A very important win for Megahalems

[MIAHALLEN]

Many comparisons have been made with the newer Prolimatech Megahalems heatsink. But none that I have read are as significant as this one:

http://www.silentpcreview.com/article1018-page1.html

IMHO, SPCR has an impeccable testing methodology, and their results can be taken to the bank

 

[EarthDog]

VN article there...that thing is a nice monster for air cooled rigs!

 

[dejo]

+1, untouchable as far as I can tell. love mine

 

[GIXXERGUY6]

+1, untouchable as far as I can tell. love mine

what are you clocked to and what do you have?

what are your temps?


nice find MIAHALLEN

 

[moocow]

Wish I had a Megahalems so I could compare it against my TRUE and venomous-X. =(

 

[dejo]

I have an I7 920 that would do 4680 on my evga with HT on, temps get unmanagable at that but have benched all day long at 4530 with max in the 80's. will idle with no power management at 38 to 40 in a warm room. My Scythe MugunII cant keep up and it is pretty good in its own right, just not on the same level. If any of you have never held a nice air cooler (mega or true) they just feel solid. the fins arent flimsy like the rest, and the performance is way better.

 

[MIAHALLEN]

Whenever I test on air...this is the cooler I use...mounting mechanism is awesome.

I posted this not to re-hash everyone's love affair with the Mega...but because SPCR's testing methodology is very good, and I'm glad they have upgraded to an i7 system...those guys will continue to use the same system in a VERY well controlled environment for YEARS to compare all the latest cooling.

The only cavet here, is that they are only concerned with low noise, so their fastest fan speed is 1600RPM. So super high CFM testing is not done.

 

[bing]

Thanks Mia for the link.

Mega in my own brief testing, scales quite well with higher cfm fans, and agreed, it has the best mounting mechanism.

Imo that simple yet powerful 3 solid thick aluminium bars at Mega mounting mechanism is lot better than the new venomous-X mounting system.

It is beaten only by the two towers type giants like Noctua D14 or TR Cogage Arrow, but then these are different league, which is not fair to compare to.




.

 

[baditude_df]

The first thing I notice about that test is that's one cool running i7. Are those the load temperatures???
It's almost a 20degC step change across the board from the results that I saw when comparing the TRUE original with the Mega.
Looks like he used speedfan to monitor his core temperatures. Someone should really give him the link to Realtemp.
And I couldn't seem to find what the ambient air temperature was during his testing. Did anyone catch that?

Granted he's at 3.6GHz for the overclocked test and I was at 4GHz, but he pumping 1.4v to my 1.3v(bios).
Similar result pattern overall though, as he is seeing ~1deg difference between the TRUE and MEGA at full fan speed, and 3-4deg at slowest fan speed.
But whereas he's seeing a 9 and 12 deg spread through the 7-12v fan speed range, I only saw a 3 and 6 deg spread.
And whereas he's seeing a greater temperature drop through the fan speed range with the TRUE, I saw a greater drop through the fan speed range with the Mega.

Here's a screencap showing both Speedfan and RealTemp stock with no calibration or modifications.

 

[Ancient_1]

I believe they adjusted for ambient. So you could add ambient to the temps and see actual.

I was a little dissapointed they didn't user a fan stronger than 1100 rpm tho.

 

[baditude_df]

I believe they adjusted for ambient. So you could add ambient to the temps and see actual.

I was a little dissapointed they didn't user a fan stronger than 1100 rpm tho.

That would make sense given the gap and that ambient was likely ~20deg, but what is the purpose of correcting for ambient in this case?

edit: Also, where in the comparison does it say that he did that?

 

[burebista]

IMHO, SPCR has an impeccable testing methodology, and their results can be taken to the bank

I was banned there only to see that everything I told them is put in their new testing methodology.
I'm happy that I've not died in vain on SPCR.

edit: Also, where in the comparison does it say that he did that?

They said in every table: °C rise Comparison

 

[baditude_df]

They said in every table: °C rise Comparison

I'm afraid I have no experience with this, could you please explain?
How does degC rise comparison translate to correction for ambient?

I can see that there is a rise in temperature from lowest to highest fan speeds with each heatsink and that constitutes a "degC rise comparison".
But it is the actual numbers that I am concerned with. Why are they in the 50s instead of the 70s?

Thx.

 

[johan851]

How does degC rise comparison translate to correction for ambient?

degC rise means the rise in temperatures above ambient in degrees celsius. So the number they show is minus ambient temperatures.

It's actually very useful to present the data this way. Now I could, for example, run over to Skinee labs to see how much above ambient a certain radiator at a certain fan speed will make my water, and then the difference between water temps and CPU temps with a given waterblock...

 

[baditude_df]

degC rise means the rise in temperatures above ambient in degrees celsius. So the number they show is minus ambient temperatures.

It's actually very useful to present the data this way. Now I could, for example, run over to Skinee labs to see how much above ambient a certain radiator at a certain fan speed will make my water, and then the difference between water temps and CPU temps with a given waterblock...

Thx for the explanation. Can't say as I find it useful personally, but I do understand.

If you can believe it, this is the first time I have ever seen data presented this way. As if that wasn't obvious.
Probably because I do most testing and comparison for myself instead of reading reviews.

Begs the question though, why don't all reviewers use this method?

 

[burebista]

degC rise means the rise in temperatures above ambient in degrees celsius. So the number they show is minus ambient temperatures.

Also you can trust HWC testing methodology. Not so precise as SPCR in measuring dBA but otherwise very solid and constant.

 

[baditude_df]

Ok, on to the next question. Speedfan. Can't say as I'm too happy with it's out of the box performance with i7(or 775 for that matter), and I was kind of suprised to see this guy using it for his review.
On average Speedfan 4.40 it's reading 10-15degC lower on the core temperatures than my trusty Realtemp 3.36. Anyone else tried both and gotten the same discrepancy?

 

[burebista]

Yeah, Speedfan is good only for his name: changing fan speeds.
I'm a looong time Speedfan user but for temps I always trust RealTemp (or Coretemp for AMD).
I understand a difference between Speedfan and others on 775 because that pesky TJMax value but on Nehalem/Lynnfield/Clarkdale is no excuse because TJMax value is written in MSR so no more guess.
TBH RealTemp plugin for RivaTuner draws nice graphs if you miss SpeedFan's graphs.

BTW bad, RealTemp is on 3.50 RC6.

 

[MIAHALLEN]

Ok, on to the next question. Speedfan. Can't say as I'm too happy with it's out of the box performance with i7(or 775 for that matter), and I was kind of suprised to see this guy using it for his review.
On average Speedfan 4.40 it's reading 10-15degC lower on the core temperatures than my trusty Realtemp 3.36. Anyone else tried both and gotten the same discrepancy?

Pat...no disrespect to you or your excellent testing by posting this article Your review was great!

Mike has been around for quite a while, and has used Speedfan for years. I've always seen it as a program that really needed to be learned, I've never been impressed out of the box either, but MikeC has been using it for years, and I'm sure he knows how to get the temps calibrated correctly. Also, I've always appriciated hes degC rise over ambient measurements as I feel most reviewrs don't take ambients as seriously as they should...JoeC recently posted his concerns about this as well here:
http://www.overclockers.com/forums/showpost.php?p=6367632&postcount=11
which I'm pretty sure you already read

You also have to keep in mind that his priority is for low noise, and for quite results, his testing is excellent

 

[reap3r]

Woo my megahalems arrived today... but unfortunately i had to source the panaflo fans from a different supplier which haven't arrived yet.. so i can't install it yet grrr! I'm curious to see just how big a difference this will make over my CM V8