Jump to content

Does Loop Order Matter?

Good day,

I've been watching some videos on water-cooling and noticed how people tend to say loop order does not matter, as long as the res is above the pump (makes sense)

But what I've always wondered is that if my GPU block was directly before my CPU block, the warmer water from the GPU will be used to cool the CPU.

And as I know, fluid temperature (in this case, the water temperature) will affect the temp of the components quite directly.

 

There are arguments saying that the water temp will equalize throughout the loop, but I don't think that's how it works.

If a water temperature is the same before and after the GPU block, it doesn't remove any heat from the GPU as far as I understand (assuming no other characteristic of the water changes)

Therefore, warmer water should be coming out from the GPU block, and thus providing the CPU block with warmer water.

Though it can be neglected if the delta in temperature is low, this can be achieved with high flow rate.

 

If there are in depth investigation, with numbers such as water temp before and after the block, and how it affects the next block, I would love to have a read on it to solve my question.

I would like to understand more before i delve into the world of water cooling

 

Thanks

-sigh- feeling like I'm being too negative lately

Link to comment
Share on other sites

Link to post
Share on other sites

7 minutes ago, Moonzy said:

There are arguments saying that the water temp will equalize throughout the loop, but I don't think that's how it works.

This is how thermo physics work, what you think is irrelevant. It will take a couple of minutes I guess until the whole loop equalise.

Link to comment
Share on other sites

Link to post
Share on other sites

So in my Loop It didnt matter at all if I went CPU to GPU or reverse. I put a rad between those two components and it was still the same. Watch JayceTwoCents watercooling myths video. He explained it pretty well. So your temp will equalize.

 

 

|| CPU: I7 6950X @4.6 (1.35V) || Cooler: CUSTOM EK LOOP || Motherboard: ASUS RAMPAGE V EDITION 10

|| GPU: 2x Vega 64 Strix OC @1762 / 1100 MHz || Memory: 16GB G.Skill Royal RGB 3200 Mhz ||

HDD: Segate Barracuda 3TB || SSD: Samsung 970 Evo 512 GB || PSU: Corsair HX 1200 W || Case: Fractal Design Define R6 Gunmetal || Fans: Corsair HD120 (x4) / BeQuiet Silent Wings 3 (x4) || Monitor: ACER XF27HU ||

Second Monitor: BENQ RL2455HM || Mouse: Logitech G502 Pretus Core || Keyboard: Logitech G810 Orion Spectrum || Headphones: Sennheiser IE80 ||

 

                                                                                                          Buildlog expirience swapping to x99: 

 

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

Loop order makes ZERO difference, literally ZERO.

Main Rig:-

Ryzen 7 3800X | Asus ROG Strix X570-F Gaming | 16GB Team Group Dark Pro 3600Mhz | Corsair MP600 1TB PCIe Gen 4 | Sapphire 5700 XT Pulse | Corsair H115i Platinum | WD Black 1TB | WD Green 4TB | EVGA SuperNOVA G3 650W | Asus TUF GT501 | Samsung C27HG70 1440p 144hz HDR FreeSync 2 | Ubuntu 20.04.2 LTS |

 

Server:-

Intel NUC running Server 2019 + Synology DSM218+ with 2 x 4TB Toshiba NAS Ready HDDs (RAID0)

Link to comment
Share on other sites

Link to post
Share on other sites

Its been tested over and over and like everyone says loop order won't change any temps at all.

Link to comment
Share on other sites

Link to post
Share on other sites

50 minutes ago, Moonzy said:

-

You can see it in the fluid temperatures in this example too.

 

Long story short your typical watercooling pump will be pushing about 4.7 L of fluid per minute, and since your typical loop, including the reservoir hovers at about 1.5 L or so; any segment of water is completely cycled through the loop a few times per minute. So the amount of direct contact that a segment has with a particular region of the loop (waterblock, radiator, reservoir, etc) is really short. The water is moved so fast through the system that  the loop heats up as a whole, and cools down as a whole.

Link to comment
Share on other sites

Link to post
Share on other sites

So the actual answer is loop order does matter to a small degree. It depends on your flow rate and how much wattage your parts are dumping into the loop, but for the most part you won't see an inlet vs. outlet difference of more than a few degrees C.  It's highly unlikely that those few degrees will actually matter to your overall system.

 

People like Jayztwocents oversimplify things to cater to a dumb/immature audience.  The dude's recent video was whether you can watercool with beer, for fuck's sake.  He ain't a genius.

