3DMark Comparison: Time Spy vs Port Royal — Which to Run?

3DMark Comparison: Time Spy vs Port Royal — Which to Run?3DMark is the industry standard for synthetic GPU benchmarking, and within its suite two tests are among the most widely used: Time Spy and Port Royal. Both measure graphics performance, but they target different APIs, workloads, and hardware features. This article explains what each test measures, how they differ, when to run one or the other, how to interpret results, and practical tips for producing reliable comparisons.

What each test is and what it measures

• Time Spy

  • Designed for DirectX 12 (DX12).
  • Uses traditional rasterization with modern GPU features such as asynchronous compute and explicit multi-threading.
  • Measures raw rasterization performance, command submission efficiency, multi-core CPU interaction with the GPU, and DX12 feature utilization.
  • Produces an overall score plus separate Graphics and CPU scores.

• Port Royal

  • Designed specifically to test real-time ray tracing performance using DirectX Raytracing (DXR) on supported hardware.
  • Uses hardware-accelerated ray tracing where available (e.g., NVIDIA RTX, AMD RDNA ⁄₃ ray accelerators).
  • Measures RT core (hardware) throughput, ray traversal and shading costs, denoising effectiveness in the render pipeline, and hybrid rendering impacts.
  • Produces an overall score focused on ray tracing workloads.

Key technical differences

• API and feature set: Time Spy uses full DX12 rasterization features; Port Royal requires DXR support and focuses on ray-tracing pipelines.
• Workload type: Time Spy stresses shader throughput, rasterization, and CPU/GPU synchronization; Port Royal stresses BVH traversal, ray-triangle intersection, RT core utilization, and ray shading.
• Hardware bottlenecks: Time Spy often exposes raw shader ALU/texture throughput and memory bandwidth limits; Port Royal can be dominated by ray tracing hardware and RT-specific memory/cache behavior.
• Score comparability: Scores between Time Spy and Port Royal are not directly comparable — they measure different workloads and should be used to evaluate different aspects of a system.

When to run each test

• Run Time Spy when:

  • You want a broad measure of DX12 rasterized gaming performance.
  • Comparing GPU shader throughput, memory bandwidth, and CPU impact on frame submission.
  • Testing systems without ray-tracing hardware or when ray tracing performance is not a priority.

• Run Port Royal when:

  • You want to evaluate real-time ray tracing performance specifically.
  • Comparing systems or GPUs that advertise ray-tracing hardware (e.g., NVIDIA RTX series, AMD RDNA 2+).
  • Measuring the impact of RT features, denoising, and hybrid rendering strategies.

Practical scenarios and recommendations

• Gamers with RTX/RDNA2+ GPUs who play titles with ray tracing (Cyberpunk 2077, Control, Metro Exodus Enhanced) should run Port Royal to understand real-world RT performance and whether enabling RT will be playable at their target resolution and settings.
• Competitive players or users focused on rasterized performance (most esports/titles without RT) should prioritize Time Spy as it better reflects conventional gaming workloads.
• Benchmarkers comparing architecture generational gains should run both tests: use Time Spy to compare raster improvements and Port Royal to measure gains in RT hardware and driver-level RT optimizations.
• Laptop testing: run both but pay attention to thermal throttling. Port Royal may push sustained power delivery differently because of RT hardware utilization; compare scores alongside thermals and power draw.

Interpreting results and variables to control

Control these variables for reliable, repeatable results:

• Driver version — use the same GPU driver for all comparisons.
• Power/thermal settings — use consistent power profiles and cooling; laptop OEM profiles can drastically alter scores.
• Background processes — disable unnecessary software that can affect CPU submission or GPU drivers.
• Resolution and preset — Time Spy and Port Royal each have standard presets (e.g., Time Spy, Time Spy Extreme) — use the same preset across GPUs for fair comparison.
• GPU clocks and boost behavior — note if overclocking or factory boosts are enabled; document them.

What scores tell you:

• Time Spy Graphics score -> general raster throughput and DX12 efficiency.
• Time Spy CPU score -> how well the system feeds the GPU under DX12 workloads.
• Port Royal score -> relative real-time ray tracing capability; pay attention to RTX/DLSS (or FSR + RT workflows) in real games for practical conclusions.

Example comparison table

Common pitfalls and sources of confusion

• Assuming a higher Time Spy score means better ray tracing — false. A GPU can excel at rasterized workloads but lag in RT if it lacks RT hardware or has less optimized RT pipelines.
• Expecting Port Royal to reflect final in-game RT performance exactly — Port Royal is synthetic and focuses on RT throughput; real games include other systems (physics, AI, world complexity) that affect frame rates.
• Ignoring driver updates — RT performance has shifted significantly with driver and API improvements; re-run Port Royal after major driver updates.

How to use both tests to make decisions

• If buying a GPU for raster performance: prioritize Time Spy scores at your target resolution and compare price-per-score.
• If buying for RT-capable games: prioritize Port Royal and pair it with real-game RT benchmarks (with DLSS/FSR toggled) to see effective playable frame rates.
• For content creators or developers: use Time Spy to ensure baseline raster performance and Port Royal to guide RT optimization, BVH design, and denoising choices.

Tips for repeatable benchmarks

• Run each test 3–5 times and use the median score.
• Record ambient temperature and power limits for laptops/desktops.
• Use the same Windows power plan and GPU power target.
• If testing overclocked hardware, also test stock settings to gauge stability and real-world gains.

Conclusion

Use Time Spy for measuring traditional DirectX 12 rasterized performance and system feeding efficiency. Use Port Royal when your goal is to evaluate real-time ray tracing capability on hardware that supports DXR. They complement each other: Time Spy shows how a GPU handles conventional gaming workloads; Port Royal reveals how it handles modern ray-traced rendering. Choose the test that matches the workload you care about — or run both for a complete view.