by Sergey Lepilov
07/14/2008 | 04:10 PM
Air coolers from Thermalright Inc. belong to a separate class of highly efficient coolers. Neither the fact that Thermalright solutions are pretty rare in the market these days, nor their high price and no fans bundled with the coolers, not even often imperfect finish of the cooler base can stop real enthusiasts. Overclockers are ready to spend hours looking for these solutions, pay extra for them and then work hard on perfecting the base finish for the sake of superior cooling efficiency. And I have to admit that the result is 100% justified, as in the end there are no coolers out there that could compete with Thermalright solutions.
Our today’s article will be devoted to two coolers from Thermalright that feature minor constructive differences from the solutions we have already reviewed before. Nevertheless, our today’s review will answer some questions you have asked us before. Namely, if Thermalright Ultra-120 A is considerably less efficient than its elder brother, Ultra-120 eXtreme? Or, if the new modified SI-128 SE model is significantly better than its younger brother, SI-128. Is paying extra justified in both cases and are these coolers comparably efficient?
Read our review to answer these and many other questions you might have about Thermalright cooling solutions.
The boxes that all Thermalright coolers come in never have any information n them. The thick brown cardboard is decorated only with the company name and cooler model. That’s it:
The cooler is wrapped in clear plastic and sits securely in a polyurethane foam casing inside the cardboard box. So, it is very unlikely to get damaged during transportation. Next to this casing there is a small box with accessories including the following components:
Like all other Thermalright coolers, Ultra-120 A comes without any fans in the box.
The heatsink of Ultra-120 A cooler is an exact copy of the heatsink on Ultra-120 eXtreme, although it would be more correct to say it the other way around, because Ultra-120 A was designed and released prior to eXtreme version.
This tower heatsink measures 63.4 x 132 x 160mm and weighs 745g (45g less than the heatsink of Ultra-120 eXtreme). The key difference from the “elder brother” is fewer heatpipes: Ultra-120 A has only four of them, while Ultra-120 eXtreme has six. Heatpipes are 6mm in diameter. The heatpipes, heatsink plates and the cooler base are covered with a thin layer of nickel alloy.
Heatpipes bear an array of 52 aluminum plates, each 0.3~0.35mm thick. They are spaced out at 1.6~1.7mm distance:
The manufacturer doesn’t indicate the size of the heat dissipating area. According to our calculations, it should make around 8050sq.cm. The heatsink plates are slightly bent, looking like airplane propeller blades. According to the manufacturer, this shape not only increases the heat dissipating surface area, but also makes the airflow from the fan work more efficiently.
The heatsink sides are not covered with anything, so we assume part of the airflow will exit there (and it proved to be exactly this way):
The heatpipes are non-linear inside the heatsink, just like by Ultra-120 eXtreme model:
So, Thermalright engineers not only made a smaller heatsink, they also managed to more evenly distribute the heat over its plate array:
The aluminum heatsink plates are soldered to the heatpipes.
The heatpipes lie in special grooves in the cooler base and are also soldered to it. The thinnest part of the base measures 2mm.
The base is 40x50mm and is finished like all other Thermalright solutions: you can see and feel the machine marks on it:
Those of you who like mirror-shining polish will have to work on this base pretty hard. As for the evenness of the base surface of our Thermalright Ultra-120 A cooler, it is OK. It is not ideal, of course, but at least this cooler features the best base of the five models we got our hands on. Nevertheless, one of the base corners was lower than the rest. As you can see, the thermal compound imprint on the processor heat-spreader is not perfect:
Later on we checked it on the glass surface and confirmed that this is where the base is slightly bent upwards. Nevertheless, since our 45nm test CPU – Intel Core 2 Extreme QX9650 - has very small contact area I decided not to level out the base. In fact, the finishing touches for the base of Thermalright coolers, such as leveling them out and polishing off nicely, could be a topic for a separate article, and we are going to continue talking about it in our upcoming reviews.
The cooler installation is very simple and is described in detail on the official company web-site. I don’t think I need to repeat it here. The installation guide offers the following schematic description of the installation procedure:
I would only like to add that you have to remove the mainboard from the system case only if you are trying to install Ultra-120 A onto an LGA 775 platform. With AMD 8 processors you use a standard backplate, so no need to remove the mainboard from the case. The cooler is pressed very tightly against the processor heat-spreader and the plate that holds the cooler base not only from the top but also on the sides prevents it from shifting or rotating on top of the CPU.
The distance from the cooler base to the lowest heatsink plate is 41.7mm, so I cannot think of any heatsinks in the area around the processor socket that could hypothetically prevent you from installing Thermalright Ultra-120 A cooler. Maybe a tall chipset cooler could cause a problem, though, in this case all you have to do is rotate the Ultra-120 A by 90 degrees. Look at the photos below:
I would like to say a few words about attaching the fans to the heatsink. First, they included only one pair of retention wire clips with the cooler (although you can fasten two fans to this heatsink). Moreover, these wire clips can barely hold the fans in place almost sliding out of the retention holes. S, if you have the same problem, just turn the clips the other way around and attach them from the opposite side:
In this case it is harder to fasten the fan, however, once you succeed, the clips will stay locked dead in place and you will be able to use one clip per fan. In conclusion I have to add that you have to stick the vibration-absorbing silicon strips between the heatsink and the fan (unfortunately, there are only two of them bundled with the cooler).
