ZALMAN CNPS10X Flex CPU Cooler Review

FrgMstr

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,598
ZALMAN CNPS10X Flex CPU Cooler Review - ZALMAN is back with a new CPU air-cooler. This one sports a plethora of supported CPUs, tons of heatpipes, and a sleek new design while not weighing in so big that you must be concerned about it fitting inside your standard chassis. But now days air cooling comes down to performance and cost with so many in the market place.

ZALMAN continues its tradition of creating solid heat sinks designed for a myriad of uses. The CNPS10X Flex isn’t going to win the top spot but that’s not the point here. What we have is a versatile cooler that fits a large variety of sockets that can be tuned for quiet operation or maximum cooling performance based on the fan you pair it with. Probably the most overlooked detail here is the size of the CNPS10X Flex. While most coolers today are getting bigger and bigger to perform better, the CNPS10X Flex is downright slim by comparison. This makes it even easier to fit on most motherboards without fear of incompatibility. This is where this cooler earns the flex part of its name.
Click to expand...
 
This looks alot like my OCZ Vendetta 2 which is an awesome HSF for my Q6600. Looks to me like the price/performance ratio is a little better with the OCZ solution.

As far as it 'not being that big', it looks plenty big from the pictures I'm looking at. I wonder if I could get it inside my Antec 900 with the side intake fan in place. I can just barely do it with the Vendetta. Impossible in my CM690.
 
Seems like a good solution. Good review.
I'm still running my Tuniq Tower which seems to struggle on cooling my i7 @ 3.5GHz. Max temps are around 60C across the cores(on average) loaded in game w/ two 8800s in the case as well.
 
snaggletooth said:
As far as it 'not being that big', it looks plenty big from the pictures I'm looking at.
Click to expand...

By comparison

It's kind of like having sex with a fat girl then going back to a healthy sized girl. She's not skinny but damn does she feel skinny by comparison.

You can call the Noctua NH-D14 the fat chick.
 
Great review. I rarely read the entire review for cpu coolers but this time I decided to read front to back. What blew me away is the testing methods used! Microchannel in the heat spreader for the temp sensor? Jesus, I hope nobody argues with the testing methods used [H]ere.
 
Kenworth said:
Great review. I rarely read the entire review for cpu coolers but this time I decided to read front to back. What blew me away is the testing methods used! Microchannel in the heat spreader for the temp sensor? Jesus, I hope nobody argues with the testing methods used [H]ere.
Click to expand...

Some do but it's all good.

Without a vigilant community that keeps us on our toes the [H] wouldn't be the best.
 
I have been playing around with one of these coolers and it has done pretty well so far on a i7-870 at stock (still learning to OC on i7). I'm running it with a pair of 120mm fans and it beat out an asetek liquid cooler (120mm rad version) with the same fan setup by about 4c at idle and load. It is a little shorter than my U12P. The fins are a lot closer together than the U12P so you probably want to pair this heatsink with some fans with good static pressure for best results.
 
Good review. Love it how here you feel it necessary to say the 1.45Vcore is excessive and purely for testing purposes...

However, I do take exception to your conclusion; certainly the CNPS10X Flex is cheap...in comparison to the other Zalman heatsinks. However, with the Cogage TRUE Spirit right up there in terms of both cost and performance, the Zalman heatsinks just pale in comparison.
 
How would this compare to the Thermalright MUX-120? I have an i7 860 that I'm trying to find the right heatsink for (and it's helped that the motherboard I picked for it has 775 mounting holes, EVGA P55 SLI) but I don't want to spend a lot extra for just a few degrees. (eg one of these $60 heatsinks where you then have to turn around and buy a fan too)

From what I understand Intel sent the Thermalright MUX-120s out to reviewers to do the i7 750 reviews with, but I can't really find reviews of that heatsink itself. It falls right in my price range ($30-50, fan included), but I'm not sure how well it performs.
 
Sad, it's overpriced and performs worst than heatsinks that has been out for quite a long time.

The only thing going for it is the compatibility, but no one is going to buy this for the 940/939/754 sockets.

I have no idea how this thing got a silver award, in my book it's a big [H] Fail.
 
I would have liked to see the Zalman tested with a 38mm thick fan like a Panaflo or Delta.
 
fullvietFX said:
So the 9900 is better than this?
Click to expand...

