(858) 613-1818Quote list 0

10 GPU PCIe AI server, 5U

Ten PCIe GPUs, RTX PRO 6000 Blackwell Server Edition or L40S, in a 5U server with dual EPYC 9005 or Xeon 6, the highest GPU count per air cooled node.

GPU: NVIDIA RTX PRO 6000 Blackwell Server Edition or L40S

Illustration of a 5U GPU server with ten PCIe GPUs on edge seen through the lid window and a front fan wall
GPUs
10 x RTX PRO 6000 Blackwell Server Edition (960 GB total) or 10 x L40S
CPU
2 x AMD EPYC 9005 or 2 x Intel Xeon 6
Memory
Up to 2 TB DDR5 RDIMM
Storage
2 x M.2 boot, up to 8 x U.2 or E3.S NVMe
Networking
Up to 5 x 400 Gb/s ConnectX-7 (one per GPU pair), 2 x 25 GbE
Power
Redundant PSUs, about 7.5 kW per node planning figure

Overview

The 10 GPU server is the highest PCIe GPU count we offer in an air cooled node. Ten RTX PRO 6000 Blackwell Server Edition cards give 960 GB of GDDR7 and 40 PFLOPS of FP4, which makes a single node a serious inference platform for many concurrent models, or a rendering and simulation engine with ten independent accelerators.

The 5U chassis isolates the GPUs in their own thermal zone with high static pressure fans, so all ten cards can run at 600 W. PCIe Gen5 switches group the GPUs into five pairs, and each pair can have its own 400 Gb/s adapter for GPUDirect RDMA into a fabric. Choose L40S instead when power is limited or when Ada Lovelace validation is required.

Hosts are dual AMD EPYC 9005 or Intel Xeon 6 with DDR5 to 2 TB, mirrored boot and up to eight NVMe drives. Plan for about 7.5 kW per node and eight nodes per rack at 40 kW. Each system is burned in with all ten GPUs loaded and delivered with per slot thermal data in the acceptance report.

Use cases

  • High density inference per node
  • Multi model serving
  • Rendering and simulation farms
  • Video processing at scale
  • Virtual workstations

Specifications

Compute
GPUs10 x NVIDIA RTX PRO 6000 Blackwell Server Edition (96 GB GDDR7) or 10 x L40S (48 GB GDDR6)
GPU interconnectPCIe Gen5 switches, five GPU pairs each with a direct path to one adapter, no NVLink
FP4 performance40 PFLOPS per node with RTX PRO 6000
CPUs2 x AMD EPYC 9005 series or 2 x Intel Xeon 6
System memory24 or 32 x DDR5 RDIMM, up to 2 TB
Storage
Boot2 x M.2 NVMe, mirrored
DataUp to 8 x U.2 or E3.S NVMe
RAIDSoftware RAID or hardware RAID
Networking
GPU fabricUp to 5 x 400 Gb/s ConnectX-7, InfiniBand or RoCE
Frontend2 x 25 GbE onboard, optional 100 GbE
ManagementDedicated 1 GbE BMC port, TPM 2.0
Physical
Form factor5U rackmount, air cooled, GPUs in a dedicated rear thermal zone
Power suppliesRedundant 3,000 W class PSUs, 200 to 240 V input
Planning powerAbout 7.5 kW per node with 600 W GPUs, about 5 kW with L40S

Figures describe the system class. Exact configuration, weights and power draw are confirmed on the quotation.

What's included

  • Rails, power cords and cable management arm
  • Burn in with multi hour GPU stress and NVLink validation, acceptance report
  • NVIDIA end user registration handled with the order
  • Operating system imaging, driver and fabric manager baseline on request
  • 3 year warranty, options to 5 years with advance replacement

Compliance

Compare it with the next class up or down

The nearest classes we quote are the 8 GPU RTX PRO 6000 PCIe AI server and the 4 GPU PCIe AI server, 2U. If neither is quite right, tell us the workload and we specify the system between them.

Talk to an engineerRequest a quotation