Giga Computing Backs Enterprise AI With AMD’s 6th-Gen EPYC

Enterprise AI deployments are increasing demand for CPU density, memory bandwidth and connectivity as inference, virtualization and cloud workloads converge in the data center. Giga Computing, a GIGABYTE subsidiary, is positioning its expanded partnership with AMD as infrastructure for that shift. The combination is intended to improve GPU utilization and lower total cost of ownership, helping enterprises scale AI services while giving hardware suppliers a path to monetize compute through recurring, service-based offerings rather than one-off system sales.

On July 23 at AMD Advancing AI 2026 in San Francisco, Giga Computing said its server lineup would support 6th Gen AMD EPYC CPUs and unveiled its first system for the EPYC 9006 platform. The SP7 design supports up to 256 cores and 512 threads per socket, PCIe Gen 6 and 16-channel DDR5; liquid-cooled racks can exceed 40,000 CPU cores. SP7 systems are due to ship alongside AMD in November 2026, while SP8 systems, aimed at space- and power-constrained scale-out workloads, are scheduled for March 2027.