[Review] CPU Block Roundup #2 Begins: EK Supreme and EK Supreme LT Tested
Preface
This is a quick version of my EK Supreme and EK Supreme Testing....it's part one of my Roundup #2 of waterblock testing. I have some cool things in store for testing (including ways to significantly improve the performance of some blocks) and will be testing a lot of blocks. Anyway, I'll cut the boring parts out of the test logs and post up the EK results in this thread
If you have any questions, check out the full review (link above)...if they're not answered, ask away here
Test Results
First up, finding the best orientation for each block:
Here we can see the Supreme pulling ahead of the Supreme LT, with both showing similar losses when oriented incorrectly. The best orientation for each is positioning the barbs parallel to the top of the board (or perpendicular to the socket latch). Specifically with the LT, having flow go from right to left (from the RAM slots to the I/O ports) nets a small but noticeable improvement. This is likely because Bloomfield processors have the die off-center, closer to the RAM slots.
Now that we know which orientations are best for each block, let's run the full flowrate spectrum and see how they respond to flowrate.
Very High Pumping Power: All three MCP355 pumps and the D5 are on at full speed--this has a very similar PQ curve to a pair of RD-30s at 20V.
High Pumping Power: Two MCP355s with EK V2 tops are on at full speed. The other two pumps are off.
Medium High Pumping Power: A single MCP355 with XSPC V3 top is on at full speed. The other three pumps are off.
Medium Pumping Power: The stock D5 is on at full speed and setting 5. The other three pumps are off.
Low Pumping Power: A single MCP355 with XSPC V3 top is on at minimum speed (~7.7V, ~2450RPM). The other three pumps are off.
Very Low Pumping Power: The stock D5 is on at minimum speed--setting 1. The other three pumps are off.
Note: I do 5 mounts at "Medium High" then take the best config of a block and test the whole flow spectrum (after a TIM curing session) then realign that curve with average of the 3 median mounts to give you the "Adjusted" data.
Other Graphs
More graphs for your enjoyment...let's start with reusing the flow vs. temperature data, but including pump heatdump (i.e., CPU vs. air temps). I have two iterations of it: CPU temperatures vs. my air temperatures and a setup with my water-to-air delta included twice more. The latter is to mimic a setup with one third the radiator power of my setup (roughly a 120x3 radiator with 1600RPM fans).
Note: these results are derived from adding the water-to-air delta three times to my water temps. I add them three times to emulate the radiator power of a loop with 1/3rd the radiator power mine has. I use 2xMCR320s with push-pull 2200RPM Yate Loons and the data emulates the conditions of a loop with a single 120x3 radiator with ~1600RPM fans.
Here we can see both blocks showing benefit (even if small) for all pumping powers tested on my testbed. Until you back down the radiator power that is...at that point both blocks really only benefit up until using dual DDCs. Past that and the heatdump from the pumps outweighs the improved block performance.
Conclusion
First things first, I'm really excited that I finally get some data where the flow vs. temperature curves are crossing! As for how the blocks compare, it's pretty data-centric. At extremely low pumping power, the LT is ahead, but at every other pumping power, the original Supreme is as good or better. The mounting systems are identical, so that's a wash. Aesthetics aren't for me to decide, so again a wash. Yes, the Supreme is more restrictive, but not by a staggering amount; the added restriction is definitely worth the improved temperatures. Pretty simply, the Supreme is the better block in my book.
I do want to point out the mounting system though--I'm not fond of it. I don't mind assembling the screw system for mounting it, but it's a noticeable drawback behind the likes of Koolance, Swiftech, and D-Tek. Furthermore, the thumbnuts EK uses are pretty bad. They work fine, but after 30+ mounts with them, the sides and pads of my index fingers were totally raw--either they need to be made taller to reduce pressure on the fingers, or just be made less sharp. No other thumbscrew or thumbnut has left me clamoring for another one, but these did.
When compared against each other, the two products fit into their respective price brackets well--the Supreme is the better block while the LT is the lower cost, lower performing offering. Where things get interesting will be seeing how the two blocks compare to their competition. More reviews are rolling out shortly, so stay tuned.
It is interesting that you got the results you did. While your review seems to agree with EK's stance on the blocks it disagrees with other reviews (review?).
It is interesting that you got the results you did. While your review seems to agree with EK's stance on the blocks it disagrees with other reviews (review?).
Yes... one thing was on orientation.
i remember telling people orientation on the supreme was important.
Another well done set of tests and stellar presentation!
my good friend!
Quote:
Originally Posted by faster3200
Excellent review and write up once again.
It is interesting that you got the results you did. While your review seems to agree with EK's stance on the blocks it disagrees with other reviews (review?).
Well, even though Vapor and I have two different benches and each have our own Supreme and Supreme LT blocks our data sets correspond quite well. Yeah I know, my data hasn't been released yet and Vapor has been dogging me about that. So after the SR1 review comes out, I'll drive the CPU block write-up to completion with lapped 920 data to come later.
Anyhow, I don't want to derail the thread just wanted to point out that my data corresponds to Eric's testing and not some of the others I've seen.