Thermalright Ultra-120 A is priced at about $44, which is about $10 cheaper than Ultra-120 eXtreme. Our today’s tests will show if paying extra $10 will actually pay back, and now let’s take a closer look at the second testing participant.
The elder brother of the Thermalright SI-128 cooler we have already reviewed before is shipped in the same exact box as the above discussed Ultra-120 A. However, it is bundled with different accessories (from left to right and from top to bottom):
The “SE” (Second Edition) version differs from SI-128 even less than Ultra-120 A from Ultra-120 eXtreme. To be more exact, we found only three differences. We are going to dwell on them a little later and now let’s recall what this cooler looks like:
The heatsink measures 125 x 145 x 91.5mm and weighs 510g. Four copper heatpipes 8mm in diameter come out of the copper base. They hold an array of 78 thin aluminum plates resting on two wire stands on the other side:
The plates are ~0.25mm thick and are spaced out at about ~1.3mm. The calculated heatsink effective surface area is ~6786sq.cm. The entire cooler is nickel-plated:
The first distinguishing feature of the new SI-128 is the perforated heatsink plates:
There are 36 square holes stamped in each plate. The manufacturer claims that they not increase the airflow turbulence, but also reduce the plates resistance thus improving the heatsink efficiency with low-speed fans. They specifically stress the latter: thanks to the perforated plates you do not need high-speed fans. So, Thermalright SI-128 SE should become a great choice for those who need an efficient and quiet cooler.
Four 8-mm heatpipes lie in special grooves in the cooler base and are soldered to it:
I would like to remind you that the first modification of this cooler that we tested earlier didn’t have the heatpipes soldered to the base: they used some thermal glue (as we found out after the article was up). Different contact between the heatpipes and the cooler base is actually the second distinguishing feature of the SE model. However, I have to say here that the last shipments of the regular SI-128 may also have come with soldered heatpipes already.
The base finish quality of the Thermalright SI-128 SE is exactly the same as that of the Ultra-120 A we have just discussed above:
However, the evenness of the base surface is much poorer. It turned out a little dented, so the main contact with the processor heat-spreader will be right in the center:
Nevertheless, we didn’t level out or polish off the cooler base. I will explain a little later why. Moreover, as I have already said, we are going to have a special article discussing the issues with the evenness and finish quality of Thermalright coolers.
Thermalright SI-128 SE installation may seem complicated, but in reality it is very simple, although it requires quite a bit of time. The schematic of the entire procedure looks as follows:
When the cooler is already installed it looks like this:
The cooler is pressed firmly against the CPU heat-spreader. It has nothing in common with the formerly used plastic clips similar to the boxed coolers retentions.
The backplate prevents the mainboard from bending. Actually, new cooler retention is the third distinguishing feature of the new SI-128 SE.
At first glance you can’t really tell what cooler modification is installed inside the system case:
Our tests will show if the new modification is more efficient than the old one. Here I would only like to add that Thermalright SI-128 SE is currently selling for $40-$45.
The Thermalright coolers we have just discussed and their competitors were tested in two modes: in an open testbed when the mainboard sits horizontally on the desk and the cooler is installed vertically, and in a closed testbed with the mainboard in vertical position.
Our testbed was identical for all coolers and featured the following configuration:
All tests were performed under Windows XP Professional Edition SP2. SpeedFan 4.34 Beta 44 was used to monitor the temperature of the CPU, reading it directly from the CPU core sensor:
The mainboard’s automatic fan speed management feature was disabled for the time of the tests in the mainboard BIOS. The CPU thermal throttling was controlled with the RightMark CPU Clock Utility version 2.35.0:
The CPU was heated up with OCCT (OverClock Checking Tool) version 2.0.0a in a 23-minute test with maximum CPU utilization, during which the system remained idle in the first 1 and last 4 minutes of the test:
I performed at least two cycles of tests and waited for approximately 20 minutes for the temperature inside the system case to stabilize during each test cycle. The stabilization period in an open testbed took about half the time. Despite the stabilization period, the result of the second test cycle was usually 0.5-1°C higher. The maximum temperature of the hottest CPU core of the four in the two test cycles was considered the final result (if the difference was no bigger than 1°C – otherwise the test was performed at least once again).
The ambient temperature was checked next to the system case with an electronic thermometer that allows monitoring the temperature changes over the past 6 hours. During our test session room temperatures varied between 24.0 ~ 24.5°C. It is used as a staring point on the temperature diagrams. Note that the fan rotation speeds as shown in the diagrams are the average readings reported by SpeedFan, and not the official claimed fan specifications.
