V-Ray Benchmark Guide 2026: CPU, GPU, and RTX Scores Explained

V-Ray Benchmark is one of the most widely used tools for comparing rendering hardware. Whether you're evaluating a new GPU, comparing CPUs for your workstation, or estimating performance on our cloud farm, understanding how benchmark scores translate to real-world render times is essential. This guide walks through the three main benchmark tests, how to interpret the results, and what they mean for cloud rendering efficiency.

What Is V-Ray Benchmark?

V-Ray Benchmark is a free tool from Chaos Group that measures the rendering speed of CPUs and GPUs running V-Ray. The benchmark downloads a standard test scene and renders it under identical conditions across different hardware configurations. The final score tells you how many samples per second (or rays per second for GPU) a given machine can produce.

We run V-Ray Benchmark regularly on our farm hardware to establish baseline performance metrics. When clients ask us "how fast will my job render on your farm?", benchmark scores are one of the first data points we check.

The Three V-Ray Benchmark Tests

V-Ray Benchmark 6.x includes three separate tests, each measuring a different rendering path. Understanding what each test does helps you pick the right benchmark for your workflow.

CPU Benchmark (V-Ray CPU with Unbiased Path Tracing)

The CPU benchmark runs V-Ray's path tracing algorithm on multi-core processors. The test scene renders at a fixed resolution and quality level, and the tool measures how many samples per second (vsamples) the CPU can complete.

What the score means: A higher vsamples score indicates faster CPU rendering. If your machine achieves 1,000 vsamples and another achieves 2,000 vsamples, the second machine should render the same frame roughly twice as fast (assuming identical settings).

Example: A dual-socket workstation with high core count will score significantly higher than a consumer laptop on this test. On our farm, machines with dual Xeon processors show consistent vsamples scores in a predictable range, which helps us estimate per-frame render times for incoming jobs.

Real-world correlation: CPU benchmark scores correlate well with actual V-Ray CPU render times, but the relationship is not perfectly linear. Scene complexity, shader count, and geometry detail all affect the final result. A scene that achieves 1,000 vsamples in benchmark might render slower or faster depending on material complexity.

GPU CUDA Benchmark (V-Ray GPU with CUDA)

The GPU CUDA benchmark measures NVIDIA GPUs using the CUDA rendering engine. It runs the same scene but on GPU hardware instead of CPU cores. The score is measured in vpaths per second (CUDA path tracing throughput).

What the score means: This test is useful if you run V-Ray with GPU acceleration enabled on workstations or if you're comparing NVIDIA GPUs for render farm deployment. Higher vpaths scores mean faster GPU rendering.

Example: An NVIDIA RTX 5090 will score much higher on the GPU CUDA test than older GPUs like the RTX 3090. The performance difference reflects the extra memory and compute cores in newer generation cards.

When to use this benchmark: Run the GPU CUDA test if your workflow uses V-Ray GPU rendering on consumer or workstation GPUs. For cloud rendering purposes, this test is less common—most render farms optimize for production-grade hardware, and the GPU RTX benchmark (next section) is more relevant.

GPU RTX Benchmark (V-Ray GPU with RTX Ray Tracing Cores)

The GPU RTX benchmark uses the dedicated ray tracing hardware in modern NVIDIA GPUs. Instead of path tracing via CUDA, it uses RTX ray tracing cores for faster, specialized ray intersection. The score is measured in vrays per second.

What the score means: This is the newest and most optimized benchmark for modern GPUs. Higher vrays scores indicate faster GPU rendering via RTX acceleration. This benchmark shows what happens when you use dedicated ray tracing hardware instead of general compute.

Example: On our farm, RTX 5090 GPUs run this benchmark with score values that reflect the specialized hardware. RTX cards are significantly faster at ray tracing than older compute-based methods, which is why this benchmark became the standard.

Real-world impact: RTX ray tracing often reduces render time per frame by 30–60% compared to CUDA path tracing, depending on the scene. Complex lighting and volumetric effects see the biggest gains.

Benchmark Versions and Compatibility

V-Ray Benchmark 6.x is the current standard and is compatible with V-Ray 6.0 and later. Older benchmarks (V-Ray Benchmark 5.x) use slightly different test scenes and measurement units, so scores from different versions are not directly comparable.

Always check which benchmark version you're running, especially if you're comparing scores from multiple sources or older hardware reviews. The changelog for V-Ray Benchmark is available on the Chaos Group website.

How to Interpret Benchmark Scores

Benchmark scores are useful for relative comparison, but they have limits. Here's what they do and don't tell you:

What benchmark scores are good for:

• Comparing two CPUs or GPUs in the same product line
• Estimating if an upgrade will improve render speed
• Establishing baseline performance on new farm hardware
• Ranking hardware within a budget category

What benchmark scores don't account for:

• Scene complexity (materials, lighting, volumetrics)
• Memory usage and VRAM requirements
• Denoising efficiency (post-render optimization)
• Plugin overhead (Forest Pack, Tyflow, Multiscatter, etc.)
• Network bottlenecks on cloud render farms

A machine that scores high on benchmark might render slower on your specific job if that job has heavy volumetric effects or uses memory-intensive plugins. Conversely, a lower-scoring machine with more VRAM might finish faster on a VRAM-limited scene.

