In March of this year, E4 announced their new ThunderX-based ARKA servers for Cloud, Big Data, and HPC workloads, powered by 64-bit ThunderX ARM processors and the NVIDIA Tesla Accelerated Computing platform. The high-performance ARKA series combines the ThunderX 48-core ARMv8 CPUs with one or more NVIDIA Tesla GPU accelerators in a modular rack-optimized architecture. Despite the typically long sales cycles for HPC deployments, E4 has already begun closing orders for these new one and two-socket systems.
ThunderX SOCs are uniquely suited to drive performance for these systems, with 48 custom cores, high memory bandwidth and capacity, onboard workload accelerators and integrated network and storage capability. Focused on workloads such as high-energy physics and molecular dynamics, the ThunderX-based systems will leverage the high-performance, energy efficient Tesla GPU accelerators to deliver best-in-class performance with lower cost and power consumption than delivered by comparable systems.
"E4's reputation is built on delivering cutting-edge custom solutions for our customers", stated Cosimo Gianfreda, CTO, E4 Computer Engineering. "We wanted powerful and highly-scalable CPUs for our new ARKA servers, that deliver outstanding performance combined with reduced TCO. We chose Cavium and ThunderX, and customers are already buying these new servers based on the increased value they bring."
"Our ThunderX ARMv8 CPUs were designed specifically for high-performance: Our unparalleled 48 cores, high memory bandwidth and capacity, and integrated workload accelerators are a great match for HPC and Hyperscale workloads in E4s ARKA server family", stated Steve Cumings, Director, Market Development, Cavium. "The value of accelerated performance and reduced TCO is highlighted by the orders E4 customers are already placing. We look forward to continued delivery of cutting-edge platforms together with E4 Computer Engineering."
"GPU accelerators provide 64-bit ARM systems with the processing muscle to power a range of HPC workloads", stated Roy Kim, group manager of Accelerated Computing at NVIDIA. "The combination of advanced ThunderX ARMv8 CPUs and NVIDIA Tesla GPUs enable E4 to deliver a new class of highly efficient, high-performance solutions to tackle the toughest HPC and datacenter computing challenges."