F5F Stay Refreshed Hardware Desktop No, running RAM below its specification does not reduce timing; it typically increases latency and slows performance.

No, running RAM below its specification does not reduce timing; it typically increases latency and slows performance.

No, running RAM below its specification does not reduce timing; it typically increases latency and slows performance.

G
GuaruGamer
Junior Member
45
09-15-2017, 07:58 AM
#1
Hey there, I saw you're thinking about a good deal for your laptop. The bundle looks cheaper than buying parts separately, but there are some technical details to consider. Your chipset only supports DDR4 2666, so even with the 3200 RAM, it won’t actually run faster than that. The CAS latency might stay similar or change little, depending on the specific model. Since your BIOS is limited, the 3200 kit probably won’t let you adjust timing beyond what’s built in. If you want the best performance, sticking with the 2666 kit might be better unless you’re okay with a slightly slower maximum speed. The extra $5 could make a difference if the 2666 kit actually hits its full potential, but it’s hard to say without testing it.
G
GuaruGamer
09-15-2017, 07:58 AM #1

Hey there, I saw you're thinking about a good deal for your laptop. The bundle looks cheaper than buying parts separately, but there are some technical details to consider. Your chipset only supports DDR4 2666, so even with the 3200 RAM, it won’t actually run faster than that. The CAS latency might stay similar or change little, depending on the specific model. Since your BIOS is limited, the 3200 kit probably won’t let you adjust timing beyond what’s built in. If you want the best performance, sticking with the 2666 kit might be better unless you’re okay with a slightly slower maximum speed. The extra $5 could make a difference if the 2666 kit actually hits its full potential, but it’s hard to say without testing it.

X
211
09-15-2017, 07:58 AM
#2
Typically the schedules would shorten. Even without that, the variation wouldn’t be something you’d really observe.
X
xXDamoDoggieXx
09-15-2017, 07:58 AM #2

Typically the schedules would shorten. Even without that, the variation wouldn’t be something you’d really observe.

N
Narwhal_Plays
Member
82
09-15-2017, 07:58 AM
#3
Uncertainty remains without testing the actual components. Most laptop RAM operates at industry-accepted JEDEC speeds and timings. If a profile for slower rates exists, it should work. For instance, my machine supports 3200, covering all standard speeds and their corresponding timings. The main concern is whether every module includes these profiles by default. JEDEC timing tables list various frequencies and speeds, but it's unclear if all devices use the same defaults. Performing a check would clarify if lower-speed options are available. Minor differences may occur, but significant issues are unlikely unless critical profiles are missing.
N
Narwhal_Plays
09-15-2017, 07:58 AM #3

Uncertainty remains without testing the actual components. Most laptop RAM operates at industry-accepted JEDEC speeds and timings. If a profile for slower rates exists, it should work. For instance, my machine supports 3200, covering all standard speeds and their corresponding timings. The main concern is whether every module includes these profiles by default. JEDEC timing tables list various frequencies and speeds, but it's unclear if all devices use the same defaults. Performing a check would clarify if lower-speed options are available. Minor differences may occur, but significant issues are unlikely unless critical profiles are missing.

D
Darphy
Junior Member
41
09-15-2017, 07:58 AM
#4
Interesting point. Those who know my 5900X will recall the RAM problems I faced. This machine didn’t like the original Ballistix I used—they claimed it was DDR4-3200 with 16-18-18-36 timings at 1.35V. DOCP recognized that spec. My Asus Tuf B550-PLUS suggested Patriot PV432G320C6K, also 16-18-18-36 at the same voltage. These were similarly labeled as DDR4-3200. The board’s DOCP picked up the Patriot DIMMs with odd timings like 16-20-20-40. I had to manually adjust them to 16-18-18-36. That shouldn’t matter, huh? Here’s the odd part: the system completely won’t boot when using the 3200 MHz clock with these modules. It works fine at 2400, but anything higher causes a blank screen until it realizes something’s wrong and jumps into BIOS. I haven’t seen any real RAM performance drops at 2400, just this puzzling behavior. So maybe the auto-detection of 16-20-20-40 is handling this situation? Updated January 15, 2023 by An0maly_76 Added more details
D
Darphy
09-15-2017, 07:58 AM #4

Interesting point. Those who know my 5900X will recall the RAM problems I faced. This machine didn’t like the original Ballistix I used—they claimed it was DDR4-3200 with 16-18-18-36 timings at 1.35V. DOCP recognized that spec. My Asus Tuf B550-PLUS suggested Patriot PV432G320C6K, also 16-18-18-36 at the same voltage. These were similarly labeled as DDR4-3200. The board’s DOCP picked up the Patriot DIMMs with odd timings like 16-20-20-40. I had to manually adjust them to 16-18-18-36. That shouldn’t matter, huh? Here’s the odd part: the system completely won’t boot when using the 3200 MHz clock with these modules. It works fine at 2400, but anything higher causes a blank screen until it realizes something’s wrong and jumps into BIOS. I haven’t seen any real RAM performance drops at 2400, just this puzzling behavior. So maybe the auto-detection of 16-20-20-40 is handling this situation? Updated January 15, 2023 by An0maly_76 Added more details

