QCT was an early collaborator with Facebook and early pioneer of OCP from its inception. Come visit our virtual booth as we continue our support of the open hardware ecosystem, and get a chance to talk to our representatives!

Watch Our Speech Sessions

Data Centers Today, Trends, and Challenges

Presented by Dustin Wu, Director of Engineering at QCT.

In this session, Dustin will discuss the future trends of data centers, the possible challenges which include commodity TDP increases, System Level Air cooling limitations, and future data center architectures which address edge servers to support 5G workloads, hardware, resource utilization, and high-density system designs.

Evolving Adaptive Platform for High Performance Computing

Presented by Stephen Wheat, Senior Director of Business Group.

Stephen will first talk about converged workload trends, followed by workload behavior and infrastructure. In this session, he will also discuss QCT’s platform on demand (QCT POD) – an evolving adaptive platform for converged HPC and AI workloads.

Check Out Our Latest Demo Systems

Performance Upgraded for vRAN/ MEC Edge Server

Distinguished, flexible 1/0 design supports up to five PC1e expansion slots in a 2U server for GPU or FPGA accelerator cards.

NEBS GR-63-CORE and GR-1089-CORE compliant edge server which can be deployed in extreme environments.

400mm(1U)/ 420mm(2U) ultra-short chassis, which fits various environments; including placement close to walls; while offering a small footprint.

Intel Edge Server
QuantaGrid SD2H-1U
  • Innovative Performance-Optimized Infrastructures for Telecom Ecosystem
  • Ultimate I/O Flexibility Design for Edge Compute Server
  • High Flexibility Edge Server with Ultra Short Chassis
Intel Edge Server
QuantaGrid D52Y-2U
  • NUMA-Balanced Infrastructure for Telecom Ecosystems
  • Performance Upgrade for vRAN/MEC Edge Server
  • Optimized Thermal Design Supports Higher Operating Temperatures to Reduce OPEX

1U Front IO Server Powered by the Latest Intel® Xeon® Scalable Processors Family

1st ODM Server Project at Uber
  • Designed for high compute performance and low latency data storage access
  • Modular design offers various mixtures of expansion and NVMe storage options for different workload requirements
  • Front 10 architecture significantly improves thermal design, reliability and service efficiency

General Purpose Compute Servers for all DC Workloads

Superior compute performance with exceptional power efficiency.

Flexible SKUs with short chassis options for regional DC deployment.

Sophisticated in-house design, engineering and manufacturing.

Defying Physical Limitations
QuantaGrid D52B-1U
  • Ultimate Compute and Storage Density
  • Flexible and Scalable I/O Options
  • Tool-less HDD/SDD Tray and PCIe Riser Module
  • Quick Deployment and Maintenance
Computing in the Edge
QuantaGrid D52BE-2U
  • Balanced Compute Solution in 2U Edge Server
  • Flexible and Scalable I/O Options
  • Tool-less HDD/SDD Tray and PCIe Riser Module
  • Optimized NFVi and vRAN Applications
High-Performance Ultra-Dense Multi-Node Server
QuantaPlex T42S-2U (4-Node)
  • Unleashing Computing Performance
  • Rapid I/O with Additional NVMe Storage Tier
  • Shared Infrastructure for Increased Density and Energy Utilization
  • Easy Service and Maintenance

Accelerated Servers for Heterogenous Computing

2x 1ntel® Xeon Scalable Processors with 4x GPU/FPGAs, an ideal CPU to GPU ratio to accelerate the most versatile workloads such as small model training, inference, HPC and VDI.

Support up to 8x Tesla® V100 SXM2, 10x double width GPUs/FPGAs, 20x single width GPUs/FPGAs to conquer the most demanding AI challenges.

Accelerating Supercomputing Performance
QuantaGrid D52BV-2U
  • Compact and 0ptimized 2U Accelerator Server Architecture
  • 2x 100Gb/s High bandwidth Networking 0ption to scale out performance efficiently
  • NVIDIA® NGC Ready Server
  • Quick Deployment and Maintenance
All-in-one Box Prevails over AI and HPC Challenge
QuantaGrid D52G-4U
  • NVIDIA® NGC Ready Server
  • Up to 8x NVIDIA® Tesla® V100 with NVLink™ support up to 300GB/s GPU to GPU communication
  •  Up to 10x dual-width 300 Watt GPU or 16x single-width 75 Watt GPU support
  • Diversify GPU topology to Conquer Any Type of Parallel Computing Workload
  • Up to 4x100Gb/s High Bandwidth RDMA-enabled Network to Scale 0ut with Efficiency
  • 8x NVMe Storage to Accelerate Deep Learning

Hybrid Storage Designs Ideal for SDS, Big Data and Cloud Storage Workloads

High capacity storage servers with NVMe tier for all IO intensive requirements.

Optimized HDD capacity with SSD performance that minimizes CAPEX with best IOPS/dollar.

Ultra-Dense Extreme Performance Storage Server
QuantaPlex T22P-4U
  • Available with Either 1 or 2 server nodes(*1)
  • Extremely High Density up to 78 HDDs
  • 0ptional 2 Hot-Swappable NVMe for Caching
  • 1 M.2 on Board Per Node for 0S Redundancy
  • Dual SAS Path Design to Release the Bottle Neck of SAS Card
The Ultimate Scale-Out Building Block for Software-Defined-Storage
QuantaGrid D52T-1ULH
  • Ultra-Dense 1U storage server optimized for tiering hot and cold data.
  • Efficient computing power with dual Intel ® Xeon® Scalable Processors. 
  • Flexible I/O expansion slots.
  • Patented inner cable chain & hot-swappable HDD carriers for enterprise-level serviceability.

Open Data Center Networking

High-performance spine switches for data center and cloud environments.

The 400G switch for Hyper Scale Data Center
QuantaMesh BMS T9032-IX9
  • Addresses the Growing Demand for More Bandwidth in Cloud Networks
  • Delivers Four Times the Throughput Compared to the Previous 100G Switches
  • Ideal for High Speed Network Environments
  • SONiC Ready
Networking Switch
QuantaMesh BMS T7128-IXF
  • IXF is the First Programmable 128 x100G Switch in the Industry.
  • Providing Industry-Leading Performance in Increasing Bandwidth Network Environments.
  • Features Include a Programmable Compiler, Extensive in Band Telemetry, Large Scale Forwarding Database, Load Balancing, and Congestion Management.
  • Flexible Line Cards in the Front Panel Allows Enterprises to Smoothly Migrate Their Infrastructure Deployment to 400G in the Future.

Contact QCT

I'm interested in...

© 2020 All Rights Reserved

Top