Lockdown Protocol

Lockdown Protocol, a first-person shooter released on 2024-03-11, was measured with the Intel Core Ultra X7 368H paired with the NVIDIA GeForce RTX 5090 Mobile, and the resulting dataset covers three resolutions and four preset tiers. The data is measured rather than estimated, which makes it possible to ask a sharper question than usual: does a flagship mobile GPU attached to a low-power mobile CPU deliver flagship performance, or does the 25 W processor clip its wings? The short answer is nuanced. Frame rates remain high at every tested combination — the worst result is 60 avg FPS at 4K Ultra — but the scaling pattern between 1080p and 4K suggests the CPU holds the pairing back at lower resolutions.

Settings Recommendations

The measured rows point to different "best" presets depending on the resolution, and the data invites a trade-off analysis rather than a single verdict.

At 1920x1080, Medium is the standout: 194 avg FPS with a measured floor of 165 and ceiling of 223. That is a tighter and more predictable experience than Low's 274 average, because Low buys raw throughput the CPU likely cannot fully exploit in an FPS where consistency matters. High at 1080p lands at 170 avg FPS — still comfortably above typical high-refresh territory — while Ultra drops to 127. For a competitive FPS, the Medium row is the recommendation: the 165 minimum indicates dips stay above common high-refresh thresholds.

At 2560x1440, Medium again posts the only full min/avg/max row: 121 / 142 / 164. High sits at 124 avg and Ultra at 93. Given that Medium's minimum of 121 nearly matches High's average of 124, Medium is arguably the sweet spot at 1440p — you sacrifice roughly 18 avg FPS versus High but gain a documented floor that stays above 120.

At 3840x2160, Ultra dips to 60 avg FPS, exactly at a common baseline threshold, while High delivers 80 and Medium 92 (78 min, 106 max). Medium at 4K therefore offers the best documented experience, with a minimum of 78 that stays clear of sub-60 territory. Low at 4K reaches 129 avg FPS if maximum smoothness is the goal.

Across all resolutions, one pattern repeats: Medium is the only preset where min and max were captured, and its floors — 165, 121, and 78 from 1080p through 4K — suggest the preset delivers predictable frame delivery where it was profiled.

GPU Role

The RTX 5090 Mobile is a large piece of silicon by mobile standards: the GB203 chip, built on TSMC's 5 nm process, packs 45,600 million transistors onto a 378 mm² die at a density of 120.6M per mm². The shader array counts 10,496 shading units, 328 TMUs, 112 ROPs, 82 RT cores, and 328 tensor cores, producing a pixel rate of 169.7 GPixel/s, a texture rate of 496.9 GTexel/s, and 31.80 TFLOPS of FP32 — with FP16 running at the same 1:1 rate.

Clocks are modest for a flagship: 990 MHz base, 1515 MHz boost, and memory at 1750 MHz (28 Gbps effective). The 24 GB of GDDR7 on a 256-bit bus yields 896.0 GB/s of bandwidth — a figure that comfortably exceeds the CPU-side memory bandwidth of 153.6 GB/s, an asymmetry worth noting when considering where data starvation might occur in this pairing.

Does the VRAM matter for Lockdown Protocol? The measured results indicate the game is not bandwidth- or capacity-bound at 4K High: 80 avg FPS is a respectable but not extraordinary number for a GPU of this class, and the drop from 1080p (170) to 4K (80) at High — a 53 percent decline — shows the GPU doing proportionally more work as pixels quadruple. The percentile field places this GPU at the 84th percentile against all GPUs, and its nearest rivals by average score include the NVIDIA GeForce RTX 4070 Ti (deltaPct 0.8), the NVIDIA RTX 5880 Ada Generation (deltaPct -1.8), the AMD Radeon Pro 5500 XT (deltaPct -0.5), and the Intel Arc A730M (deltaPct -1). In other words, aggregate benchmark standing puts it near desktop-class hardware, yet in this game the results look more like a strong mobile tier — which raises the question of whether Lockdown Protocol's workload, or the 95 W TDP envelope, is the limiter.

The API support — DirectX 12 Ultimate (12_2), Vulkan 1.4, OpenGL 4.6 — means no feature-tier limitations at any preset the game offers.

FAQ

Q: Can this combo run Lockdown Protocol at 4K?

