F5F Stay Refreshed Hardware Desktop Issue with 4 stick RAM not starting on X570 Xtreme at full speeds

Issue with 4 stick RAM not starting on X570 Xtreme at full speeds

Issue with 4 stick RAM not starting on X570 Xtreme at full speeds

Pages (2): Previous 1 2
G
geglman
Member
87
01-18-2020, 12:42 PM
#11
Aorus doesn't stand out for strong memory assistance features compared to MSI.
G
geglman
01-18-2020, 12:42 PM #11

Aorus doesn't stand out for strong memory assistance features compared to MSI.

G
Goku_Jerome
Senior Member
428
01-18-2020, 01:36 PM
#12
The main models perform well, and the Gaming 7 on X470 is quite solid too.
G
Goku_Jerome
01-18-2020, 01:36 PM #12

The main models perform well, and the Gaming 7 on X470 is quite solid too.

W
WonsGHG
Junior Member
17
01-18-2020, 03:15 PM
#13
I successfully achieved 12-12-12-28-42 at 3133 with minimal issues, no freezes or glitches. I maintained a voltage of 1.45v for extra margin and kept the SoC in auto mode, reaching around 1.075. The system performed stably during AIDA stress tests, and temperatures stayed around 42°C. I might experiment further with timing adjustments, though the frequency cap remains at 3133. This suggests a potential hardware or layout concern—possibly related to the motherboard or memory controller. Switching to newer AMD-optimized models could help, as Royal kits are reportedly better suited initially. I could potentially reach 3600 c16 or 3200 c14 right out of the box, which would be better than my current 3133 c12. Raising the frequency to 3600 to 1400 seems appealing, though it’s a trade-off. Generally, higher frequencies with tighter timings tend to perform better, but it’s a nuanced decision.
W
WonsGHG
01-18-2020, 03:15 PM #13

I successfully achieved 12-12-12-28-42 at 3133 with minimal issues, no freezes or glitches. I maintained a voltage of 1.45v for extra margin and kept the SoC in auto mode, reaching around 1.075. The system performed stably during AIDA stress tests, and temperatures stayed around 42°C. I might experiment further with timing adjustments, though the frequency cap remains at 3133. This suggests a potential hardware or layout concern—possibly related to the motherboard or memory controller. Switching to newer AMD-optimized models could help, as Royal kits are reportedly better suited initially. I could potentially reach 3600 c16 or 3200 c14 right out of the box, which would be better than my current 3133 c12. Raising the frequency to 3600 to 1400 seems appealing, though it’s a trade-off. Generally, higher frequencies with tighter timings tend to perform better, but it’s a nuanced decision.

I
iKegreenS_
Posting Freak
878
01-18-2020, 03:40 PM
#14
Alternatively, the issue might stem from the memory controller itself. You might attempt to boost performance by overclocking with just two sticks; at 32GB, it could run significantly faster if the problem lies there. You can partially recover lost speed by adjusting the FCLK and pushing it to instability until it breaks down. Synchronized FCLK should match half the memory data rate—like 1567 for a 3133MHz memory—but actual stable speeds for Zen 2 CPUs range from 1733 to 1933MHz, depending on manufacturing variations.
I
iKegreenS_
01-18-2020, 03:40 PM #14

Alternatively, the issue might stem from the memory controller itself. You might attempt to boost performance by overclocking with just two sticks; at 32GB, it could run significantly faster if the problem lies there. You can partially recover lost speed by adjusting the FCLK and pushing it to instability until it breaks down. Synchronized FCLK should match half the memory data rate—like 1567 for a 3133MHz memory—but actual stable speeds for Zen 2 CPUs range from 1733 to 1933MHz, depending on manufacturing variations.

L
lill_laagen
Junior Member
2
02-01-2020, 11:47 PM
#15
The situation has taken an even stranger turn. No boot ever succeeded at 3200, regardless of any settings I attempted—3200 was unattainable, I never reached higher. I thought it might be pointless, but for a bit of experimentation, I switched the system to 3266 c14 and it worked. It seems my testing is still ongoing. I’ll also attempt to separate FCLK, though I’m not sure if it’s a wise move.
L
lill_laagen
02-01-2020, 11:47 PM #15

The situation has taken an even stranger turn. No boot ever succeeded at 3200, regardless of any settings I attempted—3200 was unattainable, I never reached higher. I thought it might be pointless, but for a bit of experimentation, I switched the system to 3266 c14 and it worked. It seems my testing is still ongoing. I’ll also attempt to separate FCLK, though I’m not sure if it’s a wise move.

P
PGVortex
Member
146
02-02-2020, 06:05 AM
#16
Ensure decoupling only when the FCLK remains stable at least 200MHz above its value with coupling.
P
PGVortex
02-02-2020, 06:05 AM #16

Ensure decoupling only when the FCLK remains stable at least 200MHz above its value with coupling.

I
Ice_Blinger
Junior Member
25
02-02-2020, 03:04 PM
#17
I reached 3733 c14-14-14, which feels pretty impressive given the four double-sided sticks. I tried posting with 1.45 on memory but it didn’t come through. It went up to 1.47, which is right at the limits—will need to lower it slightly for regular use and run some stability tests. I’m still puzzled about the boot failure at 3200; it looks like a strange bug.
I
Ice_Blinger
02-02-2020, 03:04 PM #17

I reached 3733 c14-14-14, which feels pretty impressive given the four double-sided sticks. I tried posting with 1.45 on memory but it didn’t come through. It went up to 1.47, which is right at the limits—will need to lower it slightly for regular use and run some stability tests. I’m still puzzled about the boot failure at 3200; it looks like a strange bug.

U
Ulrik08
Junior Member
12
02-02-2020, 08:38 PM
#18
Provided a solid overview of AMD optimization techniques.
U
Ulrik08
02-02-2020, 08:38 PM #18

Provided a solid overview of AMD optimization techniques.

Pages (2): Previous 1 2