Cooler Master V8 CPU Cooler Review; Overclocking Results Inside

[rgathright]

Here is my latest review on the Cooler Master V8 CPU Cooler. I used an Intel Extreme QX9650 and ran it through a gruelling round of overclocking to see how good the heatpipe cooler really was. I also used a Scythe SCMN-1100 MINE Rev.B 3 Heat Pipes CPU Cooler for comparison.

http://www.bayreviews.com/computers/19854

http://www.bayreviews.com/cache/multithumb_thumbs/fe6f2abe85add50425f587555ad1554d.jpeg

[JerWA]

No offense, but your review contains some information that's wrong.

The V8 only has 8 heat pipes... hence the name. http://www.coolermaster.com/product.php?product_id=5279


Modularized by 4 separate set of aluminium fins and linked by 8 diagonal up-rise heat pipes, the V8 structural design allows speedy heat transfer to a large heat dissipation surface

Secondly, when you say the heat could circle back to the processor, that's almost impossible because it's counter to how heat pipes work. The problem would be that, were you not cooling the fins, the heat transfer medium inside the heat pipe would simply stay at the fin side of the heat pipe. This would stop the heat pipes from transferring any more heat away from the CPU but would not bring it back. In fact you'd still be cooling, via thermal transfer in the metal itself, just very very poorly.

*Edit: I should also note that even with no active air flow, the chances are still very good (almost 100%) that the ambient temperature around the fins would be sufficiently lower than the heat level at the CPU that you'd still actually have functioning heat transfer in the heat pipes. That's how passive coolers work.

For the effect to reverse, the heatpipe would have to have no prevention methods in place at all (possible? I think all heatpipe coolers now use wicks or other methods so that orientation doesn't matter), and the fins would have to be hotter than the CPU. Given that the thermal transfer is never going to be 100% efficient, that shouldn't be possible.

http://en.wikipedia.org/wiki/Heat_pipe
http://architecture.mit.edu/house_n/web/resources/tutorials/House_N%20Tutorial%20Heat%20Pipes.htm

Anywho I'll shut up now.

[rgathright]

No offense, but your review contains some information that's wrong.

The V8 only has 8 heat pipes... hence the name. http://www.coolermaster.com/product.php?product_id=5279


Yes and thanks for noticing.

I however, felt it was important to point out that 16 pipes were extending from the processor core.

As for discussing the thermal characteristics of this design, it is way beyond today's viewing audience. In the real world, I should have placed four, simultaneous temperature probes on the heatsink to get measure its performance.

Looking beyond my inabilities, look at the performance chart. The improvements showed that this design is superior to the competition. :)

[JerWA]

Yes and thanks for noticing.

I however, felt it was important to point out that 16 pipes were extending from the processor core.

As for discussing the thermal characteristics of this design, it is way beyond today's viewing audience. In the real world, I should have placed four, simultaneous temperature probes on the heatsink to get measure its performance.

Looking beyond my inabilities, look at the performance chart. The improvements showed that this design is superior to the competition. :)

So you felt it was important to point out something that's wrong, and because people are too dumb to understand the truth you decided to caution them about something that's counter to how heat pipes work.

Oky doky then.

[Shadow]

Here's my review, I have 2 of them, one on an i7 950, the other on a Q9650.

1. It's big, it's very big.
2. It works just as good as they say it does.
3. It's quiet, only maxes out at a tad over 2000 rpm.
4. It's a solid design and a solid build.
5. Blows away just about any other cpu cooler out there, you get what you pay for.
6. The one on the 9650 has been running 24/7 for over 2 years without a problem.
7. What are you waiting for? Go get one!

:p

[JerWA]

http://frostytech.com/articleview.cfm?articleID=2314

http://frostytech.com/top5heatsinks.cfm
^ Top 10, dunno why they still say top 5.

Anywho, it's not even in the top 10. It's down near 40th place in fact.

$60 on NewEgg.

The Spire (#1 on that chart) is $5 less, though not carried by Newegg as far as I'm aware. It coolers better (3.4C at the same thermal load), and is quieter (at max, by 3dB), than the V8 using only 1 of the 2 included fans (the 2nd one really doesn't help much so no point).

Unfortunately right now there are no bargain champions. For a long time an almost-always-on-sale-for-$30 cooler was in the top 3. You could still get one and be better off than the V8.