The noise level of each cooler was measured according to our traditional method described in the previous articles with the help of an electronic noise meter – CENTER-321. The subjectively comfortable noise level was considered 34.5dBA and is marked with a dotted blue line in the diagram. The ambient noise from the system case without the CPU cooler didn’t exceed 32.5dBA when measured at 1m distance.
As you may have already guessed, our today’s heroes will be competing not only against one another. We have also added the results for Thermalright Ultra-120 eXtreme and Thermalright SI-128 (both with their original retentions) from the previous test sessions, which we have been using as performance reference for quite some time (for tower and top-coolers, respectively). I would like to remind you that both these coolers have even base surface without any dents and didn’t require any additional polishing. All Thermalright coolers were equipped with one or two Scythe SlipStream 120 fans working at ~860RPM and ~2000RPM.
And in order to spice up the race with a non-Thermalright participant, we added the results for a very successful ZEROtherm Nirvana NV120 Premium cooler:
We tested this cooler in two modes: with the fan rotating at ~1280RPM with moderate noise and at maximum fan rotation speed of ~2780RPM and very high level of noise.
Using the “weakest” cooling system with the fan in quiet mode we managed to overclock our quad-core processor to 3900MHz without losing stability or getting into the throttling mode. The processor Vcore was increased to 1.5625V in the mainboard BIOS (1.53~1.54V according to the monitoring utility).
Since there are too many results for our coolers in different configurations, we split the diagrams into two: for the open testbed and closed system case.
At first let’s talk about the results obtained inside the system case:
Although the results obtained inside a closed system case are pretty relative, they are still very interesting. First of all, I have to say a few words about Thermalright SU-128 SE. this pretty lightweight and compact cooler with only one fan turned out more efficient than such monsters as Thermalright Ultra-120 eXtreme with two fans and ZEROtherm Nirvana NV120 Premium ay maximum fan rotation speed of 2780RPM! Remarkable results, but unfortunately only with a 120-mm fan on the side of the system case. As for the comparison between the SE and the regular SI-128 cooler, the former is about 4-5ºC more efficient, where 2ºC are delivered by the new retention (we checked it by using the retention from SE on the non-SE version). The remaining 2-3ºC seem to result from perforated heatsink plates and heatpipes soldered to the base.
As for Thermalright Ultra-120 A and Ultra-120 eXtreme, the extreme model is the winner here, although it is only 1-3ºC better. With two quiet fans rotating at ~860RPM these coolers prove equally efficient.
Now let’s check out the results in an open testbed:
As you can see, tower coolers regain their leadership, although SI-128 SE works as efficiently as Ultra-120 eXtreme in the corresponding acoustic modes. The latter is again just a tiny bit better than its younger brother – Ultra-120 A.
Since all Thermalright coolers were tested with one or two identical Scythe Slip Stream 120 fans, the acoustic measurements actually turned into measuring the level of noise generated by the fan(s). And since ZEROtherm Nirvana Nv120 Premium cooler features its own fan, its results were taken separately. We measured the level of noise generated by our today’s testing participants from a 3cm, 1m and 3m distance. The results are given on the diagram below:
As you can see, the coolers can be actually used only with slowly rotating fans, like a couple of SlipStream 120 at ~860RPM or ZEROtherm Nirvana NV120 Premium fan at ~1280RPM that yielded just a tiny bit to the competitors. All other combinations may only be used for short-term benching, as they prove excessively noisy.
First I’d like to say a few words about the “younger” brother – Thermalright Ultra-120 A. although it has two heatpipes less than Ultra-120 eXtreme, it has minimal influence on the results. Looks like two extra heatpipes on the sides of the Ultra-120 eXtreme’s base do not make any significant contribution into the final result. That is why paying $10 more for the eXtreme version is hardly justified. However, if you are an overclocker who hunts for every additional degree, then you should still go with the eXtreme model.
And now about the cooler that truly amazed us with its efficiency – Thermalright SI-128 SE. We knew that SI-128 was just a little behind Ultra-120 eXtreme. Now we can state that its modified version outperforms the former leader. See for yourselves: Ultra-120 eXtreme and SI-128 SE are equally efficient in an open testbed, while in a closed case SI-128 SE is even better than the competitor. SI-128 SE also boasts lower weight and more affordable price point: Ultra-120 eXtreme is considerably more expensive, which may swing the votes in favor of the SI-128 SE. Besides, SI-128 SE directs the airflow towards the mainboard PCB thus cooling down the components around the processor socket, although it has very little influence on the processor overclocking potential. As always, it is up to you to decide. As for me, I have found my new ideal. All I will do now is work on the impeccable finish of its base.
In conclusion we would like to wish Thermalright Inc. to pay a little more attention to the base surface quality. I believe that if you are buying a cooler priced like an inexpensive CPU, you do not really want to bother finishing off the base with all related consequences.