That's not a yes or no type of question.
It depends on what you want and what fan(s) you intend to use.

The 9900 tends to be on the loud side of the spectrum under load, but it's one of the better coolers, especially when you factor in its size.
If the noise is not a factor for you, the 9900 might be a better choice.

Factor in the price of a good fan and the difference in price also reverses.

Then you have to factor in what CPU you plan to use and whether or not you're going to O/C or not.
 
Ualdayan said:
From what I understand Intel sent the Thermalright MUX-120s out to reviewers to do the i7 750 reviews with, but I can't really find reviews of that heatsink itself. It falls right in my price range ($30-50, fan included), but I'm not sure how well it performs.
Click to expand...

The samples sent out in no way reflect retail product. That is what Thermalright told us directly. We are still waiting for the real deal to show up.
 
Just a suggestion for future articles. Ya'll are measuring the T.case temperature because of the integrated thermacouple. Have you guys considered adding a few things to your standard notes? (not that anyone would read it:() It might be of some benefit to add this from page 42:
6.5 Absolute Processor Temperature
Intel does not test any third party software that reports absolute processor
temperature. As such, Intel cannot recommend the use of software that claims this
capability. Since there is part-to-part variation in the TCC (thermal control circuit)
activation temperature, use of software that reports absolute temperature can be
misleading.
See the processor datasheet for details regarding use of IA32_TEMPERATURE_TARGET
register to determine the minimum absolute temperature at which the TCC will be
activated and PROCHOT# will be asserted.
Click to expand...

Also noting the difference between T.case (case being heat spreader) and T.core which is what is reported by something like "real temp"?
 
Random Guy said:
Is there a reason the Prolimatech Megahalem is excluded from this review?
Click to expand...

What for is probably going to end up sucking.

The Noctua is currently the king of air cooling (NH-D14) followed by the Megahalems yet it sucks only on hardocp reviews.
Hardocp, you need to improve your heatsink reviews....it is impossible the D14 is doing so badly when it kicks the crap out of any other heatsink in every other review out there.
The Megahalem is the only one close to it.

Regards
 
shaolin95 said:
What for is probably going to end up sucking.

The Noctua is currently the king of air cooling (NH-D14) followed by the Megahalems yet it sucks only on hardocp reviews.
Hardocp, you need to improve your heatsink reviews....it is impossible the D14 is doing so badly when it kicks the crap out of any other heatsink in every other review out there.
The Megahalem is the only one close to it.

Regards
Click to expand...

once is not enough really? I think they have to factor in the price too.. even it works great price-to-performance ratio is a factor too...
 
We stand by our results and are confident of them. If you have a specific suggestions as to how we should improve our testing, we are all ears.

Noctua was not happy with our recent testing of its product, and it repeatedly deferred to "other reviews" that did not fall in line with ours. When asked for definitive testing proof from Noctua itself, if could not even begin to support its claims of our testing being off or its product performing as Noctua stated it would.

When a company that is supposed to be some sort of leader in the heatsink biz cannot even back up its own claims with data, well, I am not simply going to start re-retesting products because they point to folks that are testing with their personal computer and Realtemp in their mom's basement.

We put all of our cards on the table as to how we test. We retest to verify results. I am sorry that our results are not what you want them to be.

If you wanted another cookie cutter review, you came to the wrong place.
 
Another TRUE clone/variant. I'd like to see some real innovation other than the number of heatpipes surrounded by a skyscraper's heighth of cooling fins.

The old CNPS 9 series "donut" coolers were something entirely different. How about some kind of self-contained water cooling setup.
 
burebista said:
That's not fair. As long as that basement have same ambient temperature during all tests and their CPU/frequencies/Vcore is the same their tests are as valid as yours.
Click to expand...


