Table of Contents
Introduction
Unlike the rest of the new GPUs NVIDIA has launched over the last six months with lots of fanfare, the GeForce RTX™ 5060 was released with very little marketing buzz. There was almost no announcement. There were almost no launch reviews. We were left wondering whether NVIDIA thought these cards had no competition or that they couldn’t stand up to it. And on the other side of the fence, AMD launched their new Radeon™ 9060 XT, announced at Computex, which could pose a serious challenge to NVIDIA’s cards in the budget market segment.
In this article, we will be comparing the NVIDIA GeForce RTX™ 5060 Ti, NVIDIA GeForce RTX™ 5060, and AMD Radeon™ RX 9060 XT; pitting them against each other in a variety of professional content-creation applications. Historically, 60-class cards have offered some of the best price-to-performance of their generation, so we are excited to see how they stack up.
Image
This generation, AMD has implemented several high-level architectural improvements that should boost performance across various use cases. The AMD Radeon RX 9060 XT is based on AMD’s RDNA 4 architecture, which features several improvements to the memory subsystem, scalar units, and scheduler. Though less distinct from the compute unit in NVIDIA’s designs, AMD also has next-generation Ray Tracing cores, improved dual media engines, and, for the first time, dedicated matrix compute accelerators (AI accelerators). Similarly, NVIDIA’s 50-series GPUs feature more and improved CUDA cores, fourth-generation Ray Tracing and fifth-generation Tensor cores, and the latest NVIDIA NVENC/NVDEC media engines.
Below, we have listed the most relevant GPU specifications from AMD, Intel, and NVIDIA. For more information, visit Intel Ark, NVIDIA’s 40-series GeForce page, NVIDIA’s 50-series GeForce page, or AMD’s Radeon RX Page.
On paper, the 5060 Ti looks to be a straight upgrade over the 4060 Ti. It has the same amount of VRAM, more shader units, higher clock speed, more memory bandwidth, and benefits from the new Blackwell architecture and next-gen CUDA, RT, and Tensor cores. And it does this while costing less: compared to the 4060 Ti, the 16 GB model is $70 cheaper, while the 8 GB model is $20 cheaper. Similarly, the 5060 manages to surpass the 4060 in every way save VRAM—both are 8 GB cards—for the same price.
Although we are not including AMD’s last-gen 7600 XT on the table above, the 9060 XT is likewise improved over it. More memory bandwidth (322 GB/s vs. 288 GB/s), higher frequency (3.13 GHz vs 2.76 GHz), and improved cores and architecture. It does merely match the VRAM and compute units, but it’s not slouching on upgrades. The 9060 XT has more than twice the transistor count of the 7600 XT.
However, we have one big bugbear with this generation: VRAM. And you will hear us continually bring it up again in this article. Simply put, manufacturers need to stop making 8 GB GPUs. For most users, 8 GB is insufficient to do work in professional applications or play games. An application with very low VRAM requirements may technically be fine, but it is unacceptable at this price point. Intel was able to produce a 12 GB GPU for $250. Even the B570 has the good graces to at least have 10 GB for a mere $220. For modern workloads, we feel that 8 GB is simply insufficient
Test Setup
Test Platform
| CPUs: AMD Ryzen™ 9 9950X |
| CPU Cooler: Noctua NH-U12A |
| Motherboard: ASUS ProArt X670E-Creator WiFi BIOS Version: 2604 |
| RAM: 2x DDR5-5600 32GB (64 GB total) |
| PSU: Super Flower LEADEX Platinum 1600W |
| Storage: Samsung 980 Pro 2TB |
| OS: Windows 11 Pro 64-bit (26100) Power Profile: Balanced |
NVIDIA GPUs
| NVIDIA GeForce RTX 5060 Ti (16 GB) NVIDIA GeForce RTX 5060 NVIDIA GeForce RTX 4060 Ti (16 GB) NVIDIA GeForce RTX 4060 NVIDIA GeForce RTX 3060 Ti Driver: 576.52 |
AMD GPU
| AMD Radeon RX 9060 XT (16 GB) Driver: 25.6.1 (32.0.21013.1000) |
Intel GPU
| Intel Arc B580 Driver: 101.6881 |
Benchmark Software
| After Effects 25.3 – PugetBench for After Effects 1.0 |
| DaVinci Resolve 20.0– PugetBench for DaVinci Resolve 1.2.0 |
| Topaz Video AI 6.2 |
| Unreal Engine 5.5 |
| Blender 4.4.0 |
| V-Ray 6.00.01 |
| OctaneBench 2025.2.1 |
| Llama.cpp 5122 |
For our GPU testing, we have shifted to an AMD Ryzen™ 9 9950X-based platform from our traditional Threadripper™ platform. The 9950X has fantastic all-around performance in most of our workflows. This should let the video cards be the primary limiting factor where there is the possibility of a GPU bottleneck. For testing, we used the latest available GPU drivers. Our software packages are pretty typical for our GPU reviews, including most of the PugetBench benchmarks for Adobe applications, PugetBench for DaVinci Resolve, our in-development Unreal Engine and llm benchmarks, and industry standard Blender, V-Ray, Octane, and Topaz Video AI benchmarks.