Workstation:  13700k @ 5.5Ghz || Gigabyte Z790 Ultra || MSI Gaming Trio 4090 Shunt || TeamGroup DDR5-7800 @ 7000 || Corsair AX1500i@240V || whole-house loop.

LANRig/GuestGamingBox: 9900nonK || Gigabyte Z390 Master || ASUS TUF 3090 650W shunt || Corsair SF600 || CPU+GPU watercooled 280 rad pull only || whole-house loop.

Server Router (Untangle): 13600k @ Stock || ASRock Z690 ITX || All 10Gbe || 2x8GB 3200 || PicoPSU 150W 24pin + AX1200i on CPU|| whole-house loop

Server Compute/Storage: 10850K @ 5.1Ghz || Gigabyte Z490 Ultra || EVGA FTW3 3090 1000W || LSI 9280i-24 port || 4TB Samsung 860 Evo, 5x10TB Seagate Enterprise Raid 6, 4x8TB Seagate Archive Backup ||  whole-house loop.

Laptop: HP Elitebook 840 G8 (Intel 1185G7) + 3080Ti Thunderbolt Dock, Razer Blade Stealth 13" 2017 (Intel 8550U)

Link to comment
Share on other sites

Link to post
Share on other sites

49 minutes ago, AnonymousGuy said:

So the actual answer is loop order does matter to a small degree. It depends on your flow rate and how much wattage your parts are dumping into the loop, but for the most part you won't see an inlet vs. outlet difference of more than a few degrees C.  It's highly unlikely that those few degrees will actually matter to your overall system.

 

People like Jayztwocents oversimplify things to cater to a dumb/immature audience.  The dude's recent video was whether you can watercool with beer, for fuck's sake.  He ain't a genius.

Loop order only really matters if your flow rate is slow enough to let heat pool. If you are building a custom loop then that should be something you plan to make sure doesn't happen. 

 

So there MIGHT be a 1-2c different in water temp through out your loop, but that is so minuscule it might as well not exist unless you are at the very edge of OC stability... but if you are going to that extreme you are either going ot be using ln2 for that ultimate OC pat on the back... or if for a more daily driver approach you would be using a water chiller or phase changer which both give you more headroom.

 

For 98% of the water cooling community it doesn't make a noticeable difference and from my experience it is more often than not a 1C or lower difference. 

Link to comment
Share on other sites

Link to post
Share on other sites

  • 3 years later...
On 3/1/2018 at 1:48 PM, AngryBeaver said:

Loop order only really matters if your flow rate is slow enough to let heat pool. If you are building a custom loop then that should be something you plan to make sure doesn't happen. 

 

So there MIGHT be a 1-2c different in water temp through out your loop, but that is so minuscule it might as well not exist unless you are at the very edge of OC stability... but if you are going to that extreme you are either going ot be using ln2 for that ultimate OC pat on the back... or if for a more daily driver approach you would be using a water chiller or phase changer which both give you more headroom.

 

For 98% of the water cooling community it doesn't make a noticeable difference and from my experience it is more often than not a 1C or lower difference. 

Yeah ive been disapointed with jays videos on this topic, but as for the rest he just likes to have fun, whats worse is people who claim to have technical aptitude saying it doesn't matter no matter what the case is, and then show one case of over an hour long test its the same on average, or or say some cringe like energy in and energy out remain the same, when the point isnt the overall water temp its the localized temp even if its just a 1mm2 area on a die. or respected people like der8auer making a oversimplified video when overclocking knowledge doesnt equal "fluid dynamics" knowledge. And finally people like eq saying it doesnt matter, but only if you buy our good pump.

 

fact is unless the water is moving in infinitely fast, or you are using a superconductive liquid, there will always be small instantaneous variances. The question was does it matter, some guy says oh it probably wont matter its negligible, and then a bunch of dummies start saying oh it doesn't matter at all, when in reality even if its small, it does in fact matter.

Link to comment
Share on other sites

Link to post
Share on other sites

8 hours ago, Xii-Nyth said:

Yeah ive been disapointed with jays videos on this topic, but as for the rest he just likes to have fun, whats worse is people who claim to have technical aptitude saying it doesn't matter no matter what the case is, and then show one case of over an hour long test its the same on average, or or say some cringe like energy in and energy out remain the same, when the point isnt the overall water temp its the localized temp even if its just a 1mm2 area on a die. or respected people like der8auer making a oversimplified video when overclocking knowledge doesnt equal "fluid dynamics" knowledge. And finally people like eq saying it doesnt matter, but only if you buy our good pump.

