This UserBenchmark comparison appears to be logical.
This UserBenchmark comparison appears to be logical.
Yes, the PassMark CPU benchmark table offers more precise outcomes compared to other comparison tools.
General estimates are the best you'll get. For detailed comparisons, check HWBot for your CPU specs and benchmarks. Scores can be seen by cooling type, giving a clearer picture of averages.
Consensus on "basic operation" highlights its strengths, yet it often struggles with real-world issues like failing hard drives. The way they define problems usually lacks detail, making it hard to understand things such as a CPU reaching the 98th percentile. Even with PBO turned off, my high-end RAM seems just normal. *cries* It’s confusing how it works, but it does serve as a solid check for fundamental failures that others might miss.
You're using a standard memory test instead of the actual frequency the chip can handle. To properly evaluate your memory, use AIDA64. Open the Tools tab, select Cache and Memory benchmarks, then fine-tune settings for speed, adjust timing, and see if improvements occur. This process helps with accurate performance measurement.
Ya, that's true, I could probably run tighter timings or higher frequency, or both... i just kinda settled with the first thing that booted... But see, thats what ub is useful for, simple things, that need to be interpreted by the user... i mean i see it says latency, so I'd need to try tighten the timings probably... and moar voltage ! Hey, how do you know my motto!? PS, that's also something thats stupid and not obvious... so they're comparing for example the gpu to *all* other 3070's... but the ram... they only compare that to "gskill trident z" that heavily skews the results , if so. I bet compared to *all* DDR4 3200 this ram would be baller... ! : P
I completed the DL for this benchmark and explained why it falls short. A key point is the duration each test took. Most 2D runs were around one second, while 3D took about three seconds each. I left some sections untouched because I wasn’t sure what to change. My PC is a high-end gaming machine with a GTX 980 KPE, which factory reports at 69% but actually hits 98% in real-world use. Since it was run as-is, I expected only the card’s baseline performance. No overclocking or modifications were applied. Memory usage is what really matters here. When comparing, I checked a 3600mhz 16-16-16-36 with 1.40v—this sits at the 84th percentile and exceeds expectations by 119%. It’s strange that my underclocked RAM is outperforming an overclocked one. Comparing a 3200mhz setup to a 4267mhz setup wouldn’t make sense.