- Compact 1U 19” rack chassis for space-saving server installations
- Single-socket Intel® Xeon® scalable processor with up to 165W TDP
- 4x PCIe x16 Gen3 slots for GPU and expansion cards
- 2x 10 Gigabit Ethernet ports for fast network connections
- 2x 2.5” hot-swap SATA 6Gbit/s for flexible storage options
FEATURES
Description
SHARK A.I. 4-Slot GPU server – Universal carrier-grade platform
The SHARK A.I. 4-Slot GPU server is a universal carrier-grade platform specifically designed for high-performance edge computing, AI and high-performance computing (HPC) applications. By combining a server node with a PCIe extension and a PCIe switch, the system supports up to four NVIDIA® GPUs to efficiently process a wide range of workloads. Thanks to the flexible architecture, topologies and bandwidths between GPUs and CPUs can be customized. InfiniBander support also enables easy scaling across multiple GPU clusters.
Maximum flexibility for AI and deep learning applications
The SHARK A.I. offers support for single-root and dual-root complexes, making it ideal for various AI frameworks. Single-root setups bundle all GPU resources for large-scale deep learning processes, while dual-root configurations distribute the load across CPUs and ensure more efficient machine learning processes. Dynamic PCIe lane allocation allows specific I/O and CPU cores to be configured to ensure optimal performance for virtualized applications.
Edge High Performance Computing (EHPC) for maximum computing power
As a powerful Edge High Performance Computing (EHPC) solution, the SHARK A.I. GPU Server significantly increases the agility and flexibility of cloud infrastructures. Thanks to its modular architecture, it can implement different head nodes and customize the allocation of GPUs per virtual machine (VM). This makes the SHARK A.I. ideal for remote virtualization, scientific computing, AI-powered analytics and GPU-accelerated simulations.
Robust design with hot-swap function for easy maintenance
The SHARK A.I. features a sophisticated hardware architecture that includes hot-swappable components, redundant fan modules and redundant power supplies. GPU cards can be easily replaced or expanded by removing the top cover – for a fast, maintenance-friendly configuration without long downtimes.
With its high-performance architecture, flexible configurability and industrial-grade robustness, the SHARK A.I. 4-slot GPU server is the perfect choice for AI-powered workloads, deep learning, HPC clusters and industrial automation.
Specifications
| Product: SHARK A.I. 4-Slot GPU server | |
|---|---|
| SYSTEM | |
| CPU | Single-socket Intel® Xeon® scalable processor, max up to 165W TDP |
| Chipset | Intel® C621 |
| GPU | Supports Single & Dual Root Complex Design |
| RAM | Up to 384GB DDR4 2466MHz (12x DIMM slots) |
| Storage | 2x 2.5″ Hot-Swap SATA 6Gbit/s |
| Expansion | 4x PCIe x16 Gen3 slots (Full Height / Half Length) |
| TPM | Optional |
| INTERFACE | |
| Ethernet | 2x 10-Gigabit Ethernet ports 1x Gigabit Ethernet ports (for IPMI) Intel® X550-AT2 controller |
| USB | 2x USB 3.0 (front) 1x USB 3.0 (rear) |
| COM | 1x COM (2×5-pin header) |
| Video | 1x VGA port |
| ENVIRONMENTAL | |
| Chassis | Compact 1U 19″ rack enclosure |
| Cooling | 10x 4cm fans (redundant) |
| Power Supply | 1600W AC/DC Platinum power supply unit; 1+1 redundancy Input: 1000W: 100~270V AC / 12A 1600W: 220~240V AC/ 9.48A |
| Operating Temperature | 10°C ~ 35°C |
| Storage Temperature | -40° ~ 70°C |
| Humidity | 90% @ 35° C non-condensing |
| Dimensions (W x H x D) | 438 x 43.5 x 885mm |
| Certifications | FCC (DoC), CE (DoC), CB/LVD, RCM, VCCI RoHS 6/6 conform |