__________________ skinnee labs test schedule: In Testing: Too much to list. Writing Reviews: finally getting my backlog cleaned up too. Planned Tests: Plan, what the heck is a plan.
Last edited by skinnee; 09-23-2009 at 10:40 PM.
Reason: man I suck w/o spell check.
using the information, can someone estimate what would be the best use for these waterblocks for core2quad. as it as dual dies, unfortunately this information is next to useless for me atm
Parallel Loop using EK MultiLink Reservoir CPU Loop: Heatkiller 3.0 CU (Nickel plated) - Feser 360 (S-Flex G) - MCP350 + EK Plexi Top - EK Reservoir
NB/SB + GPU Loop: EK 280 Nickel Plexi - Bitspower 780i Fullcover - Feser 240 (S-Flex G) - MCP355 + EK Plexi top
No clue....those were the settings for Round #1...but I'm using 1.52V @ 4.2GHz now (more heat). KL-350, SW GTZ, and Fuzion V2 will all be retested for Round 2.
These results are not comparable to preliminary results I showed before (at what appeared to be the same settings). Since the prelim results, I've gone up to triple channel and changed subsystem voltages. I expect official round #2 results to be hotter than the prelim tests.
Quote:
Originally Posted by Bojamijams
Its interesting on the last graph that it shows there IS such a thing as too much flow. Even for the supreme.
It's not that there's too much flow...it's that there's too much pump
With a 'normal' setup of MCR/PA/RX-120x3 with 1600RPM fans, the heatdump from the pumps does negatively affect CPU temps past dual DDCs. There will definitely be blocks that have their best performance at lower pumping power (the next results I'm posting will have at least 1 such block) and maybe a block or two that scales all the way to 4 pumps (probably not though)
Vapor I am still meaning to send you that stuff....... Very busy at work now. Sorry for the delay!
Not a problem...I have Heatkillers LT and LC, Sapphire Rev A, Fuzion V2, GTZ, GTZ SE, three Alphacools, Phobya, and the CPU-345 to test in addition to the CPU-350
Quote:
Originally Posted by Hondacity
excellent..@1.52v there is indeed alot of heat as compared to 1.46v
i hope it won't take you long
It won't be too long...testing is a lot faster now. Not sure the order I'll do them in though.
no wai. You have the enzotech sapphire? I want test results on it. Want to vertify hesmelaugh's results please.
Nice testing btw.
I can't believe that the lt and supreme are only 1.5 degrees away from each other at 1.5gpm.
Yeah, I'm curious about the Sapphire too....definitely a super low restriction block though, I was getting insane flowrates from when I connected it for a few min
One thing I want to point out (since I still haven't done a walkthrough of my testbed yet), is that flowrate through the radiators is controlled. Every test has the exact same flowrate through the radiators, so any gains you see are solely gains from the waterblock responding to flow.
Yeah, sorta....that thread didn't really go in depth about it...the focus was more on the results I guess, shrug. I need to get pictures up when it's not collecting data...it's ugly but it's beautiful in its own way
Working on Heatkiller 3.0 testing again...will be adding (hopefully) a couple of simple ways to improve them as well
Parallel Loop using EK MultiLink Reservoir CPU Loop: Heatkiller 3.0 CU (Nickel plated) - Feser 360 (S-Flex G) - MCP350 + EK Plexi Top - EK Reservoir
NB/SB + GPU Loop: EK 280 Nickel Plexi - Bitspower 780i Fullcover - Feser 240 (S-Flex G) - MCP355 + EK Plexi top
Considering the divider at the inlet (which I have to believe is there for C2Q), I'd think vertical is best? Unless I'm getting my core orientation mixed up on C2Q.
I love the improvement advice. The Fusion V2 results prompted me to buy one and test it out, until I realized I have no means of actually testing it .
Ah see, thats where you're expecting too much. Test it with whatever means you have. Download RealTemp, Prime/OCCT, and get something to measure ambients (be it a thermostat or anything you can just set near the machine)... test with whatever gear you have available and test out your own theories.
__________________ skinnee labs test schedule: In Testing: Too much to list. Writing Reviews: finally getting my backlog cleaned up too. Planned Tests: Plan, what the heck is a plan.
Ah see, thats where you're expecting too much. Test it with whatever means you have. Download RealTemp, Prime/OCCT, and get something to measure ambients (be it a thermostat or anything you can just set near the machine)... test with whatever gear you have available and test out your own theories.
See, I totally planned on doing this exactly. Then I realized all my spare comps were either sold or are being used and I can't cannibalize my current comp every time I want to make a change.
I think I might just end up buying some QDCs for my CPU block though. Ahhh WCing, what a whore you are and I am your .
Is the Supreme still really restrictive? Or is it less restrictive than the LT??
Supreme is way more restrictive than the LT.
I have a Supreme GOLD here and it has a different midplate (don't know how it performs yet), so maybe that change is going to be implemented over the entire line?
I have a Supreme GOLD here and it has a different midplate (don't know how it performs yet), so maybe that change is going to be implemented over the entire line?
pic please..i have the copper top..and reusing the original midplate from the acrylic top...
my good friend!
Supreme reigns supre. . .oopppss, wrong forum. 
Linear Mode