Oh yeah? Well this site called "intel.com" says different. They "claim" to have actually designed the processor (yeah right! we all know it was designed by CIA agents". Here's a linky
Intel themselves said:
6.5 Absolute Processor Temperature
Intel does not test any third party software that reports absolute processor
temperature. As such, Intel cannot recommend the use of software that claims this
capability. Since there is part-to-part variation in the TCC (thermal control circuit)
activation temperature, use of software that reports absolute temperature can be
misleading.
Click to expand...


burebista said:
First of all I must say that I'm puzzled about your methodology for measuring CPU temperature. Maybe that was OK long time ago on P4 era but now you can rely on DTS sensors (especially for Nehalem/Lynnfield). In fact DTS readings are the only one which matter for CPU thermal behavior (thermal throttle/thermal shutdown).
Your temperature measured in a hole drilled into IHS is way off compared with core temperatures and is meaningless too. If you want to know why you can dig in this mega-thread, rge and unclewebb did a wonderful job busting Intel TJMax crap.
Click to expand...
Again, there are some crazy people at Intel who have something to say about this.
6.1.2 Thermal Metrology
The minimum and maximum TTV case temperatures (TCASE) are specified in Table 6-1,
and Table 6-2 and are measured at the geometric top center of the thermal test vehicle
integrated heat spreader (IHS). Figure 6-2 illustrates the location where TCASE
temperature measurements should be made. For detailed guidelines on temperature
measurement methodology and attaching the thermocouple, refer to the appropriate
processor Thermal and Mechanical Design Guidelines (see Section 1.2).
Click to expand...
Through the referenced document (found here) those crazy nutters at Intel show on page 87 a scary drawing that shows exactly how [H]ard tests their processors.


To elaborate further WHY you don't want to use software:
Intel said:
A single integer change in the PECI value corresponds to approximately 1 °C change in
processor temperature. Although each processors DTS is factory calibrated, the
accuracy of the DTS will vary from part to part and may also vary slightly with
temperature and voltage. In general, each integer change in PECI should equal a
temperature change between 0.9 °C and 1.1 °C.
Click to expand...
This means over something as small as 30C delta from idle to load it is not uncommon to see a 3 degree error. They also do not say the error is repeatable. If it is Gaussian then your total accuracy from the DTS can be as high as +/-4 degrees. Or they could use a thermal couple and get +/-0.1 degrees repeatable accuracy like Intel recommends.



burebista said:
Second you cannot put Cogage in front of Noctua/Zalman/Megahalems/TRUE. You simply can't. That cooler it's just a TRUE mutilated. It cannot perform better like the original.
Click to expand...
Umm RTFA? They didn't.
1259173968fHpPcE2s6n_3_3.png




burebista said:
Anyway, maybe in future reviews I'll see prime95 large FFTs, high OC with high Vcore (1.45V is good enough though), HT on and a screenshot from RealTemp. Not normalized temps, not readings from sensors put on holes drilled in IHS. Until then I guess I came to the wrong place.
Click to expand...

Yep, your looking for a basement generated review. You are looking for a review that violates Intel’s specifications for thermal solution testing. If that's what you want, you have BY FAR come to the wrong place. Thankfully they don't do that kind of shit around here.
 
Last edited:
So am I to understand that every single website and user out there that has tested this Heatsink to be the BEST (like it or not) are just a bunch of morons, and retards that cannot make a good review and your lonely one showing completely weird results is THE ONE to trust?
Sorry but I think common sense and being humble should be more important than ego.
I am risking been banned here but I am just stating what is obvious.
Click to expand...
So you're saying that unless the [H] review agrees in every single way to every other review, it cannot possibly be scientifically valid? I'm sorry, but that is not acceptable, in terms of both logic and your much-vaunted common sense; that is why multiple sites review the same product, and that is why many people read multiple reviews when buying a product.
 
So then you explain to me how every other website and user review found out it is the best air cooler but this one?
If it was like the reviews have been mixed then I could understand it or if the reviews where showing a 1 C difference which is of course within human error but no, the reviews clearly show it to be 5C or more better than all competition.
The Megahalems being the only one usually closer 2-4C or so.
Now you explain me how everyone else is wrong and hardocp is right?
Please be realistic now.
Maybe Noctua has such deep pockets that was able to buy every single reviewer out there and even regular users to give it a nice review....come on now.
The review here is obviously flawed and the whole we do not do cookie cutter reviews is a poor atempt at an explanation that does not in any way, refute the evidence at hand.
 
I've gotta agree with the testing methodology. It's pretty damn solid.

Having said that, yet another cooler I was prospecting and yet another one I'm not purchasing. I guess I'm just gonna hafta go liquid and not look back.
 
