F5F Stay Refreshed Hardware Desktop Poor performance on multiple cores paired with a 7600 processor.

Poor performance on multiple cores paired with a 7600 processor.

Poor performance on multiple cores paired with a 7600 processor.

S
SeaniesShow
Junior Member
3
10-25-2017, 04:36 PM
#1
I’ve observed that my multiscore CPU results tend to lag behind others, even though most tests use default configurations and I’m running a PBO. It’s not overheating—I’ve recorded peak temperatures around 91°C—and I’ve adjusted the fan curve afterward to keep things lower. My single-core performance remains acceptable. Leaderboards are questionable because many participants push their CPUs beyond 5.1GHz, while I’m just using PBO. Benchmarks show: 7zip Compression 92.077 GIPS (others in 95-105), 7zip Decompression 96.142 GIPS (their range 100-110), Cinebench R23 multicore 14,150-14,200, singlecore 1,850-1,900, and R15 2,310. Even at lower resolutions like 720p/1080p, my FPSes were slightly better than most. I also tested games and other programs, with scores matching or slightly surpassing expectations except for Battlefield V, which might reflect a stronger GPU. I ran numerous additional tests—Cinebench R20, Jetstream 2, WebXPRT4, Blender, KeyShot, Corona, Speedometer 2, Nero AI Photo Tagger, Dolphin, CPU-Z, 3DMark, SuperPI, Y-Cruncher, Geekbench, etc.—and consistently saw my scores lagging marginally. Even when using stock settings, the difference is noticeable. I’m curious how much separation would be needed to suspect a hardware issue. I suspect the problem lies elsewhere; perhaps the cooler was misplaced or the fan curve needs fine-tuning. If I drop the cooler, temps stay under 95°C, which seems within the expected range for Zen 4 chips. Are non-X models affected similarly, especially at lower power draw?
S
SeaniesShow
10-25-2017, 04:36 PM #1

I’ve observed that my multiscore CPU results tend to lag behind others, even though most tests use default configurations and I’m running a PBO. It’s not overheating—I’ve recorded peak temperatures around 91°C—and I’ve adjusted the fan curve afterward to keep things lower. My single-core performance remains acceptable. Leaderboards are questionable because many participants push their CPUs beyond 5.1GHz, while I’m just using PBO. Benchmarks show: 7zip Compression 92.077 GIPS (others in 95-105), 7zip Decompression 96.142 GIPS (their range 100-110), Cinebench R23 multicore 14,150-14,200, singlecore 1,850-1,900, and R15 2,310. Even at lower resolutions like 720p/1080p, my FPSes were slightly better than most. I also tested games and other programs, with scores matching or slightly surpassing expectations except for Battlefield V, which might reflect a stronger GPU. I ran numerous additional tests—Cinebench R20, Jetstream 2, WebXPRT4, Blender, KeyShot, Corona, Speedometer 2, Nero AI Photo Tagger, Dolphin, CPU-Z, 3DMark, SuperPI, Y-Cruncher, Geekbench, etc.—and consistently saw my scores lagging marginally. Even when using stock settings, the difference is noticeable. I’m curious how much separation would be needed to suspect a hardware issue. I suspect the problem lies elsewhere; perhaps the cooler was misplaced or the fan curve needs fine-tuning. If I drop the cooler, temps stay under 95°C, which seems within the expected range for Zen 4 chips. Are non-X models affected similarly, especially at lower power draw?

_
___iRekt___
Senior Member
505
10-25-2017, 04:36 PM
#2
Many results seem to reflect a less refined operating system than the ones being compared. Since I don’t use 7zip often, I can’t give much insight into its evaluation. Cinebench may show large jumps—like 1000 points—depending on background processes not visible in single-core tests. My 5900X typically hits around 23500 on my benchmark OS, but with a daily OS loaded with extra files, it drops to 22000-22500 in R23.
_
___iRekt___
10-25-2017, 04:36 PM #2

Many results seem to reflect a less refined operating system than the ones being compared. Since I don’t use 7zip often, I can’t give much insight into its evaluation. Cinebench may show large jumps—like 1000 points—depending on background processes not visible in single-core tests. My 5900X typically hits around 23500 on my benchmark OS, but with a daily OS loaded with extra files, it drops to 22000-22500 in R23.

B
ButerBean13
Junior Member
15
10-25-2017, 04:36 PM
#3
It seems the issue isn't related to a failing CPU. You likely overlooked some background processes after closing them and disabling Windows Defender. Let me know if you need further help!
B
ButerBean13
10-25-2017, 04:36 PM #3

It seems the issue isn't related to a failing CPU. You likely overlooked some background processes after closing them and disabling Windows Defender. Let me know if you need further help!