A: Yes. At 3840x2160 the pairing averages 129 FPS on Low, 92 on Medium, 80 on High, and 60 on Ultra. Medium is the strongest documented choice, with a 78 minimum and 106 maximum.

Q: What is the highest frame rate measured in the dataset?

A: 274 avg FPS, at 1920x1080 on the Low preset. The 1080p Medium row peaks at 223 FPS.

Q: Is 60 FPS achievable at every resolution and preset?

A: Yes. The lowest average in the entire dataset is 60 FPS at 4K Ultra, so every measured combination clears that threshold on average.

Q: How does the CPU compare to its rivals?

A: By average benchmark score it sits in a near-dead heat with the Intel Core 7 253PE (deltaPct -0.1), Intel Core 5 223PE (deltaPct -0.2), AMD Ryzen 9 7940H (deltaPct 0.2), and Intel Xeon 6507P (deltaPct 0.2). It ranks in the 87th percentile against all CPUs.

Q: How much VRAM does the GPU have, and does the game need it?

A: The RTX 5090 Mobile carries 24 GB of GDDR7 with 896.0 GB/s of bandwidth. The measured scaling data gives no indication the game is VRAM-limited; performance falls predictably as resolution rises.

Q: Where does this CPU + GPU combination rank in Lockdown Protocol among tested combos?

A: Rank 770 out of 3714 tested combinations in this game.

Measured FPS Breakdown

The full measured table for the pairing:

| Resolution | Preset | Avg FPS | Min FPS | Max FPS |

|---|---|---|---|---|

| 1920x1080 | Low | 274 | — | — |

| 1920x1080 | Medium | 194 | 165 | 223 |

| 1920x1080 | High | 170 | — | — |

| 1920x1080 | Ultra | 127 | — | — |

| 2560x1440 | Low | 201 | — | — |

| 2560x1440 | Medium | 142 | 121 | 164 |

| 2560x1440 | High | 124 | — | — |

| 2560x1440 | Ultra | 93 | — | — |

| 3840x2160 | Low | 129 | — | — |

| 3840x2160 | Medium | 92 | 78 | 106 |

| 3840x2160 | High | 80 | — | — |

| 3840x2160 | Ultra | 60 | — | — |

Reading down each preset column reveals how much each step of fidelity costs. At 1080p, moving from Low to Ultra costs 147 FPS — more than half the Low result. At 1440p the Low-to-Ultra drop is 108 FPS, and at 4K it is 69 FPS. In relative terms, Ultra retains roughly 46 percent of Low's throughput at 1080p, 46 percent at 1440p, and 46 percent at 4K. That remarkably consistent ratio across resolutions implies the preset cost scales with pixel count in a near-linear fashion — a signature of a GPU-bound workload at higher settings, with no abrupt cliff where one setting change collapses performance.

The Medium min/avg/max spreads are also informative: 58 FPS wide at 1080p (165 to 223), 43 wide at 1440p, and 28 wide at 4K. The spread narrows as resolution climbs, which is characteristic of the GPU becoming the pacing component and smoothing frame-time variance, versus lower resolutions where CPU-driven variance has more room to express itself.

How This Combo Ranks

The pairing ranks 770th out of 3714 tested combinations in Lockdown Protocol — roughly the top fifth of the field, but not the top tier. That placement invites a curious contradiction: the GPU alone sits in the 84th percentile versus all GPUs, and the CPU in the 87th percentile versus all CPUs, yet the combination lands lower in this specific game than either component's individual standing might suggest.

Two hypotheses fit the data. First, Lockdown Protocol's ranking pool may be dominated by desktop combinations with higher-power CPUs; a 25 W, 16-thread mobile processor with a 2.00 GHz base clock is structurally different from the desktop field even when synthetic scores — Cinebench R23 multi-core 28221, single-core 3984, PassMark multithread 32956, single thread 4005 — are competitive. Second, the game itself may simply favor combinations with stronger per-thread headroom, and the near-flat 1080p Low results hint the CPU caps how fast frames can be issued regardless of GPU capability. The ranking data alone cannot settle which explanation dominates, but the resolution scaling section below adds evidence.

For context on component-level standing: the GPU's average benchmark score of 45152 places it within a percentage point of the Radeon Pro 5500 XT (45384), the RTX 4070 Ti (44795), and the Arc A730M (45592), and within two points of the RTX 5880 Ada (45972). The CPU's average of 40518 is effectively tied with its four nearest rivals, all within 0.2 percent. Individually competitive parts; collectively a mid-upper-table combo in this title.

