Thermalright Ultima 90

Overall Score

I ordered my Thermalright Ultima 90 about a week ago. When it first arrived, I saw, unsurprisingly, the typical Thermalright box: a lackluster brown cardboard box with the words "Ultima 90-I" on the side. Opening the box yielded few other surprises, as the heatsink was protected by the typical tight fitting foam insert. All the contents of the box were unharmed.

Contents include:
1. Thermalright Ultima 90 heatsink
2. 1 set each of 92mm and 120mm fan clips
3. Socket AM2 and Socket 775 mounting hardware
4. Universal metal backplate
5. Anti-vibration strips
6. Installation guide
7. Decent size tube of thermal grease

Overall, a decent package, although I would have liked to see the socket 754/939 mounting hardware included instead of as a separate item. The box is also rather plain, but is effective in protecting the contents: Form follows function.

The construction of the heatsink, I would rate as very good. It has 12 nickel plated copper heatpipes coming up from a nickel plated copper base. The fins are made of aluminum, and are soldered to the heatpipes. The dimensions of the heatsink are 115mmx55mmx139mm. This is quite a bit smaller than even its big brother, the Thermalright Ultra 120 eXtreme, which itself is smaller than many other heatpipe coolers. The Ultima only weighs a paltry 460g without a fan, which is much less than other coolers in the same price and performance category. A fan is not included, so one will need to be added by the owner. Both 92 and 120mm fans will work, and clips are included for each. The base polish is average, as even machining marks are present. I also have noticed that the base is a bit convex, but this is supposedly part of Thermalright’s design.

The test setup:

I am in a dorm room, and as such have no mechanism of controlling the environment, but have done the best possible to ensure a decent level of equality.

Stock AMD 4 pipe cooler setup:
AMD Opteron 165 socket 939 processor WITH IHS
Included fan: 6300 RPM 70 mm fan
Tests are run at 100% fan speed; so about 60dB

Thermalright Ultima 90 setup:
AMD Opteron 165 socket 939 processor WITHOUT IHS
Sunon 92mm fan with 50 CFM
Tests are run at 100% fan speed; so about 37dB

Common Parts:
MSI K8N Neo 2 Platinum Motherboard
2x 512mb Centon PC3200 DDR400 Memory
GeCube X1950 Pro AGP Graphics Card
2x 36gb Western Digital Raptor Hard Drives
120gb Western Digital SE 8mb cache Hard Drive
Antec 900 Case with side door 120mm Fan
Sunbeam Nuuo 450 watt Power Supply
Lite-On DVD-RW Drive

Thermal Paste
Arctic Silver 5 thermal paste was used for all tests. Both heatsinks were allowed a thermal paste ‘break in’ time of more than 48 hours.

The Results
This is a rather non scientific benchmark of the Thermalright, as I have modified the test system for the better between switching heatsinks: no more Integrated Heat Spreader for my Opteron! As such, these results are somewhat meaningless, and reflect the removal of the IHS as well as the Ultima. I also do not have controlled conditions to test in, so temperatures are relative. There also isn’t a dB meter available to me currently, so again, sound measurements are based on manufacturer’s specifications. Because of all of this, I will say that this is a "poor-man’s" benchmark, or a real world comparison, you might say. Does the average owner get out a dB meter and check the sound output, I think not.

Without further ado, the temps!

I did not test idle temperatures, because although somewhat important, the load temperatures are much more so.

Stock Heatsink: 1.8ghz 1.325vcore –> 41C

Thermalright Ultima 90: 1.8ghz 1.325vcore –> 35C

Stock Heatsink: 2.7ghz 1.325vcore –> 51C

Thermalright Ultima 90: 2.7ghz 1.325vcore –> 41C

Stock Heatsink: 2.844ghz 1.44vcore –> 62C

Thermalright Ultima 90: 2.844 ghz 1.44vcore –> 46C

As you can see, the Thermalright Ultima 90 slaughters the stock AMD heatsink.
Just for fun, lets see how well the Thermalright scales at heat output the AMD heatsink can’t handle:

Thermalright Ultima 90: 2.9ghz 1.55vcore –> 54C

Not too bad. I believe that my thermal paste job could be improved ever so slightly, which would help account for the slight difference in temperature differences between cores.

So here’s a synopsis of the results:


It seems to me that the Thermalright Ultima 90 is a good choice for those out there looking for a good heatsink but have limitations. What do I mean you say? I mean that those looking for a heatsink that doesn’t weigh nearly 2 lbs, doesn’t take up tons of room, or are looking for a heatsink that can be operated quietly, this is a viable choice for you. As can be seen, there is a massive difference between the stock cooler and the Ultima. For one, I was able to push the processor to 2.9ghz 1.55vcore and complete 10 minutes of Orthos, while maintaining temperatures under 55C. I didn’t try running the stock cooler at those settings, but I can only imagine that it would top out well over 65C.

Keep in mind that while keeping most variables under control, I also removed the IHS from the Opteron between testing the Stock Heatsink and installing the Ultima. This is a farily major change but, should only change my final temps by 2~4C, although some report greater gains. Why not test the stock heatsink without the IHS? This would be rather dangerous, due to the mounting hardware’s inability to apply proper mounting pressure to the core now that the IHS is removed.




  • Facebook
  • Twitter
  • Myspace
  • Google Buzz
  • Reddit
  • Stumnleupon
  • Delicious
  • Digg
  • Technorati

Leave A Response