HPE ProLiant Servers for AI Workloads: What IT Teams Should Consider in 2026
Share
Artificial intelligence is changing the way businesses think about their IT infrastructure.
For years, server planning was mostly about virtualization, databases, business applications, file services and general compute capacity. Now, many IT teams are adding AI-assisted applications, analytics, automation, inference and data-intensive workloads to that mix.
That creates a practical question:
What should you look for when choosing an HPE ProLiant server for AI-related workloads in 2026?
The answer isn't simply to buy the newest or most powerful server.
The right choice depends on what the server will actually do, how much processing power the workload needs, how much memory and storage it requires, whether you need accelerator support, and how much room you want for future expansion.
For businesses buying enterprise hardware online, Chicago Computer Supply offers HPE ProLiant servers and related infrastructure. But before choosing a configuration, it helps to understand what your AI workload actually requires.
AI Doesn't Automatically Mean You Need the Most Powerful Server
One of the easiest mistakes to make when planning AI infrastructure is assuming that every AI workload needs a specialized, high-end server.
That's not necessarily true.
A company might be using AI for:
- Internal productivity applications
- Document analysis
- Business analytics
- Enterprise search
- Data processing
- AI-assisted applications
- Predictive analytics
- Automation
- AI inference
Another organization might be training or fine-tuning much larger models.
Those two environments can have completely different infrastructure requirements.
That's why the first question shouldn't be:
"Which HPE server should we buy?"
It should be:
"What AI workload are we actually trying to run?"
1. Start With the Workload, Not the Server Model
Before looking at specifications, define the workload.
Ask your IT team:
- Are we running AI inference?
- Are we processing large datasets?
- Are we running AI applications alongside virtual machines?
- Are we using AI for analytics?
- Do we need GPU acceleration?
- Will the workload grow over the next two or three years?
- Will the same server also run traditional business applications?
This matters because an enterprise environment rarely consists of AI alone.
A server may need to support a combination of:
virtualization + databases + analytics + applications + AI
That makes flexibility just as important as raw processing power.
2. CPU Performance Still Matters
AI discussions often focus heavily on GPUs, but CPUs remain an important part of the infrastructure.
The CPU may handle:
- Application processing
- Data preparation
- Database workloads
- Virtualization
- Analytics
- Operating-system tasks
- AI orchestration
- General business applications
The HPE ProLiant Gen12 systems available through Chicago Computer Supply include different processor and configuration options.
For example, the HPE ProLiant DL380 Gen12 P89243-005 is configured with an Intel Xeon 6505P 12-core processor, 64 GB of DDR5 registered ECC memory and two 480 GB hot-swap SSDs.
There are also higher-core configurations.
The P90005-005 HPE ProLiant DL380 Gen12, for example, uses a 32-core Intel Xeon 6530P and 64 GB of DDR5 memory.
That illustrates an important point:
The server model alone doesn't tell you whether a configuration is right for your workload.
The processor, memory, storage and expansion options all matter.
3. How Much Memory Will Your AI Workload Need?
Memory can become one of the first bottlenecks in data-intensive workloads.
AI applications may work with:
- Large datasets
- Models
- Databases
- Analytics workloads
- Multiple virtual machines
- Containers
- Cached data
So don't only ask how much memory the server has when you purchase it.
Ask:
How much memory can the platform support if our workload grows?
The HPE ProLiant DL380 Gen12 platform supports substantial DDR5 memory expansion. The P89243-005 configuration listed by Chicago Computer Supply includes 64 GB DDR5 ECC memory, while the platform is described as supporting significantly larger memory configurations.
That expansion capability can matter more than the amount of RAM installed on day one.
4. Storage Is More Than Just Capacity
AI workloads can generate and process large amounts of data.
Depending on the application, your storage environment may contain:
- Training datasets
- Business data
- Model files
- Analytics data
- Application data
- Logs
- Documents
- Search indexes
That means you need to consider both capacity and performance.
For example, the HPE ProLiant DL380 Gen12 P90004-005 configuration available from Chicago Computer Supply includes two 960 GB SATA SSDs and supports up to eight hot-swap 2.5-inch drives.
A different configuration may make more sense if your priority is higher compute density, additional storage or a different drive architecture.
The right question is therefore not simply:
"How many terabytes do we need?"
It's:
"How much data will we process, and how quickly does the application need to access it?"
5. Don't Forget PCIe Expansion
AI-related workloads can increase the need for expansion.
