How Windows VBS and HVCI Affect Hardware Benchmarks

This article provides a comprehensive overview of how Windows Virtualization-Based Security (VBS) and Hypervisor-Protected Code Integrity (HVCI) negatively impact computer hardware benchmarking. While these features are designed to protect systems from malware by isolating memory, they introduce a virtualization layer that taxes the CPU and memory. Consequently, enabling VBS and HVCI leads to lower synthetic benchmark scores, reduced gaming frame rates, and blocked access to hardware monitoring tools, forcing testers to disable them to measure true hardware performance.

Understanding VBS and HVCI

Virtualization-Based Security (VBS) is a Windows feature that uses hardware virtualization to create an isolated, secure region of system memory. The normal operating system cannot access this "virtual secure mode," which protects critical system processes from malicious attacks.

Hypervisor-Protected Code Integrity (HVCI), commonly listed as "Memory Integrity" in the Windows Security settings, is a specific security feature that runs inside the VBS environment. HVCI checks all kernel-mode drivers and system binaries in that isolated memory space to ensure they are securely signed and unmodified before they are allowed to execute. Both features are enabled by default on clean installations of Windows 11.

The Root Cause of Performance Loss

Benchmarking software is designed to push hardware to its absolute limits by executing instructions as fast as the silicon allows. VBS and HVCI disrupt this direct access.

When these features are active, the Windows hypervisor sits between the hardware and the operating system. Every time an application or game executes code, allocates memory, or calls on a driver, the hypervisor intercepts the action to verify its security. This constant checking creates CPU overhead and increases memory latency, diverting system resources away from the benchmark task.

Impact on Different Benchmark Categories

The performance penalty of VBS and HVCI varies depending on the type of hardware being tested, typically resulting in a 5% to 15% performance drop across the board.

CPU Benchmarks

Synthetic CPU benchmarks, such as Cinebench, Geekbench, and 3DMark CPU Profile, rely on rapid instruction execution and memory access. With VBS enabled, multi-core and single-core scores drop noticeably. Tasks that require heavy virtualization or frequent kernel-level calls suffer the most, as the CPU spends valuable clock cycles running security checks rather than rendering or calculating benchmark workloads.

Gaming and GPU Benchmarks

Gaming benchmarks are heavily impacted by HVCI. Modern games constantly stream assets, compile shaders, and make rapid calls to graphics and audio drivers. Because HVCI must verify these continuous operations, games experience a bottleneck at the CPU level. This results in lower average frame rates (often 5% to 10% lower) and significantly worse 1% low frame rates, which manifests as stuttering during the benchmark run.

Memory and Storage Benchmarks

Because VBS isolates memory using address translation, it directly harms RAM performance. Benchmarks like AIDA64 show increased memory latency and reduced read, write, and copy bandwidth when VBS is active. Storage benchmarks like CrystalDiskMark can also show slightly lower random read/write speeds due to the overhead added to the storage controller drivers.

Interference with Benchmarking Tools

Beyond raw performance loss, VBS and HVCI actively break certain benchmarking and overclocking utilities.

Many hardware monitoring tools (like HWiNFO, CPU-Z, or motherboard-specific overclocking software) use low-level kernel drivers to read CPU voltages, clock speeds, and temperatures directly from the hardware registers. HVCI often flags these legitimate drivers as security risks, blocking them from loading. As a result, benchmarkers may see missing voltage readings, inaccurate clock speed reporting, or an inability to apply overclocking profiles from within Windows.

The Problem of Inconsistent Baselines

For hardware reviewers and enthusiasts, VBS creates a major consistency problem. While clean installations of Windows 11 turn VBS on by default, systems that were upgraded from Windows 10 often leave it turned off.

If a tester compares a new CPU on a fresh Windows 11 install (VBS On) against an older CPU tested on an upgraded OS (VBS Off), the data is invalidated. The newer hardware will appear artificially slower due to the software-induced penalty.

Conclusion for Hardware Testers

To accurately measure the raw capability of a CPU, GPU, or RAM kit, VBS and HVCI must be disabled. Leaving them on measures how well the hardware handles virtualization overhead rather than its actual maximum performance. However, because these features provide robust protection against ransomware and rootkits, users should re-enable them after benchmarking is complete for safe daily use.