SCANDIANO, Italy, Nov. 20, 2019 — E4 Computer Engineering, an industry leading server platform design manufacturer, today announced the validation of the E4 RK008, a Marvell ThunderX2-based node with two NVIDIA V100 Tensor Core GPUs.
The E4 RK008 supports the latest generation Marvell ThunderX2 processors and two NVIDIA V100 GPUs in a 2U node. The E4 RK008 offers professional-grade performance, security, reliability and affordability within an unmatched power consumption envelope, unique for single entry servers.
“Scientists and engineers need to enhance their competitiveness by deploying a powerful but power-savvy server platform,” said Cosimo Gianfreda, CTO of E4 Computer Engineering. “The E4 RK008 offers superlative performance by providing in a single-entry server platform two Marvell ThunderX2 processors, two NVIDIA V100 GPUs, large memory capacity, enhanced I/O capabilities, and the latest SSD storage technology. The E4 RK008 is optimized for HPC and AI applications, code development and any compute-intensive workload by providing the best balance of performance and power consumption.”
“By utilizing Marvell ThunderX2 and NVIDIA V100 GPUs within a single-entry server platform, our customers will also enjoy professional-grade performance in large file processing, storage and virtualization,” said Francesco Morsiani, CEO of E4 Computer Engineering. “The new E4 RK008 increases performance, reliability and security for entry servers, and delivers cost-effective processing power for customers running HPC and AI application workloads.”
“E4 has also integrated the E4 RK008 in ARMIDEA (ARM Integrated Development of Exascale class Applications), an 8-node strong cluster featuring a complete set of development tools and system software,” continued Gianfreda. “A number of applications are currently being ported and validated by scientific institutions, industries and ISVs. The results for the standard reference codes (HPL, HPCG) will be available later this year.”
“Marvell and E4 have a longstanding collaboration on our ThunderX program and we are excited to see E4 continue to drive innovation and application performance with the additional support of NVIDIA GPUs in the E4 RK008 system,” said Gopal Hegde, vice president and general manager, Server Processor Business Unit at Marvell Semiconductor, Inc. “Customers will benefit from the integration of the industry’s most widely adopted Arm-based server processor with the leading GPU for HPC and AI applications.”
“NVIDIA’s GPU-accelerated data center platform now works seamlessly with Arm-based servers,” said Paresh Kharya, Director of Product Management for Accelerated Computing at NVIDIA. “GPU acceleration in Arm-based systems like the new E4 RK008 provides HPC customers a path to address some of the world’s most complex research challenges.”
The E4 RK008 is available from December 2019. To learn more about the E4 RK008, please contact [email protected].
About E4 Computer Engineering
Since 2002, E4 Computer Engineering has been innovating and actively encouraging the adoption of new computing and storage technologies. Because new ideas are so important, we invest heavily in research and hence in our future. Thanks to our comprehensive range of hardware, software and services, we are able to offer our customers complete solutions for their most demanding workloads on: HPC, Big-Data, AI, Deep Learning, Data Analytics, Cognitive Computing and for any challenging Storage and Computing requirements. E4. When Performance Matters.
Source: E4 Computer Engineering