested clusters are delivered with a complete cluster management software stack to simplify deployment and management. They can be custom configured for right-sizing.
A 1U rackmount dual socket Intel® Xeon® Server with support for four NVIDIA Tesla SXM2 GPUs with NVLink. This configuration allows for up to 14,336 NVIDIA CUDA cores and 42.4 Tflops single precision floating point performance.
A cost effective 1U dual socket Intel® Xeon® server that’s optimal for mixed workload GPU-accelerated HPC. With support for up to 4 PCIe GPU cards, Apollo pc40 provides up to 37.2 Tflops of single precision performance per server.
Flexible HPC for scientific workloads
Optimized for deep learning, life sciences, and weather simulation, the 2U HPE Apollo kl20 platform offers >12 x86 Tflops and up to 272 cores. Coming soon.
CPUs per System
- Two Intel® Xeon® Skylake processors (sx40)
- Two Intel® Xeon® Skylake processors (pc40)
- Four Intel® Xeon® Phi servers (kl20)
- Two 2000W Redundant Power Supplies (sx40, pc40, and kl20)
- 2x SFF SATA drives (sx40)
- 2x SFF SATA drives (pc40)
- 12x LFF Drives (three per server) (kl20)
- Three PCIe 3.0 x16. One PCIe 3.0 x8 LP. (sx40)
- Two additional PCIe slots: One PCIe 3.0 x16. One PCIe3.0 x8 LP. (pc40)
- 8 PCIe 3.0 x17 LP (two per server) (kl20)
- 1U (sx40)
- 1U (pc40)
- 2U (kl20)
GPUs per System
- Up to four NVIDIA P100 SXM2 GPUs with NVLink (sx40)
- Four NVIDIA® Tesla P40 GPUs (pc40)
- 21.2 Tflops of double precision or 42.4 Tflops single precision (sx40)
- 37.2 Tflops of single precision performance per server (pc40)
- >12 x86 Tflops (kl20)
- 16 DDR4 DIMM slots (up to 2666 MT/s) (sx40)
- 16 DDR4 DIMM slots (up to 2666 MT/s) (pc40)
- 24 DDR4 DIMMs, 2400 MT/s (Six DIMMs per server) (kl20)
Why HPE for HPC?
Our strength as the leading High Performance Computing solution provider is our ability to offer platforms with the latest technology in cost effective, easily managed systems that operate seamlessly with our suite of HPC focused products. Our HPC solutions empower innovation at any scale, building on our purpose-built HPC systems and technologies solutions, applications and support services. The results? More power to differentiate your business, drive research and find answers like never before.
Let’s talk about HPE Apollo 10
Let’s work together to find the ideal HPC solution for your unique workloads. Talk to an Apollo 10 expert from HPE today.
HPE announces computing innovations to accelerate deep learning
Accelerating deep learning insights with new GPU-based solutions
Accelerating business innovation with HPC and AI at HPE Discover 2017 Las Vegas
What is deep learning?