Since AMD's conversion to a 90 nm fabrication process, its Athlon 64 CPUs have developed into real power misers. Their modest power intake also reduces cooling requirements. To investigate this product area further, Tom's Hardware tested six current coolers for AMD CPUs.

While the AMD Boxed cooler did quite well in the tests, it was not the heatpipe version that AMD is currently shipping with their 4400+ (and above) CPUs. I suspect that the heatpipe would have faired even better than the one in the tests. I wonder why the disparity in this article?

i too noticed this. the one i got with mine is a 4 heatpipe one i think. even so that zalman gets the same temps as i get with my 400 and its stock cooler. pretty good that. i wonder if availability of the heatpipe one varies from region to region. it may be thqat the older one still ships with some procs.

I have the 4 heat pipe version of the AMD cooler (it came with my Opteron 165) and I have replaced it with a CNPS 9500 LED. While the AMD cooler preformed well for a stock cooler it was much louder than the 9500 and the 9500 proved me with an almost 15 C drop in temp with my 20% OC. I am now running a pretty contant 35C with my 165 running at ~2400 MHz (from 1800 stock).

I've heard that the best 939 cooler to date is the Scythe Ninja. Then why THG didn't include this cooler to prove how good it is compare to other coolers?

I would like to compair the CNPS9500 to the CNPS7700 (Zalmans large heatsink, which was the previous best heatsink ever =] ) and others. Any hope for a Thermal Resistance Benchmark on these coolers?

edit: For those wondering what thermal resitance is, (as i understand it) is a benchmark that shows the cooling ability of a heatsink independent of the system it is tested on, that way you can get a base number you can compair

example http://www.tomshardware.com/2005/0 [...] age30.html

Never heard of that... interesting.

Interesting article, though the results weren't unexpected by any means. I have never heard anything bad about the recent Zalman heatsinks (other than are huge lol). Top notch cooling performance. If I were using air I would be all over that Zalman. But alas I am addicted to water.

What would be interesting to see is a CPU waterblock comparison to see if the more "advanced" blocks are worth their salt.

I wonder how the ThermalTake Big Typhoon compares in this test. Currently I have a X2 3800+ hooked up to it overclocked to 2.6Ghz and I don't see temps higher than 50. Just as a note, summer has definitely hit here.

Here's a link to THG's old tests of "Mammoth CPU Coolers" but they used a p4 for their tests so I don't know if the heat disipation needed is the same or not.

http://www.tomshardware.com/2005/0 [...] age30.html

I actually purchased the Scythe Ninja based on reading reviews. It posted in the top 5 on most review sites, and many gave it the number one spot.

However, after going through its installation I would not recommend this cooler to anyone.

The installation was horrific.

It does not fit in the Coolermaster CM830, which is a high end (and large) case, without remove the fan doors.

Nor does it work with the Asus A8N32-SLI, because of the proximity of CPU to RAM slots. The cooler is so large that it completely covered 1 of the ram slots, and partially covered another. So it was blocking the heat dissipation of the ram, causing the ram to overheat.

I wound up replacing it with the AMD stock cooler.

Dark Spider

All I know it is the best air cooling device I've purchased for a CPU period.
My temps idle around 24C and load around 36C...this was coming from a Fata1ity 7700 too

The only problem.....the physically size

I second that motion.

Is it just me, or did anyone else wonder how 2,250 RPM (i.e., 77% of the available RPM range from 1,200 to 2,500 RPM) was considered the medium fan speed for the Zalman CNPS9500 LED? 77% is medium?

The only other cooler to have all fan speeds stated (i.e., the Vapochill Micro with 1,200/1,650/2,250 RPM) used a mere 43% of the available RPM range for its medium speed. Also, while the low speed of the Gigabyte Rocket-Pro was not stated, it had a difference of 1,700 RPM between its full and medium speeds.

Why the (seemingly) large discrepancy for the Zalman?
Sure, the noise level appears to increase exponentially for this cooler, but you were testing the medium *fan speed*, not the medium *noise level*, no? Let the noise level graph indicate this interesting behaviour (as opposed to distorting the temperature graphs).

Man no freezer 64 pro? What gives. It's a good little cooling solution, being easy to install and giving great performance for the dollar. Its half the price of the Zalman and gives approximately (within a few degrees) the same performance. No credit to the underdogs this time around.

