TT Big Typhoon vs. Zalman CNPS 9500 vs. Scythe Ninja
Leonardo
Wake up and smell the glaciersEagle River, Alaska Icrontian
Thermaltake Big Typhoon
versus
Scythe Ninja
versus
Zalman CNPS 9500
Reference Heatsink: Zalman CNPS 7700-Cu
No, this isn't a rasslin' (pro wrestling) promo, it's just observations on heatsinks I'm testing. No mulit-page review follows, just a couple comparisons.
Preparing to convert systems 1 and 3 (in my signature below) to dual core rigs, I thought it wise to procure major league air cooling. You see, not only will the machines be overclocked and running at 100% load all the time, they will also be Intel D820 rigs. For those of you not familiar with this CPU, it's a wild energy pig that belches out flames. In short, it runs very, very hot.
Having been throroughly impressed with the Big Typhoon (system 4), I ordered another one. While I was at it, I saw the Scythe Ninja at Newegg and thought I'd try it. I read a comparison review a few months back between the Ninja and the Typhoon. Both were given very high marks. So I thought I'd try it. And then, I found a used Zalman 9500 for a great price at a trading forum. So, what's Leo to do? Uh, test 'em all!
OK, what follows is not scientific. The main variables involved were two different motherboards, cases, and the corresponding computers being in different parts of my office, but the mini-test did have a consistent room temperature throughout, and each comparison was done against the same type of cooler, Zalman CNPS 7700CU. Temperature readings were taken after sufficient time for the computers to fully heat the CPU. It doesn't take long with two instances of Folding@Home running on each processor. Temperature measurements for the Zalman 9500 and the Scythe Ninja were taken well before the Arctic Silver thermal interface paste had a chance to settle, or "cure".
First test - System 3:
Configuration: as you can see in my signature, except when the test began, when the computer contained a Zalman CNPS 7700CU.
Results: with the Zalman 9500 now mounted, the computer at stated OC and at full load in hyperthreading, the CPU and system temperatures are lowered respectively by 5*C and 3*C.
Observations on mounting the 9500 in on a Socket 478 board:
Difficulty Level - very easy (characteristic of Socket 478)
Heatsink noise level - quieter (to my ears) than the 7700
Aesthetics - the 9500 is as much art as engineering. It's beautiful in a windowed case
Mounting method and hardware - excellent, stable
===========
Second test - System 1:
Configuration: as you see in my signature, except when the test began, when the computer contained a Zalman CNPS 7700CU. Although the Ninja is designed to run passively if the user desires, I installed a 120mm very low CFM Antec fan.
Results: with the Scythe Ninja now mounted, the computer at stated OC and at full load in hyperthreading, the CPU and system temperatures are lowered respectively by 5*C and 4*C.
Observations on mounting the Ninja on a Socket 478 board:
Difficulty Level - difficult. The springloaded clips that are affixed to the body of the Ninja were pretty tough to seat in the socket 478 motherboard's mounting bracket. But when each catch (clip) did seat, I could hear a snap. No question as to whether the catches were in place. I did not like the pressure I had to apply onto the motherboard to engage the clips. But then, there was no bare CPU core underneath to be potentially crushed as in the K7 Athlon days.
Heatsink noise level - even with a fan mounted, this unit is wonderfully quiet. When I first started up the computer after mounting the Ninja, I did a double-take, having to look at the heatsink to make sure the fan was spinning. And that was with the case side off. If the Big Typhoon can be considered silent, then the Scythe Ninja is more silent. (More silent? "Silent" is an absolute. How can an absolute vary? )
Aesthetics - very industrial look. I like it.
Mounting method and hardware - once mounted, the Scythe is locked onto the motherboard very firmly. It's not nearly as heavy as it looks.
======================
Keep in mind, this comparison doesn't really compare the three coolers in the thread title to each other. It compares the 9500 to a 7700 in a specific configuration and compares a Ninja to a 7700 in a separate, different configuration. This wasn't meant to be exhaustive, but merely to give whoever reads the comparo at least some frame of reference, the baseline being a Zalman CNPS 7700CU in an overclocked system under full load.
