Overclock.net banner
Status
Not open for further replies.
1 - 20 of 33 Posts

·
Overclocker
Joined
·
5,589 Posts
Discussion Starter · #1 · (Edited)
Ncore is a prototype water block that has made its way to us for testing, and it is interesting enough for use to give it some test bench time. This V1 model of the Ncore uses the LGA 1151 socket to mount it to your delidded Intel processor. This block is tiny compared to almost any other air or water cooler on the market. Let's see what it is about.
https://www.hardocp.com/article/2018/03/05/ncore_v1_no_frame_water_block_prototype_sneak_peek




:eek:

A waterblock for cpu die itself. What an interesting original idea for delidder's! I honestly don't see the need for silver compound. Regular compound should do. A very thin layer I might add. I have to admit I did cringe a bit when he closed the socket. As the W/B slide into locking position I did wonder if it would scratch the die. Which brings up another question. What's the best way to apply tim? On the die or on the cold plate of the waterblock? Because I wouldn't add them to both. You run the risk of scarping tim off onto the die. As you close the socket the waterblock will slide on top of the die.

Since it will compete against the XSPC Raystorm how long will it be before we start seeing competition? XSPC I'm looking at you :p
 

·
Premium Member
Joined
·
6,916 Posts
I will be interesting to see. A lot of tests show better water cooling results with the IHS still on than running on the naked die. Something about better heat spreading so that more can be pulled off by the block
 

·
Eh, Wha?
Joined
·
9,819 Posts
Well, that's something.
 

·
Moderately Ancient
Joined
·
371 Posts
I love seeing products that throw years of conventional wisdom out the window and try something radically different. I'd really like to see a much finer machined finish on the area where the CPU core makes contact because of the sliding action from the retention system. I also worry a bit about the strength of those little copper tabs that hold the CPU block down since copper is such a soft metal. I look forward to seeing some results
 

·
Premium Member
Joined
·
6,916 Posts
I love seeing products that throw years of conventional wisdom out the window and try something radically different. I'd really like to see a much finer machined finish on the area where the CPU core makes contact because of the sliding action from the retention system. I also worry a bit about the strength of those little copper tabs that hold the CPU block down since copper is such a soft metal. I look forward to seeing some results
The IHS on these CPUs are made of copper, and thousands of people have used them to hold the IHS down after a delid just fine, so this shouldnt be any different.
 

·
Overclocker
Joined
·
5,589 Posts
Discussion Starter · #7 ·
But I seriously doubt that silver tim is needed here. As a matter of fact I have to question if Vaseline or some others safe lubricant is more ideal to protect the die.
 

·
TEC Veteran
Joined
·
2,010 Posts
TIM is still needed on the die. I would use CLP/CLU, and apply it to both the die and the underside of the block since it is so thin.

Keep in mind in the late 90s/early 2000s many CPUs did not have IHSs - just 4 foam circles on each corner. As dies got smaller and CPUs got thinner, they were added to better spread the pressure of the heat sink and prevent installation errors from ruining CPUs. I have cracked naked dies before by running modified mounts for higher pressure, so I normally prefer an IHS for safety, but due to how this mounts I would be comfortable running a naked chip. You cannot over tighten the mount unless you stupidly add spacers on the block tabs.

On chiller setups though, many have found better performance by leaving the IHS on since it increases contact area and makes it easier to pull the heat off the chip. Not sure where the coolant temp breaking point is where no IHS becomes worse than with IHS, but it is below ambient so not an issue for 99% of users.
 

·
Premium Member
Joined
·
1,108 Posts
This does look interesting, but that sliding action when the IHS meets the die is concerning. It will be interesting to see just how difficult TIM application is to get right, as well as how the block performs.
 

·
Eh, Wha?
Joined
·
9,819 Posts
Doesnt the whole CPU package slide a bit when the clamp is closed?
Absolutely not. The lid has plenty of wiggle room, but the socket itself does not. Furthermore, the amount of clamping force is enough to keep it solidly pinned in place.

