Scythe Andy Samurai Master vs. Thermaltake MaxOrb
by Wesley Fink on June 4, 2007 5:00 AM EST- Posted in
- Cases/Cooling/PSUs
Conclusion
As the numbers in our CPU cooler tests have risen, certain conclusions have become more obvious. After testing some 23 separate coolers and many more configurations the heatpipe tower has emerged as the best performing design among the coolers. A tall heatpipe tower with horizontal fins attached to a number of vertical heatpipes is the best air-cooling performance you can buy these days. These towers all use side-blowing fans to further dissipate the heat. Most are designed to use a single fan, but some can use two or more fans in a push-pull configuration.
Top air-cooler performance solidly belongs to the Thermalright Ultra-120 eXtreme at 3.94GHz, with the Tuniq Tower 120, the regular Thermalright Ultra-120, the Scythe Ninja B Plus with SilenX fan, the OCZ Vindicator with SilenX, and the Scythe Infinity with dual push-pull Scythe fans all right behind and tied at 3.90GHz. This is a remarkable group of performers that definitely deliver value for your money.
We have tried to keep an open mind about the supposed advantages of down-facing fans. The arguments and logic are persuasive - a fan or fans blowing down should also cool your motherboard components better, and that should mean better performance. The argument is logical enough, but unfortunately the execution leaves a lot to be desired. So far we have tested three down-facing designs - the Cooler Master GeminII, Scythe Andy Samurai Master, and Thermaltake MaxOrb. That's quite an illustrious group, but none of these three could really compete with our top tier of coolers.
We are left to ask the question if down-facing cools better, then why can't these down-facing designs compete with the best heatpipe towers in performance? We have no auxiliary cooling in our test bed, so the down-facing designs should shine in better performance. Unfortunately they don't.
It is sad to say that we considered the Cooler Master GeminII performance disappointing in our last review, when it is definitely the best of the down-facing coolers. That is not praise for the GeminII; it is just an example of what a disappointment this design really is. The GeminII turned out to be the best of a group of underperformers.
There are definitely good things to say about the MaxOrb and Andy Samurai Master once you get past the disappointing performance. The MaxOrb is elegant and exceptionally light for a top cooler. This means it should travel well in a LAN Party PC for example. The MaxOrb also has one of the slickest and best-working installation systems we have seen. You can truly mount the MaxOrb without any thought of having to remove your motherboard. We can't say that about many of these big high-end coolers.
The Andy Samurai does not fare quite as well. It is much heavier and not a cooler you should use in a computer on the go. Replacing the medium output quiet Scythe fan does absolutely nothing to improve performance. Fans with double the airflow still leave the Andy Samurai Master at a maximum overclock of 3.81GHz. Mounting is also supposed to be a snap without removing the board, but we found the wide overhangs made it almost impossible to mount the Andy Samurai Master without removing the motherboard from the case. Fortunately Scythe has two other heatpipe tower designs that do make it to the top of our performance charts, so you can always buy one of those if you are set on the Scythe brand.
In the end the only conclusion to be reached is that the heatpipe towers with side fans are a superior design to the down-blowing coolers. They cool to lower temperatures and allow higher overclocks than the down-facing designs - even those with super high-output fans and their own heatpipes. The conventional wisdom and logic for down-facing fans just doesn't pan out in real-world performance.
The real advantage of comparative testing in a consistent test bed is you can see patterns like this emerge over time. This is easily missed with tests that use a different test bed for almost every review, or that only compare a few items on the same test bed. Large numbers consistently tested allow you to look more deeply at a group as we are doing here with CPU coolers on a Core 2 Duo test bed.
The down-facing coolers are just fine for routine cooling. Some perform very well indeed at stock speeds. Others compare well if noise is your first concern. However, in the important cooling performance and overclocking areas, the down-facing coolers are consistently outperformed by heatpipe tower designs. The best values in cooling performance and overclocking are the heatpipe towers with side fans that top our performance charts.
As the numbers in our CPU cooler tests have risen, certain conclusions have become more obvious. After testing some 23 separate coolers and many more configurations the heatpipe tower has emerged as the best performing design among the coolers. A tall heatpipe tower with horizontal fins attached to a number of vertical heatpipes is the best air-cooling performance you can buy these days. These towers all use side-blowing fans to further dissipate the heat. Most are designed to use a single fan, but some can use two or more fans in a push-pull configuration.
