GIGABYTE Shares Sessions at GTC and Shows Support for Arm CPU and NVIDIA GPU Validation
September 20, 2022 – GIGABYTE Technology, (TWSE: 2376), an industry leader in high-performance servers and workstations, today announced GIGABYTE sessions at global AI conference, NVIDIA GTC, that will showcase a conversational AI application and an Arm-based server ecosystem powered by NVIDIA GPUs and DPUs. Furthermore, the NVIDIA-Certified Systems program has expanded to include Arm CPUs and the NVIDIA® H100 Tensor Core GPU. In turn, GIGABYTE servers will be tested and validated with the aforementioned hardware to run the NVIDIA AI Enterprise software suite to support bare-metal and virtualized deployments. GIGABYTE will also support and qualify servers for the new NVIDIA L40 GPU that offers the best in universal compute and graphics.
- [A41392]Exploring the Future with GIGABYTE’s Ampere Arm Server and NVIDIA GPU Compute
Speaker: Patrick Kennedy - Lead Analyst and ServeTheHome (STH) Founder
The era of Arm-based NVIDIA GPU servers has arrived. We'll explore GIGABYTE systems powered by NVIDIA GPUs and Ampere Altra processors. Learn how to use today's technologies to build a differentiated GPU compute platform. See the key differences of this Arm-based platform from traditional compute platforms. We'll look at the benefits of Arm-based GPU compute and show you how to get started using this emerging technology with high-performance NVIDIA GPUs to solve next-generation problems.
- [A41366]AI Development Platform Helps on Conversational AI
Speaker: Kuan-Li Pengy – Senior Engineer at Myelintek Inc.
Session with Q&A: September 22 @ 10am CST
We'll demonstrate NVIDIA NeMo and NVIDIA Riva by operating the conversational AI-development platform. We'll show the process including creating an integrated development environment, loading the pre-trained models and the application logic, and the model serving. Learn what the conversational AI and application would be. We'll also point out the advantages of using the AI-development platform, such as hardware management, monitoring and allocation, the integration of the developer tools, software, and templates, and the model serving and retraining with pipelines. This helps AI teams save time and money, and focus more on their domain knowledge. We'll also talk about the kinds of servers that best support unique features of NVIDIA technology, so that you can find a good choice for your AI applications.
Optimized, High Performing Certified Systems with Arm CPU and NVIDIA H100
In early 2021, NVIDIA announced its NVIDIA-Certified Systems program, which validates x86 servers with NVIDIA GPUs, DPUs, and network adapters for the best configuration to run a diverse range of accelerated workloads with optimum performance, reliability, and scalability.
In the months following the program’s launch, GIGABYTE was selected as the only server provider for the NVIDIA Arm HPC Developer Kit and released the G242-P32, which is used for heterogeneous CPU/GPU system development.
NVIDIA has now expanded the Certified Systems program to include Arm-based servers designed for Ampere® Altra® and Altra® Max processors. NVIDIA has validated GIGABYTE G242-P33 servers with NVIDIA A100 Tensor Core GPUs using Arm CPU architecture in the program. Now, customers can deploy systems that have gone through rigorous testing and adhere to NVIDIA’s design best practices for performance, security, and scalability.
Support for NVIDIA L40 – Unprecedented Visual Computing Performance for the Data Center
In late 2020, NVIDIA introduced the A40 GPU, delivering the best-in-class visual computing performance for a wide range of data center workloads, including with NVIDIA Omniverse™ Enterprise, cloud rendering, NVIDIA CloudXR, and NVIDIA RTX Virtual Workstations. The new NVIDIA L40 GPU includes numerous enhancements that deliver tremendous increases in graphics, compute, and AI performance to accelerate visualization, rendering, virtualization, simulation, and AI workloads in the data center. The L40 will be available in GIGABYTE servers in the near future.