Depending on the environment, you may eventually need:
- Additional networking
- Storage controllers
- High-speed storage
- Accelerators
- Other PCIe devices
The HPE ProLiant DL380 Gen12 configurations available through Chicago Computer Supply include PCIe Gen5 expansion capabilities, depending on the configuration.
This is important because infrastructure requirements can change.
You may not need every expansion option today.
But if you expect the workload to grow, having room to expand can prevent an expensive replacement later.
6. Consider Rack Density
Not every organization has unlimited rack space.
That makes server form factor an important part of the buying decision.
For example:
HPE ProLiant DL380 Gen12
The DL380 Gen12 is a 2U rack server designed for enterprise workloads, virtualization, databases, analytics and hybrid infrastructure. Chicago Computer Supply currently lists multiple DL380 Gen12 configurations.
HPE ProLiant DL360 Gen12
The DL360 Gen12 is a 1U rack server, making it attractive where rack density is an important consideration. Chicago Computer Supply currently lists DL360 Gen12 configurations with 64 GB DDR5 memory and Intel Xeon processors.
This creates a useful distinction:
DL380 → more physical space and expansion flexibility
DL360 → greater rack density
Neither is automatically "better."
It depends on the environment.
7. Power and Cooling Need to Be Part of the Decision
AI-related workloads can increase infrastructure demands, especially when workloads become compute-intensive.
Before installing a new server, IT teams should consider:
- Available rack power
- Power redundancy
- Cooling capacity
- Airflow
- Rack density
- Future expansion
- Electricity costs
For example, the Chicago Computer Supply configuration P89243-005 includes dual hot-plug 800W power supplies, while the P90004-005 configuration uses dual 1000W power supplies.
That doesn't mean you should simply choose the configuration with the largest power supply.
It means the server's power requirements should be considered alongside the facility's capacity and the workload's actual needs.
8. Enterprise AI Still Needs Enterprise Security
AI workloads don't exist in a separate world.
The server is still connected to:
- Business applications
- Databases
- Corporate networks
- User accounts
- Storage systems
- Sensitive business information
Security therefore remains a fundamental part of server selection.
The HPE ProLiant Gen12 configurations available through Chicago Computer Supply include features such as HPE iLO 7, Secure Boot, TPM support and Silicon Root of Trust-related security capabilities, depending on the configuration.
For an IT team, the question isn't only:
"Can this server run our workload?"
It should also be:
"Can we securely manage and maintain this server throughout its lifecycle?"
9. Remote Management Matters More Than It Seems
When a business has multiple servers, physically visiting each machine for maintenance isn't practical.
Remote management can help IT teams:
- Monitor hardware
- Perform configuration tasks
- Manage firmware
- Troubleshoot problems
- Monitor server health
- Reduce unnecessary physical intervention
HPE iLO 7 is included in the HPE ProLiant Gen12 configurations listed by Chicago Computer Supply.
For distributed IT environments, that can become a significant operational advantage.
10. Think About AI as Part of the Broader IT Environment
This is where many infrastructure discussions go wrong.
AI shouldn't necessarily be treated as a completely separate infrastructure project.
Suppose an organization already has:
- ERP systems
- CRM
- Databases
- Virtual machines
- File servers
- Business applications
and now wants to introduce AI.
The new server may need to support both the existing environment and the new workload.
In that situation, a flexible enterprise server may make more sense than selecting hardware solely around an AI use case.
The HPE ProLiant DL380 Gen12 is an example of a platform positioned for a broad range of enterprise workloads, including virtualization, databases, analytics and hybrid cloud infrastructure.
HPE ProLiant DL380 Gen12 vs DL360 Gen12
For organizations considering HPE Gen12 servers, the physical form factor and workload requirements can help narrow the decision.
|
Consideration |
DL380 Gen12 |
DL360 Gen12 |
|---|---|---|
|
Rack size |
2U |
1U |
|
Best suited to |
Flexible enterprise deployments |
Space-conscious deployments |
|
Rack density |
Lower |
Higher |
|
Expansion flexibility |
Strong |
More compact |
|
Typical use cases |
Virtualization, databases, analytics, hybrid workloads |
Enterprise compute, virtualization, databases |
|
AI consideration |
Useful where broader compute and expansion are required |
Useful where rack density matters |
Chicago Computer Supply currently lists both DL380 Gen12 and DL360 Gen12 configurations, so businesses can evaluate the form factor alongside their workload requirements.
