Roblox

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

The AMD Ryzen 7 9700X is an 8-core, 16-thread desktop processor built on the Zen 5 architecture (Granite Ridge) and fabricated on a 4 nm process at TSMC. It operates with a base clock of 3.80 GHz and a boost clock of 5.50 GHz, with a 65 W TDP. The chip features a cache hierarchy of 80 KB L1 per core, 1 MB L2 per core, and a shared 32 MB L3 cache. It supports DDR5 memory over a dual-channel bus, providing 89.6 GB/s of bandwidth. This processor also includes Radeon integrated graphics, though the benchmark data here relies on the discrete GPU.

For a game like Roblox, which is typically not heavily threaded, the CPU’s single-core performance is often the deciding factor. The 3DMark single-thread score of 1280 and the Cinebench R23 single-core score of 4459 indicate strong per-thread capability. In contrast, the multi-threaded scores are also robust, with a Cinebench R23 multi-core result of 31588 and a Geekbench multi-core score of 17002. The PassMark single-thread score of 4654 further confirms that this CPU can handle the per-frame logic and physics that Roblox’s sandbox environments demand. The data shows that the processor sits at the 89th percentile versus all CPUs, with an average benchmark score of 36029. Compared to its nearest rivals, the Ryzen 7 9700X is effectively tied with the Intel Core i5-14500HX (0.2% delta) and slightly ahead of the AMD Ryzen 7 PRO 5845 (0.6% delta), while trailing the Intel Core Ultra 5 225H by -0.8%. This performance envelope means that in Roblox, the CPU is unlikely to be a bottleneck at any resolution, as the measured FPS remains high even at 1080p where CPU load typically scales.

How This Combo Ranks

The combination of the AMD Ryzen 7 9700X and the NVIDIA GeForce RTX 5070 achieves a combo rank of 358 out of 1800 tested combinations in Roblox. This places the pairing in the top 20% of all tested combos, which is a strong showing given the game’s relatively modest hardware requirements. The rank reflects the synergy between the 8-core CPU and the Blackwell-architecture GPU. While the CPU provides ample headroom for frame generation, the GPU’s raw throughput is what ultimately pushes the average FPS past the 300+ mark at lower resolutions. The data indicates that this combo is not the absolute top performer—there are 357 other combinations that rank higher—but it is far from a mid-pack result. The rank is likely influenced by the fact that Roblox can be sensitive to both CPU single-thread speed and GPU fill rate, and this pairing delivers on both fronts without a glaring imbalance. For context, the GPU alone sits at the 83rd percentile versus all GPUs, with an average benchmark score of 41687, and the CPU at the 89th percentile, so the combo rank of 358/1800 is consistent with two high-performing components working in concert rather than one dragging the other down.

Resolution Scaling

The measured FPS data shows a clear and steep decline as resolution increases from 1920x1080 to 3840x2160, which indicates that the GPU becomes the limiting factor at higher pixel counts. At 1080p with High settings, the combo produces 437.2 FPS, but this drops to 292.9 FPS at 1440p and further to 163.1 FPS at 4K. This represents a 62.7% reduction in FPS from 1080p to 4K at High settings, which is a significant scaling factor that points to heavy GPU workload. The same pattern is visible across all settings: at Low settings, FPS falls from 544.2 at 1080p to 363.1 at 1440p and 206.5 at 4K, a 62.1% drop. The Ultra settings show the largest absolute decline, from 351.7 at 1080p down to 133.6 at 4K, a 62.0% reduction. These numbers suggest that the CPU is not the primary constraint, as the FPS scaling is almost perfectly linear with the pixel count increase (4K has 4x the pixels of 1080p, but the FPS drop is roughly 2.3x to 2.6x, not 4x). This indicates that the RTX 5070’s memory bandwidth and shading throughput are what govern performance at higher resolutions, while the Ryzen 7 9700X has enough headroom to keep feeding frames even at 544 FPS at 1080p Low. The data implies that for competitive play at 1080p, the combo is heavily CPU-bound at times, but at 4K, the GPU is the clear bottleneck.