 

fact is unless the water is moving in infinitely fast, or you are using a superconductive liquid, there will always be small instantaneous variances. The question was does it matter, some guy says oh it probably wont matter its negligible, and then a bunch of dummies start saying oh it doesn't matter at all, when in reality even if its small, it does in fact matter.

Every time when I hear people say it doesnt matter at all the context is that it wont matter at all in the performance of the PC, is that what you are also saying. I believe it is but if its not and you are saying loop order matters from a temperature standpoint than I would like to see what kind of evidence you use to make that claim. I only ask because I dont think Ive ever seen someone actually prove that.

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

der8auer made a good video on it....so i dont see why it would be wrong. even even showd if the temps are highter in the long run it dose not matter.

 

why the long test is because the water has to heat up so an aio will beat a cpu cooler as it take time to heat up but when the test stops it takes longer to cool down vs a cooler. so most test with aio should be at least be half hour test.

 

if you could some how have a really good fans and a really good rad then can drop the water temp fast enough before going to the other part it might be posable but how fast the water moves thow the rad it cant transfer enough heat there for slowly increasing in temp.

 

but there are other ways of dropping temps before thinking about extreme things.

 

there a balance vs flow rate and transfer.

Edited by thrasher_565

I have dyslexia plz be kind to me. dont like my post dont read it or respond thx

also i edit post alot because you no why...

Thrasher_565 hub links build logs

Corsair Lian Li Bykski Barrow thermaltake nzxt aquacomputer 5v argb pin out guide + argb info

5v device to 12v mb header

Odds and Sods Argb Rgb Links

 

Link to comment
Share on other sites

Link to post
Share on other sites

On 9/23/2021 at 9:16 PM, thrasher_565 said:

there a balance vs flow rate and transfer.

That's called temp gradient. The rate at which thermal is transfered from one place to another, then dissipated and collected to be re-heated. 

 

____

@ Topic OP post necro'd

 

Loop order. 

Pump to CPU water-block. If the pump stops turning, you'll immediately know.

The rest of the build won't matter unless you really want to crunch numbers to shave off a few degree. 

 

Rule of thumb, be able to dissipate greater than what you can produce. Radiator over-head is not a bad thing.

Quiet loops will also require radiator over-head. You can move the same BTU on less CFM.

Always study your waterblocks. More copper will be better with big BTU output. That means study plated blocks for thickness. Thicker is better. This helps thermal gradient. Full copper gives you dissipation inside and outside of the block. ***Hint for overclocking peeps; you can put a fan on a full copper water block and help dissipate heat. A plated block with acrylic top is an insulator. A fan will do very little here.

 

Remember then end factor in any water loop is the ambient temps. If your "case temp" or Tcase temp is high cause it's summer time, then you will have higher temps. More rads won't change this issue. 

Link to comment
Share on other sites

Link to post
Share on other sites

8 minutes ago, ShrimpBrime said:

That's called temp gradient. The rate at which thermal is transfered from one place to another, then dissipated and collected to be re-heated. 

 

____

@ Topic OP post necro'd

 

Loop order. 

Pump to CPU water-block. If the pump stops turning, you'll immediately know.

The rest of the build won't matter unless you really want to crunch numbers to shave off a few degree. 

 

Rule of thumb, be able to dissipate greater than what you can produce. Radiator over-head is not a bad thing.

Quiet loops will also require radiator over-head. You can move the same BTU on less CFM.

Always study your waterblocks. More copper will be better with big BTU output. That means study plated blocks for thickness. Thicker is better. This helps thermal gradient. Full copper gives you dissipation inside and outside of the block. ***Hint for overclocking peeps; you can put a fan on a full copper water block and help dissipate heat. A plated block with acrylic top is an insulator. A fan will do very little here.

 

Remember then end factor in any water loop is the ambient temps. If your "case temp" or Tcase temp is high cause it's summer time, then you will have higher temps. More rads won't change this issue. 

hmm i understand some of it haha i get the acrylic block vs full copper and adding a fan to the cpu block and still benefit even if its just cools the vrms a bit that why the artic aio has a fan on it. (thow its small)

 

i no there were detract mount block with out the his but i guess you need to no people to get one of thows.

 

liquid metal is probably the cheapest thing to drop temps imo if you no the risk.

 

i no linus made a video about how much air can azrob heat has he buitl a one u server for mc and put rads against each other.

 

Major Hardware made a few videos about other options of cooling like blowing cool air at the water or as it fall it mixes? 

 

also did a video about adding bumbles in to the water cant remember it he got it to work or not.

