HPE BL460c Gen10: NVIDIA Tesla P6 MXM GPU Upgrade

diyasjournal
8 hours ago
4 min read

Many organizations still use HPE BL460c Gen10 blades and upgrade them with NVIDIA Tesla P6 MXM GPUs to extend their usable life. This adds up to four MXM GPUs and is mainly used to keep retired BladeSystem environments running for VDI or remote graphics.

HPE BL460c Upgrade with NVIDIA Tesla GPUs

CLICK TO GET AN OFFER

✔ Up to 5-Year Warranty • Pay Only After Testing

HPE BL460c Gen10 server, NVIDIA Tesla P6 MXM GPU, Hewlett Packard Enterprise blade server, enterprise data center hardware, GPU accelerated computing, virtualization, AI machine learning, VDI workloads, high performance server components, HPE ProLiant BL460c, MXM graphics module, refurbished server hardware, refurbished GPU server, enterprise IT infrastructure, server-parts.eu refurbished solutions

HPE BL460c Upgrade with NVIDIA Tesla GPUs: NVIDIA Tesla P6 MXM GPU Upgrade

Market situation: HPE BL460c with NVIDIA Tesla GPU

The HPE ProLiant BL460c Gen10 is marked RETIRED / obsolete in official HPE documentation. The server is no longer sold new.
GPU upgrades on this platform are almost always installed-base or refurb-market projects.
Parts are typically sourced as spares or refurbished kits, often from decommissioned environments.
Vendor support is fragmented. Hardware, firmware, hypervisor, and NVIDIA software often fall under different support boundaries.

This is not a growth platform. It is a life-extension platform.

Compatibility overview: HPE BL460c with NVIDIA Tesla GPU

This upgrade applies only to the BL460c Gen10 blade and HPE BladeSystem enclosures.

Supported enclosures: HPE BladeSystem c3000 and HPE BladeSystem c7000
GPU form factor: MXM Type-B (not PCIe)
GPU model: NVIDIA Tesla P6
Maximum GPUs per blade: 4× Tesla P6 (validated in third-party hypervisor HCLs for passthrough and vGPU)

The GPUs connect through the HPE Graphics Expansion Blade and mezzanine passthrough boards, not through a standard PCIe slot.

HPE BL460c Upgrade with NVIDIA Tesla GPUs: Full parts list

To upgrade an HPE BL460c Gen10 blade server with NVIDIA Tesla P6 MXM GPUs, the following components are required:

Part number	Description
P00399-001	HPE Graphics Expansion Blade
882358-001	Mezzanine Pass-Thru GPU Adapter
882356-001	Mezzanine Pass-Thru Board
712767-001	GPU Expansion Short Cable
712768-001	GPU Expansion Long Cable
851463-001	MXM Heatsink Module
880813-001	NVIDIA Tesla P6 MXM GPU

All parts must be matched correctly. Mixing revisions or missing cables is a common cause of failed upgrades.

Complete 4× GPU Upgrade (NVIDIA Tesla) / HPE BL460c Gen10 Blade

One BL460c Gen10 blade can support 4× GPUs.. For 1× blade with 4 GPUs, the required parts are:

Qty	Part number	Description
1×	P00399-001	HPE Graphics Expansion Blade
2×	882358-001	Mezzanine Pass-Thru GPU Adapters
2×	882356-001	Mezzanine Pass-Thru Boards
1×	712767-001	GPU Expansion Short Cable
1×	712768-001	GPU Expansion Long Cable
4×	851463-001	MXM Heatsink Modules
4×	880813-001	NVIDIA Tesla P6 MXM GPUs

Exact kit contents may vary slightly by refurb bundle. Always verify before installation.

HPE BL460c Upgrade with NVIDIA Tesla GPUs: Tesla P6 MXM explanation

The NVIDIA Tesla P6 is a Pascal-generation data-center GPU designed for dense systems like blade servers.

Key technical points:

Architecture: Pascal (GP104)
Memory: 16 GB GDDR5
Form factor: MXM Type-B
Power (TDP): 90 W

For planning purposes, 90 W per GPU must always be used. There is no documented lower-power production variant.

The Tesla P6 was designed primarily for:

Virtualized graphics (vGPU)
Remote workstations
Light GPU compute workloads

HPE BL460c Upgrade with NVIDIA Tesla GPUs: Platform and Hardware Requirements

CPU requirements for Upgrading HPE BL460c Gen10 by NVIDIA Tesla GPU

The HPE BL460c Gen10 uses CPU-dependent mezzanine access.

Expansion slot 1 requires CPU in socket 1
Expansion slot 2 requires CPU in socket 2

For a 4× GPU configuration, two CPUs are mandatory. A single-CPU blade cannot access all required mezzanine paths for the graphics expansion blade.

Power & cooling requirements for Upgrading HPE BL460c Gen10 by NVIDIA Tesla GPU

Power and cooling are the most common limiting factors for this upgrade.

HPE guidance for HPE BL460c Gen10:

HPE c7000 enclosure: 10 fans populated
HPE c3000 enclosure: 6 fans populated

These are minimum recommendations.

Operational notes:

Validate enclosure power budgets using Onboard Administrator / OneView, not estimates.
GPU-heavy blades may force power derating at enclosure level.
Adjacent blades matter. Thermal crosstalk from neighboring high-TDP blades can affect stability.
Airflow policy must allow for sustained GPU load, not idle conditions.

BIOS & firmware requirements for Upgrading HPE BL460c Gen10 by NVIDIA Tesla GPU

Enclosure firmware (OA / Virtual Connect) must meet minimum supported levels.
Blade system ROM, iLO firmware, and mezzanine firmware should be aligned using a supported HPE Service Pack for ProLiant (SPP) baseline.
Do not mix firmware generations across blades in the same enclosure.

From a practical perspective:

Lock firmware versions early.
Avoid chasing “latest” updates on a retired platform.
Validate GPU visibility in BIOS before OS installation.

HPE BL460c Upgrade with NVIDIA Tesla GPUs: Supported and Unsupported Use Cases

Supported use cases: NVIDIA Tesla P6 MXM GPU

VDI with vGPU for light to mid-range 3D workloads
Remote visualization (CAD viewing, GIS, engineering desktops)
Light GPU compute that fits within 16 GB memory and Pascal-era performance

Unsupported use cases: NVIDIA Tesla P6 MXM GPU

Modern AI training workloads
Large-model inference
Deployments that require current NVIDIA vGPU software branches

Important lifecycle caveat:

Tesla P6 vGPU support is limited to older NVIDIA vGPU software branches.
NVIDIA has documented end of vGPU software support for Tesla P6 in July 2026.
After this point, no new features or security updates should be expected.
Compatibility with newer hypervisors is already constrained before this date

HPE BL460c Upgrade with NVIDIA Tesla GPUs: What matters for enterprise buyers

Support reality: This upgrade sits outside modern HPE and NVIDIA accelerator roadmaps. Validation typically relies on third-party HCLs (for example, older branches of Citrix XenServer) and field experience.
Software lifecycle risk: vGPU driver end-of-life may define the real platform end-date, not hardware failure.
Spare strategy: Buy extra GPUs, cables, and passthrough boards while they are still available.
Thermal testing: Always test a single blade under sustained load before scaling.
Exit plan: Treat this as a bridge solution with a defined retirement timeline.

For organizations keeping BladeSystem environments alive for VDI or remote graphics, this upgrade still works. But it must be approached as a controlled, time-boxed lifecycle extension, not a future-proof architecture.