Doom

CPU Role — cores, clocks, and how they relate to this game's results

The Intel Core Ultra 9 185H is a 16-core, 22-thread mobile processor built on Intel's Meteor Lake architecture using a 7 nm process. It carries a base clock of 3.90 GHz and a boost clock of 5.10 GHz, with a 45 W TDP typical of high-end mobile parts. The cache hierarchy consists of 112 KB L1 per core, 2 MB L2 per core, and 24 MB of shared L3. Memory support is DDR5, dual-channel, with a bandwidth of 89.6 GB/s. The CPU's integrated graphics are Arc Xe-LPG Graphics with 128 execution units, though the measured frame rates in Doom are driven by the discrete GPU in this combo.

The CPU's synthetic benchmark profile places it in the 85th percentile among all CPUs. Its Cinebench R23 multi-core score is 17984.5, while the single-core score is 1810.5. Geekbench results show 11657 multi-core and 1794 single-core. Passmark multi-thread performance is 29376, with a single-thread score of 3697. The average benchmark score across all tests is 35670. Its nearest rivals in aggregate scoring are the AMD Ryzen AI 7 PRO 350 (35719, delta -0.1%), Intel Core 7 250H (35728, delta -0.2%), AMD Ryzen 7 PRO 5845 (35802, delta -0.4%), and AMD Ryzen 7 7700X (35909, delta -0.7%). These are extremely tight margins — the largest gap among the listed rivals is 0.7%, meaning the Ultra 9 185H sits in a cluster of similarly performing processors.

For Doom, a game from 2016, the CPU demands are modest relative to modern titles. The data shows that at 1080p Low, the combo reaches 227 FPS average, which is far beyond display refresh rates of typical monitors. The fact that the CPU is not the bottleneck at lower resolutions is evident from the scaling pattern: as resolution rises to 1440p and 4K, frame rates drop substantially, which points to GPU-bound behavior. The CPU's 16 cores and 22 threads provide ample headroom for the game's main thread and any background tasks, and the high single-core boost of 5.10 GHz ensures that the game's primary simulation thread is well served. In benchmark terms, the single-core Cinebench R23 score of 1810.5 is strong for a mobile chip, and Passmark single-thread at 3697 confirms that per-core performance is not a limiting factor in this title.

GPU Role — VRAM, clocks, and how they relate to this game's results

The NVIDIA GeForce RTX 4070 is a desktop-class GPU built on the Ada Lovelace architecture with a 5 nm process at TSMC. It uses the AD104 chip with 35,800 million transistors on a 294 mm² die. The GPU has 5888 shading units, 184 texture mapping units, 64 ROPs, 46 ray tracing cores, and 184 tensor cores. Clock speeds are a base of 1920 MHz and a boost of 2475 MHz. Memory is 12 GB of GDDR6X on a 192-bit bus, with a bandwidth of 504.2 GB/s and an effective memory clock of 21 Gbps. The pixel rate is 158.4 GPixel/s, texture rate is 455.4 GTexel/s, and FP32 compute is 29.15 TFLOPS. The card draws up to 200 W and is rated for a 550 W system power supply. It supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

The GPU's benchmark results place it in the 81st percentile among all GPUs. Its 3DMark Steel Nomad DX12 score is 3854. Geekbench OpenCL is 154858 and Vulkan is 174152. Passmark G3D is 26927, with a GPU compute score of 14720. The average benchmark score is 37648. Nearest rivals in aggregate scoring include the NVIDIA Tesla P4 (37628, delta 0.1%), AMD Radeon RX Vega 56 (37507, delta 0.4%), NVIDIA GeForce RTX 4080 Mobile (38135, delta -1.3%), and AMD Radeon PRO W6400 (37157, delta 1.3%). The RTX 4070 is essentially at parity with the RTX 4080 Mobile in average benchmark score, trailing by 1.3%, and is slightly ahead of the Vega 56 by 0.4%.

In Doom, the GPU's role is dominant. The measured frame rates show a clear GPU-bound pattern. At 4K Ultra, the combo averages 48 FPS, while at 1080p Low it reaches 227 FPS. The 12 GB VRAM is more than sufficient for a 2016 title, even at 4K Ultra with high-resolution textures. The 504.2 GB/s memory bandwidth supports the game's streaming requirements without issue. The data indicates that the RTX 4070's raw throughput, as measured by its 29.15 TFLOPS FP32 and 455.4 GTexel/s texture rate, is the primary determinant of frame rates in this title, particularly at higher resolutions and settings. The GPU's 81st percentile ranking among all GPUs aligns with the measured performance: capable of high refresh rates at 1080p and 1440p, and playable frame rates at 4K with adjusted settings.

FAQ

Q: What is the highest average frame rate recorded for this combo in Doom?

A: The highest average frame rate is 227 FPS, achieved at 1920x1080 with Low settings.

