Barony

Barony is a roguelike that, despite its genre, still puts measurable load on the hardware it runs on. The data for the AMD Ryzen 9 8945HX and AMD Radeon 890M combo shows a system that is decisively GPU-bound at higher resolutions, while the CPU's immense power becomes relevant primarily at 1080p. The measured FPS figures reveal a clear hierarchy of performance that shifts depending on where the bottleneck lies, making this a case study in how an integrated GPU scales against a flagship mobile processor.

Resolution Scaling

The transition from 1080p to 4K tells a clear story about where the performance ceiling is for this combination. At 1080p with High settings, the system produces 90 FPS. Moving to 1440p drops that to 57 FPS, and finally at 4K it falls to 30 FPS. That is a 33 FPS loss from 1080p to 1440p, and another 27 FPS loss from 1440p to 4K. The scaling is steep and consistent, which points directly at the Radeon 890M as the limiting factor. The GPU's compute resources are being saturated as pixel count rises.

The Low settings preset shows similar behavior. At 1080p Low, the system hits 141 FPS. At 1440p Low, it drops to 89 FPS. At 4K Low, it manages 47 FPS. Notice the pattern: the percentage drop from 1080p to 1440p is roughly 37%, and from 1440p to 4K it is roughly 47%. These are large, linear declines that are characteristic of a GPU that is running out of pixel throughput. The CPU is not the constraint here; if it were, we would see the FPS differences between resolutions compress significantly.

Look at the Medium preset for a more granular view. At 1080p Medium, the average is 105 FPS with a minimum of 89 and a maximum of 121. At 1440p Medium, the average falls to 66 FPS with a range of 56 to 76. At 4K Medium, the average is 35 FPS with a range of 29 to 40. The minimum and maximum values scale down proportionally with the average, which reinforces that the entire frame render is being throttled by the GPU. There is no CPU-related stutter or frame time inconsistency visible here; it is a smooth, predictable degradation as resolution increases.

The Ultra preset confirms the trend. High yields 90 FPS at 1080p, Ultra yields 65 FPS at the same resolution. At 1440p, High gives 57 FPS while Ultra gives 41 FPS. At 4K, High gives 30 FPS while Ultra gives 21 FPS. The delta between High and Ultra is consistent at roughly 25-30 FPS across all resolutions, which suggests that the extra settings in Ultra are adding a fixed amount of GPU work regardless of pixel count. This is a sign that the game's Ultra preset introduces effects that are not purely resolution-dependent, but the GPU is still the component that absorbs the cost.

FAQ

Q: Can this system run Barony at 4K with playable frame rates?

A: Yes, but only at Low settings. The data shows 47 FPS at 3840x2160 Low, which is above the 30 FPS threshold. At Medium, it drops to 35 FPS, and at High it falls to 30 FPS. Ultra is not viable at 4K, producing only 21 FPS.

Q: What is the best resolution for a smooth 60 FPS experience?

A: 1080p is the only resolution that consistently hits 60 FPS. At 1080p Low, the average is 141 FPS. At 1080p Medium, it is 105 FPS. At 1080p High, it is 90 FPS. Only Ultra fails to reach 60 FPS at 1080p, averaging 65 FPS.

Q: Is the Ryzen 9 8945HX holding back the Radeon 890M in this game?

A: At higher resolutions, no. The FPS drops from 1080p to 4K are steep and linear, which indicates the GPU is the bottleneck. The CPU's capabilities, such as its 16 cores and 32 threads, are not being fully utilized when the GPU is struggling to render frames at 1440p and 4K.

Q: How does the Ultra preset compare to High in terms of FPS cost?

A: The cost is significant. At 1080p, High gives 90 FPS and Ultra gives 65 FPS, a 25 FPS penalty. At 1440p, High gives 57 FPS and Ultra gives 41 FPS, a 16 FPS penalty. At 4K, High gives 30 FPS and Ultra gives 21 FPS, a 9 FPS penalty. The cost is larger at lower resolutions where the GPU has more headroom.

Q: What kind of frame time consistency can be expected on Medium settings?

