Authour: James Prior
Editor: Charles Oliver
Date: November 4th, 2010
To test the efficacy of Tuniq's TX-4 TIM, we used an AMD Phenom II X2 550. With our ASUS M4A89GTD PRO/USB3, we compared stock AMD heatsink performance in overclocked and core unlocked configurations. These tests used stock AMD pre-applied TIM to compare with Tuniq TX-4. We also compared the performance of AMD's GPU pre-applied thermal paste, in the new AMD Radeon HD 6850 graphics card. Below you can see the test configuration:
Test System Specifications
The standard AMD heatsink is made of aluminum with a 4-pin PWM 70mm fan installed. The base of the heatsink is aluminum, and actually smaller than the heatspreader of a standard AMD Athlon II or Phenom II processor. The included thermal interface material is spread evenly over the base, ready for squishing out when you mount the heatsink to the socket. AMD's standard latch secures the heatsink to the socket, holding the cooler quite securely in place - no wiggle or slop.
Stock AMD Heatsink & Fan
Stock AMD HSF Base
AMD Phenom II
For the CPU tests we used three configurations of our AMD Phenom II X2 550 Black Edition. The processor was run at stock 3.1GHz, overclocked to 3.7GHz, and finally core unlocked to a quad core and overclocked to 3.4GHz. CPUID's HWMonitor was used to record minimum and peak core temperatures, at desktop idle and under prime95 load. CPU voltage was increased to 1.425v with CPU load line calibration enabled for the overclocked settings. ASUS EPU was set to maximum power savings for idle condition, and to high performance for running the CPU load tests. The results represent the lowest and highests results recorded, following cold boot to desktop and idling for 10 minutes, a ten minute Prime95 burn, and 10 minute cool-down period. The Cooler Master HAF 922 case was closed and upright, under a desk in a room with ambient temperature of 72F/22C.
The mainboard CPU thermal protection shutdown the system when testing the 4-core overclock configuration, with both TIMs. It lasted a little longer with the Tuniq TX-4, but the increase in thermal transfer wasn't enough to compensate for the lack of cooling capacity from the stock AMD heatsink.
For our GPU tests we added AMD's new Radeon HD 6850 graphics card. Again, we used CPUID HWMonitor to record low and high temperatures, from Windows 7 desktop idle (single display) and Furmark 1.8.2 run for 120 seconds at 1920x1080 resolution with 4xMSAA, Xtreme Burning Mode and Displacement enabled, with Post Processing enabled. The CPU was set to default core count and clock speed, with ASUS EPU set to high performance mode. AMD's standard fan profile was used.
The TX-4 equipped Radeon HD 6850 took longer to reach it's peak temperature than the stock one, and so was a little quieter for part of the test.
Summary & Conclusions
Clearly, Tuniq's TX-4 is superior to the standard thermal paste shipped with modern processors. The significant temperature decreases will help equipment last longer, as lower long-term operating temperatures will extend the life of the component. It can also help lower noise, as less active cooling will be required to maintain the same temperatures. The included TIM spreader is a nod to the 'credit card' spread method of TIM application, but quite difficult to use with the CPU in socket. The syringe delivery method works well, making it easy to apply specific amounts of TIM and the cap prevents excessive drying of paste still in the tube. The supplied 1ml is plenty for several applications for CPUs and GPUs, being around the same amount as you'd find in a 5g tube of a competitor's product.
content not found