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 are evaluating a new GPU, comparing CPUs for your workstation, or planning a cloud V-Ray pipeline for 2026, 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 across V-Ray 6 and V-Ray 7, and what those numbers mean for cloud rendering efficiency. If you are also weighing DCC choice, our V-Ray in Blender vs 3ds Max comparison covers feature parity and ecosystem differences.

What Is V-Ray Benchmark?

V-Ray Benchmark is a free standalone tool from Chaos that measures the rendering speed of CPUs and GPUs running V-Ray. The benchmark ships its own reference scenes and renders them under identical conditions across different hardware configurations. The final score tells you how many samples per second (vsamples, for CPU) or paths per second (vpaths, for GPU) a given machine can produce, and uploads the result to a public leaderboard at benchmark.chaos.com.

At Super Renders Farm, 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 — especially for GPU V-Ray jobs, where hardware generation (RTX 3000 vs 4000 vs 5000 series) drives most of the spread.

The Three V-Ray Benchmark Tests

Current V-Ray Benchmark builds include 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 are 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 bandwidth, CUDA core count, and SM architecture improvements 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 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 has become the reference for GPU V-Ray work.

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: V-Ray 6 and V-Ray 7

V-Ray Benchmark tracks the V-Ray product line, so the build you run should match the V-Ray version in your pipeline. In 2026, two versions matter:

• V-Ray 6.2 (mature long-support line) — the V-Ray Benchmark 6.x build pairs with this. Most production studios on 3ds Max, Maya, SketchUp, and Revit are still on this line, and our CPU render nodes primarily run V-Ray 6 jobs for archviz and product visualisation. Scores here are directly comparable across any machine running the same Benchmark 6.x build.
• V-Ray 7 (current major release, for hosts that support it) — brings substantial GPU path changes: new RTX kernels, updated denoiser, improved AI upscaling, and better volumetric performance. V-Ray 7 Benchmark scores are typically higher than V-Ray 6 on the same hardware, especially on RTX 4000 and 5000 series cards, because the renderer makes fuller use of the tensor cores and newer RT cores. V-Ray 7 scores are not directly comparable to V-Ray 6 scores — treat them as separate leaderboards.

Older benchmarks (V-Ray Benchmark 5.x and earlier) use different test scenes and measurement units, so scores from those versions are not comparable to either current build. Always check which benchmark version produced a score before quoting it, especially when referencing third-party hardware reviews. The changelog for V-Ray Benchmark is published on the Chaos website, and the full feature breakdown for the latest release is covered in our guide to what's new in V-Ray 7 for 3ds Max.

How to Interpret Benchmark Scores

Benchmark scores are useful for relative comparison, but they have limits. Here is what they do and do not 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 do not 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 evaluate hardware end to end. For a broader look at render farms that support V-Ray, see our dedicated roundup. Specifically, we use benchmark data to:

1. Establish hardware baselines. When we deploy new machines, benchmark scores help us categorise 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. Optimise hardware allocation. CPU-heavy jobs get routed to our CPU machines; GPU jobs get RTX-optimised 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 are getting.

Our Farm Hardware in Benchmark Context

We operate over 20,000 CPU cores across our farm, primarily dual-socket Xeon processors optimised for sustained multi-frame rendering. Our GPU fleet is standardised on NVIDIA RTX 5090 machines with 32 GB VRAM, the current consumer-class flagship for GPU V-Ray workloads. Official partnerships with Chaos Group (V-Ray, Corona), Maxon (Cinema 4D, Redshift), and AXYZ design (Anima) provide verified licensing for the software running on those nodes, so renders execute in a properly licensed, reproducible environment.

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

For a parallel view using a different industry-standard benchmark — Cinebench R24, which scores CPU and GPU performance across the same hardware tiers our farm runs — see our render farm hardware benchmark with Cinebench scores for 2026.

Our GPU fleet, running RTX 5090 hardware, sits at the high end of the V-Ray GPU RTX benchmark leaderboard for single-card scores. This translates to strong per-frame speed for V-Ray GPU and RTX-optimised rendering, and it is one of the reasons GPU V-Ray cloud workflows in 2026 are substantially faster than running the same scene on a mid-range local card.

Benchmark Scores for Common Hardware

Below is a reference table showing approximate V-Ray Benchmark scores for commonly used hardware in 2026. These are representative ranges; actual scores depend on V-Ray Benchmark build version (V-Ray 6 vs V-Ray 7), driver version, BIOS settings, and bench environment. Use them for relative comparison, not as absolute quotes.

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, core of our CPU fleet
Intel i7-13700K	1,200–1,500	2023	Consumer workstation, single-socket
AMD Ryzen 9 7950X	1,800–2,100	2023	Consumer enthusiast, strong value
Single Intel Xeon Platinum 8490H	2,200–2,500	2024	High-core-count single socket

GPU Hardware Comparison

For a deeper look at GPU choices for 3D rendering in 2026, see our dedicated GPU guide.

Hardware	V-Ray GPU RTX vrays (approx)	VRAM	Year	Notes
NVIDIA RTX 5090	20,000–24,000	32 GB	2025	Our GPU fleet standard; high end of the single-card RTX leaderboard
NVIDIA RTX 4090	12,000–14,000	24 GB	2022	Previous-gen flagship; still very capable for V-Ray GPU
NVIDIA RTX 6000 Ada	14,000–16,000	48 GB	2024	Enterprise GPU with larger VRAM envelope
NVIDIA L40S	10,000–11,000	48 GB	2023	Data centre GPU, balanced compute + memory

V-Ray Benchmark GPU Results (2026)

Benchmark results continue to evolve as V-Ray 7 builds land and new drivers ship. For the current public leaderboard by GPU model and Benchmark build, check Chaos V-Ray Benchmark.