A: The Medium preset has recorded minimum and maximum FPS values. At 1080p, the range is 89 to 121 FPS. At 1440p, the range is 56 to 76 FPS. At 4K, the range is 29 to 40 FPS. The spread between min and max is tight, suggesting stable frame pacing without major hitches.

Q: Is this a CPU-heavy or GPU-heavy game for this hardware?

A: The benchmark data indicates it is GPU-heavy for this specific combination. The CPU has a 95th percentile ranking among all CPUs, while the GPU has a 45th percentile ranking. The FPS scaling with resolution confirms that the GPU is the primary limiter, especially above 1080p.

GPU Role

The AMD Radeon 890M is the deciding factor in this benchmark. It operates with a base clock of 400 MHz and a boost clock of 2900 MHz, with 1024 shading units, 64 TMUs, and 32 ROPs. Its memory is System Shared, meaning it uses the same DDR5 pool as the CPU, and its bandwidth is marked as System Dependent. This is a mobile integrated GPU with a 15 W TDP, and its performance in Barony reflects that constraint.

The benchmark results place the 890M at the 45th percentile among all GPUs, with an average benchmark score of 9210. Its nearest rivals in the database are the AMD Radeon Vega 8, the NVIDIA GeForce GTX 960, the NVIDIA GeForce GTX 465, and the NVIDIA GeForce GTX 850M. The deltaPct values are tight: -0.1% versus the Vega 8, -0.7% versus the GTX 960, -0.9% versus the GTX 465, and -1% versus the GTX 850M. This places the 890M in a performance class that is roughly equivalent to a dedicated desktop GPU from a previous generation, which is notable for an IGP.

In Barony, the GPU's VRAM situation is worth noting. Since it uses System Shared memory, the amount of VRAM available is dynamic and depends on the system's total RAM and allocation. The fact that the GPU can produce 47 FPS at 4K Low suggests that the shared memory bandwidth is sufficient for the game's texture loads at lower settings, but the steep drop to 30 FPS at High settings indicates that the GPU's compute units, not memory, are the limiting factor. The pixel rate of 92.80 GPixel/s and texture rate of 185.6 GTexel/s are fixed by the clocks and ROP/TMU counts, and those are what the game is hammering.

The FP32 performance of 5.939 TFLOPS is modest by modern standards, and Barony's shader complexity at Ultra settings is clearly exceeding what the GPU can sustain at higher resolutions. The data shows that the 890M is capable of 141 FPS at 1080p Low, which is an excellent result for an IGP, but it cannot maintain that headroom as the pixel workload increases. The GPU is the component that dictates the playable settings envelope for this system.

Measured FPS Breakdown

The measured FPS data provides a full matrix of performance across three resolutions and four settings presets. At 1920x1080, the Low preset delivers an average of 141 FPS. The Medium preset delivers an average of 105 FPS, with a minimum of 89 FPS and a maximum of 121 FPS. The High preset delivers an average of 90 FPS. The Ultra preset delivers an average of 65 FPS. This shows that at 1080p, even the most demanding preset is playable, and the Low preset is well beyond the refresh rate of most displays.

Moving to 2560x1440, the FPS numbers drop across the board. Low averages 89 FPS. Medium averages 66 FPS, with a minimum of 56 FPS and a maximum of 76 FPS. High averages 57 FPS. Ultra averages 41 FPS. At this resolution, the Low and Medium presets remain playable for smooth motion, but High and Ultra require a tolerance for lower frame rates. The 1440p Medium result of 66 FPS is a strong showing for a 15 W IGP.

At 3840x2160, the performance becomes marginal. Low averages 47 FPS. Medium averages 35 FPS, with a minimum of 29 FPS and a maximum of 40 FPS. High averages 30 FPS. Ultra averages 21 FPS. The data shows that 4K is only viable for Low settings if you want to stay above 30 FPS. The Medium preset's minimum of 29 FPS indicates that even it will occasionally dip below the 30 FPS line, which could be noticeable in fast-paced roguelike combat.

