F5F Stay Refreshed Hardware Desktop Initial adjustments to RAM timing boost performance speed

Initial adjustments to RAM timing boost performance speed

Initial adjustments to RAM timing boost performance speed

Pages (2): 1 2 Next
T
TTLego
Junior Member
15
03-04-2021, 08:19 AM
#1
You're interested in understanding the performance potential of a Samsung B-die CL16-16-16-36 4X8GB kit running at 3600Mhz. It sounds like you're aiming for better speed without pushing XMP settings. Expecting tighter timings could mean improved frame rates or lower latency, depending on the workload and your system's capabilities.
T
TTLego
03-04-2021, 08:19 AM #1

You're interested in understanding the performance potential of a Samsung B-die CL16-16-16-36 4X8GB kit running at 3600Mhz. It sounds like you're aiming for better speed without pushing XMP settings. Expecting tighter timings could mean improved frame rates or lower latency, depending on the workload and your system's capabilities.

B
brad2011
Member
51
03-04-2021, 01:48 PM
#2
It should be straightforward at 3600MHz with XMP support; otherwise, simply increase the voltage. The CAS won’t actually drop the first value, though the remaining parameters might.
B
brad2011
03-04-2021, 01:48 PM #2

It should be straightforward at 3600MHz with XMP support; otherwise, simply increase the voltage. The CAS won’t actually drop the first value, though the remaining parameters might.

T
tomtom19200
Junior Member
8
03-23-2021, 12:43 PM
#3
CAS comes first, and adding 14 brings it down below the stock value. I understand now.
T
tomtom19200
03-23-2021, 12:43 PM #3

CAS comes first, and adding 14 brings it down below the stock value. I understand now.

D
DarkBoy__YT
Posting Freak
898
03-26-2021, 01:45 PM
#4
It really depends on the kit you're using. My 3600 cl14 b-die set isn't great for high speeds beyond its limits, but my 3200 cl14 b die kit really shines when pushing overclocking boundaries. Keep in mind that your RAM controller and motherboard quality also play a big role.
D
DarkBoy__YT
03-26-2021, 01:45 PM #4

It really depends on the kit you're using. My 3600 cl14 b-die set isn't great for high speeds beyond its limits, but my 3200 cl14 b die kit really shines when pushing overclocking boundaries. Keep in mind that your RAM controller and motherboard quality also play a big role.

T
TheWarlord23
Member
194
03-26-2021, 02:52 PM
#5
Sure, I understand you're not going to boost the XMP settings too much, just a slight adjustment to the clock speeds.
T
TheWarlord23
03-26-2021, 02:52 PM #5

Sure, I understand you're not going to boost the XMP settings too much, just a slight adjustment to the clock speeds.

C
chiefjumbo
Member
183
03-26-2021, 03:55 PM
#6
I understand your perspective, but the quality of your sample affects the timing options you can achieve. For primary timings, there are specific considerations; for tertiary timings, it depends on your platform. If you're using a Ryzen processor, the DRAM Calculator from 1usmus is useful, though simpler methods like testing at 1T timing may work but can be less stable with multiple sticks based on your setup.
C
chiefjumbo
03-26-2021, 03:55 PM #6

I understand your perspective, but the quality of your sample affects the timing options you can achieve. For primary timings, there are specific considerations; for tertiary timings, it depends on your platform. If you're using a Ryzen processor, the DRAM Calculator from 1usmus is useful, though simpler methods like testing at 1T timing may work but can be less stable with multiple sticks based on your setup.

J
jakers1912
Junior Member
16
03-27-2021, 10:10 PM
#7
I suggest lowering each value by hand unless it leads to issues. A DRAM calculator helps save time but isn't always reliable.
J
jakers1912
03-27-2021, 10:10 PM #7

I suggest lowering each value by hand unless it leads to issues. A DRAM calculator helps save time but isn't always reliable.

T
Tsuyoshi_
Junior Member
8
03-28-2021, 05:53 AM
#8
You're aiming for top-tier performance with the RYZEN 5000 series and a high-end X570 board. That means skipping budget components entirely.
T
Tsuyoshi_
03-28-2021, 05:53 AM #8

You're aiming for top-tier performance with the RYZEN 5000 series and a high-end X570 board. That means skipping budget components entirely.

M
Masupa
Member
59
03-28-2021, 01:57 PM
#9
Some schedules need to fit into a pattern of multiples? If that makes sense, I believe the key is consistency, not just reducing them easily.
M
Masupa
03-28-2021, 01:57 PM #9

Some schedules need to fit into a pattern of multiples? If that makes sense, I believe the key is consistency, not just reducing them easily.

K
K1ngVince
Member
157
03-28-2021, 04:37 PM
#10
Never encountered the term before. Certain parameters exist only within specific ranges (for example, CAS needs even values when Gear Down is active), but there’s no concept of "multiples of another" in memory timing.
K
K1ngVince
03-28-2021, 04:37 PM #10

Never encountered the term before. Certain parameters exist only within specific ranges (for example, CAS needs even values when Gear Down is active), but there’s no concept of "multiples of another" in memory timing.

Pages (2): 1 2 Next