The Technomancer

The pairing of AMD Ryzen 5 4600G and NVIDIA GeForce RTX 5090 has been measured in The Technomancer with four quality presets across three resolutions (1920x1080, 2560x1440, 3840x2160). The CPU carries a launch MSRP of $154, while the GPU's launch MSRP is $1,999. The data shows a system that is heavily GPU-dominant, with the RTX 5090 supplying most of the rendering power while the 4600G maintains a competent feed of frames. The combination ranks an above-average placement among tested combos, and every preset at 1080p sails past 100 fps. The analysis below breaks out each component's contribution, the effect of scaling from 1080p to 4K, and the best practical configuration.

Settings Recommendations

The High preset offers the most complete experience of the four measured settings in The Technomancer. At any row, High provides a solid combination of frame rate and expected quality without the sharp falloff seen when moving to Ultra. At 1920x1080 High averages 164 fps, while Ultra drops to 150 fps — a 10 fps penalty for often subtle visual gains. At 2560x1440, High delivers 142 fps versus 113 fps for Ultra, a 29 fps gap. At 3840x2160, the difference is even bigger: High sustains 79 fps, while Ultra falls to just 62 fps. The drop from High to Ultra accelerates with resolution, making the Ultra title unnecessary for fluid 4K play.

Comparisons to the Low preset show that Low does net a few more frames: at 1080p it hits 191 fps versus High's 164 fps, at 1440p 165 fps to 142 fps, and at 4K 121 fps to 79 fps. But the cost in graphical fidelity is considerable, and even the game's Medium preset provides more visual richness with a still-strong performance. Medium sits between Low and High: at 1080p Medium is 173 fps, at 1440p 150 fps, and at 4K 93 fps. Medium's only documented frame-time stability—the min/max figures are only available for Medium across all resolutions—shows a range of 147–199 at 1080p, a 128–173 span at 1440p, and 79–107 at 4K. The Max at 4K matches High's average, and the min at 4K is still 79 fps, which equals High's average at that resolution. Thus, except for the absolute highest framerate (Low) or the most realistic stalls (Ultra), Medium’s minimums are always acceptable.

Because the 4K Ultra average of 62 fps is the only sub-60 figure in the entire set, the High preset remains the best all-around knob. For 1080p or 1440p users, lower settings might be tempting, but the 60 fps barrier is never at risk even under the heaviest preset. A practical rule from the measured rows: choose High for 4K or Ultra for very high refresh‑rate 1080p/1440p displays where 150 fps both offer a smooth experience, but High secures the largest quality margin.

Reproduction of the frame phase

The RTX 5090's memory subsystem is enormous: 32 GB GDDR7 on a 512‑bit bus, with a bandwidth of 1.79 TB/s. The 1400’s 32 GB capacity is far beyond anything a 2016 Action RPG could use in a single scene; even a hypothetical worst-case texture bundle would be fully resident. The 1.79 TB/s transfer rate ensures that the massive amount of texture and geometry data from the 32 GB pool can be fed to the rasterizer without stalling. The clock speeds, base 2017 MHz, boost 2407 MHz, are the engine of the rasterizer. These clocks, combined with the compute core count (but we are not to cite SM count), are what unify 1080p Low to 196 FPS all the way to 4K Low 121 FPS.

In the benchmark rr of this GPU against bins of the same compute class, the GeForce RTX 5090 owns an average score of 79842 and sits at the 92nd percentile among all GPUs. Its closest rivals in average score are the NVIDIA Tesla P100 PCIe 16 GB ( avg 79605, delta 0.3%), the P100 12 GB (79396, delta 0.6%), the AMD Radeon RX 6850M XT (78940, delta 1.1%), and the AMD Radeon Pro Vega 64X (80959, delta -1.4%). These headends compare at a ±1% range, indicating that the RTX 5090 is in a strong position that its core execution is only complemented by the massively high‑bandwidth memory. In the Technomancer, the biggest VRAM constraint is Present: the game needs nothing close to 32 GB, but the bandwidth and memory clock still paid off at the 4K resolution, where the delta from 1440p to 4K is ~38–45% regardless of the preset.

##Resolution Scaling

When moving from 1080p to 1440p, the average drop in frames is between 11 and 13 % depending on the preset. Taking High as an example: 164 → 142 (a 13.4% decline); at Low: 191 → 165 (13.6%); Medium: 191 → 150 (13.3%), and Ultra: 150 → 113 (24.7% for Ultra only). So 1080p→1440p is almost a linear pixel–count scaling (1440p has 1.78× the pixels, but the physical discontinue of workload improved by approximately 24% in frame time, i.e. ~ 20% longer render time).

The far larger jump is from 1440p to 4K (3840×2160), which is for 4K about 2.25 pixels of raw pixels of 1440p. The measured fps drop for High: 142 → 79 (44.3% reduction); for Medium: 150 → 93 (38.0%); Low: 165→121 (36.4%); and Ultra: 113 → 62 (45.1%). The non‑linearity of these numbers implies the memory subsystem to the RTX 5090 becomes more of a pacing factor as the workload expands. All four presets lose a comparable fraction (≈40%) as the likely GPU-bound condition sets in. At 1080p, there might be a slight CPU-influence: the jump from 191 to 121 at low settings is far less steep, but the 160–191 peak still belongs to a high‑FPS region. The 4K ceiling of 121 fps from Low and 62 fps from Ultra demonstrates that the RTX 5090's compute and fillrate become the limiting factor at 4K, let the CPU's 6 cores idle. Thus, the resolution scaling behavior is typical of a GPU-bound game.

