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


Fast Track your AI Workflows

In this special guest feature, our friends over at Inspur write that for new workloads that are highly compute intensive, accelerators are often required. Accelerators can speed up the computation and allow for AI and ML algorithms to be used in real time.

Organizations around the world are looking for the next competitive edge in a wide range of industries in order to out innovate their competitors or be the first to discover a new scientific phenomenon. One of the most promising technologies that will assist these industries in creating new more optimized products, better drugs to combat disease and a clearer understanding of the world around us involves the use of Artificial Intelligence (AI) and Machine Learning (ML). When combined with High Performance Computing (HPC) algorithms, order of magnitude leaps in productivity can be achieved. Processes and workflows are being created to incorporate AI/ML in many industries, but many organizations are unsure of how to add this exciting new technology without the need to have a multitude of experts on site. Leading suppliers to these fields demonstrate their leadership by:

  • Offering industry leading products
  • Showing a history of working with clients as experts
  • Participating with other organizations to promote standards and the advancement of systems.

At the core of any HPC/AI/ML workflow is the hardware building block. Even standardizing on CPUs that run the x86-64 instruction set is not enough as there are many servers offered by many vendors, basically all offering the same compute performance, network connectivity and storage capacity. For new workloads that are highly compute intensive, accelerators are often required. Accelerators can speed up the computation and allow for AI and ML algorithms to be used in real time. And, not any accelerator will do. Companies that offer accelerators are constantly improving their products, from adding new features and the on-going performance improvements. A leading supplier of servers must be able to offer the latest components from their own suppliers in order to offer innovative and state-of-the-art solutions quickly.

Inspur is a leading supplier of solutions for HPC and AI/ML workloads. By innovating at the server level, partnering with leading CPU and accelerator vendors and developing the densest packaging available, Inspur offers a wide range of the most advanced servers for a variety of use cases. Inspur has been able to design servers that combine the power of the latest CPUs with the latest Nvidia GPUs. By incorporating a large number of GPUs in a compact enclosure, organizations can reduce their:

  • Power & Cooling – using less power for both the hardware and associated cooling saves money.
  • Real Estate – very dense servers save on the square feet needed for a desired computing capacity, thus saving data center real estate costs.
  • Management – less servers require less cabling and software updates are easier to manage.

Inspur is a leading manufacturer of high-density servers aimed at the HPC and AI/ML computing professionals. Inspur teams with Nvidia to offer the latest platforms for those who require the highest performing servers. Inspur is a recognized leader in the HPC and AI/ML fields with industry and solution experts around the world. Inspur provides an AI full stack of software that helps customers become productive quickly.

Inspur offers a full range of integrated hardware and software to enable fast creation of solutions for developers. Included in the full-stack are the AI servers, chips and accelerators, and AI resource suite, algorithm toolkit where popular software products have been optimized, and finally, vertical solutions for a variety of industries. Below is a diagram of how the portfolio fit together in order to provide a full solution, depending on users’ needs.

Inspur has been involved in the HPC and AI/ML industry for many years. With its team of experts, Inspur can work with users to architect and implement an entire data center for their most demanding needs. In addition, Inspur is a leading participant in Standard Performance Evaluation Corporation (SPEC). Inspur is a member of the Open System Steering Committee which provides guidance and other services for the entire organization. Inspur was also re-elected as the chair for the SPEC Machine Learning subgroup and established ML test benchmark, demonstrating ongoing leadership in new and critical technology areas.

Visit www.inspursystems.com for more information.

Resource Links: