![[H]ardOCP](/images/blank.gif){kind=small}
- Date:
- Friday , August 17, 2007
- Author:
- Kyle Bennett
- Google +1
BFGTech ThermoIntelligence Cooler Works?
A look at BFGTech's ThermoIntelligence technology and whether or not it really does what it says it is going to do for your video card's temperatures.
Fan Specs
While putting together our review of the ThermoIntelligence, one more thing about the coolers needs to be covered and that is the fans that are used. Of course airflow and volume can make a big difference when it comes to cooling. Generally, the more airflow you have, the better cooling you are going to have. Both the stock and ThermoIntelligence fans are the typical blower types that have come to be the norm in video card cooling. Interestingly enough, we started experimenting with blower fan usage on video cards back in 1999, albeit our execution was not near as pleasing to the eye.
Checking out the fan specifications you will see one glaring difference. The ThermoIntelligence cooler uses amperage of about 50% more than the stock fan. (0.19A vs. 0.28A) This gives the ThermoIntelligence fan a higher RPM rate. We did not find a tool to allow us to measure the difference, but the proof is in the pudding.
One of the biggest issues with custom coolers is fan noise and that is also addressed on BFG’s ThermoIntelligence page. At full load and at idle, neither the stock cooler nor the ThermoIntelligence unit was bothersome or even audible unless you specifically listened for it in our offices. Noise with the ThermoIntelligence should not be an issue except for the most discerning audiophiles or HTPC users. BFG states that their ThermoIntelligence unit is actually a hair quieter than the stock fan on the 8600 GTS.
In an acoustically controlled environment with a background acoustic noise level of 19.2<20 dB(A), the NVIDIA reference fansink measured 39.9 dB(A) and the BFG ThermoIntelligence fansink measured 39.2 dB(A)
BFG Results
The graph below shows the temperature results that BFG has posted on their site. They give some specifics on their testing, but they fail to say exactly how the temperature data was collected.
They also focus on SLI testing. We only had one card, so SLI testing was out of the question.
HardOCP Results
In order to get our results we tested on an open bench and held the ambient temperature at 81.3 degrees F and did not allow for a variance in the ambient temp of more than 0.3 degrees F. We used NVIDIA’s “NVMonitor” temperature gauge to collect our GPU temperatures. Be warned, this program can be sort of finicky in how it refreshes the temperature. We used a Sperry DT-506 digital thermometer to measure temperatures on the back of the card centered on the GPU to verify temperature fluctuations in NVMonitor. We also used our Sperry DT-506 to record the exhaust temperatures exiting the heatsink units as a further control to make sure our results were reliable.
We used a NVIDIA reference design cooler removed from a BFG 8600 GTS (pictured on the previous pages) and our “OC2” 8600 GTS video card clocked at 720MHz core and 1110MHz memory. We switched the coolers back and forth several times to ensure we had repeatable results. Do be warned that when comparing “identical” video cards and cooling systems in identical environments, you can get very different results simply due to the GPU itself. All like GPUs do NOT run at the same temperatures under the same conditions. Hence the reason we used the same card for all testing and the reason your temps might be different at the same ambient temperature if you are comparing.
As identified by NVTemp, our ThermoIntelligence cooler gave us 6C lower GPU temperature at idle and 11C better GPU temperature under full load. Our data is showing even better results than what BFG has posted on their site. Actually much better.
Given that this is the same video card under the same load with different coolers, it would be logical that the ThermoIntelligence cooler is removing the heat more efficiently than the stock cooler and this is exactly what is shown by the recorded exhaust temperatures.
Things to Keep in Mind
As mentioned, our tests were performed on an open bench with a very controlled ambient temperature. Inside a closed case your results could be very different given that both these cards exhaust inside the case along with most other computer components. If you have a poorly ventilated case, one in which the temperature is very high, an aggressively cooled video card is not going to do you much good. If you keep good airflow in your case you should be able to see results very much in line with ours.
Conclusions
The BFGTech ThermoIntelligence video card cooling unit does exactly what they claim it will do, and that is give you a cooler video card than you would have buying a reference design 8600 GTS heatsink and fan unit. In fact, our results showed single card testing do much better than even what BFG claimed! There is currently a price premium to pay for BFG’s out-of-the-box overclocked ThermoIntelligence video cards though. In fact, you can find stock 8600 GTS cards for as much as $105 cheaper after MIR. Of course the less expensive card does not run at highly overclocked GPU and memory clock rates, does not warranty those OCed speeds, and of course does not have a ThermoIntelligence cooler. Not everyone will find value in ThermoIntelligence, but many will, and be assured that early adopters will pay a higher price. Undoubtedly though, BFG makes no bones about what the ThermoIntelligence cooler will do for your video card, and puts its money where its mouth is.
Discussion
Please join our HardForum discussion about the ThermoIntelligence cooler.