Unfortunately, we had to skip testing Premiere Pro due to some updates from Adobe around handling H.265 media. That change prevents PugetBench for Premiere Pro from working with application versions that fully support the new NVIDIA Blackwell GPUs. As such, we were unable to collect Premiere Pro results for this review. If you want more information on this, we have a blog post available explaining the situation in detail.
Raw Results Tables
We choose our benchmarks to cover many workflows and tasks to provide a balanced look at the application and its hardware interactions. However, many users have more specialized workflows. Recognizing this, we like to provide individual results for benchmarks as well. If a specific area in an application comprises most of your work, examining those results will give a more accurate understanding of the performance disparities between components. Otherwise, we recommend skipping over this section and focusing on our more in-depth analysis in the following sections.
Motion Graphics: Adobe After Effects
Traditional After Effects workloads are primarily CPU-bound, with most GPUs being sufficient for the basic processing needed. Although many After Effects users also use programs like Premiere Pro, where a GPU is more important, a budget card may have been all that was needed. However, with the advent of native 3D workflows in After Effects, GPUs are becoming more important for the application.
The Overall score (Chart #1) looks at all facets of After Effects. Here, the 5060 Ti is 13% faster than the 4060 Ti, while the 5060 leads the 4060 by 15%. We lack last-gen comparisons for the 9060 XT, but it is the slowest overall card we tested, with the (cheaper) 5060 leading by over 50%. However, the Overall score lacks the nuance needed to understand the results. 2D (Chart #2) and Tracking (Chart #4) scores were virtually identical between all the GPUs. All the difference was due to the 3D workloads.
In our 3D tests (Chart #3), the 5060 Ti is the fastest of the tested cards, with a 28% performance advantage over the 4060 Ti and 6% faster than the 5060. The 5060 is similarly performant, with a 38% uplift over the last-gen 4060. Mirroring the Overall score, the 9060 XT is dead last, 10% behind even the $250 Intel Arc B580.
For budget systems, the 5060 is the standout card for After Effects. We have concerns around the minuscule 8 GB of VRAM, but this will be sufficient for most After Effects work. Heavy 3D users may run into VRAM issues, but they will likely want a more robust GPU anyway, with less emphasis on CPU performance. The AMD Radeon 9060 XT is likely best avoided, unless there are no plans to ever use After Effects 3D workflows.
Video Editing / Motion Graphics: DaVinci Resolve Studio
Next up is DaVinci Resolve. Starting with the Overall score (Chart #1), the 5060 Ti leads the chart. The new GPU is 37% faster than the 4060 Ti and 24% faster than the 3060 Ti. This may seem odd to some readers, but we found that the architectural changes made with NVIDIA’s 40-series—likely the reduced memory bandwidth—severely hampered performance in some applications, including DaVinci Resolve. The 5060 is 32% faster than the 4060 and 16% faster than the 9060 XT.
Moving on to codec-specific tests, we begin with LongGOP media encoding and processing (Chart #2). Despite NVIDIA’s new support for H.264/HEVC 10-bit 4:2:2 media, AMD leads the pack. The 9060 XT is 14% faster than the 5060 Ti and 5060. Those cards, which scored identically, led their respective last-gen counterparts by 31%. There are some quirks to this data, and if you work with specific media types, we strongly recommend you look at the raw result tables above. Specifically, AMD is faster in every LongGOP test except processing (but including encoding) the aforementioned HEVC media type, where it is only half as fast.
The next codec types we test are Intraframe codecs (Chart #3). Though typically CPU-based, we do see some effect from the GPU. Namely, a 40-series card or Intel Arc GPU will perform somewhat worse. We’re not sure why this is the case, but there may be some additional CPU overhead associated with these cards.
