AI training platform refreshed with 2nd-gen Intel Xeon scalable CPUs

April 05, 2019 //By Julien Happich
AI training platform refreshed with 2nd-gen Intel Xeon scalable CPUs
Adlink Technology is collaborating with Intel to refresh its AI training platform ALPS-4800 with the new 2nd Gen Intel Xeon Scalable processors.

The development represents Adlink’s active involvement in delivering Intel Select Solution for Network Function Virtualization Infrastructure (NFVI). The refreshed products are expected to be available in Q4 of 2019.

With improved Intel Turbo Boost technology 2.0 frequencies of up to 4.4 GHz and significantly enhanced memory support for Intel Optane DC persistent memory and increased DDR4 frequencies, Intel’s 2nd Gen Xeon Scalable processors not only provide an advanced compute core designed to excel wherever enormous performance for rapid data analysis is needed, but also enable a broad range of AI and machine learning applications with Intel Deep Learning Boost that adds new Vector Neural Network Instruction (VNNI).

Intel Speed Select Technology, another important new feature, provides an ideal solution for dynamic service delivery based on workload needs. In addition, the new Intel Xeon Scalable processors deliver hardware-enhanced security to enable more effective data protection, greater workload integrity and suppression of malicious exploits with reduced performance overhead. Adlink’s ALPS-4800 is the company’s latest AI training platform, validated with up to 8 GPU accelerators in a 4U server design.

More than just a hardware server system, the ALPS-4800 is an optimized, workload-conscious acceleration system for Machine Learning (ML), Deep Learning (DL) and High Performance Computing (HPC) applications for smart factory, smart city, mobile robotics, defense, medical, and digital security and surveillance (DSS).

Integrating a server node powered by dual Intel Xeon Scalable processors and PCIe expansion box with PCIe switching, the ALPS-4800 platform is highly configurable and scalable to support the combination of GPUs required to meet the needs of a variety of applications, support different workloads and demands, multiple topologies and bandwidths between GPUs and CPUs by making only simple cable routing adjustments.

Adlink Technology - www.adlinktech.com


Vous êtes certain ?

Si vous désactivez les cookies, vous ne pouvez plus naviguer sur le site.

Vous allez être rediriger vers Google.