Some more observations. System temperatures in both systems were lowered significantly by installing the new heatsinks (versus the 7700s). Why? The wide, 120mm Zalman 7700 is mounted with the fan parallel to the motherboard. It would seem the wash from the very wide fan blowing down would serve to help cool board circuitry and the norhtbridge. Well, I believe it does, but there are distinct advantages to the designs of both the Ninja and the Zalman 9500. Both heatsinks have the fan (if so fitted on the Ninja) oriented parallel to the computer case's exhaust fan. This is very efficient. With both heatsinks, the sink's fan blows cool air through the heatsinks cooling veins. Like a pipe or tunnel, the air continues straight through the heatsink out through the case exhaust fan.
Although I am highly impressed with all three of the big boy air coolers - the Zalman CNPS 9500, the Scythe Ninja, and the Thermaltake Big Typhoon, my respect for Zalman 7700 is probably increased. The three coolers of this thread's title are considered the bad boys of CPU air cooling, a position the Zalman 7700 occupied a couple years ago. Considering that the 7700 is now the older generation, that generation works very well.
So Leo, you may ask, when will you compare the Big Typhoon to the Zalman 7700. Well, I already did that when I had first built my system 4. The temperature delta between the Typhoon and the the 7700 was even greater than that of the other two coolers compared to the 7700.
Every review at tech sites it seems, must proclaim a winner, the best, or what have you. I just won't do that to you. Firstly, my comparison methodology was not structured well enough (only took two hours!) to provide the precision of measuring necessary to spout a superlative such as "the best". Secondly, each of these coolers has its own merits. Allow me to provide you what I see as best attributes of each cooler:
Big Typhoon - probably the most effective cooler I've ever put in a computer. And it's remarkable that such a powerful heatsink is as quiet as a sleeping kitten.
Scythe Ninja - the King of Quiet. I have no doubt this cooler would effectively cool any modern CPU without a heatsink fan. Yes, without! Overclocking, especially a D series Intel dual core, is another matter. But with a fan installed, this beast is indomitable. I would imagine the Ninja in passive configuration could handle an overclocked AMD X2 or dual core Opteron! If money for a heatsink purchase is a discriminating factor, the Ninja is definitely your choice. If you already have a 120mm fan in the parts bin, the Ninja absolutely cannot be beat for it's price.
Zalman CNPS 9500 - I already wrote it above. The Zalman is a work of art and is simply a pleasure to look at. Fine craftsmanship is always a joy. Also, it sure doesn't hurt anything that the 9500 is also major league in cooling ability.
I hope you enjoyed my small article.
Leo
Thermaltake Big Typhoon
Scythe Ninja
Zalman CNPS 9500
Zalman CNPS 7700CU
versus
Scythe Ninja
versus
Zalman CNPS 9500
Reference Heatsink: Zalman CNPS 7700-Cu
No, this isn't a rasslin' (pro wrestling) promo, it's just observations on heatsinks I'm testing. No mulit-page review follows, just a couple comparisons.
Preparing to convert systems 1 and 3 (in my signature below) to dual core rigs, I thought it wise to procure major league air cooling. You see, not only will the machines be overclocked and running at 100% load all the time, they will also be Intel D820 rigs. For those of you not familiar with this CPU, it's a wild energy pig that belches out flames. In short, it runs very, very hot.
Having been throroughly impressed with the Big Typhoon (system 4), I ordered another one. While I was at it, I saw the Scythe Ninja at Newegg and thought I'd try it. I read a comparison review a few months back between the Ninja and the Typhoon. Both were given very high marks. So I thought I'd try it. And then, I found a used Zalman 9500 for a great price at a trading forum. So, what's Leo to do? Uh, test 'em all!
OK, what follows is not scientific. The main variables involved were two different motherboards, cases, and the corresponding computers being in different parts of my office, but the mini-test did have a consistent room temperature throughout, and each comparison was done against the same type of cooler, Zalman CNPS 7700CU. Temperature readings were taken after sufficient time for the computers to fully heat the CPU. It doesn't take long with two instances of Folding@Home running on each processor. Temperature measurements for the Zalman 9500 and the Scythe Ninja were taken well before the Arctic Silver thermal interface paste had a chance to settle, or "cure".
