F5F Stay Refreshed Hardware Desktop Can't lower my latency below 16 ms

Can't lower my latency below 16 ms

Can't lower my latency below 16 ms

Pages (3): Previous 1 2 3 Next
P
PlayForCC
Member
217
07-24-2019, 12:09 PM
#11
Just to confirm, you're assessing the testing process for RAM stability. It's challenging to verify stability, which is why overclocking RAM isn't commonly recommended. Ryzen has made it more accessible, but it involves greater complexity than CPU overclocking and requires more thorough validation. Diagnosing RAM problems can be difficult, so ensure stability before proceeding; otherwise, unexpected issues may arise later that aren't related to RAM. memtesthelper works well on Windows, and running extensive tests could confirm reliability. The most reliable method appears to be using Linux, though it demands more experience.
P
PlayForCC
07-24-2019, 12:09 PM #11

Just to confirm, you're assessing the testing process for RAM stability. It's challenging to verify stability, which is why overclocking RAM isn't commonly recommended. Ryzen has made it more accessible, but it involves greater complexity than CPU overclocking and requires more thorough validation. Diagnosing RAM problems can be difficult, so ensure stability before proceeding; otherwise, unexpected issues may arise later that aren't related to RAM. memtesthelper works well on Windows, and running extensive tests could confirm reliability. The most reliable method appears to be using Linux, though it demands more experience.

K
kieranwebster
Member
62
07-25-2019, 09:39 AM
#12
Running memtest 64 for two hours might help, but it could be more effective.
K
kieranwebster
07-25-2019, 09:39 AM #12

Running memtest 64 for two hours might help, but it could be more effective.

M
Mister_Token
Member
202
07-27-2019, 01:55 AM
#13
Absolutely, you’d put in a lot of effort. It’s hard to tell if it’ll finish within the time limits based on your past tests.
M
Mister_Token
07-27-2019, 01:55 AM #13

Absolutely, you’d put in a lot of effort. It’s hard to tell if it’ll finish within the time limits based on your past tests.

R
Repertition
Member
186
07-27-2019, 02:48 AM
#14
You're planning to run MemTest 64 for several hours, right? Just to be sure, does that testing process put any strain on your RAM?
R
Repertition
07-27-2019, 02:48 AM #14

You're planning to run MemTest 64 for several hours, right? Just to be sure, does that testing process put any strain on your RAM?

S
Strikers928
Junior Member
1
07-27-2019, 02:46 PM
#15
Yes, there isn't a clear way to say "tire it out." Everything should be okay... Are you certain you require 1.4v? Is the stock setting XMP 1.35? Probably not, since you might not need that much. You'll find out by testing. Try running it for 8 hours at 1.4. If it works, lower to 1.37, if it still works, go down to 1.35. If neither setting is satisfactory, increase the value step by step—like from 1.37 to 1.38, etc.—aiming for the lowest stable setting.
S
Strikers928
07-27-2019, 02:46 PM #15

Yes, there isn't a clear way to say "tire it out." Everything should be okay... Are you certain you require 1.4v? Is the stock setting XMP 1.35? Probably not, since you might not need that much. You'll find out by testing. Try running it for 8 hours at 1.4. If it works, lower to 1.37, if it still works, go down to 1.35. If neither setting is satisfactory, increase the value step by step—like from 1.37 to 1.38, etc.—aiming for the lowest stable setting.

C
Cefreak113
Senior Member
484
07-27-2019, 10:59 PM
#16
I mentioned 1.4 volts because I planned to start with that setting, fine-tune each timing until it works, then check if I could increase the memory speed—currently around 3600, possibly up to 3700 or 3800 before it stops booting. Once everything was configured, I intended to gradually lower the voltage until it worked again, then raise it back to the last known working level. After that, I’d run a stress test for ten hours to ensure stability. It sounds like a long process, but it should only take about an hour or two after restarting and adjusting the timing. I won’t keep the voltage above 1.4 volts for extended periods.
C
Cefreak113
07-27-2019, 10:59 PM #16

I mentioned 1.4 volts because I planned to start with that setting, fine-tune each timing until it works, then check if I could increase the memory speed—currently around 3600, possibly up to 3700 or 3800 before it stops booting. Once everything was configured, I intended to gradually lower the voltage until it worked again, then raise it back to the last known working level. After that, I’d run a stress test for ten hours to ensure stability. It sounds like a long process, but it should only take about an hour or two after restarting and adjusting the timing. I won’t keep the voltage above 1.4 volts for extended periods.

J
163
07-29-2019, 12:40 PM
#17
It sounds like you're on the right track. That adjustment process took a couple of weeks, and I appreciate you sharing the details!
J
Jazzy_Games123
07-29-2019, 12:40 PM #17

It sounds like you're on the right track. That adjustment process took a couple of weeks, and I appreciate you sharing the details!

I
IMayBeDead
Senior Member
696
07-29-2019, 06:28 PM
#18
It seems you're asking about why a process takes so long—possibly two hours—and whether adjusting settings helps. You also want to understand the connection between memory overclocking and latency changes. The timing might be affected by how the system handles memory access speeds, which can influence core performance.
I
IMayBeDead
07-29-2019, 06:28 PM #18

It seems you're asking about why a process takes so long—possibly two hours—and whether adjusting settings helps. You also want to understand the connection between memory overclocking and latency changes. The timing might be affected by how the system handles memory access speeds, which can influence core performance.

B
Beder822
Member
80
07-29-2019, 08:09 PM
#19
It slows things down since adjusting all secondary and tertiary settings is time-consuming. With around 20 parameters, optimizing for peak performance is challenging. It took weeks to fine-tune everything and run tests. Increasing timings and boosting voltage lets you push higher MHz, but you can’t maintain the same speed with those settings. RAM overclocking is tough—it demands significant learning and practice to get it right. I’d check this out, at least a few times; it really clarifies a lot. https://github.com/integralfx/MemTestHel...aster/DDR4 OC Guide.md
B
Beder822
07-29-2019, 08:09 PM #19

It slows things down since adjusting all secondary and tertiary settings is time-consuming. With around 20 parameters, optimizing for peak performance is challenging. It took weeks to fine-tune everything and run tests. Increasing timings and boosting voltage lets you push higher MHz, but you can’t maintain the same speed with those settings. RAM overclocking is tough—it demands significant learning and practice to get it right. I’d check this out, at least a few times; it really clarifies a lot. https://github.com/integralfx/MemTestHel...aster/DDR4 OC Guide.md

P
Parzival10
Member
180
07-29-2019, 08:31 PM
#20
Sure, I understand. You're focusing on adjusting only the latency settings and possibly some core frequencies. If you were to use a DRM calculator, it could help with that?
P
Parzival10
07-29-2019, 08:31 PM #20

Sure, I understand. You're focusing on adjusting only the latency settings and possibly some core frequencies. If you were to use a DRM calculator, it could help with that?

Pages (3): Previous 1 2 3 Next