Ncore V1 No Frame Water Block Prototype Testing @ [H]

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,233
Ncore V1 No Frame Water Block Prototype Testing

The Ncore V1 is a prototype CPU water block for custom built PC cooling loops. It has a couple of big draws when it comes to enthusiast features, namely it is a direct-die cooler, and the other feature is that no mounting hardware is required. We have put the Ncore V1 through some cooling and have results to share.
 

keshley

Weaksauce
Joined
May 4, 2016
Messages
109
I'm thinking it's the small fin area that's preventing it from outperforming the RayStorm outright. I know the EK kit to direct mount makes a notable difference in temps, and that's using the same exact block you'd normally use, so all things being equal, must be the fin arrangement.
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,233
I'm thinking it's the small fin area that's preventing it from outperforming the RayStorm outright. I know the EK kit to direct mount makes a notable difference in temps, and that's using the same exact block you'd normally use, so all things being equal, must be the fin arrangement.
I would agree. The fin footprint is tiny.
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,864
Hmm.
I backed his kickstarter because my lid isnt flat on the underside causing a large rise in temps on some cores and fancied curing that (by getting rid of the lid), having the best water block and a bit of fun.
But I can cure my problem with a $13 copper lid made by someone else. I already have a very good water block.
I am looking at that again pending what NUDEcnc has to say.

ps
Thanks Kyle for your testing.
 
Last edited:

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,864
The difference with NUDEcncs results could be down to Kyles use of CLU between the IHS and CPU die.
I bet NUDEcnc didnt use CLU under the IHS when doing comparative testing with his delidded CPU.
 

NUDEcnc

NUDEcnc Inventor
Joined
Jan 2, 2017
Messages
14
Hi again

I wanted once again thank you Kyle for testing the Ncore. Unfortunately, I have to admit that for some reason test performed by Kyle do not reflect my results ☹.

This is the video with my results made by as comparison for reviewers.


Ncore was designed for hardOC (hard overclocking)

For example according to Intel's white sheet for Kaby lake: http://www.intel.com/content/www/us...esktop-s-processor-lines-datasheet-vol-1.html Page 114 lists max voltage as 1.52v.

I personally run my 6700K @1.52V to achieve 4800Mhz. (I did use CLU)

Ncore was created for this type of scenario. @1.35V CPU voltage output is low so heat transfer is not a bottleneck. I do believe that Ncore would be much more competitive at higher voltages.

Regards Arek
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,233
Hi again

I wanted once again thank you Kyle for testing the Ncore. Unfortunately, I have to admit that for some reason test performed by Kyle do not reflect my results ☹.

This is the video with my results made by as comparison for reviewers.


Ncore was designed for hardOC (hard overclocking)

For example according to Intel's white sheet for Kaby lake: http://www.intel.com/content/www/us...esktop-s-processor-lines-datasheet-vol-1.html Page 114 lists max voltage as 1.52v.

I personally run my 6700K @1.52V to achieve 4800Mhz. (I did use CLU)

Ncore was created for this type of scenario. @1.35V CPU voltage output is low so heat transfer is not a bottleneck. I do believe that Ncore would be much more competitive at higher voltages.

Regards Arek
Prime95 SmallFFT load? Load makes all the difference in the world on heat on these CPUs.
 

IKV1476

Lurker
Joined
Dec 26, 2005
Messages
336
Wonderful article. A good solid performance from a prototype.

I do wonder though about the performance of the copper itself. Is it possible that this block, and the raystorm, are limited by coppers ability to transfer heat (heat capacity)? Both had the same flow rate and nearly identical temps. Maybe the blocks are maxxed out so to speak. Or maybe a retest with less radiator?
Just thinking out loud here.
This block is so fasinating, I really want to see it kick ass.
 
  • Like
Reactions: Nenu
like this

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,864
Wonderful article. A good solid performance from a prototype.

I do wonder though about the performance of the copper itself. Is it possible that this block, and the raystorm, are limited by coppers ability to transfer heat (heat capacity)? Both had the same flow rate and nearly identical temps. Maybe the blocks are maxxed out so to speak. Or maybe a retest with less radiator?
Just thinking out loud here.
This block is so fasinating, I really want to see it kick ass.
I also want it to kick ass.
But the issue may have been covered, the lack of fin mm2.
I'm sure NUDEcnc is doing a lot of work to find what went wrong here.
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,233
I also want it to kick ass.
But the issue may have been covered, the lack of fin mm2.
I'm sure NUDEcnc is doing a lot of work to find what went wrong here.
I don't think anything went wrong. First off, that XSPC RayStorm block is far from a slouch and totally outmatch the Ncore in terms of mass and fin footprint. Considering how tiny the Ncore did, I was very much impressed with it.
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,864
Hmm.
I backed his kickstarter because my lid isnt flat on the underside causing a large rise in temps on some cores and fancied curing that (by getting rid of the lid), having the best water block and a bit of fun.
But I can cure my problem with a $13 copper lid made by someone else. I already have a very good water block.
I am looking at that again pending what NUDEcnc has to say.

ps
Thanks Kyle for your testing.
I've decided to stick with it because I dont want to use liquid metal and dont need to with the Ncore block.
Its substantially better than a delid with normal paste.
 