First test - System 3:
Configuration: as you can see in my signature, except when the test began, when the computer contained a Zalman CNPS 7700CU.
Results: with the Zalman 9500 now mounted, the computer at stated OC and at full load in hyperthreading, the CPU and system temperatures are lowered respectively by 5*C and 3*C.
Observations on mounting the 9500 in on a Socket 478 board:
Difficulty Level - very easy (characteristic of Socket 478)
Heatsink noise level - quieter (to my ears) than the 7700
Aesthetics - the 9500 is as much art as engineering. It's beautiful in a windowed case
Mounting method and hardware - excellent, stable
===========
Second test - System 1:
Configuration: as you see in my signature, except when the test began, when the computer contained a Zalman CNPS 7700CU. Although the Ninja is designed to run passively if the user desires, I installed a 120mm very low CFM Antec fan.
Results: with the Scythe Ninja now mounted, the computer at stated OC and at full load in hyperthreading, the CPU and system temperatures are lowered respectively by 5*C and 4*C.
Observations on mounting the Ninja on a Socket 478 board:
Difficulty Level - difficult. The springloaded clips that are affixed to the body of the Ninja were pretty tough to seat in the socket 478 motherboard's mounting bracket. But when each catch (clip) did seat, I could hear a snap. No question as to whether the catches were in place. I did not like the pressure I had to apply onto the motherboard to engage the clips. But then, there was no bare CPU core underneath to be potentially crushed as in the K7 Athlon days.
Heatsink noise level - even with a fan mounted, this unit is wonderfully quiet. When I first started up the computer after mounting the Ninja, I did a double-take, having to look at the heatsink to make sure the fan was spinning. And that was with the case side off. If the Big Typhoon can be considered silent, then the Scythe Ninja is more silent. (More silent? "Silent" is an absolute. How can an absolute vary? )
Aesthetics - very industrial look. I like it.
Mounting method and hardware - once mounted, the Scythe is locked onto the motherboard very firmly. It's not nearly as heavy as it looks.
======================
Keep in mind, this comparison doesn't really compare the three coolers in the thread title to each other. It compares the 9500 to a 7700 in a specific configuration and compares a Ninja to a 7700 in a separate, different configuration. This wasn't meant to be exhaustive, but merely to give whoever reads the comparo at least some frame of reference, the baseline being a Zalman CNPS 7700CU in an overclocked system under full load.
Some more observations. System temperatures in both systems were lowered significantly by installing the new heatsinks (versus the 7700s). Why? The wide, 120mm Zalman 7700 is mounted with the fan parallel to the motherboard. It would seem the wash from the very wide fan blowing down would serve to help cool board circuitry and the norhtbridge. Well, I believe it does, but there are distinct advantages to the designs of both the Ninja and the Zalman 9500. Both heatsinks have the fan (if so fitted on the Ninja) oriented parallel to the computer case's exhaust fan. This is very efficient. With both heatsinks, the sink's fan blows cool air through the heatsinks cooling veins. Like a pipe or tunnel, the air continues straight through the heatsink out through the case exhaust fan.
Although I am highly impressed with all three of the big boy air coolers - the Zalman CNPS 9500, the Scythe Ninja, and the Thermaltake Big Typhoon, my respect for Zalman 7700 is probably increased. The three coolers of this thread's title are considered the bad boys of CPU air cooling, a position the Zalman 7700 occupied a couple years ago. Considering that the 7700 is now the older generation, that generation works very well.
So Leo, you may ask, when will you compare the Big Typhoon to the Zalman 7700. Well, I already did that when I had first built my system 4. The temperature delta between the Typhoon and the the 7700 was even greater than that of the other two coolers compared to the 7700.
Every review at tech sites it seems, must proclaim a winner, the best, or what have you. I just won't do that to you. Firstly, my comparison methodology was not structured well enough (only took two hours!) to provide the precision of measuring necessary to spout a superlative such as "the best". Secondly, each of these coolers has its own merits. Allow me to provide you what I see as best attributes of each cooler:
Big Typhoon - probably the most effective cooler I've ever put in a computer. And it's remarkable that such a powerful heatsink is as quiet as a sleeping kitten.
