Thief

How This Combo Ranks

The Intel Core Ultra 5 235 paired with the AMD Radeon RX 9070 secures the 562nd position out of 2,851 tested combinations in Thief, placing it comfortably within the top 20% of all configurations. This rank reflects a strong pairing that delivers high frame rates across all tested settings, though the data reveals that the GPU is frequently the limiting factor at higher resolutions rather than the CPU.

Benchmark results indicate this combo outperforms a substantial majority of alternatives, with the ranking driven primarily by the GPU's raw throughput in a game that scales heavily with graphics hardware. The CPU's percentile ranking against all processors sits at the 89th mark, while the GPU's percentile against all graphics cards is lower at the 69th mark, suggesting the processor has more headroom relative to its peers than the graphics card does in this specific workload.

The nearest rivals for the CPU include the AMD Ryzen AI 9 HX 375 with an average score of 46,030 (a 0.1% delta), the Intel Core i9-13900HX at 46,098 (a -0.1% delta), and two AMD EPYC server parts that bracket the same performance tier. This clustering indicates the Ultra 5 235 sits at a performance plateau where tiny differences separate several very different processor families, all within a 0.3% band. The GPU's nearest rivals tell a similar story: the NVIDIA GeForce GTX TITAN Z scores 23,736 (a 0.6% delta), the AMD Radeon RX 6800S scores 24,063 (a -0.8% delta), and the NVIDIA GeForce RTX 3080 Mobile scores 23,628 (a 1.1% delta), showing the RX 9070 is tightly grouped with older and mobile parts in aggregate benchmarks despite its modern architecture.

In practical terms for Thief, this rank means the combo will handle the game with ease at most configurations. The data shows frame rates ranging from a low of 53 FPS at 4K Ultra to a high of 221 FPS at 1080p Low, which places the combo in a sweet spot where esports-level refresh rates are achievable at lower settings while 4K remains playable at reduced quality presets.

CPU Role

The Intel Core Ultra 5 235 brings 14 cores and 14 threads to the pairing, a configuration without hyperthreading that still provides ample parallelism for Thief's engine. The base clock of 3.40 GHz boosts up to 5.00 GHz, giving the processor strong single-thread performance that matters for a 2014 stealth title built around sequential game logic and AI pathfinding. The architecture is Arrow Lake-S on a 3 nm TSMC process, with 17,800 million transistors on a 243 mm² die, and the cache hierarchy includes 192 KB of L1 per core, 3 MB of L2 per core, and 24 MB of shared L3.

The CPU's benchmark scores reveal a processor that excels in both single and multi-threaded workloads. In Cinebench R23, it scores 14,769 multi-core and 2,085 single-core; in Cinebench R20, the scores are 6,202 multi-core and 875 single-core. The Passmark suite shows a multithread score of 37,816 and a single-thread score of 4,516, with specialized tests like data compression at 390,711, floating point math at 117,951, and integer math at 87,948. Physics simulation scores 2,570, and extended instructions hit 32,752.

For Thief specifically, the CPU's role is to feed the GPU with draw calls and manage game state, and the data suggests it does so without bottlenecking even at 1080p Low where the GPU is producing 221 FPS. The jump from 1080p Low to 1080p Ultra drops average FPS from 221 to 132, a 40% reduction, which is a GPU-bound scaling pattern rather than a CPU limit. At 4K Low the combo still achieves 97 FPS, and at 4K Ultra it falls to 53 FPS, further confirming that the processor is not the constraint in this title.

The dual-channel DDR5 memory support with 102.4 GB/s bandwidth provides sufficient memory throughput for the game's streaming needs, and the PCIe Gen 5 interface with 20 lanes ensures the GPU has full bandwidth available. The integrated Arc Xe-LPG Graphics 24EU is irrelevant for this discrete GPU configuration, but it does mean the platform has fallback display capability. The 65W TDP classifies this as a mainstream power envelope, and the lack of an unlocked multiplier indicates this is not a tuning-focused part, though the data shows it does not need to be for Thief.

FAQ

Q: What is the best resolution to play Thief with this combo?

A: The data shows 2560x1440 at High settings delivers 95 FPS average, which balances visual quality and smoothness. At 4K High, the average drops to 56 FPS, which is playable but leaves less headroom for frame drops, while 1080p High hits 139 FPS for maximum smoothness.

Q: How much FPS does the combo lose when moving from 1080p to 4K at the same settings?

A: At High settings, the average FPS drops from 139 at 1080p to 95 at 1440p to 56 at 4K. This represents a 60% reduction from 1080p to 4K, indicating a strong GPU-bound scaling curve.

Q: Is the CPU or GPU the limiting factor in Thief?

A: The data indicates the GPU is the primary limiter. At 1080p Low, the combo produces 221 FPS, and at 4K Ultra it drops to 53 FPS, a scaling pattern consistent with GPU saturation. The CPU's 89th percentile ranking versus the GPU's 69th percentile further supports this conclusion.

Q: Can this combo handle Ultra settings?

A: Yes, but with caveats. At 1080p Ultra, the average FPS is 132, which is smooth. At 1440p Ultra, it drops to 91 FPS, still playable. At 4K Ultra, the average falls to 53 FPS, which is borderline for fast-paced stealth gameplay.

Q: How does the RX 9070 compare to its nearest rivals in aggregate benchmarks?

A: The RX 9070's average benchmark score is 23,877. It sits 0.6% ahead of the GTX TITAN Z, 0.8% behind the RX 6800S, 1.1% ahead of the RTX 3080 Mobile, and 1.2% behind the RTX 2080 SUPER. These deltas are minimal, indicating near-parity with that group.

Q: What is the CPU's single-thread performance relative to its multi-thread performance?

