Windows 11 25H2 vs 26H2 preview: an early performance comparison

Table of Contents

• What is new in Windows 11 26H2?
• Why is 26H1 not included?
• Setup and scenarios
• Hypothesis and results
• Conclusion

Microsoft is preparing Windows 11 26H2 as its next annual feature update, and it is already available for testing through the Windows Insider Program. For every new OS release, the critical question for the EUC community remains the same: what is the performance impact for virtual desktop scenarios? Rolling a new release out to production without understanding its effect on resource consumption, user density, and application responsiveness is a gamble that could negatively affect both the user experience and the scalability of your environment.

This GO-EUC research takes an early look at the Windows 11 26H2 preview and compares it directly against the current 25H2 release, to judge ahead of general availability what the overall difference between the two versions is.

What is new in Windows 11 26H2?

Windows 11 26H2 is the second major annual update in Microsoft’s standard Windows 11 update cadence. Because 26H2 shares the same servicing branch as 24H2 and 25H2, the update is delivered as a small enablement package (eKB) that activates features already present but inactive on the device. This means devices running 24H2 or 25H2 can move to 26H2 with a single restart, while the build number changes to 26300. For image management and VDI lifecycle stability, this is a welcome approach, as it avoids a full OS replacement.

Microsoft does not provide a full downloadable ISO at this point in time and therefore, a clean ISO was created using the UUP Dump website. UUP Dump is a community-driven tool that lets you build official, customized Windows installation ISOs. It works by fetching raw update files directly from Microsoft’s Unified Update Platform (UUP) servers and compiling them into a complete, ready-to-use ISO image.

At the time of writing, 26H2 is available to Windows Insiders in the Experimental channel. Highlights of the preview cycle so far include the Low Latency Profile, which boosts CPU frequency during UI interactions to improve responsiveness, along with File Explorer refinements and various reliability fixes. More information can be found on the Windows IT Pro Blog.

It is important to emphasize that 26H2 is still a preview release. The results in this research reflect a point-in-time snapshot of the following builds:

• Windows 11 25H2 (OS Build 26200.8655)
• Windows 11 26H2 preview (OS Build 26300.8772)

As Microsoft continues to iterate on the preview, the final general availability release may behave differently.

Why is 26H1 not included?

Readers may notice that Windows 11 26H1 is not part of this comparison, despite being a more recent release than 25H2. This is a deliberate choice.

Windows 11 26H1 is a specialized release built specifically for new ARM-based devices, initially those powered by Qualcomm Snapdragon X2 Series processors. It ships preinstalled on these new devices only: it is not offered through Windows Update, cannot be installed as an in-place update on existing devices, and is not intended for broad deployment across the existing Windows 11 ecosystem. Feature-wise, it contains the same feature set as 25H2, adapted for the new silicon.

In addition, 26H1 is built on a completely different Windows core branch than 24H2, 25H2, and 26H2. As it does not share the same servicing branch, devices running 26H1 cannot move to 26H2 through the enablement package and will instead follow an independent upgrade path to a future Windows release. Microsoft explicitly advises organizations to continue deploying the broadly released versions, and more details can be found on the Windows IT Pro Blog.

For the vast majority of enterprise and VDI environments, which run on x86 hardware, the realistic upgrade path is from 24H2 or 25H2 directly to 26H2 via the enablement package. Comparing 25H2 against the 26H2 preview therefore reflects the decision most organizations will actually face, whereas 26H1 is simply not applicable to this scenario.

Setup and scenarios

In this research, Login Enterprise is used to generate a consistent user load, running the standard Knowledge Worker 2022 workload. This workload simulates a typical office user working with applications such as Microsoft Office and the Microsoft Edge browser, which is also used for viewing PDF documents, making it representative of the average knowledge worker profile.

The following two scenarios are part of this research:

• Windows 11 25H2
• Windows 11 26H2 preview

Both scenarios run as a single virtual machine on a KVM hypervisor with the guest drivers and guest agent installed. The hypervisor itself is equipped with a single-socket AMD EPYC 7282 16-core processor, 256 GB of memory, and NVMe-backed storage. Each virtual machine is configured with 8 vCPUs, 16 GB of memory, and a 64 GB disk.

It is important to note that this research does not measure single-server scalability (SSI) or host scalability. With only one virtual machine running per test, there is no resource contention from neighboring machines, and no conclusions can be drawn about how many virtual desktops a host can support on either Windows version. The goal of this comparison is to isolate the resource footprint of the operating system itself; translating these results into user density figures would require a dedicated multi-VM scalability test, which is a good candidate for follow-up research once 26H2 reaches general availability.