Scythe Ninja - the King of Quiet. I have no doubt this cooler would effectively cool any modern CPU without a heatsink fan. Yes, without! Overclocking, especially a D series Intel dual core, is another matter. But with a fan installed, this beast is indomitable. I would imagine the Ninja in passive configuration could handle an overclocked AMD X2 or dual core Opteron! If money for a heatsink purchase is a discriminating factor, the Ninja is definitely your choice. If you already have a 120mm fan in the parts bin, the Ninja absolutely cannot be beat for it's price.
Zalman CNPS 9500 - I already wrote it above. The Zalman is a work of art and is simply a pleasure to look at. Fine craftsmanship is always a joy. Also, it sure doesn't hurt anything that the 9500 is also major league in cooling ability.
I hope you enjoyed my small article.
Leo
Thermaltake Big Typhoon
Scythe Ninja
Zalman CNPS 9500
Zalman CNPS 7700CU
0
Comments
For those folks who are wondering, here's some costs for the 3 heatsinks new:
Thermaltake Big Typhoon - $49.99 + free shipping, ZZF
Zalman CPNS 9500 - $54.99 + free shipping, ZZF
Scythe Ninja - $39.99 + $4.99 shipping, Newegg
Scythe Ninja w/120mm fan - $49.99 + $4.99 shipping, Newegg
Update on the Ninja:
I replaced the fan on the Ninja with a three-speed, quiet Antec 120. I also replaced the rear exhaust fan with an Antec three-speed, heavy-duty ball bearing model (different than on the heatsink, but also very quiet). The Ninja is still nearly silent - I can only hear it with the case side cover off. Temperature delta now at full load is now 7*C lower than with the Zalman 7700. I am becoming an enthusiastic proponent of the Scythe Ninja. I cannot tell you which I prefer, the Ninja or the Typhoon. I suppose the Typhoon wins in the impressive looks category.
Please, if anyone here has been able to personally compare the XP120 to the other mentioned heatsinks, please post your comments! Of course, a review link would be good too.
I like your take on the situation, not declaring a "winner" but listing the applications for each HSF.
OK, some more observations. All the computers in my home office, Systems 1, 3, and 4 have all been upgraded now to Pentium D820 systems (Smithfield dual core, 2.8GHz rating per core). All are overclcocked between 700 to 1000MHz. Each computer has both cores running full load. The three computer cases are all different, but all have very similar air flow - very good air flow. The computers' CPUs' core temperatures are all within *1C of each other. System 3 has the Scythe Ninja with a medium CFM 120mm fan, 1 has a Big Typhoon with an Antec Tricolor 120 set at speed two (2000RPM), and 4 has a Big Typhoon with the stock Thermaltake 120. None of these heatsinks can be heard over the case intake and exhaust fans of the cases, which are all relatively quiet.
If maximum air cooling is your goal, you can't go wrong with either a Big Typhoon or Scythe Ninja. I would imagine that the Ninja would work with most applications in passive mode - completely without a fan! It rests right next to the case's 120mm exhaust, air being pulled through at a high rate. It's cooling vanes are spaced apart broadly enough to permit rapid air flow.
The Typhoon with the Antec Tricolor is very pleasing to the eye. Looking into the case you see swirling green, blue, and red with a backround of burnished copper. If you're into the eye candy - windowed case setups, it's really quite a site. The best though, was the Zalman CNPS 9500. Even my wife thought it was "pretty".
I have to agree with GHoos and you, Leo. You're simply not equipped to make the scientific analysis.
For example- I've seen reviews in other trades that might disagree with some of your temperature findings (Maximum PC for one). Perhaps the most common mistake I see in touting coolers is in postings like
"I ran my PC at 29C loaded!". Period. For all I know, that could be some drunk yahoo in Greenland playing solitaire on his back porch in January.
So, while I KNOW you've done far better than that, I've appreciated what you've done in trying to make clear that you are giving us the qualitative "feel" for these systems. Thanks, Leo!