Q: How does the Intel Core Ultra 9 185H compare to its closest rival in average benchmark score?

A: The AMD Ryzen AI 7 PRO 350 scores 35719, which is 0.1% higher than the Ultra 9 185H's 35670. The Intel Core 7 250H scores 35728, 0.2% higher. The AMD Ryzen 7 PRO 5845 scores 35802, 0.4% higher, and the AMD Ryzen 7 7700X scores 35909, 0.7% higher.

Q: What is the frame rate difference between 4K Ultra and 4K Low settings?

A: At 3840x2160, Ultra settings average 48 FPS while Low settings average 80 FPS. This is a 32 FPS difference between the two extremes.

Q: Does the RTX 4070 outperform the RTX 4080 Mobile in average benchmark score?

A: No. The RTX 4080 Mobile has an average score of 38135, which is 1.3% higher than the RTX 4070's 37648.

Q: What is the combo's rank among all tested combos in Doom?

A: The combo ranks 2033 out of 3712 tested combos in Doom.

Q: What is the minimum FPS recorded in any measured row?

A: The only rows with min FPS values are Medium settings. At 3840x2160 Medium, the min FPS is 59. At 2560x1440 Medium, the min FPS is 109. At 1920x1080 Medium, the min FPS is 168. The lowest min FPS across all rows is 59 at 4K Medium.

How This Combo Ranks

The combo of Intel Core Ultra 9 185H and NVIDIA GeForce RTX 4070 ranks 2033 out of 3712 tested combos in Doom. This places it in the middle of the distribution — approximately the 55th percentile of all tested combinations. The rank reflects the combined capabilities of both components, but in a game like Doom, the GPU carries most of the performance weight.

The CPU's 85th percentile ranking among all CPUs is notably higher than the combo's overall rank, which suggests that the CPU is not the limiting factor in this pairing. Conversely, the GPU's 81st percentile ranking among all GPUs is also above the combo's rank, indicating that the combination's mid-pack position in Doom is influenced by the game's specific engine characteristics and the tested settings mix rather than by either component being weak.

The measured frame rates support this interpretation. At 1080p Ultra, the combo delivers 137 FPS, which is well above playable thresholds. At 1440p Ultra, it delivers 89 FPS. At 4K Ultra, it drops to 48 FPS. These numbers suggest that the combo is well balanced for 1080p and 1440p gaming, but at 4K with maximum settings, the GPU reaches its limits.

Compared to the CPU's nearest rivals, which all score within 0.7% of the Ultra 9 185H, the CPU choice makes little difference in Doom's performance. The GPU's nearest rivals similarly cluster within 1.3%, meaning the RTX 4070 is positioned among a tightly grouped set of GPUs. The combo's rank of 2033 out of 3712 reflects this mid-pack positioning: neither component is exceptional within its own class, but together they deliver solid performance in this title.

Measured FPS Breakdown

The measured FPS data covers three resolutions and four settings levels, with average FPS recorded for all rows and min/max FPS recorded only for Medium settings.

At 1920x1080, the combo delivers its highest frame rates. Low settings average 227 FPS. Medium settings average 198 FPS, with a min of 168 and max of 228. High settings average 167 FPS. Ultra settings average 137 FPS. The spread from Low to Ultra at 1080p is 90 FPS, showing that settings have a significant impact even at the lowest resolution tested.

At 2560x1440, frame rates remain strong. Low settings average 147 FPS. Medium settings average 128 FPS, with a min of 109 and max of 148. High settings average 108 FPS. Ultra settings average 89 FPS. The spread from Low to Ultra at 1440p is 58 FPS. The step from 1080p to 1440p costs roughly 35% of the frame rate at each settings level: Low drops from 227 to 147 (80 FPS loss), Medium from 198 to 128 (70 FPS loss), High from 167 to 108 (59 FPS loss), and Ultra from 137 to 89 (48 FPS loss).

At 3840x2160, frame rates fall to their lowest measured values. Low settings average 80 FPS. Medium settings average 69 FPS, with a min of 59 and max of 80. High settings average 59 FPS. Ultra settings average 48 FPS. The spread from Low to Ultra at 4K is 32 FPS. The step from 1440p to 4K costs roughly 45-46% of the frame rate at each settings level: Low drops from 147 to 80 (67 FPS loss), Medium from 128 to 69 (59 FPS loss), High from 108 to 59 (49 FPS loss), and Ultra from 89 to 48 (41 FPS loss).

The Medium settings rows are the only ones with min and max values. At 1080p Medium, the min of 168 and max of 228 bracket the average of 198, with a 60 FPS range between min and max. At 1440p Medium, the min of 109 and max of 148 bracket the average of 128, with a 39 FPS range. At 4K Medium, the min of 59 and max of 80 bracket the average of 69, with a 21 FPS range. This narrowing range at higher resolutions further confirms the GPU-bound nature of the workload — as resolution increases, frame time variance compresses because the GPU is consistently saturated.