To ensure consistent and reliable results, Microsoft Defender is completely disabled so it does not influence the benchmark results. Please note that this is not considered a best practice in a production environment, but it is a common approach in benchmarking scenarios where the goal is to isolate and measure the impact of a specific component or configuration change. Beyond that, no additional third party services are running, and the only other change to the images is the installation of PowerToys, with Awake applied to keep the machine always on and awake during the test runs.

Apart from the Windows version, all hardware, software, benchmark settings, and virtual machine configurations remain identical throughout the research.

During each benchmark run, telemetry is collected from the virtual machine, including CPU utilization, memory usage, disk reads and writes per second, disk queue length, and the number of running processes.

Hypothesis and results

As 26H2 is delivered as an enablement package on the same servicing branch as 25H2, no major architectural or kernel-level changes are expected. Based on previous GO-EUC research comparing 23H2, 24H2, and 25H2, where the differences between releases were marginal, a similar outcome could reasonably be expected here: a comparable overall footprint, with perhaps a slight increase in resource usage due to newly enabled features.

Let’s start with the CPU utilization from the virtual machine perspective.

The results align closely with the hypothesis. The average CPU utilization on the 26H2 preview is approximately 8.4%, compared to 7.8% on 25H2, a difference of roughly half a percentage point in favor of the current release. Both versions follow the same CPU pattern throughout the workload, with peaks during the more demanding phases of the workload reaching 86.8% on 25H2 and 80.8% on the 26H2 preview. A difference this small falls within the normal run-to-run variation for this type of test, and there is no indication of a structural CPU regression or improvement in either direction.

The number of running processes shows a similar picture.

The 26H2 preview runs an average of 153 processes during the test, compared to 166 on 25H2, roughly 13 fewer on an identically configured image with the same workload. It is worth noting that preview builds do not always ship with the full set of features and services enabled, so this small reduction may not survive into the general availability release. This is something to re-validate once 26H2 reaches Release Preview or general availability.

The memory footprint follows the same trend.

The average memory usage comes in at 24.8% on the 26H2 preview versus 25.3% on 25H2. To put this into perspective, a 0.5% difference on a 16 GB virtual machine equates to less than 100 MB of memory, which is negligible in practice. For consolidation ratios and user density planning, both versions can be treated as having an identical memory footprint.

The next charts zoom in on the storage perspective.

Disk reads are nearly identical between both versions, which is expected since read activity is primarily driven by the workload, and the workload remains consistent across both tests. Disk writes show a slight increase on the 26H2 preview, averaging 136 writes per second compared to 118 on 25H2, with a higher peak of 2,364 writes per second versus 1,813 on 25H2. This could be related to additional logging or telemetry activity in the preview build, but the difference is small enough that it is unlikely to be noticeable on modern NVMe-backed storage.

The current disk queue length confirms this: with an average of 0.079 on both versions, the storage subsystem is never under pressure during the test, and there is effectively no difference between the two releases from a storage latency perspective.

Conclusion

Based on this early comparison, the Windows 11 26H2 preview shows a resource footprint that is effectively identical to 25H2 when running an identical knowledge worker workload. CPU utilization differs by roughly half a percentage point, memory usage is within half a percent, disk reads are unchanged, and disk writes show only a slight increase on the preview build. The 26H2 preview also runs slightly fewer processes on average, but the difference is small and well within what could change before general availability.

This is exactly the outcome the hypothesis predicted. As 26H2 is delivered as an enablement package on the same servicing branch as 24H2 and 25H2, no major architectural or kernel-level changes were expected, and none were observed. For EUC architects and VDI administrators, this is arguably good news: based on these early results, moving from 25H2 to 26H2 should be a low-risk transition from a performance perspective, with no capacity planning surprises to account for.

As always, a degree of caution is required. This research compares a shipping release against a preview build, and preview builds do not always represent the final product. Features that are dormant or absent in the Experimental channel today may still arrive at general availability, and the slight increase in disk write activity hints that telemetry and servicing behavior is still in flux. In addition, this research measures a single virtual machine in isolation, so conclusions about user density and host scalability would require a dedicated multi-VM test.

The key takeaway is that the 26H2 preview currently performs on par with 25H2, which is precisely what an enablement package release should do. This research will be worth repeating once 26H2 reaches Release Preview or general availability, to confirm that the final release maintains this neutral footprint.

Photo by Jason Dent on Unsplash

Question to our readers: With Windows 11 26H2 showing a performance footprint on par with 25H2, would you adopt it early to stay on the latest servicing timeline, or would you wait for the first cumulative updates as usual?