Overview

The FitServer G4A60 V5 is the latest 4U dual-socket GPU computing-accelerated rack server launched by FiberHome Communications. It adopts the high-performance Intel C622 chipset and supports the latest 2nd Gen Intel® Xeon® Scalable processors. This server model is equipped with 24 DDR4 DIMM slots, supporting a maximum memory capacity of 6TB; it can accommodate up to 24 2.5-inch hot-swappable SAS/SATA hard drives; it features 10 full-height, full-length dual-width GPU slots, along with an additional 1 PCIe 3.0 x16 slot and 1 PCIe 3.0 x8 slot; it has 2 onboard 10GbE network interfaces and supports multiple network options including Ethernet, InfiniBand, and OPA (Omni-Path Architecture); it supports IPMI 2.0 for remote management; and it is powered by 2000W 2+2 Titanium redundant power supplies.

As a new-generation high-density AI acceleration computing server, the G4A60 V5 is mainly applied in the fields of enterprise virtualization, automated inference, deep learning, and high-performance computing (HPC). By leveraging GPUs to handle computationally heavy and time-consuming code segments, it accelerates applications running on CPUs, helping users complete more computing tasks, process larger datasets, and reduce application runtime.

Application Environments

Suitable for application scenarios such as enterprise GPU virtualization, automated inference, deep learning, and high-performance computing.

Applied in mainstream HPC applications, primarily targeting traditional HPC fields including bioinformatics, computational chemistry, computational finance, computational fluid dynamics, and astrophysics.

Used in AI-related applications such as sensory and computer vision, face recognition, machine learning, medical imaging, life sciences, numerical analysis, and weather & climate research.

Product FeaturesHigh Performance

High Performance

Supports 2 units of 2nd Gen Intel® Xeon® Scalable processors (Skylake/Cascade Lake series). A single processor can have a maximum of 28 cores, delivering powerful computing capabilities to provide higher performance for various user applications.

Equipped with 24 DIMM memory slots and an integrated 6-channel memory controller. The memory operates at a frequency of up to 2933MHz, with a maximum total capacity of 6TB; it also supports up to 12 Intel Optane Memory modules as an optional configuration.

Supports up to 10 full-height, full-length dual-width GPU cards; supports GPU Direct RDMA to enable diverse application scenarios.

Supports a maximum of 12 PCIe 3.0 expansion slots, providing users with flexible accessory connectivity options.Offers multiple network choices, including high-speed interconnect interfaces such as 1G/10G/40G QDR InfiniBand/56G FDR InfiniBand, plus 1 dedicated 1000M network port for remote management.

High Reliability

Supports up to 24 hot-swappable 2.5-inch SATA/SAS hard drives, enabling large-capacity local storage.
Features 8 high-power intelligent temperature-controlled fans, which ensure heat dissipation and stability for mainstream industry GPU computing acceleration cards while reducing overall server noise.
Adopts full shielding design to resist electromagnetic interference (EMI) and electrostatic discharge (ESD).
Supports Intel® Trusted Execution Technology (Intel® TXT), providing reliable protection for virtual and cloud environments.
Incorporates Intel® Advanced Encryption Standard New Instructions (Intel® AES-NI) to enhance encryption capabilities, enabling faster and more robust enterprise-level data protection.

Supports 2+2 fully redundant power supply configuration, providing higher security guarantees for the server’s stable operation.

Low Power Consumption

Supports frequency reduction function, which reduces system energy consumption under low-load conditions.
Adjusts fan speeds comprehensively based on the temperature of key components, saving energy while reducing noise and improving fan reliability.
Supports ACPI (Advanced Configuration and Power Interface) energy management, providing automated server power control and helping complete more tasks with fewer resources.