Thief

The Intel Core i5-9400 paired with the NVIDIA GeForce RTX 5070 Ti delivers a smooth experience in Thief, but the data reveals a clear hierarchy in settings. At 1080p, the Low preset achieves the highest average frame rate at 117.5 FPS, while Medium drops to 105.5 FPS, High to 98.3 FPS, and Ultra to 97.8 FPS. The gap between High and Ultra is negligible at just 0.5 FPS, making High the more sensible choice for visual fidelity without sacrificing performance. At 1440p, Low again leads at 105.3 FPS, with Medium at 94.4 FPS, High at 86.1 FPS, and Ultra at 83.9 FPS. The 1440p Ultra result still clears 60 FPS comfortably, but the 2.2 FPS difference between High and Ultra at this resolution suggests diminishing returns. At 4K, the picture changes: Low hits 90.6 FPS, Medium 80.8 FPS, Ultra 72.1 FPS, and High 69.9 FPS. Here, Ultra actually outperforms High by 2.2 FPS, which is an anomaly worth noting. For the best balance across all resolutions, Medium provides a stable baseline, but High at 1080p or 1440p offers the most consistent experience per the measured rows, with Ultra only marginally better at 4K.

CPU Role

The Intel Core i5-9400 is a six-core, six-thread processor from the Coffee Lake generation, with a base clock of 2.90 GHz and a boost clock of 4.10 GHz. It draws 65 W TDP and sits on the Intel Socket 1151 platform. Its benchmark profile shows a Cinebench R23 multi-core score of 7969 and a single-core score of 1125, placing it in the 50th percentile among all CPUs. The processor’s average benchmark score is 2305, which is nearly identical to its nearest rivals: the Intel Core i3-10305 scores 2300 (0.2% delta), the Intel Core i3-1125G4 scores 2299 (0.2% delta), and the Intel Xeon E-2134 scores 2295 (0.4% delta). Only the Intel Core i7-8809G edges ahead at 2319, a -0.6% delta. This puts the i5-9400 in a tight cluster, but its six physical cores without hyper-threading (six threads total) are relevant for Thief, a 2014 title that is not heavily threaded. The game’s frame rates at 1080p, where CPU overhead is more pronounced, show a maximum of 117.5 FPS on Low, which is solid but not exceptional. At 4K, the CPU’s influence wanes, as the GPU becomes the bottleneck; the FPS drops from 90.6 at Low to 69.9 at High, a spread of 20.7 FPS that reflects the GPU’s workload rather than CPU limits. The processor’s 9 MB of shared L3 cache and dual-channel DDR4 memory support at 42.7 GB/s bandwidth are adequate for this era of game, and the data suggests that the i5-9400 does not throttle the RTX 5070 Ti’s performance at higher resolutions, though at 1080p it may cap the ceiling slightly, as evidenced by the relatively modest 117.5 FPS peak.

GPU Role

The NVIDIA GeForce RTX 5070 Ti is the dominant component in this combo, built on the Blackwell 2.0 architecture with the GB203 chip at a 5 nm process node. It packs 16 GB of GDDR7 memory on a 256-bit bus, delivering 896.0 GB/s of bandwidth. The GPU’s base clock is 2295 MHz with a boost of 2452 MHz, and memory runs at 1750 MHz (28 Gbps effective). It has 8960 shading units, 280 texture mapping units, and 96 ROPs, with 70 RT cores and 280 tensor cores. Its compute output is 43.94 TFLOPS for FP32, and it supports DirectX 12 Ultimate, Vulkan 1.4, and OpenGL 4.6. In synthetic benchmarks, the RTX 5070 Ti scores 6604 in 3DMark Steel Nomad DX12, 231226 in Geekbench OpenCL, and 227554 in Geekbench Vulkan. Its Passmark G3D score is 32974, with a GPU compute score of 20203. The GPU ranks in the 87th percentile among all GPUs, with an average benchmark score of 52086. Its nearest rival is the NVIDIA GeForce RTX 4080, which scores 52085 (0% delta), making them effectively identical in raw compute. The AMD Radeon RX 7700S is slightly ahead at 52505 (-0.8% delta), while the NVIDIA RTX A1000 trails at 50826 (2.5% delta) and the AMD Radeon Pro W6600M is ahead at 53414 (-2.5% delta). In Thief, the GPU’s VRAM and bandwidth are more than sufficient, as the game’s 2014 engine does not stress 16 GB. The measured FPS shows that at 4K, the GPU is the limiting factor: from Low to Ultra, the average drops from 90.6 to 72.1 FPS, a 20.4% reduction. At 1080p, the GPU is less taxed, so the CPU’s role grows, but the RTX 5070 Ti still pushes frame rates above 97 FPS even at Ultra. The data indicates that the GPU’s headroom is vast, and the bottleneck in this pairing is almost entirely the CPU at lower resolutions, while the GPU takes over at 4K.