Measured FPS Breakdown

At 1920x1080, the combo delivers outstanding frame rates across all settings. With Low settings, the average FPS is 544.2, which is the highest recorded in the dataset. Medium settings yield 494.4 FPS, High settings produce 437.2 FPS, and Ultra settings drop to 351.7 FPS. The gap between Low and Ultra is 192.5 FPS, or roughly 35.4% of the Low value, showing that graphics settings have a substantial impact even at 1080p. These numbers are well above the refresh rate of any mainstream monitor, so the limiting factor here is likely the CPU’s ability to prepare frames, though the Ryzen 7 9700X’s strong single-thread performance keeps this from being a problem.

Moving to 2560x1440, the FPS figures remain high but show more separation between settings. Low settings average 363.1 FPS, Medium settings hit 331.5 FPS, High settings achieve 292.9 FPS, and Ultra settings fall to 236.1 FPS. The delta between Low and Ultra is 127.0 FPS, or 35.0% of the Low value. This scaling is consistent with the 1080p results, suggesting that the GPU is starting to take on more of the workload, but the CPU still has plenty of headroom.

At 3840x2160, the FPS drops to a range that is still playable but much lower. Low settings produce 206.5 FPS, Medium settings yield 189.7 FPS, High settings deliver 163.1 FPS, and Ultra settings bottom out at 133.6 FPS. The difference between Low and Ultra here is 72.9 FPS, or 35.3% of the Low value, which is nearly identical to the percentage drop at lower resolutions. This consistency indicates that the GPU’s shading units and memory bandwidth scale linearly with the workload, and the RTX 5070’s 12 GB of GDDR7 memory on a 192-bit bus with 672.0 GB/s bandwidth is sufficient to handle 4K without running out of VRAM, as evidenced by the fact that Ultra settings still exceed 133 FPS.

GPU Role

The NVIDIA GeForce RTX 5070 is built on the Blackwell 2.0 architecture using the GB205 chip, manufactured on a 5 nm process at TSMC with 31,100 million transistors on a 263 mm² die. It operates at a base clock of 2325 MHz and a boost clock of 2512 MHz, with a memory clock of 1750 MHz (28 Gbps effective). The GPU features 12 GB of GDDR7 memory on a 192-bit bus, providing a bandwidth of 672.0 GB/s. It has 6144 shading units, 192 TMUs, 80 ROPs, 48 RT cores, and 192 tensor cores, resulting in a peak FP32 throughput of 30.87 TFLOPS. The card is rated for a 250 W TDP and requires a 600 W power supply.

In Roblox, the GPU’s role is to handle the rendering of blocks, textures, and lighting effects, which are typically not as demanding as modern AAA titles but still benefit from higher fill rates. The RTX 5070’s pixel rate of 201.0 GPixel/s and texture rate of 482.3 GTexel/s are more than sufficient for the game’s stylized visuals. The measured FPS data shows that the GPU is the primary driver of performance scaling with resolution, as the FPS drops are consistent with the increase in pixel count. The GPU’s benchmark scores reflect its capability: a 3DMark Steel Nomad DX12 score of 5077, a Geekbench OpenCL score of 185269, and a PassMark G3D score of 29137. These place the GPU at the 83rd percentile versus all GPUs. Compared to its nearest rivals, the RTX 5070 is virtually tied with the AMD Radeon Pro 5300 (0.2% delta) and the NVIDIA GeForce RTX 3090 (0.6% delta), while slightly trailing the NVIDIA Tesla M40 (-0.5%) and the AMD Radeon Pro 580X (-0.7%). The data shows that the RTX 5070’s performance in Roblox is strong, but the GPU’s 12 GB VRAM is not a limiting factor here, as even at 4K Ultra the combo maintains 133.6 FPS, indicating that the card’s memory capacity and bandwidth are well-matched to the game’s requirements.