CPU Role

The Core Ultra X7 368H is a Panther Lake-H part on Intel's 3 nm node, carrying 16 cores and 16 threads — one thread per core, no SMT doubling — with a 2.00 GHz base and 5.00 GHz boost, inside a 25 W TDP and an Intel BGA 2540 socket. Cache is generous per core: 192 KB of L1 and 2.5 MB of L2 per core, with 18 MB of shared L3. Memory support is dual-channel LPDDR5X at 153.6 GB/s of bandwidth, and PCIe connectivity is Gen 5 with 4 lanes from the CPU — a narrow link by discrete-GPU standards, though the GPU itself uses a PCIe 5.0 x16 bus interface on the platform side. The integrated Arc B390 graphics are, naturally, idle in this pairing.

How does this translate to Lockdown Protocol? The single-thread figures — 401 in Cinebench R15, 1673 in R20, 3984 in R23, 4005 in PassMark — describe a processor with strong burst throughput, and the 5.00 GHz boost clock explains it. FPS titles typically lean on a handful of fast threads for draw-call submission, and the 1080p Low result of 274 avg FPS represents the ceiling of what this CPU can feed the GPU. Notably, that ceiling is well below what a flagship-class GPU could theoretically render at 1080p Low; the flatness of Low across resolutions (274 at 1080p, 201 at 1440p, 129 at 4K) shows the frame rate falling as GPU load rises, which is expected — but the 1080p number itself is modest relative to the GPU's 3DMark Steel Nomad DX12 score of 5871 and Geekbench Vulkan score of 198405.

The multi-thread results — R15 2844, R20 11852, R23 28221, PassMark multithread 32956 and physics 2857 — are solid but describe throughput Lockdown Protocol is unlikely to fully use. The PassMark physics score of 2857 is the most game-relevant synthetic here, and it correlates with a capable but not exceptional simulation-side performer. The 87th percentile standing confirms the chip is strong for mobile; the 25 W envelope simply bounds how long and how hard those 16 cores can sustain load alongside a 95 W GPU in the same chassis.

Resolution Scaling

The scaling data tells the clearest story in the entire dataset. At the High preset, average FPS runs 170 at 1080p, 124 at 1440p, and 80 at 4K. From 1080p to 1440p — 78 percent more pixels — the frame rate falls 27 percent. From 1440p to 4K — a further 125 percent more pixels — it falls 35 percent. From 1080p straight to 4K, with 4 times the pixels, throughput drops 53 percent.

If the GPU were the sole limiter, a fourfold pixel increase would be expected to cut frame rates far more deeply. It does not. Instead, the modest declines indicate the RTX 5090 Mobile retains substantial headroom at 1440p and even 4K High, and that the frame rate is being propped up — capped, effectively — by something upstream. The CPU is the obvious candidate: as resolution rises, the GPU's per-frame cost grows while the CPU's stays roughly constant, so the measured FPS converges toward the CPU's delivery ceiling rather than collapsing toward the GPU's render limit.

The Medium rows reinforce this. Medium averages 194 at 1080p, 142 at 1440p, 92 at 4K — proportional declines of 27 and 35 percent, nearly identical to the High preset's pattern. Even Ultra, the heaviest tier, only drops 27 percent from 1080p (127) to 1440p (93) and 35 percent from 1440p to 4K (60). The uniformity of these ratios across presets is striking: whether the GPU is rendering a light or heavy workload, resolution scaling behaves almost identically. That is the signature of a CPU-limited regime at the lower resolutions, with the GPU only beginning to assert itself at 4K Ultra — where 60 avg FPS finally suggests the render cost, not the processor, is pacing the output.

The practical implication: raising resolution in Lockdown Protocol on this combo is nearly free in relative terms. A player moving from 1080p High to 4K High sacrifices 90 avg FPS but quadruples pixel density, and the 4K Medium row (92 avg, 78 min, 106 max) still outpaces many displays' refresh capabilities. Conversely, dropping resolution to chase frame rates yields diminishing returns — 1080p Low's 274 is the ceiling, and it is set by the Core Ultra X7 368H, not the GeForce RTX 5090 Mobile. For this game on this hardware, the data argues for playing at the highest resolution the display supports and tuning presets to taste; the CPU will set the pace either way.