Settings Recommendations

The measured data shows a clear hierarchy across settings presets. At every resolution, Low delivers the highest average frame rate, followed by Medium, then High, then Ultra. The ordering is consistent, and the relative gaps between settings are informative.

At 1080p, all four settings deliver playable frame rates. Low averages 227 FPS, Medium 198 FPS, High 167 FPS, and Ultra 137 FPS. Even Ultra at 1080p exceeds 120 FPS, making the highest preset a viable choice for gamers with high-refresh displays. Medium offers a 31 FPS advantage over Ultra while maintaining a 198 FPS average, and High splits the difference at 167 FPS. For competitive play, Low's 227 FPS is the clear choice, but for visual fidelity, Ultra's 137 FPS remains well above the 60 FPS threshold.

At 1440p, the recommendations shift. Ultra averages 89 FPS, which is still above 60 FPS but below 120 FPS. High averages 108 FPS, which crosses the 100 FPS threshold. Medium averages 128 FPS, and Low averages 147 FPS. The data suggests that High is the balanced choice at 1440p: it delivers 108 FPS average, which is smooth on 100 Hz or 120 Hz displays, while retaining most of the visual quality of Ultra. Medium's 128 FPS is the better choice for gamers prioritizing frame rate over visual detail, and Low's 147 FPS is for maximum responsiveness.

At 4K, the situation changes. Ultra averages 48 FPS, which falls below the 60 FPS standard for smooth gameplay. High averages 59 FPS, just shy of 60. Medium averages 69 FPS, clearing the 60 FPS threshold. Low averages 80 FPS, providing comfortable headroom. The data indicates that Medium is the recommended preset at 4K: it delivers 69 FPS average with a min of 59, meaning the minimum frame rate stays at the edge of playable. Low at 80 FPS provides more margin and is preferable if frame pacing is a concern. High's 59 FPS average is borderline, and Ultra's 48 FPS falls short of the standard.

Across all resolutions, the Medium preset offers the most balanced experience when considering the min FPS data. At 1080p Medium, the min is 168 FPS; at 1440p Medium, the min is 109 FPS; at 4K Medium, the min is 59 FPS. These minimums are all within a reasonable range of their averages, suggesting consistent frame delivery. The High and Ultra presets lack min FPS data in the measured rows, so their frame pacing cannot be assessed from this dataset.

Resolution Scaling

The frame rate progression from 1080p to 4K reveals the limiting component in this combo. At 1080p Low, the combo averages 227 FPS. Dropping to 1440p Low reduces this to 147 FPS, a 35% decrease. Dropping to 4K Low reduces it to 80 FPS, a 46% decrease from 1440p and a 65% decrease from 1080p. This scaling pattern is characteristic of GPU-bound performance: the frame rate drops roughly in proportion to the increase in pixel count.

The same pattern holds at every settings level. Medium: 198 FPS at 1080p, 128 FPS at 1440p (35% drop), 69 FPS at 4K (46% drop from 1440p). High: 167 FPS at 1080p, 108 FPS at 1440p (35% drop), 59 FPS at 4K (45% drop from 1440p). Ultra: 137 FPS at 1080p, 89 FPS at 1440p (35% drop), 48 FPS at 4K (46% drop from 1440p). The consistency of these percentage drops — approximately 35% from 1080p to 1440p and approximately 45-46% from 1440p to 4K — is a strong indicator that the GPU is the primary limiter.

If the CPU were the bottleneck, the frame rate would plateau at lower resolutions, with diminishing returns as resolution decreases. Instead, the frame rate continues to rise at 1080p, reaching 227 FPS at Low settings, which is far beyond what a CPU-bound scenario would produce in a 2016 title. The CPU's high single-core performance, as measured by its Passmark single-thread score of 3697 and Cinebench R23 single-core score of 1810.5, provides sufficient headroom at all tested resolutions.

The min FPS data at Medium settings further supports the GPU-bound conclusion. At 1080p Medium, the min is 168 FPS; at 1440p Medium, the min is 109 FPS; at 4K Medium, the min is 59 FPS. The minimum frame rates scale with resolution in the same proportion as averages, which is expected when the GPU's rendering workload scales directly with pixel count. The 4K Medium min of 59 FPS is notable — it means that even the minimum frame rate at this resolution and settings level stays at the 60 FPS playability threshold, while the average of 69 FPS provides a small buffer.

The resolution scaling data also shows that the RTX 4070's 12 GB VRAM and 504.2 GB/s bandwidth are sufficient across all tested configurations. If VRAM capacity were exceeded at 4K Ultra, the frame rate would show a disproportionate drop relative to the pixel count increase. Instead, the 4K Ultra average of 48 FPS is consistent with the scaling pattern observed at other settings levels, indicating that memory capacity is not a limiting factor in this title.