NVIDIA has decided to shift GTC 2020 on March 22-26 to an online event and cancelled the physical booths due to growing concern over the coronavirus, but that doesn’t mean QCT won’t continue to shape industry momentum.
We have put together a digital tour of what we had prepared to preview in lieu of the cancelled event and what remains a top focus for our AI/HPC ecosystem.
With the rising trend of AI & 5G, numerous innovative technologies are emerging. QCT has realized that to ensure our customers stay one step ahead of their competitors, it is our responsibility to build the infrastructure of the future now, offering future-ready solutions to let our customers focus on their core competencies. At GTC US 2020, we bring with us best-in-class products and solutions.
Visit the QCT GTC Microsite starting NOW to view our virtual kiosks which will include the following demos:
Abstract As an purpose-built system for Artificial Intelligent and High-Performance Computing workloads, QuantaGrid D52G-4U can deliver up to tensor Tflops to training deep learning model with eight NVIDIA® Tesla® V100 dual-width 10.5 inch or provides up to 293 GOPS/watt of peak INT8 performance to do inferencing with sixteen NVIDIA® Tesla® P4 and 2-socket Intel® Xeon® Scalable processor; up to 56 double precision Tflops computing power can accelerate HPC workloads such as Oil & Gas, bioinformatics, Mechanical Engineering. |
Abstract The QCT QuantaMicro X11C-8N is a front end server supporting the latest Intel® Xeon® E-2100 Mehlow family. This versatile high-density server offers flexibility with multiple nodes and accelerator options, designed to adapt to various workloads by adjusting CPU to Accelerator ratios. At a 2U 8-node form-factor, it’s compact design can be applied to dense deployments for different workloads with the use of different sleds. With compute sled, or a compute node with an accelerator, or 2x accelerators sled options, this system can scale out to meet multiple workload demands while preserving customization and flexibility. |
Abstract QuantaGrid D43K-1U is designed to unleash the power within the AMD EPYC™ 7002 Series Processors. Featuring up to 4TB memory capacity, 5 expansion slots and single width GPU support within a 1U chassis, it is the first PCIe 4.0 platform ever in the QCT product family, ready for deployment in next generation data centers. |
Abstract QuantaGrid D43KQ-2U is powered by dual AMD EPYC™ 7002 Series Processors supporting up to 10 PCIe 4.0 expansion slots and AI capabilities. It features two dual-width GPU accelerators like NVIDIA® Tesla® V100 and Quadro RTX series with optimized thermal solution. 24 U.2 NVMe drives, and up to 4TB memory capacity within just a 2U rackmount QCT server. |
Abstract QuantaGrid S43KL-1U is no ordinary single socket rackmount server. Geared up with one AMD EPYC™ 7002 Series Processor, featuring up to 2TB memory and five I/O slots, it aims to challenge the traditional dual socket data center ecosystem with flexible I/O options, single width GPU support, and PCIe 4.0 readiness. |
Abstract QCT Edge Server | QuantaGrid D52Y-2U |
Abstract QCT will demonstrate its density optimized AMD platform, QuantaGrid S43CA-2U, and show how productivity can be enhanced on VDI with NVIDIA® vGPU technology. The QuantaPlex S43CA-2U is a multi-node server that supports the next generation of powerful AMD EPYC™ processors. Each of the 4 nodes in this compact 2U chassis is capable of supporting the top bin AMD EPYC™ 7002 Series Processors, boasting a dominant 64-cores along with 16 DIMM slots that can meet the most intense computing environment needs. |
Senior Director/Consultant of Business Group, QCT
Evolving Adaptive Platform for Converged HPC/AI workloads
QCT will share its insights on HPC/AI workloads from a diversity of industries and how we solve business challenges using an adaptive platform. We will also share our observations of workload behavior and our efforts invested to simplify the management journey.
Director Sales Engineering, QCT
How QCT Edge Platforms Accelerate the AI Workloads
In the 5G era, more and more applications will benefit from the high bandwidth and low latency of 5G technology. Hence, edge computing becomes essential to most applications that require real-time processing.
QCT, as a worldwide leader in server design and manufacturing, offers complete end-to-end product lines to support 5G applications, including servers for the data centers, regional data centers, edge computing, and uCPE. These server products can support NVIDIA V100 or T4 GPUs as an accelerator to accelerate AI training and inference workloads. A quick demo will showcase how QCT is utilizing NVIDIA’s T4 GPU & AI technology in its state of the art X11C-8N server, so as to do vehicle/person/road-sign detection, car vendor identification, and car model classification, etc. This application is fully containerized, and after performance fine-tuning, each NVIDIA T4 card can handle up to 38 video streams for smooth inferencing.