The overall combo rank for this pairing in Barony is 2759 out of 2935 combos. That places it in the lower percentile of tested systems for this specific game, which aligns with the GPU being a mobile IGP. The CPU's high benchmark scores do not translate to a high rank in this game, because the game is not CPU-bound with this configuration.

Settings Recommendations

Given the measured FPS, the best experience depends on the target resolution. At 1080p, the High preset is the sweet spot. It delivers 90 FPS, which is smooth and responsive for a roguelike that demands quick reactions. Ultra is tempting at 65 FPS, but the 25 FPS penalty over High is significant, and the visual difference in a roguelike is often less impactful than the frame rate consistency. If you have a 144 Hz display, the Low preset at 141 FPS will make full use of it, but the visual fidelity loss may not be worth it for most players.

At 1440p, the Medium preset is the recommended choice. It provides 66 FPS, which is above the 60 FPS target, and it has a tight frame time range of 56 to 76 FPS. High at 57 FPS is close but falls short of the 60 FPS mark, and Ultra at 41 FPS is too low for comfortable play. Low at 89 FPS is an option if you prioritize frame rate above all else, but Medium offers a better balance between visuals and smoothness.

At 4K, the options are limited. Low at 47 FPS is the only preset that stays comfortably above 30 FPS. Medium at 35 FPS is playable but the minimum of 29 FPS means it will stutter in demanding scenes. High at 30 FPS is right on the edge, and Ultra at 21 FPS is not recommended. For 4K, the data suggests that Low is the only preset that provides a consistently playable experience, and even then, it is not a high-refresh-rate experience.

The data shows that the Medium preset is the most versatile. It is the only preset where minimum FPS values are recorded, and they show a consistent pattern: 89 FPS at 1080p, 56 FPS at 1440p, and 29 FPS at 4K. This suggests that Medium is a well-balanced preset that scales predictably. If you are willing to accept a variable frame rate, Medium at 1440p is the best compromise between resolution and smoothness.

CPU Role

The AMD Ryzen 9 8945HX is a 16-core, 32-thread processor based on the Zen 4 architecture, with a 2.50 GHz base clock and a 5.40 GHz boost clock. It has 64 MB of L3 cache and supports DDR5 memory. Its benchmark scores are exceptional: it sits at the 95th percentile among all CPUs, with a Cinebench R23 multi-core score of 42713 and a single-core score of 6030. In the database, it is ranked nearly identically to the Intel Core Ultra 9 285HX, with a deltaPct of 0.1% in favor of the Intel part. It also trades blows with the AMD EPYC Embedded 8224P and AMD Ryzen Threadripper PRO 9945WX, both at -0.4% delta, and the AMD Ryzen 9 9950X3D at 0.6% delta.

However, in Barony, this CPU power is largely irrelevant. The measured FPS data shows that the GPU is the bottleneck in almost every scenario. The evidence is in the resolution scaling. If the CPU were the limiter, the FPS would remain relatively flat as resolution increases, because the CPU's work per frame stays constant. Instead, the FPS drops sharply from 1080p to 4K, which means the GPU is running out of steam.

The CPU's role becomes more apparent at 1080p Low, where the system hits 141 FPS. At that point, the GPU is not fully saturated, and the CPU's high single-thread performance (6030 in Cinebench R23) is able to feed the GPU enough draw calls to sustain that frame rate. The 16 cores and 32 threads are overkill for a roguelike, but they ensure that there is no CPU-side frame pacing issue even at the highest settings. The data shows no minimum FPS drops that would indicate CPU stalls, which confirms the processor's headroom.

The memory bandwidth of 83.2 GB/s is relevant here because the GPU uses System Shared memory. Since the CPU and GPU share the same DDR5 pool, the CPU's memory controller and bandwidth can affect GPU performance. The fact that the 890M is able to reach 141 FPS at 1080p Low suggests that the shared memory bandwidth is not a limiting factor at lower resolutions, but at 4K High, the GPU's compute limits are hit before memory bandwidth becomes an issue. The CPU is a capable partner, but it is not the star of this benchmark. The Radeon 890M is the component that defines this system's gaming capabilities.