Why Cloud Render Farms Use Benchmark Data

On our farm, we use V-Ray Benchmark results to:

1. Establish hardware baselines. When we deploy new machines, benchmark scores help us categorize them and set realistic per-frame time estimates for clients.

2. Estimate job completion time. If a client has a sample render that took 10 minutes on their 2,000-vsamples machine, and our farm hardware averages 8,000-vsamples, we can estimate the job will complete in roughly 2.5 minutes per frame (accounting for overhead).

3. Optimize hardware allocation. CPU-heavy jobs get routed to our CPU machines; GPU jobs get RTX-optimized nodes. Benchmark data helps us match job requirements to the right hardware subset.

4. Communicate performance expectations. When clients ask "will this render faster on your farm?", we reference our hardware and benchmark comparisons. We maintain a range of benchmarks across our CPU and GPU fleet so clients understand what they're getting.

Our Farm Hardware in Benchmark Context

We operate over 20,000 CPU cores across our farm, primarily dual-socket Xeon processors optimized for sustained multi-frame rendering. Our GPU fleet includes NVIDIA RTX 5090 machines with 32 GB VRAM, designed for high-performance GPU rendering.

In benchmark terms, our CPU machines produce consistent vsamples scores in a predictable range. This consistency is one advantage of running a large farm—the hardware is standardized, so estimates are more reliable than comparing random workstations.

Our GPU fleet, with RTX 5090 hardware, ranks at the top of the RTX benchmark chart. This translates to significant per-frame speed for V-Ray GPU and RTX-optimized rendering.

Benchmark Scores for Common Hardware

Below is a reference table showing approximate V-Ray Benchmark scores for commonly used hardware. These are representative ranges; actual scores depend on driver version, BIOS settings, and bench environment.

CPU Hardware Comparison

Hardware	V-Ray CPU vsamples (approx)	Year	Notes
Dual Intel Xeon E5-2699 V4	3,500–4,200	2016	Server-grade dual-socket, good for render farms
Intel i7-13700K	1,200–1,500	2023	Consumer workstation, single-socket
AMD Ryzen 9 7950X	1,800–2,100	2023	Consumer enthusiast, good value
Single Intel Xeon Platinum 8490H	2,200–2,500	2024	High-core-count single socket

GPU Hardware Comparison

Hardware	V-Ray GPU RTX vrays (approx)	VRAM	Year	Notes
NVIDIA RTX 5090	20,000–24,000	32 GB	2025	Our farm standard; top-scoring RTX GPU
NVIDIA RTX 4090	12,000–14,000	24 GB	2022	Previous-gen flagship; still very fast
NVIDIA RTX 6000 Ada	14,000–16,000	48 GB	2024	Enterprise GPU with more VRAM
NVIDIA L40S	10,000–11,000	48 GB	2023	Data center GPU, balanced compute+memory

From Benchmark Scores to Render Times

Here's a practical example of how to use benchmark data to estimate render time on a cloud farm.

Scenario: You have a 3ds Max + V-Ray scene that renders one frame in 45 minutes on your workstation. Your workstation hardware scores approximately 1,500 vsamples on V-Ray CPU Benchmark. You want to know how long it will render on Super Renders Farm.

Step 1: Calculate your machine's efficiency.

• Frame time: 45 minutes
• Hardware vsamples: 1,500
• Efficiency: 45 / 1,500 = 0.03 minutes per vsample

Step 2: Check our farm baseline.

• Our typical CPU machine: 3,500 vsamples
• Time adjustment: 3,500 / 1,500 = 2.33× faster

Step 3: Estimate farm render time.

• Farm time estimate: 45 minutes / 2.33 = ~19 minutes per frame

This is a rough estimate. Actual time depends on scene overhead, plugin complexity, and job queue wait time. But benchmark-based estimates give you a useful ballpark for render cost calculation.

Benchmark and Cloud Rendering Cost

If render time is predictable via benchmark, then cost becomes calculable. Here's how:

1. Find our pricing. Cloud rendering cost is typically per-core-minute or per-machine-minute. Check our render farm pricing guide for current rates.

2. Estimate total compute minutes. Take your estimated per-frame time (from Step 3 above) and multiply by frame count. If your job is 300 frames at 19 minutes per frame, that's 5,700 machine-minutes.

3. Calculate cost. Multiply machine-minutes by the per-minute rate. Costs vary by machine type (CPU vs GPU), so check the pricing page for your specific use case.

