Skip to content
HomeSolutionsLenovoThinkSystem SR665 V3 Dual EPYC Turin with 3x NVIDIA H100 NVL HPC Server
Lenovo logo
Lenovo

ThinkSystem SR665 V3 Dual EPYC Turin with 3x NVIDIA H100 NVL HPC Server

Dense 2U GPU compute for large-model inference and HPC acceleration

Full manufacturer warrantyAuthorized channel48-hour quote

We help you choose, configure, and deliver the right system — no obligation.

ThinkSystem SR665 V3 Dual EPYC Turin with 3x NVIDIA H100 NVL HPC Server — Lenovo enterprise hardware
Lenovo logo
ThinkSystem SR665 V3 Dual EPYC Turin with 3x NVIDIA H100 NVL HPC Server hardware detail 1
ThinkSystem SR665 V3 Dual EPYC Turin with 3x NVIDIA H100 NVL HPC Server hardware detail 2
ThinkSystem SR665 V3 Dual EPYC Turin with 3x NVIDIA H100 NVL HPC Server hardware detail 3

Configuration at a Glance

GPU3x NVIDIA H100 NVL 94GB HBM3 (Hopper, double-wide)
CPU2x AMD EPYC 9555 (5th Gen Turin, 64 cores each, PCIe Gen5)
Memory1.5TB DDR5 (24x 64GB RDIMM, 6400 MHz), up to 6TB
Storage8x 2.5-inch NVMe Gen5 hot-swap SSDs

Tailored per engagement. Full technical overview below.

Configuration Options

Core specifications for this system. Every component is configurable to your workload — request a quote for a tailored build.

GPU / Accelerator

3x NVIDIA H100 NVL 94GB HBM3 (Hopper, double-wide)

Processor

2x AMD EPYC 9555 (5th Gen Turin, 64 cores each, PCIe Gen5)

Memory

1.5TB DDR5 (24x 64GB RDIMM, 6400 MHz), up to 6TB

Storage

8x 2.5-inch NVMe Gen5 hot-swap SSDs

Overview

This SR665 V3 pairs dual 5th Gen AMD EPYC Turin processors with three double-wide NVIDIA H100 NVL GPUs for large-model inference, fine-tuning, and HPC. Nexus Compute specifies, assembles, and validates each system, delivering it warranty-backed through authorized Lenovo channels.

Who This Solution Is For

Research teams running HPC and AI workloads
Enterprises serving large language models
Teams fine-tuning multi-billion-parameter models
HPC centers needing dense GPU compute

Business Benefits

Top-tier GPU compute

Three H100 NVL GPUs bring data-center-class acceleration to a compact 2U chassis.

Massive core count

Dual EPYC Turin processors supply abundant cores and PCIe Gen5 lanes to keep GPUs fed.

Configured and supported

Each node is tested to spec with firmware and drivers set, backed by Lenovo onsite warranty.

Typical Business Use Cases

1

Large language model inference

2

Model fine-tuning and training

3

HPC simulation and modeling

4

Scientific and engineering compute

Industry Applications

AI & Machine LearningHPCHigher Education & ResearchHealthcare & Life Sciences

Technical Overview

The Lenovo ThinkSystem SR665 V3 is a 2-socket 2U server built on 5th Gen AMD EPYC 9005 (Turin) processors with up to 160 cores each and up to 160 PCIe Gen5 lanes. Its support for three double-wide GPUs is populated here with NVIDIA H100 NVL Hopper accelerators carrying 94GB HBM3 each for large-model and HPC workloads.

GPU3x NVIDIA H100 NVL 94GB HBM3 (Hopper, double-wide)
CPU2x AMD EPYC 9555 (5th Gen Turin, 64 cores each, PCIe Gen5)
Memory1.5TB DDR5 (24x 64GB RDIMM, 6400 MHz), up to 6TB
Storage8x 2.5-inch NVMe Gen5 hot-swap SSDs
NetworkingOCP 3.0 dual-port 100GbE plus PCIe Gen5 x16 slots for fabric NICs
Form Factor2U rack
ManagementLenovo XClarity Controller (XCC) with XClarity Administrator
PowerDual redundant 2600W 80 PLUS Titanium PSUs
Warranty3-year Lenovo onsite limited warranty, 9x5 NBD

Specifications are indicative and configured to each engagement. Request a quote for a configuration tailored to your requirements.

Warranty, Support & Fulfillment

Every system ships from an authorized channel, configured and tested, with the documentation enterprise buyers need — backed by warranty and a dedicated account team.

Enterprise Warranty

Full manufacturer warranty with optional on-site, next-business-day support and extended coverage.

Authorized Channel

Sourced through Tier-1 distribution and OEM partners — never grey market. Asset & warranty records included.

Lead Time & Deployment

48-hour quotes, then configured, burn-in tested, and delivered on a committed schedule.

Nationwide Fulfillment

Coordinated logistics, rack-and-stack, and delivery wherever your infrastructure lives.

Frequently Asked Questions

Is three H100 NVL GPUs enough for training?

It suits fine-tuning, mid-scale training, and large-model inference; for full multi-node pretraining we design clustered deployments with high-speed fabric across multiple SR665 V3 nodes.

Why H100 NVL instead of L40S for this build?

H100 NVL delivers HBM3 bandwidth and Transformer Engine acceleration far beyond the L40S, making it the choice for the largest models and HPC, while the L40S SR650 V3 targets cost-efficient inference and graphics.

How do you handle power and cooling for this density?

We specify dual 2600W Titanium PSUs and validate rack power and airflow; for high-density rows we advise on thermal limits and PDU sizing before deployment.

Hardware Assistance

Configure the ThinkSystem SR665 V3 Dual EPYC Turin with 3x NVIDIA H100 NVL HPC Server with Nexus Compute

Tell us your requirements and a hardware specialist will help you specify, configure, and quote the right system — typically within two business days. No obligation.