Malpine Walis
10-08-03, 11:12 PM
I am beginning to wonder if Ed is hoping that people will kill a few CPUs on this one. This time we do not have an article to quote but only a link. Lost Circuits review (http://www.lostcircuits.com/advice/carbon_black/)
We had three threads on this the last time Ed posted about this supposedly “revolutionary” heat sink compound. The one that I started here and two in the Cooling forum. The fact is that carbon black is simply not the best thing since sliced bread. It is, in fact, sold to industry as thermal insulation for blast furnaces.
A Dr. Chung published a paper claiming that she had developed a remarkable new heat transfer material. However, there were several problems with the paper in relation to over-clocking. The most obvious one being that she did not compare her new wonder substance to any commercially available heat sink compound. What she did compare her stuff to was only such materials that are already known to be worse thermal conductors than carbon. Basically, it was a runoff between thermal insulators and the one that was the worst thermal insulator was declared to be the king of heat transfer materials.
Another problem was that her tests were performed on highly specialized laboratory equipment that is not the same thing as a processor and it did not adequately model how her product would be used as we use products in our computers. Each test only lasted a fraction of a second. Even the most extreme of the cooling crowd leaves a heat sink installed for quite a bit longer than that.
Both of these problems (and a few others) were identified in the threads that we had on the matter the last time Ed had something up. Now he is linking to an article that tells us nothing new about the matter.
In fact, if you read the article, you will find several problems here as well. Among those problems:
They did not test the carbon black mix against AS3. Rather they tested against something they identified as “Arctic Silver (original)”. I would have to guess here but might they be using a years old tube of AS1?
Nothing in the article explicitly tells us how long they tested any product for. They do admit that they used MBM5 for their tests and that they used a five second poling interval. However, looking at the graph, it would appear that each test ran for, at most, a few minutes. We all know that any version of arctic silver does not reach peak performance for the first few days.
The last point notwithstanding, the reason that AS3 does not develop peak performance for a few days is because the liquid that makes it flow so well needs to evaporate. The interesting fact here is that the liquid part of AS3 is the same as the liquid part of the carbon black mixture ( I spoke to Nevin on the phone to confirm this). If carbon black was used the way that we use AS3, after three days, all that would reside between a core and an HSF would be the carbon black itself.
It is also worth noting that the carbon black compound as received was the worst material in the test as far as processor temps go. The best material they tested was AS Ceramique, which we already know to be as good as it gets in today’s market.
I could go on but to no real point. The bottom line is that thermal insulators should not be touted as thermal conductors. This material is not something that would be good to use the way we use thermal interface materials. If you think that this is something you want experiment with, please use a processor that you can afford to sacrifice and post back to us with your long-term results.
We had three threads on this the last time Ed posted about this supposedly “revolutionary” heat sink compound. The one that I started here and two in the Cooling forum. The fact is that carbon black is simply not the best thing since sliced bread. It is, in fact, sold to industry as thermal insulation for blast furnaces.
A Dr. Chung published a paper claiming that she had developed a remarkable new heat transfer material. However, there were several problems with the paper in relation to over-clocking. The most obvious one being that she did not compare her new wonder substance to any commercially available heat sink compound. What she did compare her stuff to was only such materials that are already known to be worse thermal conductors than carbon. Basically, it was a runoff between thermal insulators and the one that was the worst thermal insulator was declared to be the king of heat transfer materials.
Another problem was that her tests were performed on highly specialized laboratory equipment that is not the same thing as a processor and it did not adequately model how her product would be used as we use products in our computers. Each test only lasted a fraction of a second. Even the most extreme of the cooling crowd leaves a heat sink installed for quite a bit longer than that.
Both of these problems (and a few others) were identified in the threads that we had on the matter the last time Ed had something up. Now he is linking to an article that tells us nothing new about the matter.
In fact, if you read the article, you will find several problems here as well. Among those problems:
They did not test the carbon black mix against AS3. Rather they tested against something they identified as “Arctic Silver (original)”. I would have to guess here but might they be using a years old tube of AS1?
Nothing in the article explicitly tells us how long they tested any product for. They do admit that they used MBM5 for their tests and that they used a five second poling interval. However, looking at the graph, it would appear that each test ran for, at most, a few minutes. We all know that any version of arctic silver does not reach peak performance for the first few days.
The last point notwithstanding, the reason that AS3 does not develop peak performance for a few days is because the liquid that makes it flow so well needs to evaporate. The interesting fact here is that the liquid part of AS3 is the same as the liquid part of the carbon black mixture ( I spoke to Nevin on the phone to confirm this). If carbon black was used the way that we use AS3, after three days, all that would reside between a core and an HSF would be the carbon black itself.
It is also worth noting that the carbon black compound as received was the worst material in the test as far as processor temps go. The best material they tested was AS Ceramique, which we already know to be as good as it gets in today’s market.
I could go on but to no real point. The bottom line is that thermal insulators should not be touted as thermal conductors. This material is not something that would be good to use the way we use thermal interface materials. If you think that this is something you want experiment with, please use a processor that you can afford to sacrifice and post back to us with your long-term results.