On this page
Upgrading server memory is one of the most cost-effective ways to boost performance — but only if the modules fit and go in correctly. Here's a safe, six-step checklist for HPE ProLiant servers.
Last updated: 2026
Adding memory to an HPE ProLiant server can improve performance, stability, and headroom for growth — provided the modules are compatible and installed to HPE's rules. This checklist walks through the six essential steps, from assessing what you already have to testing the finished upgrade.
- Check your current setup — use iLO or a physical inspection to record installed modules and free slots.
- Verify compatibility — match memory type, speed, form, and ECC to your server and CPU.
- Select your modules — choose the right module type and capacity, and plan for future growth.
- Prepare to install — back up data, set up ESD protection, and gather tools.
- Install the memory — follow the correct alignment and slot population order.
- Test and monitor — confirm the modules are recognized and watch performance afterward.
Step 1: Check your current memory setup
Before buying anything, confirm exactly what's installed and what's free. This avoids compatibility surprises and tells you how much room you have to grow.
Use HPE iLO. The iLO management interface reports each installed module without opening the server — memory type (DDR4 or DDR5), capacity per module, operating frequency, rank, HPE part number, and health status, along with which slots are populated. On Linux you can also run dmidecode, and on VMware ESXi smbiosDump, for the same detail.
Count your open slots. From iLO's memory section or a careful physical inspection (following ESD precautions), note how many DIMM slots are empty. Slot count matters: a server with four modules in a 32-slot chassis has far more upgrade room than one that's already half full.
Review your server's QuickSpecs. HPE's QuickSpecs and Memory Configurator list the maximum supported capacity and the qualified module configurations for your exact model. For the full capacity picture by generation, see our guides to maximum memory for Gen11 servers and Gen12 servers.
Step 2: Verify compatibility
Compatibility is where most upgrades succeed or fail. The module has to match your server's generation and processor across several dimensions:
- Memory type — DDR4 for Gen10 and earlier, DDR5 for Gen11 and Gen12. The generations are not interchangeable.
- Speed — set by your processor. The system runs all installed memory at the speed of the slowest module, so match speed grades. (For why this matters, see our note on memory speed for HPE servers.)
- Form — RDIMM, LRDIMM, or UDIMM, determined by your server (see Step 3).
- ECC — HPE ProLiant servers use error-correcting memory; non-ECC modules are not supported.
- Rank — keep ranks consistent across a channel for predictable performance. (See understanding memory rank.)
The safest sources are genuine HPE modules, which carry full warranty coverage, guaranteed compatibility, and HPE Smart Memory features that unlock iLO health and performance reporting. Third-party memory that meets the specs may be cheaper and easier to find, but it can void HPE warranty coverage and lose those features. HPE's QuickSpecs and Memory Configurator will confirm which modules are qualified for your server — or we can verify the exact part for you.
Step 3: Select your modules
Once you know what fits, choose the specific module type and capacity.
Module type by generation. HPE servers use three module forms, and which one applies depends entirely on your platform:
| Module type | Buffering | Typical use | Best for |
|---|---|---|---|
| RDIMM (Registered) | Address/command buffered | Mainstream Gen10, Gen11, Gen12 | Standard enterprise workloads |
| LRDIMM (Load-Reduced) | Fully buffered | Gen10 and earlier (DDR4) | Maximum capacity on older platforms |
| UDIMM (Unbuffered ECC) | None | Entry servers (ML30, DL20) | Small-business and entry servers |
The mainstream Gen11 and Gen12 servers are RDIMM-only — LRDIMM was a Gen10-and-earlier option for reaching the highest DDR4 capacities, and UDIMM appears only on the entry ML30 and DL20. Never mix module types (RDIMM with LRDIMM, or registered with unbuffered) in the same server; doing so can prevent the system from booting.
Plan for growth. Where you have a choice, fewer high-capacity modules leave more slots free for later. For example, four 64GB modules reach 256GB while keeping most slots of a 32-slot server open, whereas eight 32GB modules hit the same capacity with less room to expand. Weigh current usage, expected growth over the next 18–24 months, and budget — and remember that DDR5 (Gen11/Gen12) brings higher bandwidth and lower power (1.1V) than DDR4 (Gen10, 1.2V, up to 3200 MT/s).
Step 4: Prepare to install
A little preparation prevents the two most common upgrade mishaps: data loss and static damage.
Back up first. Save your server's configuration, databases, and user data using the 3-2-1 rule — three copies, on two types of media, with one copy offsite.
Set up ESD protection. Static electricity can destroy memory modules. Wear an ESD wrist strap attached to the chassis, work on an antistatic mat, and handle modules only by their edges — never touch the gold contacts. Keep modules in antistatic packaging until you're ready to seat them.
Gather your tools. A non-magnetic screwdriver, your server documentation (for the population order), an ESD kit, and a clean, well-lit workspace. Then shut the server down and unplug all power cords before opening it.