Top air-cooler performance solidly belongs to the Thermalright Ultra-120 eXtreme at 3.94GHz, with the Tuniq Tower 120, the regular Thermalright Ultra-120, the Scythe Ninja B Plus with SilenX fan, the OCZ Vindicator with SilenX, and the Scythe Infinity with dual push-pull Scythe fans all right behind and tied at 3.90GHz. This is a remarkable group of performers that definitely deliver value for your money.
We have tried to keep an open mind about the supposed advantages of down-facing fans. The arguments and logic are persuasive - a fan or fans blowing down should also cool your motherboard components better, and that should mean better performance. The argument is logical enough, but unfortunately the execution leaves a lot to be desired. So far we have tested three down-facing designs - the Cooler Master GeminII, Scythe Andy Samurai Master, and Thermaltake MaxOrb. That's quite an illustrious group, but none of these three could really compete with our top tier of coolers.
We are left to ask the question if down-facing cools better, then why can't these down-facing designs compete with the best heatpipe towers in performance? We have no auxiliary cooling in our test bed, so the down-facing designs should shine in better performance. Unfortunately they don't.
It is sad to say that we considered the Cooler Master GeminII performance disappointing in our last review, when it is definitely the best of the down-facing coolers. That is not praise for the GeminII; it is just an example of what a disappointment this design really is. The GeminII turned out to be the best of a group of underperformers.
There are definitely good things to say about the MaxOrb and Andy Samurai Master once you get past the disappointing performance. The MaxOrb is elegant and exceptionally light for a top cooler. This means it should travel well in a LAN Party PC for example. The MaxOrb also has one of the slickest and best-working installation systems we have seen. You can truly mount the MaxOrb without any thought of having to remove your motherboard. We can't say that about many of these big high-end coolers.
The Andy Samurai does not fare quite as well. It is much heavier and not a cooler you should use in a computer on the go. Replacing the medium output quiet Scythe fan does absolutely nothing to improve performance. Fans with double the airflow still leave the Andy Samurai Master at a maximum overclock of 3.81GHz. Mounting is also supposed to be a snap without removing the board, but we found the wide overhangs made it almost impossible to mount the Andy Samurai Master without removing the motherboard from the case. Fortunately Scythe has two other heatpipe tower designs that do make it to the top of our performance charts, so you can always buy one of those if you are set on the Scythe brand.
In the end the only conclusion to be reached is that the heatpipe towers with side fans are a superior design to the down-blowing coolers. They cool to lower temperatures and allow higher overclocks than the down-facing designs - even those with super high-output fans and their own heatpipes. The conventional wisdom and logic for down-facing fans just doesn't pan out in real-world performance.
The real advantage of comparative testing in a consistent test bed is you can see patterns like this emerge over time. This is easily missed with tests that use a different test bed for almost every review, or that only compare a few items on the same test bed. Large numbers consistently tested allow you to look more deeply at a group as we are doing here with CPU coolers on a Core 2 Duo test bed.
The down-facing coolers are just fine for routine cooling. Some perform very well indeed at stock speeds. Others compare well if noise is your first concern. However, in the important cooling performance and overclocking areas, the down-facing coolers are consistently outperformed by heatpipe tower designs. The best values in cooling performance and overclocking are the heatpipe towers with side fans that top our performance charts.
Facebook
Twitter
RSS
50 Comments
View All Comments
Wesley Fink - Monday, June 4, 2007 - link
We could change it to embedded, which we used several other times to refer to the MaxOrb fan. We just thought always calling it embedded might get boring ;-)DrMrLordX - Monday, June 4, 2007 - link
The VX recently won a large HSF shootout here:http://www.hexus.net/content/item.php?item=8757&am...">http://www.hexus.net/content/item.php?item=8757&am...
I know you're doing things differently than Hexus, but still . . . why no VX? The Big Typhoon has already been shown to be an outstanding HSF with a high cfm fan (100-110 cfm being the sweet spot). The fact that it's a top-down cooler like the two you've tested here is merely an added bonus for those who own cases with otherwise-poor airflow that can become cooling dynamos with a powerful top-down cooler (thanks to the prevalence of side air ducts/air guides since the P4 days).
Some folks have said the Scythe Andy is better than the Big Typhoon, but without a head-to-head comparison, will we ever know?
wollyka - Tuesday, June 5, 2007 - link
it will be interesting to test also the Big Typhon VX like the OP said..xsilver - Monday, June 4, 2007 - link
the VX design looks a bit dodge because of the very high centre of gravity. with practically a 1kg cooler I dont want that lump of metal falling off and smashing onto a 8800gtx or something!at least with a tuniq style cooler the centre of gravity is much more evenly distributed.
