Thank you, Siren. I'm choosing the Raystorm because it performs good enough, and it has some LED bling going on that matches my build. :up:
Printable View
Thank you, Siren. I'm choosing the Raystorm because it performs good enough, and it has some LED bling going on that matches my build. :up:
Just when I thought I was out... they pulled me back in...
http://youtu.be/G29DXfcdhBg
3 more blocks coming now lol...
Can't thank you enough for the hard work, Keep at it!
Finally got some flow rate testing done on the blocks. I realized its kinda hard to tell which is which because the legend isn't in order of flow, so here's the other plot to compare with if it helps:
http://i.imgur.com/6aAX4j0.png
http://i.imgur.com/VwRDI.png
FIXED with the legend in order of restriction :)
Very helpful charts. Thanks a bunch for the extraordinary effort. :)
Awesome work as always, thanks!!
Fixed the order of blocks in the legend and added a helpful arrow :)
Now to buy haswell and do all of this over again :rofl:
Soo..what are the 3 newcomers you mentioned if that's something you can share?
Just curious. I have a very old prototype block someone sent me many moons ago that needs testing too. The design is rather unique. From the 80s I believe and uses some sort of special ultra fine microchannel design with a unique parallel jet system. Kind of funny that it's so old yet so advanced in design. Who knows it may even still hang with the best of them.
Need to fabricate a hold down for it though, has some funky oem bracket system.
Hey Martin - Well I'm not sure what's going ahead and what's not to be honest. There was a new manufacturer's block that I had tested a prototype of and they seem to have disappeared now. I'm not sure what I'm allowed to say so I'll keep it generic for now. I'm supposed to have two revisions from existing manufacturers coming in. I heard rumors of a third and I also saw a pic of the phobya uc2, so I need to talk to AT about that last one. So possibly anywhere from 0-5 blocks coming in lol. I'm really not sure what's going to happen though because some were mentioned a while back as shipping that day but never showed up and still waiting on a response!Quote:
Originally Posted by Martinm210
Sad to hear that DT shut down too :( I really liked the quality of Erik's blocks.
Nice..success! Your work in my opinion is by far the best current data out there and as usual it has promoted competition and further advancement.
Look forward to MOAR if you have the will and ability to continue keeping the test bench static. Really enjoy seeing the pressure drop addition as well as the Dt vs pumping power charts. Icing on my data cupcake..loving it!
I see PPC's carrying a new EK-JET Supremacy J4 Jet plate J4 yields up to 1?C lower temperatures on Intel LGA-2011 platform --> H E R E
Yeah I tested the prototype here: http://www.xtremesystems.org/forums/...acy-Jetplate-4Quote:
Originally Posted by Lu(ky
When I get the real thing I'll include it here :)
Well if you want a super low restriction block you best jump on the last few IceForces because MIPS has gone out of business :(
Why do you post results in water temp and not CPU temp? The point of a CPU water block is to remove as much heat as possible from the CPU. If the radiator stays the same (method of cooling the water) and the CPU heat output stays the same (method of heating the water), then shouldn't the block that results in the highest water temp technically be the best?
Just to make sure nobody takes this the wrong way, I'm not saying anyone is wrong, just thinking out loud/starting a discussion/looking for answers... Also, if this was already covered somewhere else can someone link me?
I think it's more to do with the signal to noise. The temp sensors will not be accurate enough to measure the delta temperature between in and out of the cpu block accurately enough to seperate blocks the same way. Often it's only 1C across the block, so if one block gave you 1.01C and one gave you 1.02C it'd be really hard to know that that measurement was accurate.Quote:
Originally Posted by warmmilk
Is the cpu temp difference that small between blocks? I've never tested this before so I don't know... btw I'm talking about cpu temp as displayed by realtemp or something similar so we're on the same page.
The results on the first page show a delta of 6.97 from best to worst block, was actual cpu temp difference only in the 1 degree range?
Oh - then we're talking about completely different things. My results are cpu temps measured by realtemp relative to the coolant temperature.Quote:
Originally Posted by warmmilk
I thought you meant why not just measure the delta between coolant in and out of the block in order to get rid of any bias in the onboard temperature sensors.
So are you showing actual cpu temp anywhere? In my mind the best block is the one that results in the lowest cpu temp independent of water temp. Is there something I'm not taking into account that puts a flaw in my logic?
CPU temp with coolant subtracted is a far better indicator. If you just look at cpu temps only then you're not taking into account how your ambient air temperatures change during the course of testing. If you subtract air temperatures then you're then also dependent on your radiator and how it performs vs flow and whether it gets blocked with dust over time as well as yet more second order effects unrelated to the cpu block.Quote:
Originally Posted by warmmilk
A cpu block takes heat from the cpu and dumps it to the water. The only fair way is to measure the cpu temperature and the water temperature. As soon as you measure it relative to anything else (even "nothing" i.e. absolute zero) then you're looking at the performance of more than just the CPU block. Does this make sense?
This is why any review of a cpu block that just looks at pure cpu temperature that is not relative to coolant should be ignored even if they claim they hold ambient steady. The amount of variables mean testing errors will be way too significant. Now if you're testing a whole system vs another system, then you could measure vs ambients, but you still need to measure your local ambients coming into your radiators accurately. If you're not measuring ambient or coolant, then the numbers are basically worthless.
Oh, I see. I wasn't taking ambient temp, dust, and fan performance into account in my thought process... Thanks for explaining
So when do you think you'll be testing Haswell?
