Customer awareness about air cooling compared to liquid cooling is an area that needs attention. Liquid cooling requires existing setups, and since AI workloads are significantly different from traditional workloads, a single rack can now have more than 600 kW, which is much higher than traditional workloads. Current data centers are not ready to support heavy AI HPC clusters. If you have requirements from that perspective, you need to understand that you must build new data centers that are liquid cooling ready. This ensures you can have high-density GPUs in a single rack to save space rather than having multiple semi-empty racks with only one or two air-cooled servers.
Lead Solution Architect at Avineon India Private Limited
Real User
Top 10
Mar 11, 2026
What I found in Dell PowerEdge XE-Series servers is the disk reading; sometimes with heavy imagery, we find it to be slow. Some of our workloads utilize heavy imagery, which could be several terabytes or gigabytes in size, so when accessing imagery, we find disk read to be a bit slow. I need to check with my team, but I recently noticed some compatibility issues with disk upgrades and RAM upgrades in Dell PowerEdge XE-Series, and limitations were there. In the upcoming series, I hope to see flexibility in adding GPUs and similar features.
For Dell PowerEdge XE Servers, particularly for standalone servers, extendability would be a great improvement. This feature is currently available only with virtualization environment. Adding extendability capabilities to standalone servers would be a significant enhancement to make the solution closer to perfect.
It should be easier to extend PowerEdge XE servers. We haven't architected them with an eye toward scalability. They have a fixed number of slots for hard drives, which is enough for our present use case, However, we would have more options if we could extend that. We could deploy the servers in a broader range of use cases with different requirements. It's fine for our current deployment with only a few thousand users. We're exploring other solutions like such as VxRail or PowerFlex for larger public-facing use cases.
Senior Field Application Engineer at a tech vendor with 1,001-5,000 employees
Real User
Jan 12, 2021
In future releases, there should be more support options for the latest processor. The company might already be doing this but this is what I would like to see.
The PowerEdge XE-Series caters to AI-centric workloads by offering exceptional GPU support and processing power for demanding tasks like inferencing and large-scale training, making it ideal for organizations aiming to innovate with AI.PowerEdge XE-Series stands out in modern GPU-centric computing with its purpose-built design for AI acceleration. Known for handling extreme workloads like model fine-tuning and AI-driven applications, it supports high-density GPU configurations and advanced...
Customer awareness about air cooling compared to liquid cooling is an area that needs attention. Liquid cooling requires existing setups, and since AI workloads are significantly different from traditional workloads, a single rack can now have more than 600 kW, which is much higher than traditional workloads. Current data centers are not ready to support heavy AI HPC clusters. If you have requirements from that perspective, you need to understand that you must build new data centers that are liquid cooling ready. This ensures you can have high-density GPUs in a single rack to save space rather than having multiple semi-empty racks with only one or two air-cooled servers.
What I found in Dell PowerEdge XE-Series servers is the disk reading; sometimes with heavy imagery, we find it to be slow. Some of our workloads utilize heavy imagery, which could be several terabytes or gigabytes in size, so when accessing imagery, we find disk read to be a bit slow. I need to check with my team, but I recently noticed some compatibility issues with disk upgrades and RAM upgrades in Dell PowerEdge XE-Series, and limitations were there. In the upcoming series, I hope to see flexibility in adding GPUs and similar features.
For Dell PowerEdge XE Servers, particularly for standalone servers, extendability would be a great improvement. This feature is currently available only with virtualization environment. Adding extendability capabilities to standalone servers would be a significant enhancement to make the solution closer to perfect.
The support team must improve the speed of response. The product should be promoted more in the market.
There is room for improvement in terms of stability, particularly in addressing disk failures or connectivity issues.
It should be easier to extend PowerEdge XE servers. We haven't architected them with an eye toward scalability. They have a fixed number of slots for hard drives, which is enough for our present use case, However, we would have more options if we could extend that. We could deploy the servers in a broader range of use cases with different requirements. It's fine for our current deployment with only a few thousand users. We're exploring other solutions like such as VxRail or PowerFlex for larger public-facing use cases.
The pricing could be a bit better. Upgrading can be a bit difficult.
Dell PowerEdge XE Servers can improve by adapting HCI technology. In a feature release, it would be beneficial to have better management monitoring.
The quality and reliability could be improved.
I would like to have support for performing a detailed analysis of the server from a mobile device.
In future releases, there should be more support options for the latest processor. The company might already be doing this but this is what I would like to see.