Nutanix and GPUs: building an AI cluster

Which nodes take cards — and which do not

A hyperconverged node is a server where disks, compute and networking are fused into one block under AOS. Room for a full-size card is not always there — drive bays and fans eat it. So some NX models take up to four accelerators and others take none.

Per Nutanix node specifications; the last row are database and storage nodes

Reading GPU capability from the letter suffix does not work — the table shows it directly. There are two reliable checks: the GPU line in the model’s spec sheet, and the card list validated for that chassis. NX-9151-G9 and NX-3155-G9 accept the H100; the NX-8155-G9 with the same 2U1N form factor stops at the L40S. Power and airflow set the limit, not case size.

What blocks a fifth card

The power spread across the line is sixfold: an NX-1150S-G9 peaks at 531 VA, while an NX-9151-G9 with four cards requires 3,232 VA and rejects up to 11,021 BTU per hour — a separate conversation with your data centre about the phase and row cooling.

The second reason: server cards are passive. The L40S and RTX PRO 6000 Server Edition have no fans of their own and live off front-to-back chassis airflow. NVIDIA’s forum has a telling case: an L40S in a Dell Precision workstation idling at 100 °C. Outside a server, this cannot be fixed.

Three ways to hand a card to a VM

Then comes the limit that breaks half of all multi-card node plans. The MIG documentation states it plainly: with MIG enabled, NVLink links are off, the card loses P2P, and NCCL is unsupported. NVIDIA’s forum position: distributed training over MIG can theoretically be pushed through GPUDirect RDMA over the network, but it will be slower than taking a bigger slice or disabling MIG. Partitioning a card and fast card-to-card exchange are mutually exclusive modes.

Which card splits how far

The last row deserves attention when picking a node: the L40S is the workhorse card for the NX-9151-G9 and NX-3155-G9 — and it is exactly the one that does not slice. “How do we partition the L40S then?” is a question asked on NVIDIA’s forum by people who chose the card by TFLOPS and looked at the partitioning model afterwards.

What works on AHV

NVIDIA maintains a dedicated Nutanix AHV support page and classes the hypervisor as generic Linux with KVM. Ordinary time-sliced vGPU works. But in the consolidated feature table, MIG-backed vGPU is ticked only for vSphere, Red Hat KVM and Ubuntu KVM from release 19.0. Verify compatibility for your AOS version before ordering.

Two more things surface after deployment. Live migration of a vGPU VM requires the identical physical card, driver version and hypervisor version on the receiving host — and never works across different MIG profiles. And unlicensed behaviour: 20 minutes at full speed, then 1280×1024 at three frames per second with CUDA off. C-series compute profiles come only with NVIDIA AI Enterprise (licensed per GPU), graphics vWS per concurrent user — two separate budget lines.

How many VMs per card

Compute profiles for the RTX PRO 6000 Blackwell Server Edition

For virtual desktops NVIDIA quotes the same density for both RTX PRO cards: two to four heavy users per card, three to six medium, six to eight light. And the budget-saving caveat: the best-effort scheduler often allows 2–3× oversubscription. The monitoring rule is simple — VM video-memory fill should rarely cross 90% and average below 70%.

The last row is the price of partitioning: an individual slice is roughly five times slower than the whole card. Aggregate throughput still grows, because a small model does not saturate an A100 — and scaling is sublinear.

Where to start and how to grow

The working starting configuration is three nodes: RF2 replication runs on three, and the cluster survives losing one node without stopping VMs. Cards are not needed in every node — the pattern that most often works: three nodes for general workloads, one or two of them with accelerators for inference.

Keep the GPU pool homogeneous, and the reason is live migration. A second node with the same card gives headroom and a place for VMs to move during maintenance. A mixed cluster with an L40S in one node and an RTX PRO 6000 in another works too — but every planned maintenance turns into an evening of service stops.

What we supply

Eurokommerz supplies the Nutanix NX line together with the GPUs validated for it — L40S, L4, H100 — plus the RTX PRO 6000 Server Edition and H200 NVL, with EU invoicing and manufacturer warranty. Deployment and virtualisation engineering comes from our partner Vixen.UNO under the same contract.

FAQ