Scores dropped sharply in the homebuild system recently—any issues?
Scores dropped sharply in the homebuild system recently—any issues?
Hi everyone, roughly two weeks ago I assembled a new setup. I managed to get a Ryzen 9800X3D, a fresh motherboard, and some new RAM. The rest came from my old machine. My configuration included the Ryzen 7 9800X3D Gigabyte X870E, AORUS ELITE WIFI7, G.Skill Flare X5 F5-6000 (32GB), and a Gigabyte RTX 3070 Gaming OC Corsair RM750x. After running several 3DMARK benchmarks and stress tests, everything seemed normal. Occasionally, my scores were lower than others with identical hardware, but I hadn’t overclocked anything, so I wasn’t sure it was an issue. About two weeks later, I ran the same tests again... And here’s what happened—I dropped a lot of points! Attached are the comparisons and the actual differences. The first set is from Firestrike, the second from Timespy. The orange lines represent my builds, while the red ones show the results after two weeks. During this period, I made almost no changes except updating BIOS and GPU drivers. Everything else stayed consistent. My RAM was also functioning correctly, and I disabled GSync without affecting performance. Clearly, I’m a bit worried. Since I built it myself, I’m concerned something might be wrong. Could anyone help clarify what’s going on or fix this? It would really help a lot. Thanks in advance,
Mast3rDisast3r
The results aren't that different, so it's unclear what you're referring to by not meeting expectations. You likely wouldn't see this gap while playing. The CPU scores look similar on November 11 versus November 24. It's the GPU performance that affects your final rating. When you first tested on November 11, your setup was available; now it's closed and stored. Maybe the room wasn't well-ventilated. What temperature is the GPU at during testing? Have you updated the GPU drivers since the initial test? Using the new "NVIDIA app" could also influence the score.
I've used the same drivers and updated them, yet I haven't reached the November 11 target with the latest runs. The case was resolved in both instances and on the desk. I haven't adjusted the position—I mean, I'm now at the lower ranks with these new scores. Did the Timespy Extreme end up in the bottom third of all tests with the same hardware? That seems unusual. Or did I create a system that underperforms?
In summary, I ran a Cinebench R24 test and achieved a multicore score of 1306. This result is roughly 60 to 70 points behind what many reviewers are observing.
Same issue, previously world leader in 4060+12600kf builds on steel nomad, now consistently falling about 50-100 points behind top players. Benchmarks serve as guidance and should be repeated several times to reduce randomness, provided FPS remains stable. Edited November 25, 2024 by Tridefender