What About a Tower Server?
Not every business needs a rack-mounted data center server.
For organizations with smaller infrastructure environments, a tower form factor can be worth considering.
Chicago Computer Supply's HPE ProLiant Gen12 collection also includes the HPE ProLiant ML350 Gen12, providing another form factor for organizations that don't have the same rack-density requirements as a large data center.
This is another reason not to start the buying process with:
"Which server is the most powerful?"
Start with:
"What environment are we deploying it into?"
Does Every AI Workload Need a GPU?
No.
This is an important distinction.
Some AI-related workloads can rely primarily on CPUs and memory, particularly where the application is focused on:
- Data processing
- Analytics
- AI-assisted business applications
- Certain inference workloads
- Orchestration
- Search
- Traditional enterprise workloads with AI components
Other applications may require specialized acceleration.
The key is to determine the workload's actual requirements before choosing the hardware.
If the application requires significant accelerator capacity, the server configuration needs to be evaluated specifically for that requirement.
Don't add expensive hardware simply because the project contains the word "AI."
What Should Businesses Consider Before Buying an HPE Server for AI?
Here's a practical checklist.
Workload
- What AI application will run on the server?
- Is it inference, analytics, automation or model processing?
- Will traditional business applications run alongside it?
CPU
- How many cores are required?
- Is high clock speed important?
- Will multiple workloads share the server?
Memory
- How much RAM is needed today?
- How much might be needed later?
- Does the platform provide enough expansion?
Storage
- How much capacity is required?
- How much performance is required?
- Will the workload generate large datasets?
Networking
- What network speed is required?
- Will the server need additional networking hardware?
- Will data movement become a bottleneck?
Expansion
- Will you need additional storage?
- More memory?
- Additional networking?
- Other PCIe devices?
Power and cooling
- Can the rack support the server?
- Is redundant power required?
- Is cooling sufficient?
Management
- Can IT remotely monitor and manage the server?
- How will firmware and security updates be handled?
Future growth
- Will the workload increase?
- Can the server be expanded?
- Will the organization need additional servers later?
HPE ProLiant Servers for AI: What Should You Actually Buy?
There isn't one answer.
The appropriate choice depends on the workload.
Choose a DL380 Gen12 when:
- You need a flexible enterprise platform.
- Virtualization is important.
- You expect substantial storage requirements.
- You need expansion flexibility.
- AI is one of several workloads.
- You want a 2U enterprise platform.
Chicago Computer Supply currently lists several DL380 Gen12 configurations, including the P89243-005, P90004-005 and P90005-005.
Consider a DL360 Gen12 when:
- Rack space is limited.
- 1U density is important.
- You need enterprise compute in a compact form factor.
- The workload doesn't require the same physical expansion space as a larger chassis.
- Chicago Computer Supply currently lists multiple DL360 Gen12 configurations.
Chicago Computer Supply currently lists multiple DL360 Gen12 configurations.
Consider an ML350 Gen12 when:
- A tower form factor is more practical.
- You don't have a traditional rack environment.
- You need enterprise server capabilities in a different physical deployment model.
The ML350 Gen12 is also present in Chicago Computer Supply's HPE ProLiant Gen12 collection.
Where Chicago Computer Supply Fits Into the Buying Decision
For an organization buying enterprise hardware online, choosing a server is only one part of the process.
The more important question is whether the specific configuration matches the workload.
Chicago Computer Supply's HPE ProLiant inventory includes multiple Gen12 configurations, allowing buyers to compare different processors, memory capacities, storage configurations and form factors rather than treating every ProLiant server as identical.
For example:
- HPE ProLiant DL380 Gen12 P89243-005 — 12-core Xeon configuration with 64 GB DDR5 and two 480 GB SSDs.
- HPE ProLiant DL380 Gen12 P90004-005 — 16-core Xeon configuration with 128 GB DDR5 and two 960 GB SSDs.
- HPE ProLiant DL380 Gen12 P90005-005 — 32-core Xeon configuration with 64 GB DDR5 and an NVMe boot device.
- HPE ProLiant DL360 Gen12 P89244-005 — 1U configuration with a 16-core Xeon processor and 64 GB DDR5.
This is why configuration-level comparison matters.
Two servers can both carry the name "HPE ProLiant Gen12" while offering very different processor, memory, storage and expansion configurations.