Frankly, from this review the only one that would really offer a significant benefit over the new heatpipe style stock cooler would be the Zalman. None the less, I've been running my 4400+ with the stock cooler at 2.5ghz with absolutely no problems and its the quietest component in my system. My HDD's are even noisier. I'm sure the new heatpipe design can beat the pants off more than a few of the coolers in this review.

No doubt, alot of these coolers in the review just plain suck. I mean come on, we are all thinking it. I still give props to Zalman and VapoChill micro, but the rest of the coolers pretty much are a waste of tin (or whatever non copper element they are made of). Furthermore the Zalman's aren't that expensive. Most I've seen one run is about $60. So for roughly $20 more you can get a worth while cooler as opposed to some V8 POS that just looks like it should be placed in a sealed box and thrown in the ocean lol.

Good review, bad coolers

what happended to SI-120 i got onw with a panaflo 120mm on low (around 1000 rpm) and it cools it to 36-7 C under load and 36 idle, this is with around 1.482V and a newely applies AS5 (not cured)

granted this thing is only single core, but it is at 2.6 GHz

Edit: i goto update my sig... oh well

what was the ambient room temperature for the test....I too have a 4800 with the new hsf and i think its pretty damn good

I know, that the bad thing about the Scythe Ninja, the installation. Looking at the parts I wonder what kind of tools do you need to get this cooler attach to the mobo. Also the fan mount is horrible as well. As I have read it is very good air cooling but very poor on installation. I have a Thermaltake Big Typhoon cooling my Pentium D 840 oc to 3.8Ghz, and it is idling at 42c and around 58-60c at full load. This is at medium fan speed operation with no audible sound as I have replace the 120mm fan with the Silverstone FM121. It is fairly easy to install as well.

I say they should have included Thermaltake in the tests...ALL my cooling products are thermaltakes...
i know its not much for heat...
but i've got an Athlon XP 2800+ OC'd to 2.5GHz (500MHz+)
NEVER goes over 40C with my Thermaltake Venus 12
but then of course, it does run at 6700RPM...but i don't mind the noise.


my machine...
AMD Athlon XP 2800+ 2500MHz
AOpen AK77-600MAX w/VIA KT600
1536MB DDR RAM
524MHz 128MB AGP 8x RADEON 9600xt
Maxtor ATA133 120GB 8MB/B 7200RPM
Seagate ATA133 120GB 10MB/B 7200RPM
Western Digital ATA133 160GB 8MB/B 7200RPM
Sound Blaster Live! Platinum
Sony 48x24x48x CD-RW
NEC DVD-RW/RAM/DL
Thermaltake Volcano 12 Heatsink
Thermaltake PurePower 420W dual fan PSU
X-Dreamer II case

I agree with this opinion. In my case, actually my friend hold the Scythe Ninja in position that I can use srewdriver to attach it to the mount base. Installing the 12cm fan is also a tricky process. This fan should blow the air upward that will be sucked to outside by fans in PSU.

After it sits properly in my Asus A8N5X mobo, the temp drop drastically from 40oC to 35oC in idle status. Also the fans is very quite with 1300RPM.
Scythe Ninja is definely not recommended for small cases. One should purchase it with prior knowledge.

In my opinion, Scythe Ninja, Big Typhoon and Zalman CNPS9500 should be in the same class. Purchasing one of them and you will never look back for CPU cooling solution.

for those interested in seeing how the Zalman fares against other HSF, I've done a roundup with these:

# AMD’s new Heat pipe Stock cooling
# Thermalright HR-01
# Thermalright SI-120
# Thermaltake Big Typhoon
# Scythe NCU-2000 *
# Scythe Ninja *
# Thermalright XP-90C *
# Thermalright XP-90 *
# AMD Stock Cooling *
# Arctic Cooling Freezer 64 *
# Arctic Cooling Freezer 64 Pro *
# Asus X-Mars *
# Coolermaster Susurro *
# Noctua NH-U 12 *
# Noctua NH-U 9 *
# Sharkoon Red Shock *
# Silverstone NT02 *
# Scythe Samurai Z *
# Scythe Shogun *
# Thermalright XP-120 *
# Zalman CNPS9500CU *

the 9500CU, while doing okay, is bested by a few of the other hsf available