D-Tek Fuzion V2 full review posted

[Anemic]

Great review Martion, keep up the good work! Im looking forward to test this block out on my very hot Q6600

[Repliquant67]

Nice work
With stronger pump than a D5, the fuzion v2 should not have better results with nozzles ?
I have mine at home, one ddc with EK top and I'm waiting a sanso pump.

[Jimmer411]

So I wonder how much of an improvement this is over my classic apogee...


Martin do you have any plans to test older waterblocks like the apogee, rbx and tdx? It would be kinda nice to see how things have improved over time, and how some of those older blocks fare on newer tech.

[Xilikon]

Whoa, I didn't expect to see a bigger improvement than what I thought with the V2. I guess this would be a great recommendation for everyone who ask which is best

I'm even tempted to replace my V1 with the V2 but since 2-3C difference doesn't change my overclocking a lot, I'll wait

[[XC] NetburstXE]

Great review, thanks! I'm surprised at the improvement over the V1.

[Snyxxx]

I must be slow, I didn't have any that far off. My worst case was the V1 testing, but it was only a 1.95C delta. I think a majority of that is due to the mounting plate on V1 being loose, it's much more difficult to get the block level with the plate loose.

As far as the V2 results go, I tested three different configurations. Stock (5 mounts), Stock + Quad Nozzle (only 2 mounts before giving up), and Stock plus 4.5mm nozzle (5 mounts).

Mounting consistency was much better on the V2. Stock worst to best was only .58C.

Martin,

I think what Andy was saying is the temperature scale on your graph is coretemp-water in temp and not necessarily true delta CPU core temperature.

Does this mean the actual temperatures from coretemp actually followed the same delta? Maybe I am being slow if it is.

[junkun13]

4 degrees lower! i should upgrade to V2

[clock_ed]

Great review thanks. they are nice blocks the the pro mount kit as standard makes the block great value.

As the V2 block has a higher restriction than the V1, will this mean that a greater flow rate from a more powerful pump will yeild cooler tamps?

Also that XSPC Edge block looks like a great performer. As it has better temps than the D-tek V1 and is almost as free flowing. Therefore is this the block most people should get if they have a multi block loop?

Cheers

[MikalCarbine]

Would this be a noteworthy upgrade from a Storm for a dual core C2D. I remember the benefit from Storm->Fuzion v1 for a dual core was about 2-3C, not worth the money unless you have some to burn IMO

[Martinm210]

Thanks for all your hard work, 3c difference on average is nice.


I wonder how it compares to a Supreme.



*runs away*

I'll have a supreme coming pretty soon so I'll give a whirl.

Thanks for the testing Martin great work! Now tell me why does it have to slap the V1 all over town, give it a wedgie and take it's lunch money?

Anyone want to buy a 2 month old V1 with an EK Supreme backplate for a good price?

Hehe. It was a bigger gain than I expected too.

Great review Martin, but... WOW, 3 full degrees improvement? Isn't that quite something? With that pressure drop I guess one has to put that thing in it's own loop, but hell isn't that what everybody in here recommends? So far, D-Tek really did accomplish something here.

I don't think the restriction is very much without nozzles in there, it's more of a low/medium restriction compared to other blocks and a D5 or DDC would have plenty of pumping power. Give it a try in the flow rate estimator, you'll probably get more than 1.5GPM in most cases unless you were running a highly restrictive GPU block.

Martin, thanks so much for the review. I love the fact that all those blocks are within 3-5C of each other. Just shows how far the industry has come since we were drilling parallel holes in copper bar...makes you wonder if much more can be done. Then again, people were saying that 4 years ago!

Yeah, although I still like seeing people make their own blocks, it was much more challenging back then, now we have all these great blocks to bolt on..

Nice work
With stronger pump than a D5, the fuzion v2 should not have better results with nozzles ?
I have mine at home, one ddc with EK top and I'm waiting a sanso pump.

It's a good question, I think there is possibly some gain with nozzles, but it's just hard to measure. Stronger pumps and higher heat loads like quad cores probably make the best use of them. I figured most people run either a DDC or a D5, so that's what I chose to test with. The two nozzles tested on my dual core did not show any measurable improvement... But I would still try some nozzles with quads and stronger pumps, there could be something there I didn't capture with the D5/dual core setup.

So I wonder how much of an improvement this is over my classic apogee...


Martin do you have any plans to test older waterblocks like the apogee, rbx and tdx? It would be kinda nice to see how things have improved over time, and how some of those older blocks fare on newer tech.

I am going to try an apogee drive and I have a couple of other blocks I'd like to catch up on that have been collecting.

Whoa, I didn't expect to see a bigger improvement than what I thought with the V2. I guess this would be a great recommendation for everyone who ask which is best

I'm even tempted to replace my V1 with the V2 but since 2-3C difference doesn't change my overclocking a lot, I'll wait

Yeah, and that's a gain at 100% TAT load, so it's a fairly extreme heat load.

Martin,

I think what Andy was saying is the temperature scale on your graph is coretemp-water in temp and not necessarily true delta CPU core temperature.

Does this mean the actual temperatures from coretemp actually followed the same delta? Maybe I am being slow if it is.

I'm the slow one. What I did was log core temperatures using TAT after loading 100%. The delta I'm showing is the average of both cores subtracting out the water in temperature. This does a good job of correcting out the ambient air temperature variations which I also logged during the entire test. This does correlate to the core temperatures as you see in TAT, this is similar to "Coretemp" but a different program. It's much higher than what is registered by the motherboard as CPU temperature.

