SYS-420GH-TNGR
4U AI Training GPU Server. Dual 3rd Gen Intel® Xeon® Scalable Processors
- Dual Intel® Xeon® Scalable Processors (Ice Lake) TDP up to 270W
- Intel® C621A Chipset
- 32x DIMM slots, up to 8TB Registered ECC DDR4 3200MHz SDRAM
- 1x PCIe 4.0 x16 FHHL slot plus 2x PCIe 4.0 AIOM (OCP 3.0 superset) slots
- 24x on-board 100GbE (via 6x QSFP-DDs) for scale-out and 1x 10GbE BaseT (dedicated IPMI port)
- 4x 2.5” SAS/SATA/NVMe Hybrid Hot-Swap Bays
- 5x Removable Heavy-Duty Fan(s)
- 4x 3000W(54V+12V)/2000W(12V) redundant, Titanium Level power supplies
Key Applications
- Computer Vision
- Recommendation
- Natural Language Processing
- Purpose Built for AI/Deep Learning Training
Product Specification
Product SKUs | SuperServer SYS-420GH-TNGR (Silver) |
Motherboard | Super X12DPG-U6 |
Processor | |
CPU | Dual Socket P+ (LG-4189) |
Core Count | Up to 40C/80T |
Note | Supports up to 270W TDP CPUs (Air Cooled) |
Hardware Accelerator | |
AI Processor | 8x Habana training Gaudi HL-205 OAM |
System Memory | |
Memory | Memory Capacity: 32 DIMM slots Up to 8TB: 32x 256 GB DRAM Memory Type: 3200/2933/2666MHz ECC DDR4 RDIMM/LRDIMM |
Memory Voltage | 1.2 V |
Error Detection | ECC |
On-Board Devices | |
Chipset | Intel® C621A |
Network Connectivity | 6x 400GbE QSFP-DD 1x 10GbE BaseT Note: IPMI LAN port on AI processor tray, two AIOM slots on CPU tray. |
IPMI | Support for Intelligent Platform Management Interface v.2.0 IPMI 2.0 with virtual media over LAN and KVM-over-LAN support |
Input / Output | |
USB | 2 USB 3.0 port(s) (2 USB) |
Video | 1 VGA port(s) |
System BIOS | |
BIOS Type | AMI 32MB SPI Flash EEPROM |
Management | |
Software | IPMI 2.0 KVM with dedicated LAN Super Diagnostics Offline SuperDoctor® 5 Supermicro Update Manager (SUM) Supermicro Power Manager (SPM) Supermicro Server Manager (SSM) Redfish API |
Power Configurations | ACPI Power Management Power-on mode for AC power recovery |
Security | |
Hardware | Trusted Platform Module (TPM) 2.0 Silicon Root of Trust (RoT) – NIST 800-193 Compliant |
Features | Cryptographically Signed Firmware Secure Boot Secure Firmware Updates Automatic Firmware Recovery System Lockdown |
PC Health Monitoring | |
CPU | 7+1 Phase-switching voltage regulator Monitors for CPU Cores, Chipset Voltages, Memory |
FAN | Fans with tachometer monitoring Pulse Width Modulated (PWM) fan connectors Status monitor for speed control |
Temperature | Monitoring for CPU and chassis environment Thermal Control for fan connectors |
Chassis | |
Form Factor | 4U Rackmount |
Model | CSE-428G2 |
Dimensions and Weight | |
Height | 7" (178mm) |
Width | 17.6" (447mm) |
Depth | 32" (813mm) |
Package | 15" (H) x 27.3" (W) x 49.4" (D) |
Weight | Gross Weight: 137 lbs (62 kg) |
Available Color | Silver |
Front Panel | |
Buttons | Power On/Off button |
LEDs | Carrier Board Node LED CPU Node LED Information LED PSU Failed LED UID LED |
Expansion Slots | |
PCI-Express (PCIe) | 1 PCIe 4.0 x16 FHFL slot(s) 2 PCIe 4.0 x16 AIOM slot(s) |
Drive Bays / Storage | |
Hot-swap | 4x 2.5" hot-swap NVMe/SATA/SAS drive bays (4x 2.5" NVMe hybrid) |
M.2 | 2 M.2 NVMe OR 2 M.2 SATA3 |
System Cooling | |
Fans | 5 removable heavy-duty Fan(s) |
Fans | 5 Removable Heavy-Duty 84x203.5x82.4mm Fan(s) |
6000W Redundant Titanium Level power supplies | |
Certification | Titanium Level |
Operating Environment | |
Environmental Spec. | Operating Temperature: 10°C ~ 35°C (50°F ~ 95°F) Non-operating Temperature: -40°C to 60°C (-40°F to 140°F) Operating Relative Humidity: 8% to 90% (non-condensing) Non-operating Relative Humidity: 5% to 95% (non-condensing) |
AI Training Optimized
The Supermicro X12 Gaudi AI Training System prioritizes two key real-world considerations: training AI models as fast as possible, while simultaneously reducing the cost of training with unprecedented scale-out bandwidth.
Key Features
- GPU: 8 Gaudi HL-205 mezzanine cards
- CPU: Dual 3rd Gen Intel® Xeon® Scalable Processors
- Memory: 32 DIMMs: up to 8TB registered ECC DDR4-3200MHz SDRAM
- Drives: 4 hot-swappable NVMe/SATA hybrid hard drives
- Power: 4x 3000W Titanium Level redundant power supplies
- Networking: 24x 100GbE RDMA (6 QSFP-DDs)
- Workloads: AI Training and Inferencing