GIGABYTE to Showcase Advanced Server Solutions and Cooling Technologies for Supercomputing

In the upcoming Supercomputing Conference (SC23), GIGABYTE is poised to present its range of servers and cooling technologies, emphasizing solutions for the NVIDIA GH200 Grace Hopper Superchip and forthcoming AMD Instinct APU. At the event, GIGABYTE plans to demonstrate how it will address core challenges faced by supercomputing data centers, highlighting efficient cooling methods and AI processing capabilities designed for rapid real-time analysis and swift result delivery.

In the upcoming Supercomputing Conference (SC23), GIGABYTE is poised to present its range of servers and cooling technologies, emphasizing solutions for the NVIDIA GH200 Grace Hopper Superchip and forthcoming AMD Instinct APU. At the event, GIGABYTE plans to demonstrate how it will address core challenges faced by supercomputing data centers, highlighting efficient cooling methods and AI processing capabilities designed for rapid real-time analysis and swift result delivery.

As computing infrastructures generate more heat due to increased performance demands, GIGABYTE looks to help solve this issue with its advanced cooling solutions, including the A1P0-EA0 single-phase immersion tank and direct liquid cooling (DLC) systems. Specifically for the latter, GIGABYTE’s cold plates and cooling modules technologies like the NVIDIA Grace CPU Superchip, NVIDIA Grace Hopper Superchip, AMD EPYC 9004 processors, and 4th Gen Intel Xeon processors. These technologies aim to maintain optimal hardware performance with greater energy efficiency without expanding the data center’s physical footprint.

A1P0-EA0 Features

• Supports up to 12U EIA servers
• Minimum PUE reaching 1.02
• Up to 40kW of heat dissipation capacity
• Supports server length up to 900 mm

For AI and HPC, GIGABYTE is demonstrating its expertise with modular systems like GIGA PODs, which are already in use by leading cloud service providers. The G593-SD0 server is designed for compatibility with Intel Xeon and NVIDIA H100 GPUs and is built specifically to enhance AI processing power. Additionally, GIGABYTE is unveiling its purpose-built G383-R80 GPU server for the next-generation AMD Instinct APU, targeting AI application acceleration.

G593-SD0 Features	H263-V11 Features	H223-V10 Features
NVIDIA-Certified system for scalability, manageability, security, and performance NVIDIA HGX H100 with 8 x SXM5 GPUs NVIDIA NVLink and NVSwitch technology Up to 900GB/s GPU to GPU interconnection 4th Gen Intel Xeon Scalable Processors Intel Xeon CPU Max Series Dual processor, LGA 4677 8-Channel RDIMM DDR5 per processor, 32 x DIMMs Dual ROM Architecture Compatible with NVIDIA BlueField-2 DPUs 2 x 10Gb/s BASE-T LAN ports (Intel® X710-AT2) 1 x Dedicated management port 8 x 2.5″ Gen5 NVMe/SATA/SAS hot-swappable bays 12 x LP PCIe Gen5 x16 slots 1 x LP PCIe Gen4 x16 slot 4+2 3000W (240V) 80 PLUS Titanium redundant power supplies	Direct liquid cooling solution CPU+GPU design for giant-scale AI and HPC 2U 4-node rear access server system NVIDIA Grace Hopper Superchip 900GB/s NVLink-C2C Interconnect Up to 480GB CPU LPDDR5X ECC memory per module Up to 96GB GPU HBM3 per module Compatible with NVIDIA BlueField-2 / BlueField-3 DPUs 8 x 10Gb/s BASE-T LAN ports (Intel® X550-AT2) 4 x Dedicated management ports 1 x CMC port 16 x 2.5″ Gen4 NVMe hot-swappable bays 8 x M.2 slots with PCIe Gen5 x4 interface 4 x FHHL PCIe Gen5 x16 slots 4 x OCP 3.0 Gen5 x16 slots Triple 3000W (240V) 80 PLUS Titanium redundant power supply	CPU+GPU design for giant-scale AI and HPC 2U 2-node rear access server system NVIDIA Grace Hopper Superchip 900GB/s NVLink-C2C Interconnect Up to 480GB CPU LPDDR5X ECC memory per module Up to 96GB GPU HBM3 per module Compatible with NVIDIA BlueField-2 / BlueField-3 DPUs 4 x 10Gb/s BASE-T LAN ports (Intel® X550-AT2) 2 x Dedicated management ports 1 x CMC port 8 x 2.5″ Gen4 NVMe hot-swappable bays 4 x M.2 slots with PCIe Gen5 x4 interface 2 x FHHL Dual slot PCIe Gen5 x16 slots 2 x FHHL Single slot PCIe Gen5 x16 slots 2 x OCP 3.0 Gen5 x16 slots Triple 3000W (240V) 80 PLUS Titanium redundant power supply

