OCZ Hydroflow Mini Review

[Martinm210]

First off a special thanks to a few folks before I get started:

Tony at OCZ for the block sample and OCZ freeze sample I have yet to try out.
Gary at Sidewinder for providing me with the Q6600 that has been recieving all the abuse lately.
Martin and Mario at TFC for the TFC480ER radiator used in testing
Alex from Petras for the free MX-2.
Paul from XSPC for the XSPC ResTop pump top combo.
And everyone in the forum for putting up with my science & testing mantra...

Here is a link to my full review if you want details:
http://www.martinsliquidlab.com/OCZ_...ow_Review.html

First up the block:


The accessories package is extremely complete, not only one, but two backplates, two hold down plates, two sets of metal barbs, and everything you could possibly need...this is one of the most complete accessory packages I've seen...very nice!!


Internals:
A nice durable crack resistant black nylon top (Thanks waterlogged) and a copper triangular pin/fin arrangement base.


And a stepped base to maximize mounting pressure over the critical die location:


The mounting mechanism is very well thought out. It uses standoffs to hold the backplate in place and a combination of washers/springs and thumb/screw nuts to tighten the block down. More importantly these screws install the block with a fair amount of mounting pressure and they bottom out when done, so there is a very posotive "I'm done" with the screws when you've installed them correctly, and this ensures all four corners have the same pressure and the block is true.


Fitting compatibility = 3/4" OD barbs fit with plenty of room to spare. Pretty much any barb will work here.


Pressure drop was about equal to the Apogee GT which is probably less than average in restriction of all the blocks I've tested. This is good and means flexibility in using the block in either multiple block loops or loops with pumps of lower power..


Thermal testing, first a few words about thermal testing. It is plaqued with inaccuracys, our DTS sensors are only resolved to 1C, so trying to pick apart more resolution than that is a real challenge. In addition I've tested several blocks of the same type before and managed to squeeze different levels of performance out of one over the other. Finally, each processor is different particularly when it comes to dual core vs quad core, performance on one does not equate to the performance on another. All you can really conclude from testing is a general understanding how a block peforms on your particular setup and it may or may not be the same on another system.

With that said on my setup, I found tha the As Shipped thermal performance is about equal to the Apogee GT, although more mounting pressure could possibly improve this by a couple of degrees.


It's a nice block, has an amazing accessory package, awesome price, and performance is at least on par with the bowed Apogee GT...likely improved with more mounting pressure by the use of additional washers which are included in the package.

Cheers!
Martin

[aspire.comptech]

So, decent all around performance, good compatibility, nice bundle, and a low low price.

Looks like a good replacement block for my brothers old TDX.

[-Acid-]

good review there

nice block

[gillbot]

Very nice martin!! This is on par with what I expected, about = to the Apogee GT.

[Nate P.]

I really like the fact they stepped the base right out of the box. Great pictures and review Martin, as always!

[sirheck]

Another nice review martin.

[Movieman]

Very nice. I need to learn how to make those excellent looking graphs!

[Dangals]

informative yet concise - fantastic job Martin


nice block too

[Waterlogged]

Hmm, I was close, but not close enough.

Meh, looks like it'll be an average performer. I say this knowing how deceptive photo's can be but I can't help but notice how similar the pins are to the original Apogee in both size and quantity. This blocks only saving grace may be the step which may bring the performance in line with maybe the Apogee GTX.

Looks like I gave it just a smidge too much credit on performance.


Great test as always Martin. .....and a decent entry level (average) block for OCZ. I think I'm seeing how OCZ might go for Hydroflow v2.

Edit: Martin, the top is made from nylon, not Acetal/Delrin.

[YMAA]

Maybe it's just the picture, but is the stepped part of the base wide enough to cover both dies on a quad (if it makes a difference at all)?

[Martinm210]

Edit: Martin, the top is made from nylon, not Acetal/Delrin.

You guys are my editors...off to fixing that, I just assumed since it was black and resembled delrin.

Thanks!

[Bojamijams]

Hmm... since its so free flowing, I wonder how it'd with some nozzles.

[AndrewZorn]

and only $40!

[Martinm210]

Maybe it's just the picture, but is the stepped part of the base wide enough to cover both dies on a quad (if it makes a difference at all)?

The step is 24.3mm +- in width (Hard to measure with calipers). I don't know all the dimensions of the actual dies, but I thought that was big enough to cover current quads.

Here is the measurement in question...I thought that might come up.


I made Naeku's ACDI 25mm, but Scott was looking for 23mm on his block I have yet to do.

I think 24mm is about right..

[SaiNRuB]

Thank you Martin, very informative as always. Much appreciated!

[Kibbler]

Martin, another fantastic review, thank you.

1st of all my apologies if this was already brought up in the other thread.

From your description (scrubbing + turbulence) I imagine the flow path is like this:
(circles are inlet/outlet, lines are floooow)


Provided the mounting holes are square, what do you think will happen if we rotate the base 90deg? I (wild ass guess) think worse temps but even lower restriction?

[AndrewZorn]

I was wondering WHICH way it was supposed to be (no, I didn't read everything) and also wondered whether there would be different, yet differently rewarding ways to configure it.

[Martinm210]

Martin, another fantastic review, thank you.

1st of all my apologies if this was already brought up in the other thread.

From your description (scrubbing + turbulence) I imagine the flow path is like this:
(circles are inlet/outlet, lines are floooow)


Provided the mounting holes are square, what do you think will happen if we rotate the base 90deg? I (wild ass guess) think worse temps but even lower restriction?

Good question.

Here is a picture straight from the box, so it does flow with the direction of the 45 degree fin cuts.


Going against the direction of the pattern would likely make it more restrictive, not sure what it would do for thermal performance though.

[ryan92084]

i agree with the well thought out mounting method...the am2+ plate for the fuzion v1 gives way too easy and could really use those standoffs

[hotdun]

Martin, great review, thanks man. Looks like a solid performer that might grow to new heights with a little tweaking. Cheers.

[zanzabar]

is there a reason why the flow graph has no dtek v1 but the temp one dose

[AndrewZorn]

I think the point with the temp graph was to show what it was close to and a reference, not to make it look bad.

[skinnee]

Good write-up and review Martin...thanks again!

[Martinm210]

is there a reason why the flow graph has no dtek v1 but the temp one dose

No, I just don't like the graphs all cluttered. The D-Tek V1 is a very free flowing block, just slightly higher than the DD MC-TDX. If I show all the blocks on the same graph it just gets really hard to follow, some overlap and are identical. My intent was to show the relative restriction to a similar cross flow pin/fin type block which it's pretty close to how my GT came as shipped.

The Hydroflow is slightly different in that it flows corner to corner over a larger distance so it's very slightly more restrictive which is consistent with the GTX vs GT.

[SNiiPE_DoGG]

well hmmm, this gives me a tough decision (I think I'll wait it out for a full review) Apogee GTZ or Hydroflow??? I want to support OCZ in this endeavor and I think the block might really be able to cool with more pressure on the mount and maybe a turned base...