Should i switch to an AIO?

[DookieNukie]

Yeah have fun doing that in more than two 16 GB DIMMs.

do you like think im the first person to do this? if so im flattered

 

[dgianstefani]

do you like think im the first person to do this? if so im flattered

You would be the first to run 128 GB of RAM at 7800 MHz. Pity you don't comprehend that.

 

[DookieNukie]

You would be the first to run 128 GB of RAM at 7800 MHz. Pity you don't comprehend that.

im not going to idk why you assumed it

 

[sneekypeet]

Corsair Vengenance DDR5 is pretty good for options with/without RGB.

The RGB circuit generates potential interference and takes from the power budget of the DIMM, so going without gives you more headroom for MHz or other tuning.

The diffuser is also a piece of plastic that reduces the effectiveness of the heatspreaders.

Here we go again, just on another tangent.

As a guy who tests ram, nothing you say in this comment has real bearing, its just ideas floating around to make those who dont like RGB feel better.

 

[DookieNukie]

nice!, just FYI, cinemad4d recommends up to 32GB and octane up to 64 GB. But if you want to go 128 GB for fun, by all means.

Yeah i figured why not i never had money before so now that i do i wanted to just go all out

 

[dgianstefani]

Here we go again, just on another tangent.

As a guy who tests ram, nothing you say in this comment has real bearing, its just ideas floating around to make those who dont like RGB feel better.

I guess that's why the RGB in mice affects the polling rate, because interference isn't a thing. And the extensive testing on temperature/stability in DIMMs done by ircow and others is also not true.

I use RGB in my own PC, dominator plat DDR4. Got nothing to do with my like/dislike of RGB, it's just physics.

 

[sneekypeet]

I guess that's why the RGB in mice affects the polling rate, because interference isn't a thing. And the extensive testing on temperature/stability in DIMMs done by ircow and others is also not true.

I use RGB in my own PC, dominator plat DDR4. Got nothing to do with my like/dislike of RGB, it's just physics.

I dont see IR here making such claims as it has nothing to do with the thread.

RGB means nothing, I get the same OC with or without them. Temps may vary slightly, but still within spec. So what now?

 

[ir_cow]

the Z790 Hero lists support for DDR5-7800+

I'm just letting you know its of course partly dependent per MB, but the main restraint for 4x dual-rank memory is the CPU IMC. Probably will max out at DDR5-5600, but I wouldn't be surprised if its lower.

 

[DookieNukie]

I'm just letting you know its of course partly dependent per MB, but the main restraint for 4x dual-rank memory is the CPU IMC. Probably will max out at DDR5-5600, but I wouldn't be surprised if its lower.

from what i read people experience issues at like 5200 because of a recent bios update. i'm not too worried.

 

[Punx223]

I'm just letting you know its of course partly dependent per MB, but the main restraint for 4x dual-rank memory is the CPU IMC. Probably will max out at DDR5-5600, but I wouldn't be surprised if its lower.

I've run up to 6000 4x32GB, but it requires some serious tuning, and the right board with really good/robust layout. (GBT Xtreme in my case)

Honestly, for anyone looking for stability, 5600 CL38 on a good IMC would be the best option, or 5600 CL40 is your IMC is meh.

 

[dgianstefani]

I dont see IR here making such claims as it has nothing to do with the thread.

RGB means nothing, I get the same OC with or without them. Temps may vary slightly, but still within spec. So what now?

Well we know that the RGB signal causes interference that can affect clock generating hardware in mice for example, wired or wireless.

We also know that LED light strips use 1-3W for a CPU block, and about 60% of that power is turned into heat (still much better than the 95% of incandescent lighting).

We also know that plastic is a terrible thermal conductor.

The rest is just logic. If it's offtopic I can understand, but someone asked.

 

[A Computer Guy]

Oh, that's right, I forgot, the custom loop master race are gods among men and their loops are impervious to leaks.

You didn't know that? Now pay homage while I double check all my compression fittings.

 

[Vya Domus]

No one is trying to break RAM OC world records here, this is inconsequential.

 

[sneekypeet]

Well we know that the RGB signal causes interference that can affect clock generating hardware in mice for example, wired or wireless.

We also know that LED light strips use 1-3W for a CPU block, and about 60% of that power is turned into heat (still much better than the 95% of incandescent lighting).

We also know that plastic is a terrible thermal conductor.

The rest is just logic. If it's offtopic I can understand, but someone asked.

View attachment 281077

I never said it wasnt happening, it doesnt matter to just about everyone but you.

 

[dgianstefani]

I never said it wasnt happening, it doesnt matter to just about everyone but you.

Well that's a good attitude on a technical forum.

No one is trying to break RAM OC world records here, this is inconsequential.

No, but someone is trying to run 128 GB of RAM, which is difficult enough at JEDEC speeds on a top tier board.

 

[sneekypeet]

Well that's a good attitude on a technical forum.

And pressing your opinion as gospel is?

 

[ThrashZone]

Hi,
Asus rgb crapware "only one I know of" had a fun history with ddr4 corrupting timings and now we enter armoury crap

 

[ir_cow]

I've run up to 6000 4x32GB, but it requires some serious tuning, and the right board with really good/robust layout. (GBT Xtreme in my case)

Honestly, for anyone looking for stability, 5600 CL38 on a good IMC would be the best option, or 5600 CL40 is your IMC is meh.

Impressive it even booted! But most people aren't going to do anything besides enable XMP. I'm trying 128GB DDR5-5600 right now on the MSI Z790 ACE, it doesn't even boot. Maybe if I manually set the voltages it would... Not willing put time into it. Stability tests takes a whole day. I seriously doubt that 6000 MT/s 128GB was fully stable.

 

[dgianstefani]

And pressing your opinion as gospel is?

What part of the facts I've laid out is opinion?

 

[sneekypeet]

What part of the facts I've laid out is opinion?

Where it matters to the rest of us in this thread you have taken way off topic the entire time past your original comment?

 

[dgianstefani]

Where it matters to the rest of us in this thread you have taken way off topic the entire time past your original comment?

Custom loop and fan hardware discussion
Fan hardware
Air vs custom loop vs AIO discussion
Discussion about relative performance between different cooling options
Discussion about cost/longevity of different cooling options.

Which of these are offtopic?

Original topic was should I use air or AIO.

 

[Punx223]

Impressive it even booted! But most people aren't going to do anything besides enable XMP. I'm trying 128GB DDR5-5600 right now on the MSI Z790 ACE, it doesn't even boot. Maybe if I manually set the voltages it would... Not willing put time into it. Stability tests takes a whole day. I seriously doubt that 6000 MT/s 128GB was fully stable.

It was, but trust me not worth the effort for most

 

[ir_cow]

It was, but trust me not worth the effort for most

As you said, it requires some TLC and a MB that can handle it. Not to mention some luck of the draw in terms of CPU IMC quality. Just shows that even with experience like yourself, 6000 MT/s is out of the question for 99% of users.

 

[Punx223]

As you said, it requires some TLC and a MB that can handle it. Not to mention some luck of the draw in terms of CPU IMC quality. Just shows that even with experience like yourself, 6000 MT/s is out of the question for 99% of users.

absolutely dude, honestly 5600 is attainable by most CPUs if you are reasonable on timing, but I have had some really horrible IMCs and those may have to bin down to the 5200 range.

However, if you're talking AM5, well, you already know the score.

 

[ir_cow]

However, if you're talking AM5, well, you already know the score.

Just finally able to do 4800 on AM5 with the newest AGESA 1.0.0.4. What a pain too.