Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:

‘AI on the Fly’: Moving AI Compute and Storage to the Data Source

The impact of AI is just starting to be realized across a broad spectrum of industries. Tim Miller, Vice President Strategic Development at One Stop Systems (OSS), highlights a new approach — ‘AI on the Fly’ — where specialized high-performance accelerated computing resources for deep learning training move to the field near the data source. Moving AI computation to the data is another important step in realizing the full potential of AI.

World’s First 7nm GPU and Fastest Double Precision PCIe Card

AMD recently announced two new Radeon Instinct compute products including the AMD Radeon Instinct MI60 and Radeon Instinct MI50 accelerators, which are the first GPUs in the world based on the advanced 7nm FinFET process technology. The company has made numerous improvements on these new products, including optimized deep learning operations. This guest post from AMD outlines the key features of its new Radeon Instinct compute product line.

One Stop Systems Steps up GPU Servers for Ai and World’s First PCIe Gen 4 Cable Adapter

In this video from SC18, Jaan Mannik from One Stop Systems describes how the company’s high performance GPU system power HPC and Ai applications. At the show, the company also introduced HIB616-x16, the world’s first PCIe Gen 4 cable adapter. “The OSS booth will also feature a partner pavilion where several OSS partners will be represented, including NVIDIA, SkyScale, Western Digital, Liqid, One Convergence, Intel and Lenovo. OSS and its partners will showcase new products, services and solutions for high-performance computing, including GPU and flash storage expansion, composable infrastructure solutions, the latest EOS server, cloud computing, and the company’s recently introduced Thunderbolt eGPU product.”

Implementing PCIe Gen 4 Expansion

After a long run for PCI Express (PCIe) Gen 3, Gen 4 is fast becoming the latest de facto standard for general purpose I/O of the modern computer system. “The ability to run PCIe over cable at full performance with complete software transparency has opened up a range of new application possibilities over the past decade for CPU to I/O system re-partitioning with expansion systems uniquely situated to take advantage of the new PCIe Gen 4 bandwidth soon available on servers.”

Liqid steps up with Composable Infrastructure for HPC at SC17

In this video, Jay Breakstone and Sumit Puri from Liqid describe the company’s innovative composable infrastructure technology for HPC. “Liqid Grid enables once-static infrastructure to scale on demand to effectively manage the explosion of data associated with cloud, enterprise, HPC and AI, as well as other emerging, high-value, data-intensive applications.”

Liqid Showcases Composable Infrastructure for GPUs at GTC 2017

“The Liqid Composable Infrastructure (CI) Platform is the first solution to support GPUs as a dynamic, assignable, bare-metal resource. With the addition of graphics processing, the Liqid CI Platform delivers the industry’s most fully realized approach to composable infrastructure architecture. With this technology, disaggregated pools of compute, networking, data storage and graphics processing elements can be deployed on demand as bare-metal resources and instantly repurposed when infrastructure needs change.”

A PCIe Congestion-Aware Performance Model for Densely Populated Accelerator Servers

“MeteoSwiss, the Swiss national weather forecast institute, has selected densely populated accelerator servers as their primary system to compute weather forecast simulation. Servers with multiple accelerator devices that are primarily connected by a PCI-Express (PCIe) network achieve a significantly higher energy efficiency. Memory transfers between accelerators in such a system are subjected to PCIe arbitration policies. In this paper, we study the impact of PCIe topology and develop a congestion-aware performance model for PCIe communication. We present an algorithm for computing congestion factors of every communication in a congestion graph that characterizes the dynamic usage of network resources by an application.”

OSS Introduces Flash Appliances

Today One Stop Systems (OSS) introduced a pair of high-speed networked storage appliances that supports high-performance, shared storage services. “The OSS approach optimizes the hardware for the environment and optimizes the software for the application in the Flash Storage Array for Networks product line (FSAn). This hardware and software optimization in the FSAn product line provides the best ROI in any environment by minimizing hardware and license costs through advance array-level optimizations while maximizing the utilization of the flash array through VSI and VDI application support.”

Intel Xeon Phi Coprocessor Architecture

“High performance systems now typically a host processor and a coprocessor. The role of the coprocessor is to provide the developer and the user the ability to significantly speed up simulations if the algorithm that is used can run with a high degree of parallelization and can take advantage of an SIMD architecture. The Intel Xeon Phi coprocessor is an example of a coprocessor that is used in many HPC systems today.”

New Mellanox Networking Solutions Accelerate NVMe Over Fabrics

“We’ve seen the rapid evolution of SSDs and have been contributing to the NVMe over Fabrics standard and community drivers,” said Michael Kagan, CTO at Mellanox Technologies. “Because faster storage requires faster networks, we designed the highest-speeds and most intelligent offloads into both our ConnectX-5 and BlueField families. This lets us connect many SSDs directly to the network at full speed, without the need to dedicate many CPU cores to managing data movement, and we provide a complete end-to-end networking solution with the highest-performing 25, 50, and 100GbE switches and cables as well.”