In 15th place, it's still 2C cooler and 1dB quieter than the V8. And about half the price ($32), even not on sale. http://www.newegg.com/Product/Product.aspx?Item=N82E16835207004

Or the HDT-S1283 version for $4 more: http://www.newegg.com/Product/Product.aspx?Item=N82E16835233003

But, in an effort for fairness, here's the OCC review where it beats the core contact freezer by 2C: http://www.overclockersclub.com/reviews/cm_v8/

And the Hardware Canucks review where it loses, quite badly, to the HDT-S1283:
http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/17444-cooler-master-v8-cpu-cooler-review.html

Though at Legit Reviews the V8 beats the HDT-S1283 by 1.5C:
http://www.legitreviews.com/article/833/1/

I'm sure I can post a dozen more reviews showing a win on either side. The "why" is simply the setup. Are they testing on an open air bench? In a case? Which case? What exhaust/intake fans? Etc.

Just saying that one of these coolers is $60 and the other is $35. You could probably guarantee better cooling over all by getting the cheaper cooler and upgrading your case fans with the money saved. Just my 2 cents.

[Dandasarge]

This thread is going way to long about a dang article. You can have a lot of different configs all better at some point or another. Anyone that understands thermal property's at all knows this. Smaller pipes move heat faster bigger pipes move more heat. Almost all producers clam some kind of chimney effect but its simply thermal transfer. Sure the hot gas moves to the top everyone knows that, but after the first bit moves up then the creates pressure from the expansion think steam boiler. Just because the water on the bottom isn't boiling doesn't mean it isn't just as hot. It just mean's it wasn't hot first. After that first min its all thermal transfer. It doesn't matter how you do it. just as long as there's enough to do it.

a thermal loop is possible anytime you have a closed system. It would never happen with a fan running. The likelihood would be probability less then 1 in 100m but a side mounted CPU with a dead fan could create a loop. (this is the theory that 100% of passive solar water heaters use.) But no it wouldn't matter, is it food for thought? sure.

I personally like the V8, It's a great design. I would never spend more then 30 bucks on a heat sink though because at 60 bucks I could build one of those high tech closed loop water coolers and I like copper. (something with 100 arguments on each side) Like silver is better.

[rgathright]

Well thanks to all of you for the comments.

I first started writing computer hardware reviews because getting a clear comparison (Case temp, air flow, specifics about processor utilization) was close to impossible. Complicating matters was the fact that the writers took an elitist attitude about criticism and would often never respond.

I have taken all your points into consideration and will try my best to present a more accurate discussion in the future.:cool:

[GregK]

well done rgath!

[YoDude9999]

OMG,

I'll just put in MY two cents worth.........

After a good trusted friend of mine recommended this cooling option to me, I bought one.
In a nutshell.........

IF YOU WANT A BETTER COOLING OPTION, YOU NEED TO START LOOKING AT LIQUID COOLING OPTIONS

This heatsink is the $hit you NEED to have if you're going with air cooling. I have not yet seen ANYTHING that cools better than this heatsink for CPU applications. I know the results, I own one and those that do own one or more will attest to this product.

I can guarantee you that if you ever see a BAD review from ANYONE about this product, they DO NOT or HAVE NOT EVER OWNED one to see the results for themselves, OR, they are complete idiots and have NO idea how to install a heatsink properly!

Yo-

[JerWA]

Define "bad."

Like I posted above, including links, there are plenty of air options that are better, and cheaper.

And I highly doubt these review sites, especially ones like Frosty Tech that do mostly cooling reviews, are "idiots" who have "no idea how to install a heatsink."

Besides, aren't you the guy running everything you own >100C? I thought you didn't even use heatsinks. :P

[rgathright]

OMG,

I'll just put in MY two cents worth.........

After a good trusted friend of mine recommended this cooling option to me, I bought one.
In a nutshell.........

IF YOU WANT A BETTER COOLING OPTION, YOU NEED TO START LOOKING AT LIQUID COOLING OPTIONS

This heatsink is the $hit you NEED to have if you're going with air cooling. I have not yet seen ANYTHING that cools better than this heatsink for CPU applications. I know the results, I own one and those that do own one or more will attest to this product.

I can guarantee you that if you ever see a BAD review from ANYONE about this product, they DO NOT or HAVE NOT EVER OWNED one to see the results for themselves, OR, they are complete idiots and have NO idea how to install a heatsink properly!

Yo-

Thanks!

Again, I publish links to my reviews here on the forums to help team members find the right cooling products.

