F5F Stay Refreshed Hardware Desktop Yes, they are equivalent in terms of specifications and performance.

Yes, they are equivalent in terms of specifications and performance.

Yes, they are equivalent in terms of specifications and performance.

M
Meba11
Member
64
01-09-2016, 05:12 AM
#1
I realized I purchased the incorrect RAM type for my system. I assumed a 3200 MHz DDR4 would suffice, but it didn’t work as expected. My existing modules are 2xHX432C16FB3/8 and 2xHX432C18FB2/8, which function well at 2400 MHz but behave poorly under XMP settings. Since HyperX is now part of HP, the current match seems to be the "Kingston Fury" KF432C16BB/8. After reviewing the data sheets, nothing stands out, but I wanted to share them just in case. Attached are the documents for your reference.
M
Meba11
01-09-2016, 05:12 AM #1

I realized I purchased the incorrect RAM type for my system. I assumed a 3200 MHz DDR4 would suffice, but it didn’t work as expected. My existing modules are 2xHX432C16FB3/8 and 2xHX432C18FB2/8, which function well at 2400 MHz but behave poorly under XMP settings. Since HyperX is now part of HP, the current match seems to be the "Kingston Fury" KF432C16BB/8. After reviewing the data sheets, nothing stands out, but I wanted to share them just in case. Attached are the documents for your reference.

D
ducky198chica
Junior Member
17
01-24-2016, 12:31 AM
#2
Ensure identical components, speed, and timing for best results. Even with matching spec sheets, variations in ICs across different boards can cause problems. The same manufacturer may use different chips depending on cost, so a matched pair is usually safest. Consider your CPU's IMC and sensitivity—frequency and timing alignment will reduce risks.
D
ducky198chica
01-24-2016, 12:31 AM #2

Ensure identical components, speed, and timing for best results. Even with matching spec sheets, variations in ICs across different boards can cause problems. The same manufacturer may use different chips depending on cost, so a matched pair is usually safest. Consider your CPU's IMC and sensitivity—frequency and timing alignment will reduce risks.

V
Vesgo
Member
230
01-24-2016, 09:49 PM
#3
Sure, I understand. It seems like the matched pair isn't an option for me anymore since I already have some sticks. I'm considering selling them or buying a new kit, but it might be expensive. I've been reviewing these datasheets, and everything looks similar except for a tiny difference in stick height of about 0.1mm, which is fine. In theory, it should work, correct?
V
Vesgo
01-24-2016, 09:49 PM #3

Sure, I understand. It seems like the matched pair isn't an option for me anymore since I already have some sticks. I'm considering selling them or buying a new kit, but it might be expensive. I've been reviewing these datasheets, and everything looks similar except for a tiny difference in stick height of about 0.1mm, which is fine. In theory, it should work, correct?

B
blueyednick
Member
199
01-24-2016, 11:10 PM
#4
It usually should. As said, the last unknown is whether there's some difference in their ICs that the CPU doesn't like. Speaking of 4x8… it's often better to go with 2x16. The more memory sticks you have, the harder it gets for the IMC to drive them at higher frequencies. Depending on the CPU there's a chance it still won't work. For example Zen 1/2 often doesn't like high-frequency XMP with four sticks.
B
blueyednick
01-24-2016, 11:10 PM #4

It usually should. As said, the last unknown is whether there's some difference in their ICs that the CPU doesn't like. Speaking of 4x8… it's often better to go with 2x16. The more memory sticks you have, the harder it gets for the IMC to drive them at higher frequencies. Depending on the CPU there's a chance it still won't work. For example Zen 1/2 often doesn't like high-frequency XMP with four sticks.

R
RG48
Posting Freak
778
01-24-2016, 11:59 PM
#5
You're aware of the issue, but I'm curious—would boosting the memory controller's speed help resolve it? Otherwise, it seems like we're stuck with the existing problem.
R
RG48
01-24-2016, 11:59 PM #5

You're aware of the issue, but I'm curious—would boosting the memory controller's speed help resolve it? Otherwise, it seems like we're stuck with the existing problem.