Step 5: Install the memory
Power down and open. Shut the server down gracefully through the operating system, disconnect all AC power cords, release any keyed cover lock, and slide the cover off to reach the DIMM slots beside the processors.
Seat each module correctly. Line up the notch in the module with the key in the slot — it only fits one way — then press down with firm, even pressure until both retaining clips snap closed on their own. If the clips don't engage, don't force it: re-check the alignment and try again.
Follow the population order. HPE servers have a defined order — start with the white-labeled DIMM slot in each channel, populate channels evenly, and on dual-socket servers balance memory across both processors. Filling slots out of order can reduce performance or leave memory unrecognized, so follow the population rules in your server's QuickSpecs.
Close up and boot. Confirm every module is fully seated, reattach the cover, reconnect power, and start the server. If it doesn't boot, power down and re-seat the modules — a DIMM that looks seated sometimes needs a firmer press to make contact.
Step 6: Test and monitor
Confirm recognition. On boot, enter the BIOS/UEFI (or check iLO) and verify the total capacity, per-module details, speed, and channel configuration match what you installed. If the server shows less memory than expected, power down and re-seat the modules.
Run diagnostics. Use the HPE hardware diagnostics (via Intelligent Provisioning or the HPE support tools) to run memory tests. If errors appear, test one module at a time to isolate the culprit, and confirm each module is on your server's supported list.
Monitor afterward. For the first days after the upgrade, keep an eye on memory usage, response times, and the iLO event log. Document the final configuration — module part numbers, slot layout, and BIOS settings — to make the next upgrade faster.
Buy genuine HPE memory
Genuine HPE memory is qualified and authenticated for your specific server, which is what preserves warranty coverage, HPE Smart Memory features, and full iLO health reporting — and it's all we sell. Shop by generation to get guaranteed-compatible modules:
- HPE Gen12 server memory and HPE Gen11 server memory (DDR5)
- All HPE ProLiant server memory across generations
For a worked example of how capacity, speed, and rank come together on one platform, see our deep dive on the memory capabilities of the DL380 Gen11. Not sure which module fits your server? Contact us with your model or serial number and we'll confirm the right part before you order.
Frequently asked questions
How do I check what memory is currently in my HPE server?
The easiest way is through HPE iLO, which reports each installed module's slot, capacity, speed, rank, part number, and health without opening the server. On Linux you can also run dmidecode, and on VMware ESXi smbiosDump. For a physical check, power down and inspect the slots following ESD precautions.
How do I know if memory is compatible with my HPE ProLiant server?
Match the module to your server's generation and processor: the correct memory type (DDR4 for Gen10 and earlier, DDR5 for Gen11 and Gen12), the speed your CPU supports, the right form (RDIMM, UDIMM, or LRDIMM), and ECC. HPE's QuickSpecs and Memory Configurator list the qualified modules for each model, and we can confirm the exact part for your server.
What's the difference between RDIMM, LRDIMM, and UDIMM?
They differ in how the module buffers signals. RDIMM (Registered) buffers the address and command lines and is the mainstream enterprise choice across current HPE generations. LRDIMM (Load-Reduced) fully buffers the data lines for the highest capacity on Gen10 and earlier DDR4 platforms. UDIMM (Unbuffered ECC) has no buffering and is used only on entry servers like the ML30 and DL20. They are not interchangeable.
Can I mix different memory modules in an HPE server?
Not freely. You should never mix module types (such as RDIMM and LRDIMM, or registered and unbuffered) in the same server, and you shouldn't mix speeds or ranks within a channel, since the system runs all memory at the slowest module's speed and mismatches can prevent boot. For best results, use identical HPE-qualified modules across all channels.
Do I have to install memory in a specific order?
Yes. HPE servers have a defined population order: start with the white-labeled DIMM slot in each channel and populate channels evenly, balancing memory across both processors on dual-socket systems. Follow the population rules in your server's QuickSpecs, because installing modules out of order can reduce performance or leave memory unrecognized.
Why isn't my server recognizing all the new memory?
Usually a seating or population issue. Power down and re-seat the modules until both latches click, confirm you followed the correct slot population order, and check that every module is the same qualified type and speed. If a module still isn't recognized, test it individually and verify it's on your server's supported list.
Does third-party memory work in HPE ProLiant servers, and will it void my warranty?
Third-party memory that matches the specs may physically work, but it can void HPE warranty coverage, lose HPE Smart Memory features and iLO health reporting, and isn't guaranteed compatible. Genuine HPE memory is qualified and authenticated for your server, verified for your exact model.
The bottom line
A memory upgrade is low-risk when you work in order: confirm what's installed, verify compatibility against your server's QuickSpecs, choose the right module type and capacity, install with ESD protection and the correct population order, and test before you rely on it. Get those right and you'll add capacity without downtime or wasted spend. Send us your server model or serial number and we'll match guaranteed-compatible HPE memory before you order.