 

i no back in the day there was a zalman external res witch i think was aluminum? with a fan on top i alwes wonder if the res could help cool? but i guess its probably little surface area it probably be better to use replace the res with a rad it self

I have dyslexia plz be kind to me. dont like my post dont read it or respond thx

also i edit post alot because you no why...

Thrasher_565 hub links build logs

Corsair Lian Li Bykski Barrow thermaltake nzxt aquacomputer 5v argb pin out guide + argb info

5v device to 12v mb header

Odds and Sods Argb Rgb Links

 

Link to comment
Share on other sites

Link to post
Share on other sites

19 minutes ago, thrasher_565 said:

hmm i understand some of it haha i get the acrylic block vs full copper and adding a fan to the cpu block and still benefit even if its just cools the vrms a bit that why the artic aio has a fan on it. (thow its small)

 

i no there were detract mount block with out the his but i guess you need to no people to get one of thows.

 

liquid metal is probably the cheapest thing to drop temps imo if you no the risk.

 

i no linus made a video about how much air can azrob heat has he buitl a one u server for mc and put rads against each other.

 

Major Hardware made a few videos about other options of cooling like blowing cool air at the water or as it fall it mixes? 

 

also did a video about adding bumbles in to the water cant remember it he got it to work or not.

 

i no back in the day there was a zalman external res witch i think was aluminum? with a fan on top i alwes wonder if the res could help cool? but i guess its probably little surface area it probably be better to use replace the res with a rad it self

Diving deep into it, the very FIRST issue with ALL processors is the size of the Integrated Heat Spreader that's soldered to the silicon. I've removed a couple dozen soldered chips. And with that, I use a massive copper plate. Probably about double the size and weight of the stock IHS plates. 

 

Since socket 754 AMD's desktop processor IHS plate has not changed in size, shape or thickness to any degree that would be noticeable on your desktop temp readout. This small plate results in a very steep temp curve. Not related to a "temp spike". 

 

So think about thermal transfer and thermal storage.

Then think about time. Does it take less time to heat the larger mass of copper? (hint full copper blocks vs plated)

 

OK so that's essentially the basis of your temp gradients. A poor integrated heat spreader. Quickly heats, 40mmx40mm (only a few mm thick) isn't a lot of surface area and you are only dissipating heat on one of the 6 surfaces of that plate while one other is being heated by the cpu it's self.

 

Now, back to that core soldered to IHS plate deal. As you've seen pictures, that surface area is quiet small. Fits inside a 30mmx30mm area. I don't know the exact dimensions off hand, but this is where "Temp Spike" comes in. The surface area is so small and the transistor density is so great, that it physically cannot be prevented. 

 

All in a nut shell I guess.

Link to comment
Share on other sites

Link to post
Share on other sites

16 minutes ago, ShrimpBrime said:

Diving deep into it, the very FIRST issue with ALL processors is the size of the Integrated Heat Spreader that's soldered to the silicon. I've removed a couple dozen soldered chips. And with that, I use a massive copper plate. Probably about double the size and weight of the stock IHS plates. 

 

Since socket 754 AMD's desktop processor IHS plate has not changed in size, shape or thickness to any degree that would be noticeable on your desktop temp readout. This small plate results in a very steep temp curve. Not related to a "temp spike". 

 

So think about thermal transfer and thermal storage.

Then think about time. Does it take less time to heat the larger mass of copper? (hint full copper blocks vs plated)

 

OK so that's essentially the basis of your temp gradients. A poor integrated heat spreader. Quickly heats, 40mmx40mm (only a few mm thick) isn't a lot of surface area and you are only dissipating heat on one of the 6 surfaces of that plate while one other is being heated by the cpu it's self.

 

Now, back to that core soldered to IHS plate deal. As you've seen pictures, that surface area is quiet small. Fits inside a 30mmx30mm area. I don't know the exact dimensions off hand, but this is where "Temp Spike" comes in. The surface area is so small and the transistor density is so great, that it physically cannot be prevented. 

 

All in a nut shell I guess.

ya now im lost haha way over my head

 

ya i guess water and metal work kinda the same way with metal the thicker the more heat it can take but needs fins to dissipate the heat

 

and water the more water the longer it takes to heat up and needs to thin it out thew a rad to cool it on 2 sides.

 

and i guess the water fall air thing makes the water droplets making the air around it dissipate the heat.

I have dyslexia plz be kind to me. dont like my post dont read it or respond thx

also i edit post alot because you no why...

Thrasher_565 hub links build logs

Corsair Lian Li Bykski Barrow thermaltake nzxt aquacomputer 5v argb pin out guide + argb info

5v device to 12v mb header

Odds and Sods Argb Rgb Links

 

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×