Cloud V-Ray vs Local GPU Rendering in 2026

For GPU V-Ray work in 2026, the practical question is rarely "which card tops the leaderboard" in the abstract — it is "does it make sense to keep rendering locally, or move to cloud?". A few 2026 realities shape that answer:

• Single-card scores have plateaued at the top. RTX 5090 leads the consumer leaderboard, but the jump over RTX 4090 is smaller than the 3090 → 4090 generational leap. Adding a second 5090 locally often runs into PSU, case airflow, and motherboard lane constraints.
• V-Ray GPU scales near-linearly with cards. Two 5090s render roughly twice as fast as one; four roughly four times. Cloud farms assemble that parallelism on demand without the capital cost or the thermal headaches.
• V-Ray 7 exposes more GPU work. Newer denoiser paths and lighting work well on RTX 4000/5000 hardware but demand modern drivers and reliable VRAM — exactly what a standardised farm environment provides.

On our farm, GPU V-Ray jobs submit through the same upload → render → download flow as CPU jobs. There is no remote desktop step, no per-machine V-Ray licensing for clients to manage, and no local thermal ceiling capping frame throughput. For comparison methodology, our managed vs DIY render farm guide covers the cost and workflow trade-offs in detail.

From Benchmark Scores to Render Times

Here is 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 is 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 is 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. For a deeper cost model, see our render farm cost-per-frame guide.

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, V-Ray Benchmark build (V-Ray 6 vs V-Ray 7), and any special BIOS settings such as turbo mode or power plan. Benchmark results vary with environment.

5. Compare your score to others using the public leaderboard at benchmark.chaos.com or the 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%. For RTX 40-series and RTX 50-series cards, we recommend NVIDIA Studio Driver 560 or newer for stable V-Ray GPU 6 and V-Ray GPU 7 results.
• V-Ray Benchmark build. Scores from a V-Ray 7 Benchmark build are not comparable to a V-Ray 6 build on the same hardware — keep the build constant when benchmarking.

If you are 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 is the V-Ray Benchmark? A: At Super Renders Farm, V-Ray Benchmark is a free standalone tool from Chaos that measures how fast your hardware renders a reference V-Ray scene. It runs two tests — a CPU test scored in vsamples and a GPU test scored in vpaths — and uploads the result to a public leaderboard so you can compare your system against other artists. Higher numbers mean faster rendering on that hardware.

Q: What is a good V-Ray Benchmark score in 2026? A: For V-Ray GPU on a single consumer card, anything above 3,000 vpaths is solid production territory in 2026; RTX 4090 and RTX 5090 class cards push well beyond that. For V-Ray CPU, workstation scores above 25,000 vsamples indicate a serious dual-socket or high-end Threadripper setup suitable for heavy scenes. Lower scores are not a problem for stills or small animations — they just mean longer per-frame times.

Q: Which GPU leads V-Ray benchmarks in 2026? A: On the current V-Ray Benchmark leaderboards, the NVIDIA RTX 5090 holds the top spot for single-card GPU rendering, followed by the RTX 4090. Multi-GPU rigs and workstation cards such as the RTX 6000 Ada score higher in aggregate because V-Ray GPU scales almost linearly with additional cards. On our farm we standardise on RTX 5090 nodes because the single-card V-Ray GPU throughput at that tier handles archviz, product, and motion-design scenes comfortably.

Q: Does V-Ray Benchmark test both CPU and GPU? A: Yes. The standalone V-Ray Benchmark app includes two separate tests — a CPU test (vsamples) that uses the V-Ray production renderer and a GPU test (vpaths) that uses the V-Ray GPU engine (CUDA / RTX). You can run either test independently. The two scores are not directly comparable because they use different units and different reference scenes.

Q: How do I download V-Ray Benchmark? A: V-Ray Benchmark is a free download from Chaos at chaos.com/vray-benchmark. You do not need a V-Ray license to run it — the tool ships with its own renderer and reference scenes. Installers are available for Windows, macOS, and Linux. After the run finishes, the app offers to upload your result to the public leaderboard.

Q: Why does my V-Ray Benchmark score change between runs? A: Small run-to-run variance is normal — background processes, thermal throttling, and GPU driver state can all move the score a few percent. Larger differences usually come from a V-Ray or driver update, a different Benchmark build, or power-limit settings on the GPU. For the most consistent numbers, close other apps, let the machine cool down between runs, and use NVIDIA Studio Driver 560 or newer on RTX 40-series and RTX 50-series cards.

Q: How does cloud V-Ray rendering compare to local GPU rendering in 2026? A: For a single card on a recent CPU, local GPU rendering is fine for stills, short clips, and look-dev. Where cloud V-Ray rendering pulls ahead in 2026 is throughput: a cloud farm can put dozens of RTX 5090 nodes on one job simultaneously, so a 300-frame animation that would take a day on one local 5090 finishes in a small fraction of that wall-clock time. Cloud V-Ray also avoids local thermal throttling, per-machine licensing management, and the VRAM ceiling that caps complex 2026 archviz and product scenes on lower-tier GPUs. On our farm, V-Ray GPU jobs run through an upload → render → download flow without remote desktop and with V-Ray and plugin licensing handled on the farm side, which is the main operational difference from running locally or on IaaS providers.

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 optimise rendering efficiency across your pipeline. To see how these benchmarks apply in practice on cloud infrastructure, explore our V-Ray cloud render farm resources.