Inside

The Intel Core Ultra 9 290K Plus pairs with an NVIDIA GeForce RTX 2070 SUPER for the 2016 puzzle platformer Inside. The dataset provides a measurable snapshot of performance across three resolutions and four applied presets, offering a clear view of how this combination behaves in a low-ambient, low-fidelity title. With a combo rank of 2087 among 3716 tested configurations, this pair slots neatly into the middle tier of tested setups, and the measured FPS results reveal a load profile that tilts heavily onto the GPU once resolution scales upward. The remainder of this analysis interrogates those numbers — what the CPU brings, where the GPU takes over, and which settings yield the most sensible frame rate for a game that pushes subtle visuals rather than raw physics.

CPU Role

The Intel Core Ultra 9 285K Plus is a desktop 24-core, 24-thread Arrow Lake Refresh part built on a 3 nm TSMC process, with a base clock of 3.70 GHz and a boost clock of 5.80 GHz. It packs 36 MB of shared L3 cache, 3 MB of L2 per core, and a 192 KB L1 allocation per core. The data records it as having a 125 W TDP and a 243 mm² die, with a transistor count of 17,800 million and a memory bandwidth of 115.2 GB/s across a dual-channel DDR5-4800 interface. In the synthetic benchmarks, the CPU posts a Cinebench R23 multi-core score of 51,731 and a single-core score of 7,303, along with a Passmark multithread score of 60,860 and a Passmark single-thread score of 4,823.

None of those synthetic workloads directly map to Inside, which is a $28 puzzle platformer with simple environments and no physics-heavy interactions. However, the measured FPS data shows a pattern: at 1920x1080 with Low settings, this combo sustains an average of 310 FPS, and even at 4K Low it holds 140 FPS. The CPU is clearly not the limiting factor at any resolution in these rows. If the CPU were stalling, the 1080p and 1440p rows would show much less headroom. The average results at 1440p Low (223 FPS) and 4K Low (140 FPS) indicate the graphics pipeline handles the scaling. The processor supplies enough thread and clock headroom to avoid becoming a blocker — the 3.70 GHz base is ample for a 2016 launch title, and the 5.80 GHz boost only helps preserve a steady minimum frame window at lower resolutions where the GPU has overhead.

Its nearest rivals in the broader CPU database all score close to each other. The listed delta percentages show the 285K is 0.2 percent ahead of the Intel Core Ultra 9 285K, 1.1 percent behind the AMD EPYC 4578 multi-core score, and 1.2 percent behind the EPYC 9R, while sitting 1.3 percent above the EPYC 7F72. This is a high-end desktop part that overperforms for a game like Inside, but the actual measured FPS differences stem from the GPU side, as the aggregate data shows no CPU-bound ceiling below 310 FPS at any resolution.

FAQ

Q: What is the highest average FPS recorded for Inside with this CPU/GPU combo?

A: The highest measured average FPS is 310 FPS, achieved at 3840×2160 Low and 1920×1080 Low. Wait, purely the 3840x2160 Low row shows 140 FPS, not 310. The highest is 1920×1080 Low at 310 FPS. The 4K Low row shows 140 FPS.

Q: How does the CPU rank against its nearest rivals?

A: In the average benchmark score, the Core Ultra 9 285K Plus surpasses the Intel Core Ultra 9 285K by only 0.2 percent (84003 vs 83807). It also beats the AMD EPYC 4584X (+1.1%), but trails the EPYC 9135 (−0.2 percent) and is beaten by the EPYC 7F72 (−0.7 percent). None of these differences exceed 1.3 percent, indicating an even playing field among the top processors.

Q: What is the minimum FPS at 4K Ultra settings?

A: The 3840×2160 Ultra row reports a min FPS of 60, but the actual measured value shows 67 FPS for average. The pack does not provide a minFPS for that row; it only lists the average as 67 FPS. Looking at the Medium row at 4K, the min FPS is given as 95 FPS (with a max of 129 FPS). The Ultra row has no min/max data, so I can only state the average.

Q: Is the GeForce RTX 2070 SUPER a bottleneck at 1080p?

A: No. At 1920×1080 with Low settings, the measured average is 310 FPS, and at Medium it drops to 248 FPS. The 2080 SUPER is capable of hitting 300+ FPS in this game, so the FPS count is limited mostly by the frequency of rendering frame time, and the CPU remains untapped.

Q: What is the combination’s rank among all tested combos?

A: The combo rank is 2087 out of 3716 tested configurations. That places it in the middle percentile of the pool, indicating that while it achieves ple duringthe game, it is not an outlier in performance for Inside.

Resolution Scaling

The transition from 1920×1080 to 3840×2160 demonstrates a steady workload shift from the CPU to the GPU. At Low settings, the average FPS drops from 310 (1080p) to 223 (1440p) to 140 (4K) — that's a 1.33 drop from 1080p to 1440p, then a 1.59 factor drop from 1440p to 4K. The differing ratios at each step suggest that the render resolution adds a uniform workload burden, but the gap between 1080p and 4K on Medium shows a decay from 248 to 112, which is a 2.21 step. The high overhead of 112 FPS at 4K Medium is still well above the 60-cathode, and the Ultra settings capture a 147 FPS 1080p, 106 FPS 1440p, and 67 FPS 4K.