Last edited:

ACC

n00b
Joined
Apr 7, 2018
Messages
35
Hey Kyle, I am curious whether you played around with the flow rate by using the bolts?

Also, do you believe it is comparable in fin size to the other solutions on the market from EK ,Alphacool, XSPC, Aquacomputer, Watercool, and Swiftech? Optimus watercooling innovated in terms of fin size.
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,233
Hey Kyle, I am curious whether you played around with the flow rate by using the bolts?

Also, do you believe it is comparable in fin size to the other solutions on the market from EK ,Alphacool, XSPC, Aquacomputer, Watercool, and Swiftech? Optimus watercooling innovated in terms of fin size.
Yeah, I did. I never saw a difference of more than 1C, but admittedly I did not spend a lot of time with it. It might show to be much more of a feature with other pumps than this D5.

No, it is not comparable in terms of fin size. The footprint is tiny.
 

ACC

n00b
Joined
Apr 7, 2018
Messages
35
I was trying to find the fin size versus the other blocks on the market. If the block is smaller relative to other blocks yet performs similarly, it should have more fins per unit area I guess?
(Arek's promo video suggests that he is limited by his current manufacturing process as his CNC is not an industrial grade one.)

For example, the Alphacool Eisblock XPX states: "fin structure was refined to 0.2mm and the fin count raised to 81 fins".
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,233
I was trying to find the fin size versus the other blocks on the market. If the block is smaller relative to other blocks yet performs similarly, it should have more fins per unit area I guess?
(Arek's promo video suggests that he is limited by his current manufacturing process as his CNC is not an industrial grade one.)

For example, the Alphacool Eisblock XPX states: "fin structure was refined to 0.2mm and the fin count raised to 81 fins".
Fin size is certainly a big part of it, and I am unsure of the dimensions of Ncore's, but what I was referring to mainly was the footprint. A lot more fin surface area on XSPC block we used, and that is certainly one of the bottlenecks when it comes to CPU heat transfer.

So I think the Ncore is trading off one bottleneck for another. Direct die, great, tiny fin footprint bad. Where as on a traditional CPU water block, layers of TIM and heatspreader, bad, big fin footprint, good.
 
  • Like
Reactions: ACC
like this

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,233
Footprint - I understand you referring to the size of the base of the waterblock. If yes, I must disagree, that bigger footprint is always better
This is simulation showing how heat is spreading in theoretical waterblock:

No, I am referring to the footprint of the fins, not the footprint of the entire block.
 

ACC

n00b
Joined
Apr 7, 2018
Messages
35
Linus is a buffoon as usual :p

Arek, I really want to see it succeed. Don't take what Kyle says as a negative thing he is trying to help you make it the best it can be before it launches to the public.
 

NUDEcnc

NUDEcnc Inventor
Joined
Jan 2, 2017
Messages
14
Linus is a buffoon as usual :p

Arek, I really want to see it succeed. Don't take what Kyle says as a negative thing he is trying to help you make it the best it can be before it launches to the public.

Guys, my view on the situation is completely opposite. I'm very grateful to Kyle for testing. His review and this thread is the great way to learn what people perception of my design is.

I truly want to create great design. Kyle clearly points out that this is still a prototype. Now based on people reviews I'm able to understand what's important to you and what's not and improve the design.
I'm working on Ncore day and night.

next big design update is coming next week.
I will keep you updated.

Regards Arek
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,233
Guys, my view on the situation is completely opposite. I'm very grateful to Kyle for testing. His review and this thread is the great way to learn what people perception of my design is.

I truly want to create great design. Kyle clearly points out that this is still a prototype. Now based on people reviews I'm able to understand what's important to you and what's not and improve the design.
I'm working on Ncore day and night.

next big design update is coming next week.
I will keep you updated.

Regards Arek
Gotta like that attitude.
 

Aenra

Limp Gawd
Joined
Apr 15, 2017
Messages
191
Empirically speaking, i'd have agreed with Kyle. But since you were kind enough to provide an equation?
The larger the block's (actual) size, the larger the surface area and the larger the Delta between contact size and upper size (in the equation above, A and Dt).
So while what NUDEcnc's saying (apologies for forgetting your name) makes sense, it only amounts to one variable among many (in the equation above, L).

..right..? :)

Anyway, i hope you get this out quick, because i've heard Derbauer is making a similar block, i believe he even mentioned it in his last video, so one assumes it's imminent; bad timing, nothing beats a good PR.
So best of luck! Never lived under a bridge, but i know what it's like counting pennies outside the local supermarket (because inside, people whould see you doing it and why would you want that)
 
Top