Finally, we test with RAW codecs (Chart #4) such as RED and ARRIRAW. In these, the 5060 Ti is again the fastest, leading the 5060 by 13% and the 4060 Ti by 58%. RAW codecs are one of the worst-performing areas for this tier of 40-series cards. The 5060 similarly leads the 4060 by 40% and the 9060 XT by 25%.
After our codec-specific tests in our DaVinci Resolve suite are the GPU Effects (Chart #5). As you would expect, these are principally GPU-bound and historically favor NVIDIA. The 5060 Ti is the fastest card tested, 9% ahead of the 5060 and 36% faster than the 4060 Ti. The 5060 follows just behind it, with a 44% performance advantage over the 4060, and 37% faster than the 9060 XT. AMD does particularly poorly here, with even the B580 27% ahead of it.
Fusion (Chart #6) is another area where the GPU is usually second to the CPU, but with lower-end cards, the GPU can make more of a difference. Our testing found that the 5060 Ti was the fastest of the 60-class GPUs, 24% faster than the 4060 Ti and 10% faster than the 5060. The latter card is a mere 7% ahead of the 9060 XT but 16% faster than the 4060.
Our final category of DaVinci Resolve tests is AI (Chart #7), encompassing a variety of tasks such as masking, relighting, and audio transcription, among others. In these workflows, the 5060 Ti was 21% faster than the 5060 and 35% faster than the 4060 Ti. The 5060 here has a 29% performance advantage over the 4060, and a large 52% advantage over the 9060 XT.
Topaz Video AI
Image
Topaz Video AI constantly releases new versions, making it somewhat difficult to keep up with its most recent changes. However, in our initial Blackwell review, we found that the 50-series cards were slightly slower than the previous-generation 40-series. It appears as if this issue has been, at least somewhat, addressed.
The fastest card we tested in this review was the AMD Radeon 9060 XT, which, with a score of 399, was 20% faster than the 5060 Ti and 36% faster than the 560. Combined with its 16 GB of VRAM and comparable price to the 5060, it’s the best choice in this price bracket for Topaz. The 5060 Ti managed a 14% lead over the 4060 Ti, while the 5060 was 29% faster than the 4060.
Game Dev / Virtual Production: Unreal Engine
Our Unreal Engine benchmark renders various real-time scenes, which feature a mix of features (such as Nanite and hardware RT) at three different standard resolutions. We combine all of those into one overall score, but we have also pulled out some resolution—and feature-specific averages of test scores.
Starting with the Overall geometric mean FPS (Chart #1), the faster card was the 5060 Ti, with a minor 4% lead over the 9060 XT. This represents a 9% generational uplift for the 5060 Ti, matched by the 7% of the 5060 over the 4060. The 9060 XT was 27% faster than the 5060.
We don’t want to dig too deeply into the resolution-based groupings, save to say that we believe 8 GB of VRAM is largely insufficient for professional use. The Intel Arc B580, with 12 GB of VRAM, starts as the second-slowest card at 1080p, and moves to be behind only the 16 GB cards at 4K. This is mainly due to the 8 GB cards exceeding their VRAM buffer. We struggle to recommend an 8 GB card for most modern workflows, and would especially caution against the 8 GB versions of the 5060 Ti and 9060 XT.
Looking at Rasterized FPS and RT FPS (Charts #5 and #6), we see the most significant differentiation between the NVIDIA and AMD cards. The 5060 Ti has the best tested performance in RT workloads, leading the 9060 XT by 10% and the 5060 by 38%. However, in the Rasterized scenes, the 9060 XT is the fastest, with a 5% performance advantage over the 5060 Ti and 20% over the 5060. While we don’t expect this difference to be a selling point for most professional use cases, it is interesting to see the divergence in priorities between AMD and NVIDIA.
We think the 9060 XT is an excellent value for Unreal Engine and other game engines. Its performance is essentially on par with the more expensive 5060 Ti, and the 16 GB version provides the minimum amount of VRAM we think is necessary.
GPU Rendering: Blender & V-Ray
Moving on to offline GPU rendering, we tested with three different applications; we hope to include more in the future now that application support for Blackwell cards is improving. Currently, only one of the applications supports non-NVIDIA cards, but NVIDIA has historically had much better rendering performance anyway.
Starting off with Blender (Chart #1), we are largely unsurprised by the results. Rendering is NVIDIA’s game to lose at the moment. The 5060 Ti leads the chart, with an 11% advantage over the 4060 Ti and a 15% advantage over the 5060. The 5060 has a 20% lead over the 4060 and a massive 133% lead over the 9060 XT. In fact, the 9060 XT is even slower than the B580. Although the relatively cheap VRAM compared to NVIDIA may seem tempting, the GPU cannot keep up with the other options.