Anyhow all the numbers are TAT's logged core temperatures subtracting the water in temperature my crystalfontz was logging at the same time. Using the crystalfontz and TAT at the same time makes testing pretty easy. I just mount the block, fire up TAT and load at 100% for each core and walk away for 10 minutes. Then I come back and start logging the crystalfontz and logging on TAT both at the same time. I then walk away for half an hour and come back, shut off the logging/loading then average all the results. Then I'm ready for another mount.

4 degrees lower! i should upgrade to V2

The average of the mounts stock vs stock was actually closer to 3 degrees, but my variability was much higher in V1. The new solid hold down plates are a big improvement to helping with mounting consistency..

Great review thanks. they are nice blocks the the pro mount kit as standard makes the block great value.

As the V2 block has a higher restriction than the V1, will this mean that a greater flow rate from a more powerful pump will yeild cooler tamps?

Also that XSPC Edge block looks like a great performer. As it has better temps than the D-tek V1 and is almost as free flowing. Therefore is this the block most people should get if they have a multi block loop?

Cheers

I don't know, the c/w curve looks similar to V1, so I suspect there could be a degree or two to be had, but it all depends on the added heat by the pump. In the end it's possible you could go backwards if the negatives from the heat dump are greater than the added flow rate.

[NaeKuh]

how come for both you and zorlin does the D-tek V2 perform the same with and without nozzle?

Im not understanding this. Then it wouldnt make sense to have the nozzle in there.

[nikhsub1]

Umm, zolkorn tested a V1 IIRC, not a V2.

[Martinm210]

how come for both you and zorlin does the D-tek V2 perform the same with and without nozzle?

Im not understanding this. Then it wouldnt make sense to have the nozzle in there.

He had a V1.


I'm also only using a Laing D5 and my test was on a dual core. There might be some benefit there, but it's very small and too small to really measure. Maybe it makes a bigger difference with a dual pump or Iwaki setup, I'm not sure.

I think most of what people saw for nozzle kit gains in the V1 was actually just the washer which V2 fixed.

I've got my quad core now, so I'll be doing ALOT more testing, as it sort of means starting all over again. I've got an EK supreme now too, so I might make one last set of runs on the E6600 before switching over to quad.

[Garrett]

This block WILL replace my V1... period
Thanks again Martin

[Big_Daddy]

quad, quad, quad.. yeah.. can't wait.. Now, all I have to do is find a home for a v1 with everthing... Oh, and get a v2..

[TheThreeDegrees]

probably a silly question but if you dont ask you never know!

when you are testing the various nozzles are you lads just popping the nozzle in the chamber from the barb end without opening the whole block?
I'm pretty sure that the nozzle just sits into place when you do so

Only problem is getting the bugger out is difficult.

Just want to avoid having to reseat the block everytime.

[Garrett]

probably a silly question but if you dont ask you never know!

when you are testing the various nozzles are you lads just popping the nozzle in the chamber from the barb end without opening the whole block?
I'm pretty sure that the nozzle just sits into place when you do so

Only problem is getting the bugger out is difficult.

Just want to avoid having to reseat the block everytime.

I've been wondering the same thing, but I think most people open op the block to do so... it takes reseating, refilling everything yes... but I'm afraid it's hard to get out, maybe with a q tip and a piece of rubber or whatever?
Anyone got tips?

For those who don't know what I meant:

[Martinm210]

I've been taking the block apart to install the nozzles, that way I can inspect and clean the base at the same time if needed. Actually during testing I leave the block connected and just take the base off. Easier to take the base off than tubing and tip ties.

I think it's also important to push the nozzle all the way down until seated flush, so it helpful to see the underside.

[Kilyin]

While the 'nozzle instructions' for the Fuzion nozzle kit claim you can remove the nozzle without removing the barbs, I've found this to be impossible without taking the block apart, which defeats the whole purpose..

[g1raffe]

Time to upgrade my Apogee GTX.

Anyone know if this block will work with the backplate of the Swiftech block? If I don't have to remove my MB to change them over it will be a bonus.

[By-Tor]

Once again thanks Martinm for the hard work and also thanks to Gary @ Sidewinder.

[Martinm210]

Thanks!

I've got a new motherboard on the way, so hopefully by next weekend I'll have by quad core settled on a good overclock for some quad thermal testing. I've got about 8 blocks sitting here waiting in line..

[MINIz guy]

You sound like a very busy man, Martin. Sitting here and answering questions while you could be out enjoying your Memorial Day. I commend your work!

[smee]

Thanks!

I've got a new motherboard on the way, so hopefully by next weekend I'll have by quad core settled on a good overclock for some quad thermal testing. I've got about 8 blocks sitting here waiting in line..

Nice! What board are you getting?

BTW, what camera do you have? Your pictures are amazing!!

You sound like a very busy man, Martin. Sitting here and answering questions while you could be out enjoying your Memorial Day. I commend your work!

+1, go take a brake Martin!!!

[Martinm210]

Nice! What board are you getting?

BTW, what camera do you have? Your pictures are amazing!!

+1, go take a brake Martin!!!

I ordered this DFI LT X48 T2R board from NCIX. It seems like it would be really easy to watercool and was recommend by several members here. My 680i was a good board for the E6600, but it just doesn't like the new quad. I think I have chipset blocks to take care of the NB and SB, so I may just have to work up a PWM block for myself. I don't want to change anything once I start testing, so I want to get it wet off the get go..

Canon 20D, it's getting older these days but still a much better camera than this operator can handle.

Heading to Reno tomorrow for the whole week..