I encourage everyone to consider the computer cases I have linked here as well if you looking around. In one review, I discovered that the massive cooling fans in a $60 computer case directly affect the cpu temperature when turned off.

Finally, if you are looking for a new computer to attack any BOINC project, consider an AMD G34 based system. From a budget cruncher's point of view, the cost per core and extreme expansion options these motherboards offer are light years ahead of any Intel product on the market.

The market today is filled with high quality products, we just need people to find them. :)

[JerWA]

So are you getting kickbacks or something?

G34 is the best crunching for users on a budget? Really?

Maybe you should look into the new Sandy Bridge Intel server stuff, E3-XXXX. I don't think the AMD stuff is competitive at all, not even cost.

If you're really on a budget, why would you be shopping server side products anyways? For what it's worth you can get a cheaper E3 Intel board than a G34 AMD board.

Really, for a budget, you should be looking at desktop parts because they're always going to be cheaper. And the sweet spot right now at just about any budget is the new 2xxxK proc from Intel. The first CPU ever... yeah, ever... to GAIN efficiency when overclocked. In other words, the first CPU that actually gains processing power faster than it increases in electricity consumption. And the cheaper of the two, the 2500K, when overclocked runs on par with the $1,000 Extremes. The 2600K outruns them, using less power, and producing less heat. Yay for a generation change that matters for once.

If you've got some data that says G34 is the best budget option, please share it. Seriously.

[DrPop]

G34 is the best crunching for users on a budget?

Alright JerWA, I'll bite. Ease up just a second. I'm not sure if you know that rgathright is a dedicated CPU cruncher, like NJCaNS. He's not into GPUs. So, if you want max credits per day, actually a G34 makes some kind of sense. About the same cash outlay as a high end Sandy Bridge platform, and you can have 16 cores on that G34 rig - nothing to sneeze at, for sure.
Couple that with the fact that you can't even get the good O/Cing chipset for the Sandy Bridge until this fall (read about it on Anandtech), then for now - the G34 is probably quite competitive from that point of view...

[John P. Myers]

Alright JerWA, I'll bite. Ease up just a second. I'm not sure if you know that rgathright is a dedicated CPU cruncher, like NJCaNS. He's not into GPUs. So, if you want max credits per day, actually a G34 makes some kind of sense. About the same cash outlay as a high end Sandy Bridge platform, and you can have 16 cores on that G34 rig - nothing to sneeze at, for sure.
Couple that with the fact that you can't even get the good O/Cing chipset for the Sandy Bridge until this fall (read about it on Anandtech), then for now - the G34 is probably quite competitive from that point of view...

Well...i've seen the 2500K @ 5GHz. That's good enough for me :p The 2600K has hyperthreading where the 2500K does not. So a 2600K @ 5GHz? Yes, please. Btw, that's with AIR cooling, which is as cheap as it gets :) You can view the 2500K here. (http://www.overclock3d.net/reviews/cpu_mainboard/msi_big_bang_marshal_p67_review/4)

Edit: And here (http://www.guru3d.com/article/msi-big-bang-p67-marshal-review/13) are some 2600K crunching results. As you can see, Jerwa is right that an OCd 2600K is faster than a 980X 6-core. But you'll also notice than an OC'd 2600K is more than twice as fast as a Phenom II 1100T. If you put it in turbo mode, maybe 2 of those would equal 1 2600K. Unfortunately, 2 1100Ts cost $145 MORE than a single 2600K. Also the 2600K uses just 95W while the 2 1100Ts would suck down 250W (power bills should *always* be factored in to building a 'budget' system). Now i know this AMD chip isn't going to be used on that dual 34 board, but for whatever is used to actually have the insane amount of power increase that will be needed to compete with the 2600K (and 99% of server boards can't be overclocked so you're probably stuck with stock specs on everything, and error correcting server memory is much more expensive than standard desktop RAM, not to mention slower) well...i just don't see it.

[JerWA]

Alright JerWA, I'll bite. Ease up just a second. I'm not sure if you know that rgathright is a dedicated CPU cruncher, like NJCaNS. He's not into GPUs. So, if you want max credits per day, actually a G34 makes some kind of sense. About the same cash outlay as a high end Sandy Bridge platform, and you can have 16 cores on that G34 rig - nothing to sneeze at, for sure.
Couple that with the fact that you can't even get the good O/Cing chipset for the Sandy Bridge until this fall (read about it on Anandtech), then for now - the G34 is probably quite competitive from that point of view...