The data implies that the 1080p row is likely CPU-headroom-limited because the 24-thread processor has 310 FPS to spare, but the 4K rows drop to single-digit FPS at High (70) and Ultra (67). High and 4K, an 449 average figure of 139 FPS at 1440p likely reflects the GPU's pixel output, not the CPU’s shader or AI calculations. Since the GPU holds 8 GB of memory and a 256-bit bus (448.0 GB/s bandwidth), the 4K high row remains strong. There is no evidence that the CPU degrades in this pressure curve: the single-core boost of 5.80 GHz provides more than enough peak time to keep the 140 FPS threshold at 4K Low.

How This Combo Ranks

Placing 2087th within a set of 3716 tested combos means that 2086 configurations perform better for the same game, while 1630 of combos score lower. This places the pair in the third quartile of the overall ranking. The leading combos are higher the same scale; the RTX 3070 Mobile unit is only 1.6% faster, and the RTX 2070 SUPER slots just north of a Quadro M4000 variant. The combo tier is dominated by cards with similar Pascal/Turing architectures, so that rank generally matches expectations for a 2-generation-old GPU and an over-kill CPU which still runs the game fine. The data shows that the 8 GB memory capacity and 512 GB/s bandwidth are not strained, but the 4K rows show the GPU is the cost bottleneck that halts the FPS increase.

Measured FPS Breakdown

The measurements cover 12 data points across 3 resolutions and 4 preset qualities.

• 1920×1080: Low → 310 avg. Medium → 248 avg (min 210, max 285). High → 155 avg. Ultra → 147 avg.
• 2560×1440: Low → 223 avg. Medium → 178 avg (min 151, max 205). High → 111 avg. Ultra → 106 avg.
• 3840×2160: Low → 140 avg. Medium → 112 avg (min 95, max 129). High → 70 avg. Ultra → 67 avg.

The medium preset is the only one to record a full min/max range, and all three resolutions share a close gap between High and Ultra — for instance 1080p High (155) vs Ultra (147) has a 8-FPS difference, while 4K High (70) vs Ultra(67) has a 3-FPS offset. Low presets at 1080p and 1440p reach the 300+ FPS territory, suggesting that the frame are render-constrained only.

Settings Recommendations

If the goal is to maintain a stable 60+ frames per second on a high refresh monitor — or simply to avoid hitching in a puzzle game — the Medium preset at all three resolutions is formal best bet. At 4K Medium, the measured average is 112 FPS, double the Ultra frame rate of 67, and it provides a documented frame range of 95–129. The jump from Low to Medium costs only 31 FPS on average at 1080p (310-248), but visually the render budget is still generous. For a 1440p monitor, Medium reaches 178 FPS, also above any 144 Hz display, while 4K at 112 FPS is ideal for a high refresh, low-detail action title.

Ultra only makes sense when the equipment is specifically chosen as a retro League-of-Legends shortcut: 147 FPS at 1080p still flies, but the 4K Ultra (67 FPS) dips into a range that, while adequate for a puzzle title, won't satisfy a dedicated tournament. The measured data shows that High presets are a compromise — 155 FPS at 1080p and 70 FPS at 4K — but the additional visual quality does not justify the 45-50% jump in frame output over Medium within the rendering resolution.

GPU Role

The NVIDIA GeForce RTX 2070 SUPER runs a 12 nm TU104 die with 2560 shading units, 160 texture mapping units, and 64 ROP raster output. It ramps to a boost of 1770 MHz, outputs 1600 MHz base, and stores 8 GB of GDDR6 across a 256-bit path at 1750 MHz (14 Gbps effective), giving a bus of 448.0 GB/s. Its raw TFLOPS figure reaches 9.062 and 12.0 FP32, with a complex 40 W budget of 215 W and a 550 W equal-power supply prescript. Its percentile rank stands at 65 among all GPUs.

In Inside, the GPU is demonstrably the performance scissor. At Low settings, the 4K row reaches 140 FPS, which drops to 112 on Medium and 67 on Ultra. The 4-chip scaling from 1080p (310) to 4K Low (140) is a ratio of 2.2 times — consistent with a pure pixel-wide fillrate increase. At Medium, the gap between 1080p (248) and 4K (112) is exactly linear by pixel count (4x pixels -> 4x load). The 72 FPS difference across the other rows also holds up. The RTX cards’ differences with rivals (the Quadro M4000) are tiny – the current GPU is 1% behind the average of the Quadro M6000 in synthetic benchmark, but the actual FPS chiffres are measurable and correct. Despite the extensive RT core count (40) and tensor cores (320), the 2.0 updated game has no instance of these features in the data — it runs on the fixed-function pipeline.

Overall, the 8 GB memory can handle Inside’s minimal texture overhead without pressure. The FPS table falls continuously from 310 to 140 across resolutions on Low, which indicates the GPU's fill-rate is what caps each headline — high card face.

That analysis combines synthetic numbers with the measured game performance, giving a useful map for a system trying to hit 144 Hz at 1440p or 30 FPS at 4K.