Performance is lower than anticipated on the latest release.
Performance is lower than anticipated on the latest release.
I've put together my PC about two weeks back. The issue is that I'm seeing lower average frames than expected compared to benchmarks on YouTube for those games. I looked for tests focusing on the processor and found they showed around 90fps with a 3060. All tests were in 1080p. I've tried it on Spiderman Remastered and Cyberpunk 2077. On Spiderman, I barely notice changes when adjusting settings like crowd density or LOD. I usually get low FPS when moving quickly through the city (around 40 to 50). This happens with everything on Max settings except Raytracing on High (no DLSS used; DLSS didn't help much). In Cyberpunk, I see low FPS in the 40s near crowds and in some areas even at low density. On high settings, RT at 1080p is good, but DLSS quality didn’t make a big difference. My specs are: motherboard MSI B660M, DDR4 Mortar, BIOS E7D42IMS.110 (dated 12/13/2021), CPU Intel i5 12400F, 16GB DDR4 Patriot Viper RGB, 1000W PSU, 3080 Ti GPU, Samsung 980 PRO Evo case, Corsair 4000D airflow fans, Noctua NH-u12A fan, Cinebench scores around 1725 (single core) and 12428 (multicore). For Heaven Benchmark at max settings: 5878 FPS, 34.1 FPS minimum, 487.2 FPS maximum. A custom test near 4K showed 3803 FPS with 33.1 FPS min and 315.1 FPS max. I believe the benchmark results are normal. CPU temps reach 60°C, GPU temps up to 80°C. I'm wondering if my hardware isn't performing well or if something else is affecting the numbers. Is having only one RAM stick instead of dual channel really impacting performance? Could an older BIOS version be the reason for the low FPS? Any advice or further tests you'd recommend?
12400 causes a significant slowdown on the 3080 TI at 1080p, especially with just one stick of RAM. For optimal performance you should use dual-channel RAM, meaning two sticks to fully utilize the system. Higher-end CPUs paired with 3080s and above will struggle more in 1080p, particularly when targeting 4K resolution. At a minimum, aim for 1440p to see better results. If you want to verify this yourself, turn on DSR and run games in 4K—your GPU utilization should increase, leading to higher FPS.
I'm not sure about that. Your chip likely isn't the main issue, but using today's high-end chips with just one RAM stick can hurt performance significantly. Depending on the game, you might lose 10 to 35% of frames by cutting memory bandwidth. These CPUs were built for dual-channel operation; single sticks work but aren't ideal. I suggest adding another matching stick or swapping them for a faster setup. If you can afford it, consider a 13th-gen chip and make sure the BIOS is up to date. That said, the game isn't optimized well, so you'll need to tweak settings yourself. Some players prefer lower fidelity for higher FPS, while others enjoy detailed visuals even at 30-45 FPS. Be wary of claims about high FPS in videos—some are misleading or designed to push sales. Certain clips seem to exaggerate performance on cheaper cards like the Intel 4500 Graphics.
I think adding just one RAM stick on a 12400 at 1080p won’t improve performance much compared to a GPU over 3060 since you’re mainly limited by CPU and RAM speed. Try upgrading to another stick and run at 1440p, which matches the optimal setting for your system.
Some notes on your situation:
- I understood that my GPU would work best for 2k games when everything is set to maximum, including RT. That was the main aim.
- Right now I’m stuck with a 1080p display, which limits performance.
- The extra RAM stick could help, but its impact might be limited—usually around 10-20% at most. With XMP enabled and running at 3200 Hz, it shouldn’t be that significant.
- I tried upscaling to 2880x1800 using FSR, but FPS dropped from around 30 to 50 or less depending on resolution.
- It seems lower FPS on 1080p compared to higher resolutions isn’t unusual; most tests confirm this.
- Benchmarks show similar results for CPUs with dual channels and high-end GPUs, often reporting 60+ fps in ideal conditions.
- Your concern about seeing higher FPS numbers elsewhere is valid—your results don’t match the general expectations.
Hi, I've already purchased it. The goal is to refresh everything except the GPU in the future. I'm wondering if anyone noticed something or suggested a possibility that might explain why FPS is much lower than what most users experience (most think a single stick would help).
It was DSR to correct the misspelling, it wasn't done directly from Ge force experience, though I’ll attempt it now. Yes, he used a 3090 Ti specifically to evaluate CPU performance. That video serves as an example of a benchmark where the 12400F didn’t underperform much. You might say the 20-30 fps gap is due to GPU limitations in a CPU bottleneck situation like that one. I’m not sure I’d be able to run both at maximum 60 fps on 1080p even with two sticks. There are many videos comparing single vs dual GPUs, but the difference usually isn’t this large—it depends heavily on the game. I’ll update once I have another RAM stick.
The problem involved using just one stick of RAM. I tested Spider-man near Times Square and performance improved by 40% overall and 1%. Cyberpunk benchmarks showed a 40% jump at 1% lows and 25% average FPS. Running through the city added nearly 50% at 1% lows (up from about 35 to 58). Cinebench and Heaven Benchmark results weren't much better, but I'm satisfied with what I got. Cyberpunk average is 97, with 77 at 1% and 57-65 fps in 1%. Spiderman rapid swings around Times Square averaged 74 fps at 1%, about 61 fps overall.