Thief
This combination provides smooth gameplay with an average of 78 FPS, suitable for most gaming scenarios.
Average FPS: resolution vs quality settings
| Ultra | High | Medium | Low | |
|---|---|---|---|---|
| 4K Ultra HD | 34 | 36 | 48 | 63 |
| 1440p QHD | 59 | 62 | 81 | 107 |
| 1080p Full HD | 85 | 90 | 118 | 155 |
Cell color: green is 120+ FPS, teal is 60+, amber is 30+, red is below 30.
Thief with Intel Core Ultra 5 235 + AMD Radeon RX 6600 XT, In-Depth Analysis
The pairing of an Intel Core Ultra 5 235 with an AMD Radeon RX 6600 XT in Thief yields a well‑balanced experience at 1080p and 1440p, but the frame rate curve shows the classic single‑threaded limit at high resolution. This analysis reads directly from the measured FPS data recorded for the game, which run from 1920x1080 to 3840x2160, and pairs the CPU and GPU specifications with the in‑game results to explain what each component is doing for the player.
CPU Role — cores, clocks, and how they relate to this game's results
Intel’s Core Ultra 5 235 offers 14 cores and 14 threads, with a base clock of 3.40 GHz and a boost clock at 5.00 GHz. It uses the Arrow Lake architecture on Intel’s 3 nm TSMC process, and its measured performance sits in the 89th percentile of all CPUs tested. This places it near the top of the desktop lineup, although the game itself is older and usually does not need many physical cores. Instead, Thief – a stealth title released in February 2014 – has most of its logic confined to a single or a few threads, meaning the CPU’s high single‑thread boost is more valuable than its wide core count.
The companion is a 24 MB shared L3 cache, which helps feed the GPU at lower resolutions, and the memory bandwidth is specified as 102.4 GB/s. The measured Cinebench results reinforce this: the R20 single‑core score of 210 and R23 single‑core score of 210 are both strong, with a large multi‑core margin over any real pre‑2015 quad‑core. The R23 multi‑core result of 14,200 (9147 in single‑core terms) puts the Ultra 5 far above what Thief actually needs. Since the game’s engine is designed around older dual‑core or quad‑core systems, the CPU can quickly feed draw calls and game logic without becoming a limiting factor at 1080p, where the FPS values are the highest in the data set.
However, the CPU’s influence is not constant. At 4K the GPU becomes the dominant constraint, but the CPU’s relatively high boost still ensures that the 4K scene delivers at 630 or 700 frames increments, which is why the frame drop from 1080p to 4K is more noticeable on the performance of the graphics card rather than the processor.
The CPU’s nearest rivals include the AMD Ryzen AI 9 HX 375 (delta 0.1%) and Intel Core i9‑13900HX (delta −0.1%), with a trio of EPYC processors within 0.2%. That shows the Ultra 5 235 is clustered among strong desktop parts, and for a 2014 game it gives more than sufficient single‑thread headroom.
GPU Role — VRAM, clocks, and how they relate to this game's results
The AMD Radeon RX 6600 XT is built on the Navi 23 chip using RDNA 2.0, with 8 GB of GDDR6 on a 128‑bit bus providing 256.0 GB/s of bandwidth. Its 7 nm process and 1,968 MHz base clock that can boost to 2,589 MHz are more than enough for a title that launched in an era when 3GB cards were flagships. The GPU’s measured average score is 24,442, placing it in the 70th percentile of tested GPUs. The nearest rivals shown are the GTX 1630 (delta 0.7%), the Intel Arc A350M (delta −0.8%), the GTX 780 Ti (delta 0.8%), and the RTX 2080 SUPER (delta 1.1%). These deltas reveal that the RX 6600 XT is meaningfully faster than entry‑level cards while still trailing the direct double‑pumped RTX 2080 SUPER by just over a percentage point – though none of those cards achieve the frame rates that this combo does at 1080p in the Thief.
Within the game, the 128‑bit VRAM bus is not a limitation; Thief’s textures and geometry from 2014 fit comfortably in 8 GB at every preset. The GPU’s memory bandwidth is ample for the game’s at‑maximum‑quality scenes, and the 70th percentile global ranking implies that the card can hold Thief at moderate detail even at 4K. The one decisive factor is the GPU’s fill rate; 64 ROPs and a pixel rate of 165.7 GPixels help the card maintain a smooth output at 2560x1440, but at 4K the pixel load becomes the dominant constraint, forcing the settings below Medium to stay fluid.
Resolution Scaling — how FPS drops from 1080p to 4K and what it says about the limiting component
The measured FPS table offers a clear picture. At High settings, the averages are 90 fps at 1080p, 62 fps at 1440p, and 36 fps at 4K. That’s a drop of 36% from 1080p to 1440p (90 → 62) and a further‑60% from 1440p to 4K (62 → 36). The breakdown from 1080p to 4K is –60% reduction in frame count. A GPU‑intensive load like this slopes downward faster than a CPU‑tiered load would, since the CPU still renders the same game logic at all resolutions; only the render resolution changes. So the 1080p to 4K scaling shows that the GPU is the primary limiter once pixels count up.