How This Combo Ranks

The combination of the Intel Core i5-9400 and NVIDIA GeForce RTX 5070 Ti ranks 880th out of 1642 tested combos in Thief, placing it in the 46th percentile of all system pairings. This is a middling result, given the GPU’s high percentile (87th among all GPUs) and the CPU’s mid-pack standing (50th percentile). The ranking suggests that the CPU holds back the overall combo, as the GPU alone would place much higher if paired with a stronger processor. The data shows that the i5-9400’s average benchmark score of 2305 is only 0.2% higher than the Intel Core i3-10305 (2300) and 0.4% higher than the Intel Xeon E-2134 (2295), meaning the CPU is not a standout performer. Meanwhile, the RTX 5070 Ti’s average score of 52086 is exactly on par with the RTX 4080 (52085, 0% delta), indicating the GPU is a top-tier part. The 880th rank out of 1642 reflects this imbalance: the combo is not in the top quartile, but it is also not in the bottom half. In practical terms, the measured FPS at 1080p Ultra (97.8 FPS) and 1440p Ultra (83.9 FPS) are playable, but the 4K Ultra result (72.1 FPS) is where the combo starts to feel less capable. Compared to other tested combos, this pairing is best suited for 1080p and 1440p gaming, where the GPU’s power can shine, but it will not top leaderboards due to the CPU’s modest single-threaded and multi-threaded scores. The rank of 880 out of 1642 indicates that roughly half of all tested systems outperform it, which is expected given the i5-9400’s age and the RTX 5070 Ti’s relative newness.

Resolution Scaling

The measured FPS data shows a clear scaling pattern across resolutions. At 1080p, the average frame rates range from 117.5 FPS on Low to 97.8 FPS on Ultra, a drop of 19.7 FPS (16.8%). Moving to 1440p, the same settings yield 105.3 FPS on Low and 83.9 FPS on Ultra, a drop of 21.4 FPS (20.3%). At 4K, the range tightens: 90.6 FPS on Low and 72.1 FPS on Ultra, a drop of 18.5 FPS (20.4%). The percentage drop from 1080p to 4K is consistent across settings: Low drops from 117.5 to 90.6 FPS (22.9% reduction), Medium from 105.5 to 80.8 FPS (23.4% reduction), High from 98.3 to 69.9 FPS (28.9% reduction), and Ultra from 97.8 to 72.1 FPS (26.3% reduction). This pattern reveals that the GPU is the primary limiting factor at higher resolutions, as the frame rate decreases proportionally with pixel count. The CPU’s influence is most visible at 1080p, where the frame rates plateau — the difference between High (98.3 FPS) and Ultra (97.8 FPS) is just 0.5 FPS, suggesting the CPU is capping performance. At 4K, the GPU’s workload dominates, and the CPU becomes less relevant, as seen in the narrower spread between Low (90.6 FPS) and Ultra (72.1 FPS) compared to 1080p. The data indicates that this combo is best utilized at 1440p, where the GPU can stretch its legs without the CPU bottleneck becoming too severe. At 4K, the frame rates are still playable, but the 28.9% drop from 1080p High to 4K High shows that the GPU is working hard, and the CPU’s six cores are not the constraint there. The scaling from 1080p to 4K is not linear, but the consistent 20-29% reductions across settings point to a GPU-bound scenario at higher resolutions, with the CPU only mattering at 1080p and below.