A: In Cinebench R23, the single-core score of 2,085 is about 14% of the multi-core score of 14,769, which is typical for a 14-core part. The Passmark single-thread score of 4,516 versus multithread of 37,816 shows a similar ratio, indicating balanced core efficiency.

Measured FPS Breakdown

The measured data provides a complete matrix of performance across three resolutions and four settings presets. At 1920x1080, the combo delivers 221 FPS average on Low, 182 FPS on Medium, 139 FPS on High, and 132 FPS on Ultra. The Medium preset includes measured minimum and maximum values of 155 and 209 FPS respectively, showing a tight frame pacing window. The gap between High and Ultra at 1080p is only 7 FPS, suggesting Ultra offers minimal visual benefit for the performance cost at this resolution.

Moving to 2560x1440, the averages are 164 FPS on Low, 125 FPS on Medium, 95 FPS on High, and 91 FPS on Ultra. The Medium preset at 1440p shows a minimum of 106 and maximum of 144 FPS, which represents a 38 FPS spread. The delta between High and Ultra narrows to just 4 FPS at this resolution, making High the more efficient choice for 1440p gaming.

At 3840x2160, the averages drop to 97 FPS on Low, 73 FPS on Medium, 56 FPS on High, and 53 FPS on Ultra. The Medium preset at 4K shows a minimum of 62 and maximum of 84 FPS, a 22 FPS spread that indicates some scene-dependent variance. The difference between High and Ultra at 4K is only 3 FPS, making Ultra essentially a free upgrade over High if the visual improvements are desired, though both sit below the 60 FPS threshold that many players target.

The scaling from Low to Ultra at each resolution shows diminishing returns: at 1080p, the drop from Low to Ultra is 89 FPS (221 to 132); at 1440p, it is 73 FPS (164 to 91); at 4K, it is 44 FPS (97 to 53). This pattern indicates that as resolution increases, the GPU becomes more saturated and the relative cost of higher settings decreases in absolute terms, though the percentage drop remains significant.

Resolution Scaling

The FPS data reveals a clear resolution scaling pattern that identifies the GPU as the dominant constraint. From 1080p to 1440p at High settings, the average FPS drops from 139 to 95, a 32% reduction. From 1440p to 4K at High, it drops from 95 to 56, a 41% reduction. The total drop from 1080p to 4K at High is 60%, which aligns closely with the pixel count increase of 4.5x, though the FPS reduction is proportionally smaller, indicating some headroom in the GPU's compute capacity.

At Low settings, the scaling is steeper: 221 FPS at 1080p drops to 164 FPS at 1440p (26% reduction) and then to 97 FPS at 4K (41% reduction from 1440p). The total drop from 1080p to 4K at Low is 56%. At Ultra settings, the pattern is similar: 132 FPS at 1080p, 91 FPS at 1440p (31% reduction), and 53 FPS at 4K (42% reduction from 1440p), for a total drop of 60%.

The consistency of these reductions across settings presets—ranging from 56% to 60% total drop—strongly indicates that the GPU's raw pixel throughput is the limiting factor. If the CPU were the bottleneck, we would expect to see the scaling flatten at lower resolutions where the CPU cannot keep up; instead, the FPS continues to rise as resolution drops, suggesting the CPU has significant headroom. The CPU's boost clock of 5.00 GHz and its Passmark single-thread score of 4,516 support this interpretation, as Thief's engine benefits from high per-core performance.

The 4K Medium measurement includes a minimum of 62 FPS and maximum of 84 FPS, which means the combo can maintain playable frame rates even in demanding scenes at 4K, though the average of 73 FPS suggests occasional dips below 60 FPS are possible. At 1080p Medium, the minimum of 155 FPS ensures no perceptible stutter, and at 1440p Medium, the minimum of 106 FPS keeps the experience smooth. This data suggests that for players prioritizing frame rate stability, the Medium preset offers the best consistency across all resolutions.

Settings Recommendations

For players targeting the smoothest experience, the Low preset at 1080p delivers 221 FPS average, which is well beyond any display's refresh rate and provides maximum input responsiveness. However, the data shows that Medium at 1080p still achieves 182 FPS average with a minimum of 155 FPS, making it the superior choice for those who want both speed and visual quality—the 39 FPS cost of moving from Low to Medium is a reasonable trade for the graphical improvements.

At 1440p, the High preset at 95 FPS average represents the best balance for most users, as it sits comfortably above 60 FPS while offering significantly better visuals than Medium's 125 FPS (which is faster but less detailed). The Ultra preset at 91 FPS is only 4 FPS slower than High, so players with 1440p displays should consider Ultra if their monitor supports high refresh rates, as the visual fidelity gain comes at minimal performance cost. The Medium preset at 1440p with its 106 FPS minimum ensures no frame drops below the 100 FPS threshold, making it the safest choice for competitive play.

At 4K, the recommendations shift toward lower presets. The Low preset at 97 FPS average is the only option that exceeds 60 FPS comfortably, making it the default choice for 4K gaming. The Medium preset at 73 FPS average with a 62 FPS minimum is playable but risks occasional dips below 60 FPS in heavy scenes. The High and Ultra presets at 56 and 53 FPS respectively sit below the 60 FPS target, so they are only recommended for players who prioritize visual quality over frame rate smoothness and are willing to accept sub-60 FPS performance.

The data shows that the Medium preset offers the best overall consistency across all resolutions, with minimum FPS values that never fall below 62 FPS even at 4K. For a stealth game where reaction times matter less than in fast-paced shooters, players may prefer High or Ultra at 1440p for the best visual experience, knowing the combo can maintain 91-95 FPS averages. The verdict is clear: at 1080p, choose Medium for the best combination of speed and quality; at 1440p, High or Ultra both work well; at 4K, Low or Medium are the only viable options for maintaining smooth gameplay.