Incidentally, if your CPU package did wiggle around in the socket, you would be lucky to even boot before it crashed from instability, and even if you did, you would be likely to literally catch the CPU on fire from heat due to high electrical resistance.
 

·
Jedi Knight
Joined
·
682 Posts
Oh man, this is awesome. I can't wait to see it's performance. I loved the EK bare die kit when I had my Haswell and was really disappointed when I upgraded to Skylake to learn I wouldn't have a bare die option like that anymore.
 

·
TEC Veteran
Joined
·
2,010 Posts
EK used to make a mount kit to allow their blocks to be used bare die. Was worth ~3c vs a delidded chip with CLP/CLU.

Hoping this block would be worth at least 5c vs the top blocks on a reseated IHS since it is designed from the ground up to cool the die directly, and not just a traditional block mounted on die.
 

·
Registered
Joined
·
4,038 Posts
I will be interesting to see. A lot of tests show better water cooling results with the IHS still on than running on the naked die. Something about better heat spreading so that more can be pulled off by the block
More probably due to poor contact. It's much easier to make a good contact on a larger surface.
 

·
Smug, Jaded, Enervated.
Joined
·
1,235 Posts
Laptops have always had bare dies... I'm really surprised it has taken someone this long to do this for desktops.
It hasn't...desktop CPU's originally didn't have IHS's back in the day, 20+ years ago.

Nor is this the first time this has been done in terms of water cooling..the interesting thing about this design is its use of the CPU socket retention bracket, as previously with the EK IB naked die kit, part of the installation required removing the hold-down bracket completely, and the CPU just sort of floated on the pins in the socket, and was a little finicky to install.
 

·
Registered
Joined
·
3 Posts
Hi NUDEcnc here!
I would like to thank the forum for the warm welcoming of the design and novelty idea behind this project.

Yeah, he will release a benchmark review using the XSPC Raystorm with the IHS on the cpu. I'll update this thread once it's out.
I send 4 Ncores for a review to this English-speaking reviewers and Techlipton (as I’m Polish not English as stated in the video ). But I’m working on a second batch of 10 blocks after major design revision.
Would you @EastCoast be interested in doing independent testing for overclock.net ?
Design is in prototype stage so next revision will have many changes. Arek
 

·
The 6502 Still Rocks
Joined
·
777 Posts
Odd that the reviewer calls the top 'copper'. Has he never seen brass in his life?

It's a nice design. It will be interesting to see how it performs. I'm not sure why people think some thin copper "heat spreader" with an over-plating and Intel Pigeon Pooh is going to magically move heat around better than something with a direct copper (just like the IHS) contact cooling plate. I think the IHS does more die protecting in many cases. If the contact block is aluminum or nickel over copper with directly exposed heat pipes etc, yeah maybe it's heat spreading to some degree.

Now if the IHS is soldered to the die ... that's different.

Can't wait to see some test results.

Edit: He corrected himself on the copper v. brass in the written article. I just thought it was odd. I guess I would stumble around a bit if I had to do a video as well so my apologies to the original author.
 

·
Not a linux lobbyist
Joined
·
2,925 Posts
I will be interesting to see. A lot of tests show better water cooling results with the IHS still on than running on the naked die. Something about better heat spreading so that more can be pulled off by the block
I ran a 4770k naked by removing the mounting bracket and tossing some washers into the mounting springs of a Cooler Master 120 Seidon xl and the same thing happened to me - worse cooling than with the ihs on. But I think my aio cooler went pretty cheap on the thickness of the copper in the pump so that was probably the reason.
It wasn't exactly the best cooler out there.
This one also seemed like the copper got a little thin in the die region. Copper is cheap and they should make it at least as thick as the ihs. If they are worried about cranking on the pins, they can include a support bracket that goes in the regular cooler mounting holes. The important thing is losing an unnecessary interface. They could always rgb up a support bracket if they were worried about the looks. Everyone self conscious of such stuff loves rgb.
 
1 - 20 of 33 Posts
Status
Not open for further replies.
Top