GPU Acceleration Shifts Into High Gear

Dell Technologies is adding the latest NVIDIA A100 Tensor Core graphics processing unit (GPU) in PowerEdge server solutions starting later this summer. These GPUs deliver unprecedented acceleration and flexibility for artificial intelligence, data analytics and high performance computing to meet ever-growing processing needs. The A100 expands Dell Technology’s wide range of solutions with NVIDIA GPUs, software and technologies enabling customers to accelerate workloads from edge to core to cloud.

Faster time to discovery, insight and innovation

Data scientists, researchers and engineers are focused on solving tough analytics, scientific and industrial challenges. They are driven to find the answers, to make new discoveries and to innovate in quests for personalized medicine, conversational AI, recommendation engines and more. As a result, the demand for GPUs continues to grow rapidly. Recent projections from Orbis Research indicate the value of the GPU market will increase 31 percent per year, rising from $20.6 billion in 2019 to $104.7 billion by 2025.

At the same time, the world’s top selling Dell PowerEdge server portfolio just keeps getting better with valuable systems management, security, performance and scalability for customers’ most demanding workloads. Incorporating the latest NVIDIA technologies in Dell PowerEdge servers and solutions can boost performance in numerous AI, HPC, and analytics applications.

Accelerating workloads big and small

While many workloads continue to scale, both in size and complexity, some acceleration tasks are not as demanding, such as early-stage development or small batch inference. So, an ideal data center accelerator doesn’t just accelerate big complex workloads. It also accelerates many smaller workloads. The NVIDIA A100 accelerates workloads big and small with its multi-instance GPU feature. Each A100 can be partitioned into as many as seven isolated GPU instances to boost workload performance. This apportioning capability enables optimal resource utilization, while effectively expanding GPU access to a wider range of users and applications.

With its third-generation Tensor Core technology, the A100 now accelerates more levels of precision for diverse workloads, speeding time to insight, as well as time to market. For some workloads, NVIDIA reports more than 2X performance/watt improvement.

NVIDIA A100 Tensor Core GPUs power the NVIDIA data center platform, which includes NVIDIA Mellanox HDR InfiniBand, NVSwitch, and Magnum IO software with GPUDirect RDMA and GPUDirect Storage. This set of technologies can efficiently scale to tens of thousands of server GPUs to train the most complex AI networks at unprecedented speed. All of this is now being introduced with the A100 in a PCIe form factor in Dell PowerEdge servers and Ready Solutions.

GPU acceleration in award-winning Dell PowerEdge server solutions

At Dell Technologies, we are excited to offer NVIDIA A100 Tensor Core GPUs in PowerEdge servers and Ready Solutions. Dell Ready Solutions for HPC, AI and Data Analytics feature award-winning accelerated server platforms such as the Dell PowerEdge C4140, DSS 8440 and R7525.

  • The PowerEdge C4140 is an incredibly dense rack server designed to handle the most demanding technical computing workloads in financial services, life sciences, machine and deep learning, oil and gas exploration, and more. This 1U 2-socket server packs a punch with up to four NVIDIA GPUs, PCIe or SXM with NVLink direct GPU-to-GPU interconnect.
  • The DSS 8440 is a 4U 2-socket accelerator-optimized server that meets the performance demands for increasingly complex computing challenges. With support of up to 10x double-wide or 16x single-wide accelerators, the DSS 8440 server delivers exceptional flexibility for AI, machine and deep learning workloads.
  • The PowerEdge R7525 server featuring 2nd gen AMD EPYC processors offers performance improvements for traditional and emerging workloads. This 2U 2-socket server has more processing cores and more memory performance with faster data transfer speeds. With up to 3x double-wide GPUs, or 6x single-wide accelerators, this next generation server is a great fit for workloads such as accelerated databases and virtual desktop infrastructure (VDI).

Dell Technologies and NVIDIA at the leading edge

With continued investment in research and development to create leading edge technologies, Dell Technologies and NVIDIA maintain a strong relationship. Dell Technologies was among the first to work with NVIDIA to certify NVIDIA EGX platform-based systems for edge computing. Dell’s NGC-Ready servers, first powered with NVIDIA V100 and T4 GPUs, will be offered with the NVIDIA A100 GPUs. These systems are built to run deep learning and machine learning workloads and are tested for functionality and performance of the AI stack with GPU-optimized software from NVIDIA’s NGC registry.

And this is just the beginning. So, stay tuned for the upcoming TOP500 races, which are sure to show NVIDIA GPU-acceleration kicking into the next gear.

To learn more

For a look at how the San Diego Supercomputer Center is capitalizing on NVIDIA GPUs in its new Expanse supercomputer from Dell Technologies, see the case study “Architecting for Mixed Workloads at Massive Scale.” Find out more about Dell PowerEdge server accelerators at

About the Author: Praveen Vishakantaiah

Praveen joined Dell Technologies from Intel where he was VP of Innovation with responsibilities for artificial intelligence at the edge, next generation mobile and stationary devices, client gaming, virtual/augmented reality, enterprise and small and medium business technologies. Previous roles include president of Intel India, VP of Client Solutions and Technology and VP of Client R&D. Praveen holds MS and Ph.D degrees in Electrical and Computer Engineering from The University of Texas at Austin.