Personally Im thinking about a setup where you can have dual fans but only have the second one kick in when a certain temp. is reached or maximum load is triggered. Besides the thermalrights - which are too expensive imo, is the scythe ninja the best bet?
cujo - Monday, June 4, 2007 - link
if you've used the bt you'll find it doesn't have a high center of gravity at all. it's far ligher than say my old xp90c.i was planning on getting a ultra 120 extreme based on this review but after the spanking he regular ultra 120 got in the hexus review i think i'll stick with my bt.
i've asked at least twice in this comments section for the bt to be included but to no avail. i actually signed up an account just to ask.
the one difference between anandtech and hexus that i think anand does better is use a consistent thermal paste. that's why i'd love to see how the bt performs in anands test environment.
oh and for those who said "ha, they're using an old cpu." those old cpus ran so much hotter than the c2d's we're used to today. i figure it's a better test to use hotter cpus as opposed to modern cpus as those results will transfer over. i don't care how they cool my cpu, i care how they cool a hotter cpu than i will ever have.
how many of us like seeing psu reviews where they test it up to the 500w that each of us would probably max use. no, we want to know how it handles it's full rated spec.
Wesley Fink - Tuesday, June 5, 2007 - link
Since many have requested it, our next cooler review will be the Thermaltake Big Typhoon VX which we just received in the lab. Thanks to Frozen CPU for getting us a review sample so quickly.It is no wonder the review referenced did not get great performance from the Thermalright Ultra-120. We don't understand why, but the tested the Thermalright with a Noctua low-ouput Silent fan. Our top results were with a high output, but quiet Scythe S-Flex SFF21F. Not surprisingly the Noctua tower, using the same Noctua fan appeared about the same place in cooler performance as the Thermalright.
Perhaps this was just an oversight on the reviewers part, but it appears the Thermalright was set up to do less than top performance in that review. Output of the VX is around 90CFM on high speed, our S-Flex is about 64CFM, and the Noctua is either 30CFM or 47CFM depending on which plug was used for the review.
DrMrLordX - Wednesday, June 6, 2007 - link
Sweet. And yes, I did notice that Hexus tested the Ultra-120 with a pretty slow fan. The main advantage of the original Ultra-120, in my opinion, was that it could continue to benefit from more airflow up to (and exceeding) 140-150cfm. Most HSFs stop benefiting from increased airflow at much lower rates. It doesn't make much sense to use the Ultra-120 (or Ultra-120 eXtreme) with such a slow fan.However, Hexus also tested the Big Typhoon VX at half fan speed, which is around 43 cfm. Big Typhoons (and the VX) see notable improvement in performance all the way up to around 100-110 cfm (and 150 cfm will net you another 1C drop in temps . . . joy). So . . . take that however you will.
Also, since you're responding to reader feedback (which is great), have you taken a look at this review here?
http://forumz.tomshardware.com/hardware/DaClan-Rev...">http://forumz.tomshardware.com/hardware...-Interfa...
This is really very interesting reading, especially in light of the fact that your testers have not found any noticeable difference in TIMs that you've used. Have you considered a high-end TIM shootout of your own?
jkostans - Monday, June 4, 2007 - link
That hexus review doesn't make any sense. They used the thermal paste included with each cooler, rather than a single paste. While this will show out of the box performance, you can't directly compare the coolers to each other. I don't understand the reason for this, I buy my high performance coolers based on how the heat sink performs, not the included paste. I'd like to know who here buys a $60 cooler and doesn't use AS5 or similar paste. I usually like Hexus review too.....DrMrLordX - Tuesday, June 5, 2007 - link
Anandtech does about the same thing that Hexus does. They either use the included paste or bulk silver TIM, depending on whether or not any paste is included with the HSF.Wesley Fink - Tuesday, June 5, 2007 - link
DrMrLordX has correctly stated what we have said in many reviews. Our policy was to use the same quality silver-colored thermal paste if the cooler came with basic white paste of no paste, and the premium stuff if the cooler was supplied with a premium paste. We did this because our testing and experience shows Thermal Paste makes no difference as long as you use a good paste that doesn't dry out, but ig the mfg cared enough to supply the expensive stuff we would use it.We still believe Thermal Paste really doesn't matter, and we only changed to the same paste on each cooler review because we got tired of expalining the truth in every cooler review. Using the same paste just eliminated the questions. No, we will not tell you which paste we used because it is irrelevant. You just need to know we use the same quality paste on every cooler test these days.