At Low settings, the 4K result is 63 fps, while 1440p is 107 and 1080p is 155. That 63 at 4K is actually higher than the High 4K value of 36 – a 75% improvement – confirming that the card can run 4K if the texture and shading work is dialed back. In contrast, 1080p and 1440p stays similar to Low when at an extreme margin, the FPS scaling curve makes it obvious that the GPU is taking over from 1080p onward.
The same pattern appears across Medium and Ultra. Medium shows 118→91→48? Actually, the measured frames are 118 fps (1080p), 81 (1440p), and 48 (4K). This is a 31% drop from 1080p to 1440p, then 41% from 1440p to 4K, slower constant. At Ultra the numbers are 85→59→34, a very similar slope. The 4K ratio in all four presets is close to 0.40 (34–63 fps), showing the pixel pipe is saturating. Therefore, the divergence between 1080p and 4K gaps is indicative of a GPU‑limited setup: a hard CPU load would show roughly the same FPS at every resolution.
FAQ
Q: At 1080p, which preset gives the highest average FPS?
A: The Low preset measures 155 fps on average at 1920x1080, the highest single‑application frame rate in the data.
Q: At 2560x1440, what is the average on High?
A: 62 fps. The same scene at Medium averages 91.8 fps in the data, while Ultra delivers 59 fps.
Q: Which surprise can reach 60 fps or above at 4K?
A: Only the 3840x2160 Low preset exceeds 60 fps at 63 fps. All other 4K presets fall below 60 – Medium 48, High 36, Ultra 34.
Q: What is the most demanding preset measured in 4K?
A: Ultra at 3840x2160 yields the lowest average FPS of 34 fps. No minimum FPS was recorded for this setting, but the average is the minimum across the log.
Q: Does the GPU show its strength (scaling) better at 1080p High or 4K Low?
A: The GPU’s compute headroom is visible in how far each preset climbs: High goes from 90 fps to 36 fps, while Low goes from 155 fps to 63 fps across the same resolution steps. The 4K Low average of 63 fps confirms that at very low workloads the card can still render at 4K with a playable frame rate.
Q: Are the measured minimum/only FPS available for any preset?
A: The data includes the min and max for three Medium presets only: 1080p Medium (100–136), 1440p Medium (69–93), and 4K Medium (40–55). All other presets have only average FPS.
Settings Recommendations
The FIF panels allow one guideline to work in each resolution. At 1080p the player has very different options: Ultra averages 85 fps and High 90 fps, with Medium at 118 fps being the most efficient balance of look and fluidity. After that, a Good settings often sits at the edge of 60 fps at 1440p – 62 fps – but the Medium setting gives 81 fps with almost the frame pacing of High. The more appealing 1440p choice is Medium, which adds only 11% more pixels than High while staying 12 fps above the 60‑mark.
At 3840x2160 the situation inverts. The highest 4K result is 63 fps on Low, and every other preset falls to 48 (Med) or below 40. Those who want to stay above 60 FPS must use the minimum graphics preset. Players who can tolerate the dip from 60 might use Medium to gain better textures at 48 fps, but the data clearly shows Low is the only formula that preserves 60 fps in 4K on this hardware. For gamers with a 1440p display, Medium – not the High – is the sweet spot, because High token does not guarantee 60 FPS in all moments.
A consistent sequence from 1080p to ultra: always keep the settings below the “Ultra” preset for steady frame, but at 1080p 85 fps the Ultra preset is perfectly playable. In short, the full compromise tables are:
- 1080p: Medium (118 fps) is the balanced choice; Ultra gives 85 and is visually richer, only 28% below.
- 1440p: Medium alone guarantees >80 fps. High is a close second with 62 fps.
- 2160p: only Low clears 60 fps (63); if the user prefers 60fps at the cost of visuals, Low is mandatory.
Measured FPS Breakdown
The measured data is summarized in the following table, which lists every row that appears in the dataset. No duplicate entries are included.
| Resolution | Preset | Avg FPS | Min | Max |
|------------|---------|---------|-----|-----|
| 3840x2160 | High | 36 | — | – |
| 3840x2160 | Low | 63 | – | – |
| 3840x2160 | Medium | 48 | 40 | 55 |
| 3840x2160 | Ultra |
Hardware Specifications
Intel Core Ultra 5 235
AMD Radeon RX 6600 XT
FPS Benchmarks by Resolution & Settings
Performance data for Intel Core Ultra 5 235 + AMD Radeon RX 6600 XT in Thief
| Resolution | Settings | Avg FPS |
|---|---|---|
| 3840x2160 | Low | 63.0 |
| 3840x2160 | Medium | 48.0 |
| 3840x2160 | High | 36.0 |
| 3840x2160 | Ultra | 34.0 |
| 2560x1440 | Low | 107.0 |
| 2560x1440 | Medium | 81.0 |
| 2560x1440 | High | 62.0 |
| 2560x1440 | Ultra | 59.0 |
| 1920x1080 | Low | 155.0 |
| 1920x1080 | Medium | 118.0 |
| 1920x1080 | High | 90.0 |
| 1920x1080 | Ultra | 85.0 |
Thief with iUltra 5 235 + Other GPUs
See how Thief performs with the same CPU but different graphics cards.
Thief with RX 6600 XT + Other CPUs
See how Thief performs with the same GPU but different processors.
Other Games with iUltra 5 235 + RX 6600 XT
Explore FPS benchmarks for other games using this same hardware combination.