Here at Prowess Consulting, many of our benchmarking efforts compare vastly different devices: for example, Apple iPads compared to Microsoft Surface devices or devices with new processors against those with older processors. Those comparisons are pretty straightforward. Sometimes, though, as in the case of our newest healthcare benchmark for thin clients, the comparison is more subtle.

In our latest benchmark, “Select the Right Thin Client to Improve User Experience,” we “compare” a variety of thin clients—all with Intel processors—for use in a healthcare setting, although our findings are applicable to settings outside of healthcare as well. I say “compare” because our effort was less a comparison and more a proof of the concept that some thin clients are better suited for some needs and users. We examined which niche each device best inhabits.

The premise of our report was that all users in a healthcare setting have a baseline need to access a set of applications, such as the electronic-medical-record (EMR) platform, and basic tools, such as the web, while other users need greater functionality in addition to the baseline. And, because we frequently test Intel processors and see their value every time we do, we chose to do our testing using thin clients with Intel processors. We tested devices with processors that span the performance spectrum—from Intel Atom processors to Intel Celeron processors to Intel Core i3, i5, and i7 processors. Our goal was to determine the best processor for the best user experience without over- or under-equipping the user with a more or less powerful processor than the user needs.

Of course, the most powerful client we tested—the one with an Intel Core i7 processor—could work for any user, but why equip a nurse sitting at a nurse’s station and only accessing the EMR and some online apps with all that power if it will go to waste? Hence our testing. Which device, which processor, best matches which user?

Welcoming Some New Kids to the Thin-Client Block

We added a few interesting new flavors of thin clients to our testing. Thin clients are no longer relegated to little boxes. New to our mix were the Intel Compute Stick and the Intel NUC. Either can be used as full computer or as thin clients.

While the specific Intel NUC we tested was often the slowest of the devices with more powerful processors in our testing mix—which consisted of two Intel Core i5 processors and one Intel Core i7 processor—it almost always outperformed the entry-level processors—the Intel Atom processor, the Intel Celeron processor, and the Intel Core i3 processor.

Breaking It Down by User Need and the Realities of the User Environment

For our testing, we divided users into two groups:

  • Task workers with a baseline need: to access the EMR platform and basic tools
  • Knowledge workers who need baseline access and specialized access to tools, such as video conferencing, voice transcription, and a picture archiving and communication system (PACS)

Table 1 shows the devices we tested, our assumptions for which user group each belonged in, and the processor found in each device.

Table 1. Task-worker and knowledge-worker devices tested

User Device Processor
Task worker Intel Compute Stick Intel Atom x5-Z8300 processor
Dell Wyse 3030 LT Intel Celeron processor N2807
Lenovo ThinkCentre M700 Tiny Intel Core i3-6100T processor
Knowledge worker Dell Wyse 7040 Intel Core i5-6500TE processor
Intel NUC Intel Core i5-5300U processor with Intel vPro technology
Lenovo ThinkCentre M900p Tiny Intel Core i7-6700T processor


We tested the specialized tools, such as video conferencing, voice transcription, and a picture archiving and communication system (PACS), used by knowledge workers running locally on a thin client when possible, and running as virtualized applications. Why? Well, the reality is that, even in virtualized environments, not all applications have been virtualized. And because of the nature of these specialized apps, in some environments, they might never be virtualized. So we looked at both options.

And the Winner Is …

All of the devices we tested delivered adequate performance on most tests. As we assumed, though, the knowledge-worker devices outperformed the task-worker devices most of the time. And, for tasks targeted at knowledge workers, the devices with higher-power processors delivered the best performance—performance often significantly better than the devices with entry-level processors. For example, the Lenovo ThinkCentre M900p Tiny with its Intel Core i7-6700T processor was 63 percent faster than the Intel NUC with its Intel Core i5-5300U processor when using medical dictation software. And overall, all of the processors in the knowledge-worker devices were 21 to 155 percent faster than those in the task-worker devices, even for the baseline need to access the virtualized EMR.

Our conclusion is that organizations are best served by choosing thin clients and processors that match, but not exceed, the needs of the users who will work with them.

We’ve tested these devices, so you don’t have to; read our full report to dive deeper into our findings and specifics, which can help you apply what we found to your own environment and users’ needs.

Share this: