F5F Stay Refreshed Hardware Desktop 5900x PPT format, TDC and EDC specifications

5900x PPT format, TDC and EDC specifications

5900x PPT format, TDC and EDC specifications

Pages (3): Previous 1 2 3
T
tmt108
Junior Member
45
09-15-2016, 01:57 AM
#21
I wasn’t expecting that until I was working on keeping the top boost clock steady during single-core tests. Those numbers really matter—it’s not just for all core scenarios. Running at 5GHZ should be manageable, and I believe I can hit close to the stock settings with a small adjustment.
T
tmt108
09-15-2016, 01:57 AM #21

I wasn’t expecting that until I was working on keeping the top boost clock steady during single-core tests. Those numbers really matter—it’s not just for all core scenarios. Running at 5GHZ should be manageable, and I believe I can hit close to the stock settings with a small adjustment.

M
MattGruGru
Junior Member
45
09-15-2016, 01:57 AM
#22
I'm not referring to boosting settings. When you reach 5GHz or higher, you're no longer using the stock configuration. For regular clocks, even with low power draw (around 65W TDP) like 88 PPT, 60 TDC, and 90 EDC, my Cinebench R23 single-core performance barely improved and my boost clock stayed near 4.95GHz, often reaching over 4.8GHz—matching the rated maximum for the 5900X. After lowering it to 35W TDP (48 PPT, 30 TDC, 60 EDC), I noticed a noticeable drop in sustained single-core speed, dropping to about 4.6GHz.
M
MattGruGru
09-15-2016, 01:57 AM #22

I'm not referring to boosting settings. When you reach 5GHz or higher, you're no longer using the stock configuration. For regular clocks, even with low power draw (around 65W TDP) like 88 PPT, 60 TDC, and 90 EDC, my Cinebench R23 single-core performance barely improved and my boost clock stayed near 4.95GHz, often reaching over 4.8GHz—matching the rated maximum for the 5900X. After lowering it to 35W TDP (48 PPT, 30 TDC, 60 EDC), I noticed a noticeable drop in sustained single-core speed, dropping to about 4.6GHz.

M
MinergodPT
Junior Member
10
09-15-2016, 01:57 AM
#23
Upon interacting with the curve optimizer, you're experiencing an overclock. Just a few bounces past 4600 will reveal your R23 score; it's possible you've reached a high setting, but you need to know the actual clock speeds.
M
MinergodPT
09-15-2016, 01:57 AM #23

Upon interacting with the curve optimizer, you're experiencing an overclock. Just a few bounces past 4600 will reveal your R23 score; it's possible you've reached a high setting, but you need to know the actual clock speeds.

H
HatterHologram
Junior Member
45
09-15-2016, 01:57 AM
#24
From the AMD perspective, yes—but it feels unusual since I'm keeping the clock speed unchanged. I’d say overclocking means exceeding the rated frequency, but what I meant is that unless you increase it, you need to significantly reduce PPT to avoid affecting single-core performance. What do you mean by "effective clocks"? I’m actually repeating the 35W TDP test. Most of the time it runs at around 4.725GHz on a single core, fluctuating between 4.6GHz and 4.825GHz in Cinebench R23. The final score came to 1546, which is lower than expected but still within normal range. This shows that those numbers aren’t crucial for tasks with few threads. However, they do have a big impact on multi-core performance.
H
HatterHologram
09-15-2016, 01:57 AM #24

From the AMD perspective, yes—but it feels unusual since I'm keeping the clock speed unchanged. I’d say overclocking means exceeding the rated frequency, but what I meant is that unless you increase it, you need to significantly reduce PPT to avoid affecting single-core performance. What do you mean by "effective clocks"? I’m actually repeating the 35W TDP test. Most of the time it runs at around 4.725GHz on a single core, fluctuating between 4.6GHz and 4.825GHz in Cinebench R23. The final score came to 1546, which is lower than expected but still within normal range. This shows that those numbers aren’t crucial for tasks with few threads. However, they do have a big impact on multi-core performance.

E
ErichSteve
Junior Member
11
09-15-2016, 01:57 AM
#25
It seems like this might not be a major issue after all. What I know for sure is that each AMD CPU is unique and doesn't follow a uniform pattern.
E
ErichSteve
09-15-2016, 01:57 AM #25

It seems like this might not be a major issue after all. What I know for sure is that each AMD CPU is unique and doesn't follow a uniform pattern.

Pages (3): Previous 1 2 3