 For scalable data center infrastructure, GIGABYTE’s G493-SB0 system is designed for the latest NVIDIA GPUs and anticipates certification for NVIDIA’s OVX system designation as well. This forward-looking design is complemented by the XH23-VG0, which integrates the NVIDIA Grace Hopper architecture for extensive AI and HPC tasks. These systems align with NVIDIA’s vision for expansive, accelerated computing environments and support for advanced networking technologies.

G493-SB0 Features

• Supports up to 8 x Dual slot Gen5 GPU cards
• NVIDIA-Certified system for scalability, manageability, security, and performance
• 4th Gen Intel Xeon Scalable Processors
• Intel Xeon CPU Max Series
• Dual processor, LGA 4677
• 8-Channel RDIMM DDR5, 32 x DIMMs
• Dual ROM Architecture
• 2 x 10Gb/s BASE-T LAN ports (Intel X710-AT2)
• 1 x Dedicated management port
• 8 x 3.5″/2.5″ Gen5 NVMe/SATA hot-swappable bays
• 4 x 3.5″/2.5″ SATA hot-swappable bays
• 1 x M.2 slot with PCIe Gen3 x1 interface
• 8 x FHFL PCIe Gen5 x16 slots for GPUs
• 3+1 3000W (240V) 80 PLUS Titanium redundant power supplies

GIGABYTE continues to cater to enterprise needs with innovative server designs tailored to specific market demands. New introductions include the slim S183-SH0 server with dual Intel Xeon processors and dense solid-state storage configurations, and the versatile H253-Z10 multi-node E1.S supporting server. Focused on AI, the G293-Z43 and G293-Z23 servers are optimized for AI training and inference, respectively, showcasing GIGABYTE’s dedication to providing targeted solutions for varying computational workloads.

S183-SH0 Features	H253-Z10 Features	G293-Z43 Features
4th Gen Intel Xeon Scalable Processors Intel Xeon CPU Max Series Dual processor, LGA 4677 8-Channel RDIMM DDR5 per processor, 32 x DIMMs Dual ROM Architecture 2 x 1Gb/s LAN ports (Intel I350-AM2) 1 x Dedicated management port 32 x E1.S NVMe hot-swappable bays 2 x M.2 slots with PCIe Gen3 x4 or SATA 3.0 interface 3 x FHHL PCIe Gen5 x16 and x8 slots Compatible with SupremeRAID by Graid Technology 1+1 2000W 80 PLUS Titanium redundant power supplies	2U 4-node front access server system AMD EPYC 9004 Series processors AMD EPYC 9004 Series processors with AMD 3D V-Cache Technology Single processor per node, 5nm technology 12-Channel RDIMM DDR5 per processor, 96 x DIMMs Dual ROM Architecture 8 x 1Gb/s LAN ports (Intel I350-AM2) 4 x Dedicated management ports 2 x CMC ports 8 x E1.S Gen4 NVMe hot-swappable bays 12 x M.2 slots with PCIe Gen5 x4 and SATA interface 4 x LP PCIe Gen5 x16 slots 4 x OCP 3.0 Gen5 x16 slots Dual 3000W (240V) 80 PLUS Titanium redundant power supply	Supports up to 16 x Single slot Gen5 GPU cards AMD EPYC 9004 Series processors Dual processor, 5nm technology 12-Channel RDIMM DDR5, 24 x DIMMs Dual ROM Architecture 2 x 10Gb/s BASE-T LAN ports (Broadcom BCM57416) 1 x Dedicated management port 4 x 2.5″ Gen5 NVMe/SATA/SAS hot-swappable bays 4 x 2.5″ SATA/SAS hot-swappable bays 16 x FHFL PCIe Gen5 x16 slots for GPUs 2 x LP PCIe Gen5 x16 slots for add-on cards Dual 3000W (240V) 80 PLUS Titanium redundant power supply

Alongside these server innovations, GIGABYTE will also support the newly announced NVIDIA H200 Tensor Core GPU, which is expected to revolutionize generative AI and HPC with its advanced memory performance.

Engage with StorageReview

Newsletter | YouTube | Podcast iTunes/Spotify | Instagram | Twitter | TikTok | RSS Feed