Existing Chicago Computer Supply Resources Worth Reading
If you're planning an HPE infrastructure refresh, Chicago Computer Supply already has several related resources that can be used to build a broader research path:
- HPE ProLiant Gen12 Servers: Upgrade Guide — useful for understanding the broader Gen12 upgrade decision.
- HPE ProLiant Gen10 vs Gen11 Comparison Guide — useful when comparing the economics and capabilities of older and newer ProLiant generations.
- Preparing for the AI Workload Surge — provides broader infrastructure planning context around the growth of AI workloads.
- AI Infrastructure: Switching, Storage and Traditional Workloads — useful for understanding why networking and storage can become important alongside compute.
Final Thoughts
Choosing an HPE ProLiant server for AI workloads in 2026 isn't about buying the biggest server you can afford.
It's about matching the infrastructure to the workload.
An organization running AI alongside virtualization and databases may need a flexible enterprise platform such as an HPE ProLiant DL380 Gen12.
A business where rack density is the priority may prefer a DL360 Gen12.
Another organization may find that a tower platform such as the ML350 Gen12 better fits its physical environment. Chicago Computer Supply currently lists all three types within its HPE ProLiant Gen12 collection.
The important questions are:
How much compute do we need?
How much memory and storage will the workload consume?
Do we actually need acceleration?
How much room do we need for expansion?
Can our facility support the power and cooling requirements?
And perhaps most importantly:
Will this server still make sense as our AI workloads grow?
For businesses purchasing HPE enterprise hardware online, Chicago Computer Supply provides multiple ProLiant configurations to evaluate. The best purchase isn't necessarily the newest or most expensive model — it's the configuration that provides the right balance of performance, scalability, manageability and cost for the workload.
Frequently Asked Questions
Are HPE ProLiant servers suitable for AI workloads?
Yes. HPE ProLiant servers can support a range of workloads that incorporate AI, analytics, data processing and other compute-intensive applications. The appropriate configuration depends on the workload's CPU, memory, storage, networking and acceleration requirements.
Which HPE ProLiant server is best for AI?
There is no single best model for every AI workload. A flexible platform such as the DL380 Gen12 can be appropriate when AI runs alongside virtualization, databases and other enterprise workloads. A DL360 Gen12 may be preferable where rack density is more important. The right choice depends on the workload and deployment environment.
Does every AI workload require a GPU server?
No. Some AI-related workloads can run primarily on CPUs and system memory. GPU acceleration becomes more important for certain demanding AI workloads, but businesses should determine the actual application requirements before purchasing accelerator-heavy infrastructure.
How much RAM should an HPE server have for AI workloads?
There is no universal amount. Memory requirements depend on the model, dataset, application, virtualization environment and number of simultaneous workloads. It is also important to consider the server's maximum memory capacity so that the system can be expanded later.
Is the HPE ProLiant DL380 Gen12 good for AI?
The DL380 Gen12 is a flexible enterprise platform that can make sense when AI-related workloads run alongside applications such as virtualization, databases and analytics. Chicago Computer Supply currently lists multiple DL380 Gen12 configurations with different processor, memory and storage specifications.
What is the difference between the HPE DL380 Gen12 and DL360 Gen12?
The main physical difference is the form factor: the DL380 Gen12 is a 2U rack server, while the DL360 Gen12 is a 1U platform. That makes the DL360 attractive where rack density is important, while the DL380 can provide more physical space for certain expansion and storage configurations. Chicago Computer Supply currently lists both models.
Can an HPE ProLiant server be upgraded later?
Depending on the model and configuration, ProLiant platforms can provide expansion options for memory, storage, networking and other components. Buyers should verify the exact expansion capabilities of the SKU they are purchasing rather than assuming every configuration is identical.
What should I consider besides CPU performance?
Memory, storage, networking, PCIe expansion, power, cooling, security, remote management and future scalability all matter. For AI-related workloads, looking only at the CPU can lead to an unbalanced infrastructure design.
Is the HPE ProLiant Gen12 worth considering for a 2026 infrastructure refresh?
It can be, particularly for organizations looking for current-generation enterprise infrastructure and room for future expansion. However, the right decision depends on the workload, existing infrastructure, budget and expected growth.
Where can I buy HPE ProLiant servers online?
Chicago Computer Supply offers HPE ProLiant servers online, including multiple Gen12 configurations such as DL380, DL360 and ML350 systems.