F5F Stay Refreshed Hardware Desktop DDR4 at 3000MHz CL15 versus 3200MHz CL16 offers different performance trade-offs; the choice depends on your needs.

DDR4 at 3000MHz CL15 versus 3200MHz CL16 offers different performance trade-offs; the choice depends on your needs.

DDR4 at 3000MHz CL15 versus 3200MHz CL16 offers different performance trade-offs; the choice depends on your needs.

Pages (2): 1 2 Next
J
Jomppaboy
Member
192
12-16-2018, 12:06 PM
#1
I successfully got my DDR4 quad-channel setup working, adjusting it down from its standard 3000MHz CL16 speed to a CL15. I also discovered that maintaining the CL16 allows for a stable 3200MHz without issues. Both configurations use the default 1.35V voltage. As for the question, which option performs better—CL15 at 3000MHz or CL16 at 3200MHz—I haven’t found clear data to confirm a difference in performance.
J
Jomppaboy
12-16-2018, 12:06 PM #1

I successfully got my DDR4 quad-channel setup working, adjusting it down from its standard 3000MHz CL16 speed to a CL15. I also discovered that maintaining the CL16 allows for a stable 3200MHz without issues. Both configurations use the default 1.35V voltage. As for the question, which option performs better—CL15 at 3000MHz or CL16 at 3200MHz—I haven’t found clear data to confirm a difference in performance.

S
san004
Junior Member
9
12-16-2018, 12:06 PM
#2
Likely the higher frequency is the best option.
S
san004
12-16-2018, 12:06 PM #2

Likely the higher frequency is the best option.

2
21Vid3oGam1ng
Member
73
12-16-2018, 12:06 PM
#3
yeah pretty much no difference, imo. get more MEGAhz just cuz MEGAhz.
2
21Vid3oGam1ng
12-16-2018, 12:06 PM #3

yeah pretty much no difference, imo. get more MEGAhz just cuz MEGAhz.

T
tsnyder01
Member
171
12-16-2018, 12:06 PM
#4
Performance remains unchanged; however, the 3200MHz offers a theoretical advantage. Latency stays consistent at 5ns regardless of the choice.
T
tsnyder01
12-16-2018, 12:06 PM #4

Performance remains unchanged; however, the 3200MHz offers a theoretical advantage. Latency stays consistent at 5ns regardless of the choice.

E
Erik0Gordy
Junior Member
5
12-16-2018, 12:06 PM
#5
in real life you won't see any change. when using winrar, opt for the quicker 3200mhz setting (you might be the one who increased the speed in the newsgroup). if both need the same voltage, choose whichever works best. run some aida64 cache memory tests to compare performance at different speeds—timings or max memory usage.
E
Erik0Gordy
12-16-2018, 12:06 PM #5

in real life you won't see any change. when using winrar, opt for the quicker 3200mhz setting (you might be the one who increased the speed in the newsgroup). if both need the same voltage, choose whichever works best. run some aida64 cache memory tests to compare performance at different speeds—timings or max memory usage.

Z
ZexyZeke
Member
166
12-16-2018, 12:06 PM
#6
You're unsure about the test setup, but once you identify the preferred speed, I recommend running memtest86 a few times to ensure stability.
Z
ZexyZeke
12-16-2018, 12:06 PM #6

You're unsure about the test setup, but once you identify the preferred speed, I recommend running memtest86 a few times to ensure stability.

C
ChocoLee
Junior Member
11
12-16-2018, 12:06 PM
#7
This is actually a very good idea. I will do that! Thanks!
C
ChocoLee
12-16-2018, 12:06 PM #7

This is actually a very good idea. I will do that! Thanks!

A
anakindaur
Senior Member
576
12-16-2018, 12:06 PM
#8
Would you mind sharing your findings? I conducted some personal tests to explore the differences between DDR3 and DDR4 in this benchmark. I aimed to keep the CPU close to 4.7 with a 1:1 cache ratio, using the kit from Newegg. I ran three tests with various speeds (2800, 3000, 3200, 3400 MHz).
A
anakindaur
12-16-2018, 12:06 PM #8

Would you mind sharing your findings? I conducted some personal tests to explore the differences between DDR3 and DDR4 in this benchmark. I aimed to keep the CPU close to 4.7 with a 1:1 cache ratio, using the kit from Newegg. I ran three tests with various speeds (2800, 3000, 3200, 3400 MHz).

_
_Herurg_
Junior Member
3
12-16-2018, 12:06 PM
#9
Sure, I'll share the findings later. I won't save yesterday's work, so I'll need to redo the tests when I get back from work. I recall that 3000MHz at CL15 performed slightly better than 3200MHz at CL16, with a difference of about 0.5-0.6ns in favor of the former.
_
_Herurg_
12-16-2018, 12:06 PM #9

Sure, I'll share the findings later. I won't save yesterday's work, so I'll need to redo the tests when I get back from work. I recall that 3000MHz at CL15 performed slightly better than 3200MHz at CL16, with a difference of about 0.5-0.6ns in favor of the former.

A
Abrikozenjam
Junior Member
8
12-16-2018, 12:06 PM
#10
I conducted additional testing and benchmarks and discovered some intriguing findings. The results I achieved with the 3200MHz overclock didn't match yesterday's outcomes. Other configurations also showed promise. Keep in mind this isn't a motherboard-certified memory package. On my system, only two kits certified for over 3000MHz are available. CPU: 4.4GHz Cache: 3GHz (Stock) RAM: 3000MHz 15-15-15-35 CPU: 4.4GHz Cache: 3GHz (Stock) RAM: 3000MHz 16-16-16-36 CPU: 4.4GHz Cache: 3.5GHz RAM: 3000MHz 16-16-16-36 It's worth noting I couldn't achieve CL15 performance with the cache boosted. I experimented with system voltage, but it didn’t significantly improve stability. This was similar to what I experienced yesterday when I booted at 3200MHz while keeping RAM at CL16. The last preset feels noticeably smoother in Windows compared to others. Although the latency differences are minimal, I’m curious about the Command Rate and its effect on performance. My motherboard defaults to 2T, yet reaching 3000MHz without adjusting it to 1T is challenging. At 1T Command Rate, applications load much quicker, and Windows responds more quickly.
A
Abrikozenjam
12-16-2018, 12:06 PM #10

I conducted additional testing and benchmarks and discovered some intriguing findings. The results I achieved with the 3200MHz overclock didn't match yesterday's outcomes. Other configurations also showed promise. Keep in mind this isn't a motherboard-certified memory package. On my system, only two kits certified for over 3000MHz are available. CPU: 4.4GHz Cache: 3GHz (Stock) RAM: 3000MHz 15-15-15-35 CPU: 4.4GHz Cache: 3GHz (Stock) RAM: 3000MHz 16-16-16-36 CPU: 4.4GHz Cache: 3.5GHz RAM: 3000MHz 16-16-16-36 It's worth noting I couldn't achieve CL15 performance with the cache boosted. I experimented with system voltage, but it didn’t significantly improve stability. This was similar to what I experienced yesterday when I booted at 3200MHz while keeping RAM at CL16. The last preset feels noticeably smoother in Windows compared to others. Although the latency differences are minimal, I’m curious about the Command Rate and its effect on performance. My motherboard defaults to 2T, yet reaching 3000MHz without adjusting it to 1T is challenging. At 1T Command Rate, applications load much quicker, and Windows responds more quickly.

Pages (2): 1 2 Next