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


Mellanox HDR 200G InfiniBand Speeds Machine Learning with NVIDIA

Today Mellanox announced that its HDR 200G InfiniBand with the “Scalable Hierarchical Aggregation and Reduction Protocol” (SHARP) technology has set new performance records, doubling deep learning operations performance. The combination of Mellanox In-Network Computing SHARP with NVIDIA 100 Tensor Core GPU technology and Collective Communications Library (NCCL) deliver leading efficiency and scalability to deep learning and artificial intelligence applications.

2019: The Year of PCI Express 4.0

Computer systems are about to get a whole lot faster. This year starting at the high end of the market a transition will begin toward systems based on PCI Express 4.0. The interconnect speed will double to 64GB/sec in a 16 lane connection. Tim Miller, Vice President Strategic Development for One Stop Systems, explores the expected speed and innovation stemming from the introduction of PCI Express 4.0. 

Podcast: How Real is Ai Anyway?

In this podcast, the Radio Free HPC team discusses AI, bias in AI, and just how real AI actually is. “Ethics in AI, policy, legal framework are all big threads here. The trigger is the rather funny article from The Register: Artificial Intelligence, You Know it isn’t real, yeah?”

Building the business case for Hybrid HPC Cloud

Over the past two years, we’ve seen a dramatic acceleration in cloud adoption for HPC. Intuitively cloud computing makes sense — it enables organizations to rent rather than buy expensive HPC hardware, deploy solutions faster, reduce capital investments, and focus on core competencies. Software company Univa highlights the reason behind the recent cloud adoption boom among the HPC community, and the benefits of the hybrid HPC cloud.

Radio Free HPC Looks at NVIDIA’s acquisition of Mellanox

In this podcast, the Radio Free HPC team looks at NVIDIA’s acquisition of Mellanox. “This called for an emergency session of our crack panel. While it will be several months before the full impact of this merger is felt, the RFHPC team believes this will change both the HPC and the Datacenter markets. It also signals Nvidia’s journey towards becoming more of a systems company and gives them a better shot at the enterprise AI market.”

Achieving the Best QoE: Performance Libraries Accelerate Code Execution

The increasing consumerization of IT means that even staid business applications like accounting need to have the performance and ease of use of popular consumer apps. Fortunately, developers now have access to a powerful group of libraries that can instantly increase application performance – with little or no rewriting of older code. Here’s a quick rundown of Intel-provided libraries and how to get them.

Quobyte to showcase software defined storage for Machine Learning at GTC

Today Quobyte announced announced plans to showcase distributed storage technology at NVIDIA’s GPU Technology Conference in Silicon Valley next week. Quobyte software defined storage turns commodity servers into a reliable and highly automated data center file system. “As machine learning increasingly becomes an indispensable tool for organizations looking to make critical decisions derived from their cumulative data, there is an increasing need for unified storage infrastructure that enables faster results and the ability to fully leverage GPUs,” said Bjoern Kolbeck, Quobyte co-founder and CEO. “We are excited to have the opportunity to present to attendees of NVIDIA’s GTC how they can quickly and easily speed up their ML workloads by overcoming data challenges associated with scale, throughput and access to all data throughout the organization.”

Nvidia Certifies Colovore As DGX-Ready Data Center Partner

NVIDIA has gained traction in datacenter Machine Learning with their DGX platforms. Now Bay Area provider Colovore has signed up as a colocation partner supporting NVIDIA DGX deployments. “NVIDIA’s DGX-1 and DGX-2 platforms are leading the way in solving complex AI challenges and we are proud to partner with NVIDIA and their customers to provide the most cost-effective, flexible, and scalable data center home for these servers. With close to 1,000 DGX platforms already deployed and operating at Colovore, we have tremendous experience providing the optimal footprint for DGX and HPC infrastructure success.”

Scalable Machine Learning: The Role of Stratified Data Sharding

Srinivasan Parthasarathy from Ohio State University gave this talk at the Stanford HPC Conference. “With the increasing popularity of structured data stores, social networks and Web 2.0 and 3.0 applications, complex data formats, such as trees and graphs, are becoming ubiquitous. I will discuss a critical element at the heart of this challenge relates to the sharding, placement, storage and access of such tera- and peta- scale data.”

Announcing Google’s New TPU Dev Board for Machine Learning on the Edge

Google just launched Coral, a Beta platform for building intelligent devices with local AI. To enable this initiative, Google is making an edge version of its TensorFlow Processing Unit available for sale for the first time. “Our first hardware components feature the new Edge TPU, a small ASIC designed by Google that provides high-performance ML inferencing for low-power devices. For example, it can execute state-of-the-art mobile vision models such as MobileNet V2 at 100+ fps, in a power efficient manner.”