Your loss trust me.
In any case I have nothing to add here.
If you guys really believe the world is wrong and this reviewer is right then please PM me as I have some beautiful properties on Mars for sale..cheap!
Regards
 
Last edited:
shaolin95 said:
Your loss trust me.
In any case I have nothing to add here.
If you guys really believe the world is wrong and this reviewer is right then please PM me as I have some beautiful properties on Mars for sale..cheap!
Regards
Click to expand...

50286819.jpg
 
burebista said:
That's not fair. <snip>
Click to expand...

Your thoughts are noted and given them I would suggest you not rely on our reviews if you feel that way. Apparently anyone with a PC and a thermostat can do what you expect in a HSF review.
 
shaolin95 said:
So am I to understand that every single website and user out there that has tested this Heatsink to be the BEST (like it or not) are just a bunch of morons, and retards that cannot make a good review and your lonely one showing completely weird results is THE ONE to trust? :rolleyes:
Click to expand...

I am not sure what you understand. Here is what I understand.

We stand by our results and are confident of them.
Click to expand...

And yes, we are the one to trust. Thanks.
 
well kyle I read your review on the noctua so I cancelled it, not only it wasn't really worth the price/performance ratio the HSF itself won't fit too, it's so massive its not even funny and I don't think it makes much of a difference... good thing I got a good deal with prolimatech and NB fan so I was set...
If I bought the noctua I would be pissed as the prolima is a hard fit as it is...

img5382.jpg
 
One thing Im still not sure about the temp probe in IHS is that its placed right in the middle. Usually heatsinks are never perfectly even or perfectly concave in the center so the heatsink that makes best contact with probe is going to win out. Now whether the temps in other locations on the winner heatsink are better or worse is up for debate because there is no probe(s) there. And since you guys are doing a quad core chip I don't know whether just one probe smack dab right in the middle is good enough.

So I guess what I'm trying to say...is the single probe in center a good idea when evaluating heatsinks which are specifically designed to be concave and when the chip IHS is uneven and slightly convex?
 
DeadSkull said:
<snip> And since you guys are doing a quad core chip I don't know whether just one probe smack dab right in the middle is good enough.

So I guess what I'm trying to say...is the single probe in center a good idea when evaluating heatsinks which are specifically designed to be concave and when the chip IHS is uneven and slightly convex?
Click to expand...

I spoke to Intel engineers about this directly and we are following Intel guidelines in our reviews. If that is not good enough for you, I honestly do not know what more to tell you.

Keep in mind, I am open to suggestions, but reasons that start with "I think" are not holding much water with me. We built our HSF testing program on proof and data, not trial and error and hoping it was right. We researched the issue and did what is proven to work. I am not aware of any other site doing it that way.
 
Kyle, just a suggestion. Keep your testing bed like it is now if you like it but attach a screenshot with RealTemp. Not in every test just in Apple to Apple-Overclock comparison chart. For me is enough to see core temps for a Nehalem @3.6GHz 1.45V Vcore at 25°C ambient, prime95 Large FFTs and HT on (BTW maybe I've missed that but in your OC'ed test HT is on or off?).

And yes everyone can do a proper heatsink review if he's honest and he keep testing conditions the same (open stand, constant ambient, same fan and same RPM's, same CPU/MB/voltages/frequencies, same mounting kit pressure if it's possible, same TIM and same amount, same testing software and if it's possible CPU lapped and heatsink lapped).
It's not rocket science to do a proper heatsink review if it's done right. Time consuming yes, but not rocket science.

I have for my reference my friend's round-up. It's a so called here basement test but IMO is the best test I've seen to date.
Nobody cares about TCase. I'm not a system integrator to follow Intel specs I'm a home user and I'm doing my own cooling setup if I'm starting to think at aftermarket heatsinks, fans and case.
 
I would like to say thanks to [H] for the great reviews on everything. I do agree that it might be a nice addition to include the "core temperature" (or at least delta to TJunction) as reported in software like CoreTemp, RealTemp, HWMonitor, etc. All previously stated inaccuracies inherent to the internal temp diode notwithstanding, this data could be very useful to make a rough comparison between the reviewed coolers with something else where such an excellent temperature measuring method is unavailable.
 
Kyle i have a Question for you. im building a nnew rig using a 955 BE c3 or a 965 BE c3 but my case is not the biggest in the world. I cannot fit a True as its 2 Long. Would you recommend the cpns9900?

Im looking for a good cooler but needs to be kind of Short.
 
Last edited:
You must log in or register to reply here.
Back
Top