In V-Ray, we test with both CUDA (Chart #2) and RT (Chart #3) workflows. However, the relative performance between the two is virtually the same, so we will look at the better-performing RT mode. The 5060 Ti is, as expected, the fastest card of those tested. With a score of 4,384, it is 32% faster than the 4060 Ti and 13% faster than the 5060. The 5060 is 31% faster than the 4060. Both of these are great generational improvements. Octane, like V-Ray, only supports NVIDIA GPUs. In it, we found that the 5060 Ti was 15% faster than the last-gen 4060 Ti and 11% faster than the 5060. Likewise, the 5060 was 20% faster than the 4060.
AI: LLM (llama)
System Image
Both NVIDIA and AMD have made big claims about AI performance with their new generations of video cards. Although we won’t be doing a full article on AI performance for this tier of cards, we did want to include a few benchmarks. This benchmark specifically looks at LLM performance on the GPUs. For this, we used Llama.cpp build 5122 with a Phi-3 Mini Q4 model. The small model with a low quantization allows us to test it even on low-VRAM cards.
The first test we ran was prompt processing (Chart #1): how quickly the model can tokenize a prompt in order to start creating a response. The 5060 Ti was the fastest card at this, generating 17% more tokens per second than the 4060 Ti. It also led the 5060 by 19%. The 5060 also performed well, with 34% more tokens per second than the 4060. Unfortunately, we could not test the 9060 XT in this round of benchmarking.
After prompt processing, the model needs to generate a response. Token generation (Chart #2) measures this. We found that, unlike prompt processing, token generation for NVIDIA seems to scale primarily with VRAM bandwidth. The 5060 Ti is nominally the fastest, but it is only 3% faster than the 5060 and 7% faster than the 3060 Ti. Both 40-series cards trail well behind.
Although these are interesting results, our primary concern with any of the tested cards for AI usage is VRAM. We intentionally test with a small model for compatibility purposes, but most models today are larger than what would fit in 8 GB of VRAM. Indeed, many models even exceed a 16 GB frame buffer.
How do the NVIDIA GeForce RTX 5060 Ti and AMD Radeon RX 9060 XT Compare for Content Creation?
The NVIDIA GeForce RTX 5060 Ti and AMD Radeon RX 9060 XT are both competent cards for their price point—as long as you stick with the 16 GB models. Although it depends on the specific workflow they are used for, both can offer good performance for the dollar. Similarly, the 5060 performs well enough in most tests for its price, but we have concerns about the limited 8 GB of VRAM.
In After Effects, both NVIDIA cards lead the pack in the 3D tests, outperforming their last-gen counterparts by about 30%. The 9060 XT is unoptimized in this workflow, with one-third the performance. However, all the cards are totally fine for traditional 2D workflows. Similarly, in DaVinci Resolve, the 9060 XT is competitive in codec-specific tests and Fusion, but falls far behind in AI and GPU Effects workloads. The 5060 Ti was, overall, 37% faster than the 4060 Ti, while the 5060 was 32% faster than the 4060.
We found that the 9060 XT was the fastest card in Topaz Video AI, 20% faster than the 5060 Ti and 36% faster than the 5060. The NVIDIA cards did have fine generational uplifts, but we were particularly impressed by the combination of performance, VRAM, and price of the 9060 XT for this application. We couldn’t test the 9060 XT for our LLM benchmark, but we did find fairly standard 17% and 34% better performance for the 5060 Ti and 5060, respectively, over their last-gen counterparts.
Moving on to our Unreal Engine testing, the 9060 XT and 5060 Ti were nearly equivalent. However, the former was relatively better at rasterized scenes at lower resolution and the latter at higher-resolution, ray-traced scenes. Overall, the 5060 Ti was 9% faster than the 4060 Ti, and the 5060 was 7% faster than the 4060. Finally, in our Offline Renderers (Blender, V-Ray, and Octane), the 9060 XT was either unsupported or incredibly slow. The NVIDIA cards are just better supported, and saw solid 11 to 32% higher scores.
Whether an NVIDIA or AMD card at this price point is better depends a lot on the workflow being used. We are fairly impressed with the overall combination of VRAM and performance for the price of the AMD Radeon 9060 XT. But, if you use offline rendering applications or do heavy 3D work, it will likely be much worse value than an NVIDIA GeForce RTX 5060 Ti.
Looking for a content creation workstation?
We build computers tailor-made for your workflow.
Don’t know where to start?
We can help!
Get in touch with one of our technical consultants today.