What I'm saying is there's nothing budget about a G34 setup, at least not that I know. I'm willing to learn, hence the asking. I'm serious, not just being sarcastic heh.

But from the quick research I did I could do a sandy bridge Xeon server cheaper. Less cores? Yes. Less work? No. In fact they absolutely destroy the competition. Sadly I don't think Bulldozer is even going to be enough to fix it, but that won't work in a G34 anyways.

I'm just seeing a lot of what I consider mis information, with nothing to back it up. Yes, I'm being a bit combative, because I'm backing up everything I'm saying and that proof is conveniently being ignored.

So yeah. Just trying to keep the marketing hype to a minimum, until I see some proof. But hey, if you'd rather be misled than informed, more power to you.

I'll go back to ignoring this forum. See no evil, right. X_X

[John P. Myers]

Well...i've seen the 2500K @ 5GHz. That's good enough for me :p The 2600K has hyperthreading where the 2500K does not. So a 2600K @ 5GHz? Yes, please. Btw, that's with AIR cooling, which is as cheap as it gets :) You can view the 2500K here. (http://www.overclock3d.net/reviews/cpu_mainboard/msi_big_bang_marshal_p67_review/4)

Edit: And here (http://www.guru3d.com/article/msi-big-bang-p67-marshal-review/13) are some 2600K crunching results. As you can see, Jerwa is right that an OCd 2600K is faster than a 980X 6-core. But you'll also notice than an OC'd 2600K is more than twice as fast as a Phenom II 1100T. If you put it in turbo mode, maybe 2 of those would equal 1 2600K. Unfortunately, 2 1100Ts cost $145 MORE than a single 2600K. Also the 2600K uses just 95W while the 2 1100Ts would suck down 250W (power bills should *always* be factored in to building a 'budget' system). Now i know this AMD chip isn't going to be used on that dual 34 board, but for whatever is used to actually have the insane amount of power increase that will be needed to compete with the 2600K (and 99% of server boards can't be overclocked so you're probably stuck with stock specs on everything, and error correcting server memory is much more expensive than standard desktop RAM, not to mention slower) well...i just don't see it.

And...The mobo used in that OC test of the 2500K and 2600K was $30 cheaper than the Asus KGPE-D16.

Edit: lol and the cheapest 2 8-core G34's would cost $560, which again is $245 more than a single 2600K, which would also consume 135W less, which is enough of a chunk of power that you might consider using a PSU with a lower wattage rating, which ALSO would cost less...

[DrPop]

Jeez guys hang on a sec - JerWA, I wasn't getting combative - and I certainly hope you don't ignore this forum in the future. We're all grown-ups here, and we can have a good discussion without losing our cool, right?;)
Beside all that, you have lots of good information to contribute! Let's just start a new thread and have a real discussion on this, because I'd like to get yours and JPM's input too...
...I'll be back in a sec with a new thread on the topic. :D

[YoDude9999]

What I mean by a BAD review, is anyone that down talks the V8 because they didn't get some miraculous results by installing one.

My results are obtained from actual experiences, so you're not going to get a link to support my findings. My experience with the V8, I consider exceptional performance from the standpoint of comparing it to two other cooling options I have installed on Godzilla (X58 Classified w/ i7 920 OC'd to 3.2g) over the past couple years.

The stock cooler that comes with the 920 is great, if you've never going to crunch CPU WU's. It would allow the CPU to reach those high temps I mentioned in another post while crunching and I ran it that way for a considerable time without any disastrous results.

When it came time to upgrade Rodan (EVGA 790i 3 way SLI FTW) from the E8500 to the Q9650, the cooling option (Nochtura) on that system went over to Godzilla and the stock cooler that came with the Q9650 went to Rodan. Rodan went into cold storage for a while at that point and Godzilla became my main system. The Nochtura cooler was a major improvement over the stock cooler bringing down the temps to something more reasonable at the low 80c range. I also ran that system for a considerable time with that option until my friend picked up the V8 and after running it for some time, recommended it to me.

The V8 so far has proved itself to me by making another big improvement in the cooling capacity of the system. From the low 80c temps, my system now runs in the mid 60c and on very hot days (90-100f+) will keep the CPU down below 90c, which to me is nothing to worry seriously about considering how much torture that poor CPU has had already.

There may be even better cooling options out there, but for the money, it was well spent. I am very pleased with the results it gives. At this point, I don't see a need to attempt to upgrade it to something else, but if I ever do, it'll go to Rodan and replace the Nochtura and the Nochtura will go to another system.

Yo-