_
_Sunnydayz_
Member
50
09-15-2017, 07:58 AM
#5
RAM performance usually becomes more flexible when running at XMP settings to maintain stability or for added safety. Timings can actually get tighter if you skip XMP, but that’s not always ideal. Generally, faster clocks provide better speed, while slower ones offer more headroom. If you didn’t choose XMP, you might want to opt for the lowest-cost RAM available—unless your device supports memory overclocking. Yes, Patriot memory is involved here. Have you experimented with adjusting XMP/DOCP and fine-tuning sub timings yourself? It’s common to have multiple profiles, like one for a 3200 CL16 and another for a 3000 CL15. As frequency increases, tighter timings become less beneficial because higher bandwidth delivers greater performance gains.
_
_Sunnydayz_
09-15-2017, 07:58 AM #5

RAM performance usually becomes more flexible when running at XMP settings to maintain stability or for added safety. Timings can actually get tighter if you skip XMP, but that’s not always ideal. Generally, faster clocks provide better speed, while slower ones offer more headroom. If you didn’t choose XMP, you might want to opt for the lowest-cost RAM available—unless your device supports memory overclocking. Yes, Patriot memory is involved here. Have you experimented with adjusting XMP/DOCP and fine-tuning sub timings yourself? It’s common to have multiple profiles, like one for a 3200 CL16 and another for a 3000 CL15. As frequency increases, tighter timings become less beneficial because higher bandwidth delivers greater performance gains.

S
Sparkles779
Junior Member
16
09-15-2017, 07:58 AM
#6
The Patriot memory is stuck at 3200. The issue arises because the board's DOCP doesn't match the QVL's expectations. If I apply DOCP's auto-detection of 16-20-20-40 at 3200, would the latency decrease or adjust to 16-18-18-36 as specified in the QVL, while still operating at 3200? Edited January 15, 2023 by An0maly_76 Updated with more details.
S
Sparkles779
09-15-2017, 07:58 AM #6

The Patriot memory is stuck at 3200. The issue arises because the board's DOCP doesn't match the QVL's expectations. If I apply DOCP's auto-detection of 16-20-20-40 at 3200, would the latency decrease or adjust to 16-18-18-36 as specified in the QVL, while still operating at 3200? Edited January 15, 2023 by An0maly_76 Updated with more details.

_
_LilacSoul
Member
183
09-15-2017, 07:58 AM
#7
This seems unusual, are you certain it matches the exact same mem on your QVL? The sequence 16-18-18-36 is more precise than 16-20-20-40, so I question that would be more reliable. If it fails at the latter point, it's unlikely to run smoothly at 16-18 etc. Try using 3200@16-20-20-40, adjust your tRFC, and set it high to test stability. You might also need to relax your tFAW and tRC settings. If it's an Asus board, look for a recommended timing guide on the SPD site for tFaw, tRC, and various tRFC options.
_
_LilacSoul
09-15-2017, 07:58 AM #7

This seems unusual, are you certain it matches the exact same mem on your QVL? The sequence 16-18-18-36 is more precise than 16-20-20-40, so I question that would be more reliable. If it fails at the latter point, it's unlikely to run smoothly at 16-18 etc. Try using 3200@16-20-20-40, adjust your tRFC, and set it high to test stability. You might also need to relax your tFAW and tRC settings. If it's an Asus board, look for a recommended timing guide on the SPD site for tFaw, tRC, and various tRFC options.

K
ketman34
Posting Freak
834
09-15-2017, 07:58 AM
#8
It's a bit puzzling... The Asus QVL suggests 16-18-18-36 for this build... *refer to below*. Yet the board identifies it as 16-20-20-40. What's going on? I'm curious—someone mentioned Ryzen likes dual-rank DIMMs, and my boards seem dual, but could that play a role? It works at 2400, but it doesn't boot with 3200 timings even though the settings look right. Maybe the DOCP auto-detects 16-20-20-40 could enable 3200? Thanks for keeping this on track, OP. Edited Jan 15, 2023 by An0maly_76 Updated with more details
K
ketman34
09-15-2017, 07:58 AM #8

It's a bit puzzling... The Asus QVL suggests 16-18-18-36 for this build... *refer to below*. Yet the board identifies it as 16-20-20-40. What's going on? I'm curious—someone mentioned Ryzen likes dual-rank DIMMs, and my boards seem dual, but could that play a role? It works at 2400, but it doesn't boot with 3200 timings even though the settings look right. Maybe the DOCP auto-detects 16-20-20-40 could enable 3200? Thanks for keeping this on track, OP. Edited Jan 15, 2023 by An0maly_76 Updated with more details

R
Ranger6800
Member
241
09-15-2017, 07:58 AM
#9
Sure, I'm glad you're finding value in our conversation. Keep up the good work!
R
Ranger6800
09-15-2017, 07:58 AM #9

Sure, I'm glad you're finding value in our conversation. Keep up the good work!

C
Charliemc909
Posting Freak
898
09-15-2017, 07:58 AM
#10
I turned on DOCP during those specific settings and everything worked smoothly.
C
Charliemc909
09-15-2017, 07:58 AM #10

I turned on DOCP during those specific settings and everything worked smoothly.