Honestly probably not until late july!
looking forward to it! thanks for testing all this stuff for the community!
Your water temp is going to be just as effected by a dirty radiator. The end result is the CPU temp. The other recordings are to predict accuracy. Its arbitrary to pick one over the other where both can legitimately be claimed to have a significant impact.Quote:
Originally Posted by stren
Yes water temps will be affected by those things, but the CPU temps relative to the water temps won't be. The CPU temperatures should only be affected by the water temperature. Yes there will be small secondary ambient effects due to secondary cooling paths (socket cooling), or due to poor linearization of internal cpu temperature sensors. But those effects will be small. If you measure relative to ambient or relative to -273K then those secondary effects are much much larger.Quote:
Originally Posted by Aberration
If you are effecting water temperature, then of course you are effecting the water to cpu gap. The reasons you give of why not to measure ambient temperature directly apply to measuring the water temperature. It wouldn't be such a big deal had you not took a dig so hard at everyone else ;)
First off it's not clear if you or warmmilk meant cpu - ambient or pure cpu temps regardless of the ambient. I've seen both out there in a variety of reviews.Quote:
Originally Posted by Aberration
CPU temps will raise as the water temps raise. The delta should be maintained within a reasonably small error. I'm sorry if you took my comments personally, but feel free to correct me with reasons if you disagree. I would still argue that measuring vs coolant temperature is more accurate than measuring vs air. It removes unnecessary errors and variables that need not be there. I'm not trying to say that there are no effects, simply that those are smaller than those would be vs ambient. Also bear in mind that when you vary flow you're also varying your radiator performance too. If you haven't calculated that in when you measure vs ambient you quickly start measuring the system performance and not the block performance. Yes maybe you could get the same accuracy by perfectly controlling and measuring ambients and keeping your radiators clean and subtracting radiator performance vs flow. But even then, you would be more accurate if you measured vs coolant. If you can be more accurate, why would you choose to measure with less accuracy?
Personally I think cpu-coolant is the most accurate way to measure the block performance. However I also know from first hand experience that it isn't perfect. There are secondary effects such as cooling through the socket and common mode shifts with temperature due to presumabley poor cpu temperature linearization. Airflow blocked in a radiator will cause a shift in coolant temperature which will stack on top of any ambient variation. This can be minimized by oversizing your radiator and by minimizing ambient shifts. This is all much harder and needs to be much more tightly controlled when measuring vs ambient. That is my point, and I think your point is that if you control those well that you can still be accurate and this is true if fix the flow rate. But why not be more accurate?
As for issues of digging at others - that was not my intent. Ok maybe ignored was too strong word, and it seems like there might be some butt hurt going on, but if a reviewer is choosing to be less accurate when they could easily be more accurate then I have to question what else they are taking short cuts on. That's the point I was trying to make by saying that because there are plenty of reviews showing data at 0.1C resolution when their ambients aren't even measured but are "controlled' to an unspecified resolution and flows are not fixed. So if I've offended you then I apologize, but nothing you've said has convinced me or demonstrated that measuring vs ambients in the same setup is ever as accurate as measuring vs coolant temps.
So here's some data from my cpu testing showing the coolant - ambient temperature delta for two cpu blocks. This uses IX so TIM aging effects should be zero.
http://i.imgur.com/s0Y3YrF.png
What do we see?
We see that ambients and coolants are not a fixed offset as would be implied if you believe that measuring vs ambient is equally good to measuring vs coolant.
The delta varies with flow, and they vary between runs. The run variance could be due to ambient changes or due to time (e.g. dust).
So you have to choose which one you compare your cpu temperatures to. Do you choose coolant or the ambient temps? The water block is cooled predominantly by water, I think we can all agree on that. Therefore it makes sense to use the water temperature rather than the ambient.
What this plot is really showing is radiator performance throughout my testing. For reference it's an XSPC EX560 with 2150RPM gentle typhoons. This low 3-4C delta means that the variance betwen ambient and coolant is low. A reviewer therefore who is not oversizing their radiator, or using low speed fans and is not fixing flow will therefore see much greater variance than I do here. Therefore for those reviewers its even more important to measure vs coolant temperatures if they want to hope to achieve and legitimately claim <1C accuracy in their data :up:
Hi Stren,
Thank you once again for taking the time to post up this detailed review, it has helped no end in choosing a new block. Hence I have a Koolance cpu-380i making its way to me and I was wondering would you rotate the block 90 degrees for a i7 920 1366 socket cpu? I want to try this block out on my current rig before I upgrade my cpu in month or so.
Thanks in advanced. :)
Hey Razor - for the 920 I honestly don't know the correct orientation as I haven't tested it. I believe that there was some data someone took about orientation on bloomfield but I'm not sure which direction is best!
Stren - thanks for the quick reply, I will have a google about and see what turns up.Quote:
Originally Posted by stren
It wasn't really that important as I only plan on running the 920 for a few weeks as I am waiting on ivy-e to be released.
Edit: After a bit of googling it seems the best way to mount the 380 is with the koolance writing perpendicular to the memory slots to follow the contour of the die which makes sense. :)
koolance 380i installed and my cpu temps have dropped by 4c from my Swiftech Apogee XT version 1.0, flow rates are hugely improved which was expected. Overall I am pleased with the new block. :)
Nice :up:Quote:
Originally Posted by Razor Time
Tnx stren for a great write write up. I noticed there is a phobya UC2 available - any idea how it performs?
No not yet - I've been meaning to ask for one. Been busy trying to wrap up titan/780 block testing ;)