Product Overview
The midrange AI offering from Boston, the ANNA Ampere M1 server features the power of up to 4 NVIDIA A40 GPUs. The system supports PCI-E Gen 4 for fast CPU-GPU connection and high-speed networking expansion cards.
Ideal for AI/Deep Learning Training, Visualisation and HPC, this midrange AI offering from Boston features the power of up to 4 NVIDIA A40 GPUs. The ANNA Ampere M1 supports PCI-E Gen 4 for fast CPU-GPU connection and high-speed networking expansion cards.
With GPU Direct RDMA and 1 to 1 mapping between network interconnects and GPU. With 1 + 1 power redundancy it makes the system ideal for HPC and AI workloads. Designed with AI/Deep learning and Inference in mind, the ANNA Ampere M1 makes an ideal choice for a midrange system to start implementing advanced AI solutions.
Certified by NVIDIA for use with Nvidia AI Enterprise suite and with access to NGC.
Key Features
1U Rackmount Chassis Dual 3rd Generation Intel® Xeon® Scalable processors Up to 4TB Registered ECC DDR4 RDIMM/LRDIMM Up to 4x NVIDIA® A40 GPU’s (Double width) PCIe NVMe M.2 slot for fast boot drive 2 front Hot-swap 2.5" U.2 NVMe Gen4 drive bays
-
Chipset
Intel C621A
-
Drive Bays
2 Front Hot-swap 2.5" U.2 NVMe Gen4
-
Drive Support
1 M.2 NVMe for boot drive only
NVMe
-
Expansion Slots
1 PCI-E Gen 4 x8 AIOM networking slot
6 PCI-E Gen 4 x16 (4 internal, 2 external) slots
-
Form Factor
1U Rackmount
-
GPU Manufacturer
NVIDIA
-
GPU Quantity
Up to 4x NVIDIA A40 double width GPUs
-
Manufacturer
Supermicro
-
Memory (Maximum)
Up to 4TB: 16x 256 GB DRAM and 8x 512 GB PMem
-
Memory Type
3200/2933/2666MHz ECC DDR4 RDIMM/LRDIMM
Intel® Optane™ persistent memory 200 series
-
Network Connectivity
1 RJ45 1GbE Dedicated IPMI Management Port
2 x 10GBase-T ports
-
Power Supply
2000W Redundant Power Supplies with PMBus