Benchmark-based cost estimates are approximate, but they let you budget render jobs before submission. More accurate estimates come from test frames submitted to the farm.

How to Run V-Ray Benchmark Yourself

1. Download the tool from the Chaos V-Ray Benchmark page.

2. Install V-Ray Benchmark on your target machine (Windows, Mac, or Linux).

3. Run the benchmark. The tool will render the test scene and display vsamples/vrays scores on completion.

4. Note the environment. Record driver version, OS, and any special BIOS settings (like turbo mode or power plan). Benchmark results vary with environment.

5. Compare your score to others using online databases (e.g., V-Ray Benchmark result sharing on Chaos forums).

Advanced: Interpreting Score Variation

If you run the benchmark twice on the same machine, scores may vary by 3–7% due to:

• CPU throttling. Thermal limits or power management may throttle cores mid-benchmark.
• Background processes. Other software consuming CPU cycles reduces available compute.
• BIOS variability. Turbo clock settings, power states, and memory speed affect throughput.
• Driver updates. GPU driver versions can shift scores by 5–10%.

If you're comparing two machines, run the benchmark at least twice on each to ensure variability is accounted for. Look for the average score, not single runs.

FAQ

Q: What's the difference between benchmark scores and real-world render time? A: Benchmark scores measure throughput under ideal conditions with a standard test scene. Real-world jobs involve plugin overhead, varied materials, and often more complex geometry. Benchmark gives you a baseline; actual time requires more factors. Use benchmark for relative hardware ranking, not absolute time prediction.

Q: Can I compare CPU and GPU benchmark scores directly? A: No. CPU vsamples and GPU vrays use different units and measure different rendering paths. A GPU might score 20,000 on RTX benchmark but a CPU might score 4,000 on CPU benchmark—the GPU is not necessarily "5x faster" because they're measuring different things. Compare within hardware type: CPU to CPU, or GPU to GPU.

Q: Why does my machine's benchmark score seem lower than expected? A: Check driver version (GPU) or BIOS settings (CPU). Older drivers or power-saving modes reduce scores. Ensure background processes aren't consuming resources. If your hardware is new and score is still low, check the Chaos Group forums—there may be known driver issues.

Q: How often does V-Ray Benchmark update? A: Chaos Group releases new benchmark versions when V-Ray gets major releases (e.g., V-Ray 6.0, 7.0). Benchmark 6.x is current for V-Ray 6.x users. Staying on the latest benchmark and V-Ray version ensures you get the most accurate hardware comparisons.

Q: Does benchmark help predict render farm speed? A: Yes, but with caveats. A render farm with faster hardware (higher benchmark score) will finish your job faster, roughly proportional to the hardware speed difference. However, farm overhead (queue time, job setup) means the improvement isn't perfectly linear. For best accuracy, submit a test frame to the farm and measure actual time.

Q: Is benchmark the only metric I should consider when choosing cloud rendering? A: No. Consider uptime, support quality, API integration, and ease of use. A slightly slower farm with better uptime and faster support might be worth more than raw speed. Benchmark is one part of the decision; it's not the whole picture.

Q: What about denoising? Do benchmarks account for it? A: V-Ray Benchmark measures raw render output without denoising. In production, you often use denoising to reduce sample count and lower render time. A machine with lower benchmark scores might finish faster if you apply aggressive denoising. Benchmark measures pure rendering speed; denoising is a separate optimization step.

Q: Can I use benchmark scores to estimate GPU VRAM requirements? A: Not directly. Benchmark scores tell you rendering speed, not memory usage. A high-scoring GPU might still run out of VRAM on complex scenes. For VRAM estimates, benchmark your actual job on the target GPU with the same scene and material setup. Then benchmark becomes more predictive.

Q: Why is our farm focus on CPU rendering if GPU is faster? A: GPU rendering is fast, but CPU rendering is more flexible and cheaper at scale. Most of our workload is CPU-based (70% of jobs) because V-Ray CPU is reliable, plugin-compatible, and works well for archviz (our largest vertical). GPU is growing but doesn't dominate our mix yet. We support both equally—the fleet reflects real demand.

Q: How does benchmark relate to render farm pricing? A: Render farm pricing is often per-core-minute or per-node-hour. Faster hardware (higher benchmark score) might cost more per unit time, but finishes your job faster overall, potentially lowering total cost. Benchmark helps you estimate the trade-off: slower machine, lower per-minute cost but longer total time; faster machine, higher per-minute cost but shorter total time. Check our pricing page for specific rate comparison.

V-Ray Benchmark is a powerful tool for understanding hardware performance. Use it to make informed decisions about workstation upgrades, cloud rendering budget, and job scheduling. Combined with real-world test submissions, benchmark scores give you the data needed to optimize rendering efficiency across your pipeline. To see how these benchmarks apply in practice on cloud infrastructure, explore our V-Ray cloud render farm resources.