You wouldn't have to have a perfectly controlled lab though,to run valid testing. As long as the conditions were the same for each test, it would be simple. As long as:
the exact same computer was used for each cooler
each cooler was mounted with the same thermal interface material
the computer ran the same software for the same amount of time
the room temperature was constant (I can't do that!)
But the results of such a test would still not prove which cooler worked best. All the tester would be able to determine would be which cooler worked best for him in his given setup. Yes, I've seen a lot of very sloppy, inclusive heatsink comparisons at tech sites where it was very unclear what the variables were. But I have also seen very simple tests/comparisons that were valid, as the authors stated clearly what the variables were and took pains to point out that the results could only be considered valid for the test platform used in the test.
While I'm on that topic. Be careful with user ratings at Newegg. Half the people posting reviews seem to have no clue. Yeah, this motherboard overclocks great - does not post stock versus overclocked results. Yeah, this cooler cut my temps by 345* Kelvin - does not state what he was using before, case config, room temp, et cetera.
I had to cook dinner tonight, so I didn't get a chance to complete my thoughts in my post. Here they are-
Don't cut yourself so short on how well you've done scientifically, either. Your numbers (I've read some of your prior writings on the subject as well) actually matched pretty well with many other tests I've seen when I was fishing for a cooler (and actually, I have my misgivings about Maximum PC and trust places like Tech Report more). I found your numbers "qualitatively" to be both credible and "in the ballpark".
So I think you've done a good ... no ... G-R-E-A-T job!
(PS- I settled on the Zalman. It was the best overall "fit" for me.)
One is to be fitted to an Abit 775 series board that, according to Abit's 3rd eye, is running at 67C. I still have a spare Big Typhoon, but it does not look like it would fit, hence the trial with the Ninja.
I have fitted 2 x 9500's to our new opteron duallie, but will not be able to compare to stock as i won't be using them. On looks alone, 2 x 9500's is "Teh Bling".:)
DragonV8 - glad to see you've found good cooling configurations.
Your write-up convinced me to order the BT for an upcoming 805 build.
Jon, the stock heatpipe heatsinks that come on the 1xx Opterons are actually a pretty darn good heatsink if you change out the stock fan with a better fan like the Delta FFB series or a TT SmartFan 2. I assume your 265 Opty's come with the same heatpipe solution too. They are in a bit of demand from people over here in the US and are selling for around $20-25. So, if you have a rig that's using the older non-heatpipe version I would change it out with one of the retail hsf's you received with your Opty's. If you can find a Delta FFB0812HHE 80mm fan over there in Oz, they are a great replacement for the stock fan. It brings the performance up to an XP90's cooling performance.
Many Short-Media members will find this thread useful as they adopt high heat producing Intel dual core CPUs.
I went from 39C idle to 32C idle and 39C load!!! No I am not kidding and yes my old heat sink was installed correctly it was a coolermaster Hyper 6 (original version), so it wasn't exactly a crap heat sink. It was easy as anything to install and of corse it looks good to. Definitely gets a thumbs up from me!! if they weren't 60 bucks, I'd pick one up for the wifes rig too!
I had no trouble whatsoever. The instructions are simple to follow. In my case I had to remove the black plastic HSF mounting bracket on the top of the board and retain the faceplate at the back of the board. Mine was a metal faceplate, some I believe are made of plastic, in which case the instructions are slightly different.
I then mounted the Big Typhoon directly onto the CPU and used the same mounting holes and screws that came off the boards HSF mounting bracket to tighten it up. Prime is right about the tensioning. The mounting bracket must be tight and should display a slight convex shape.
Hope the attached pics help a bit
Yes, the Typhoon outclasses the 9500 both in cooling effectiveness and quiet operation. At full load, you can expect about a 4*C better cooling efficiency with the Typhoon over the 9500. If you won't be overclocking and have a windowed case, I'd go with the Zalman 9500. It takes up far less case room and looks great.
BTW, what case and motherboard will you be using?
Regardless, where the processor will be, directly above it will be the mesh side panel, the typhoon would be able to pull air through the mesh pretty easily since it will be close to it anyway...That actually may be beneficial in a way.