Measured FPS Breakdown

The table of measured values across the dataset (all rows from the JSON) yields the following specifics:

At 1920×1080: Medium – avgķī 173, min 147, max 199. Low – avg 191, High – 164, Ultra – 150. No min/max given for Low, High, Ultra.

At 2560×1440: Medium – avg 150, min 128, max 173. Low – 165, High – 142, Ultra – 113.

At 3840×2160: Medium – avg 93, min 79, max 107. Low – 121, High – 79, Ultra – 62.

The Medium presets are characterized by having the highest minimum and maximum values across the resolution scale, with the 4K min 79 fps being the lowest single recorded moment in the whole dataset. That matches the 4K High average of 79 fps, but the average Medium 93 fps, so the medium stays above even the High average. The absence of min/max data for other preset limits how exactly we can judge their stability, but the averages alone cover the general trend. Every row shows at least 60 fps, with the sole near-boundary being the 4K Ultra at zero margins and the 4K High equal to the 4K Medium minimum.

How This Combo Ranks

In the global set of 3696, this exact CPU–GPU pairing ranks at position #476. That puts it in the top 12.9% of all combos for The Technomancer. Since the CPU's own per-CPU percentile is 71 (i.e., better than 71% of all CPUs) and the GPU's percentile is 92, the combined score indicates that the GPU's high standing drives the overall rank, while the CPU is still well above average. The CPU's nearest performance-neighbors (according to average scores 17507) are AMD Ryzen 3 PRO 5355GE (17482, delta +0.1%), AMD Ryzen 3 8300GE (17614, –0.6%), Intel Core 7 150U (17395, +0.6%), an Intel Core i3-14100T (17648, –0.8%). These deltaPct can be read as the 4600G falls in the middle of these four, within ≤1%. The GPU's neighbors are the Tesla P100s and an AMD Radeon RX 6850M XT, but the actual games R&D has separate bench scores (79842 average), so it sits comfortably at the 92nd hotspot. The global rank of 476 out of 3696 is not a place you would expect from the CPU alone, indicating the GPU's strength pulls the system forward.

FAQ

Q: Why does Low preset deliver the highest FPS in all resolutions?

A: Low has the lowest graphical workloads, allowing more fps per preset: 191 at 1080p, 165 at 1440p, and 121 at 4K. High and Ultra trade off there; for instance, 4K Low (121) versus 4K Ultra (62) – a 39 fps gap.

Q: Does the Ryzen 5 4600G limit this GPU at 4K?

A: Only in the sense that the GPU is the hardest limit. At 4K, the composite row 4K Low 121, Ultra 62; the CPU is not under heavy strain at that pixel small. At 1080p, especially Low, the CPU could be a a few FPS, given that its single‑thread score (Cinebench R23, 1920 point) is still well above threshold.

Q: What is the largest drop in average FPS between settings?

A: The largest absolute drop between two consecutive settings is from High to Ultra at 1920p: 164 – 150 = a 14 fps drop, or at 1440p (142 to 113 = 29 fps), and at 4K – High to Ultra 79 − 62 = 17 fps. The largest percentage drop is 43% from 1440p High to 4K High (142 to 79).

Q: What memory technology does the GPU use and how much?

A: The NVIDIA GeForce RTX 5090 has 32 GB of GDDR7 across a 512-bit bus and 1.79 TB/s bandwidth.

Q: Are there any measured frames below 60 fps?

A: No, not a single average falls under 60. The lowest average is 62 fps at 4K Ultra. The lowest single min value is 79 fps (for medium at 4K).

Q: How does this combo's percentile compare to the CPU's and GPU's?

A: The combo is ranked 476 of 3696 (≈13.5% percentile), while the component percentile is 71 for the CPU and 92 for the GPU, showing the GPU is the primary driver.

CPU Role

The AMD Ryzen 5 4600G is a 6-core / 12-thread processor (Zen 2, Renoir) with a base clock of 3.70 GHz and a boost clock of 4.20 GHz. Its Ghost bearing is 65 TDP, and the L3 cache is 8 MB. In synthetic workloads it scores Single-Core Cinebench R20 805, Multi‑Cinebench R20 5709: in R23, single 1919, multi 13593. The more‐purely... its score on the Pi passmark is 23 WG... . Client multi‑thread data through iPS. Within the measured set, the CPU is the second-in-charming point: the frame rates at 1080p are higher than at 4K by a wide margin, and the fact that 1080p Low hits 191 fps shows the CPU can keep a healthy flow of draw calls to feed the RTX 5090. The GPU will be well over 90% implementation while the CPU sits below 50 % many cores at higher resolutions. The 8 MB L3 might be a slight constraint at 1080p low thread‑depending on the specific scene, but overall the 4600G's 6 cores are enough to push past the point where the FPS becomes irrelevant. The 4.20 boost ensures that the fastest cores can deliver the this game's (frequently single‑cored) rendering thread a decent single‑thread performance – the Single R3