Scythe Infinity: 5 Heatpipes and Silent 120mm Fan
by Wesley Fink on February 26, 2007 12:05 AM EST- Posted in
- Cases/Cooling/PSUs
Overclocking
As cooling solutions do a better job of keeping the CPU at a lower temperature, it is reasonable to expect the overclocking capabilities of the CPU will increase. In each test of a cooler we measure the highest stable overclock of a standard X6800 processor under the following conditions:
CPU Multiplier: 14x (Stock 11x)
CPU voltage: 1.5875V
FSB Voltage: 1.30V
Memory Voltage: 2.20V
nForce SPP Voltage: 1.5V
nForce MCP Voltage: 1.7V
HT nForce SPP <-> MCP: Auto
Memory is set to Auto timings on the 680i and memory speed is linked to the FSB for the overclocking tests. This removes memory as any kind of impediment to the maximum stable overclock. Linked settings on the 680i are a 1066FSB to a DDR2 memory speed of DDR2-800. As FSB is raised the linked memory speed increases in proportion. The same processor is used in all cooling tests to ensure comparable results.
The Scythe Infinity overclocking was just average in the stock single fan configuration - reaching 3.83GHz with complete stability. However the second fan in a push-pull configuration allowed the Infinity to match the highest overclock on air we have seen with this CPU. At 3.9GHz the Tuniq Tower 120 was still cooler, but the Infinity still managed to hold 3.90GHz with stability during our load testing.
These overclocking results were particularly interesting since we have seen some forum comments that the second cooling fan made little difference on the Infinity. Our results indicate the second cooling fan can make a very significant difference in both cooling and overclocking the Infinity.
The smaller differences others have reported may have been the result of a less demanding test configuration or perhaps it was a result of a better single cooling fan on the tested Infinity. To test this second possibility, a higher RPM and CFM Silverstone fan was mounted on the Infinity as a single fan and the highest overclock was tested. The Infinity with the beefed up fan reached 3.85Ghz - a little higher than the 3.83GHz with the stock fan - with temps about 1C lower than the "silent" fan shipped with the Infinity. Clearly a fan pushing more air improves performance by a small amount, but it still does not match the performance of two "silent" fans in a push-pull arrangement.
As cooling solutions do a better job of keeping the CPU at a lower temperature, it is reasonable to expect the overclocking capabilities of the CPU will increase. In each test of a cooler we measure the highest stable overclock of a standard X6800 processor under the following conditions:
CPU Multiplier: 14x (Stock 11x)
CPU voltage: 1.5875V
FSB Voltage: 1.30V
Memory Voltage: 2.20V
nForce SPP Voltage: 1.5V
nForce MCP Voltage: 1.7V
HT nForce SPP <-> MCP: Auto
Memory is set to Auto timings on the 680i and memory speed is linked to the FSB for the overclocking tests. This removes memory as any kind of impediment to the maximum stable overclock. Linked settings on the 680i are a 1066FSB to a DDR2 memory speed of DDR2-800. As FSB is raised the linked memory speed increases in proportion. The same processor is used in all cooling tests to ensure comparable results.
The Scythe Infinity overclocking was just average in the stock single fan configuration - reaching 3.83GHz with complete stability. However the second fan in a push-pull configuration allowed the Infinity to match the highest overclock on air we have seen with this CPU. At 3.9GHz the Tuniq Tower 120 was still cooler, but the Infinity still managed to hold 3.90GHz with stability during our load testing.
These overclocking results were particularly interesting since we have seen some forum comments that the second cooling fan made little difference on the Infinity. Our results indicate the second cooling fan can make a very significant difference in both cooling and overclocking the Infinity.
The smaller differences others have reported may have been the result of a less demanding test configuration or perhaps it was a result of a better single cooling fan on the tested Infinity. To test this second possibility, a higher RPM and CFM Silverstone fan was mounted on the Infinity as a single fan and the highest overclock was tested. The Infinity with the beefed up fan reached 3.85Ghz - a little higher than the 3.83GHz with the stock fan - with temps about 1C lower than the "silent" fan shipped with the Infinity. Clearly a fan pushing more air improves performance by a small amount, but it still does not match the performance of two "silent" fans in a push-pull arrangement.
Facebook
Twitter
RSS
39 Comments
View All Comments
Zoomer - Thursday, March 1, 2007 - link
Could you try adding more than two fans, or perhaps blocking the sides of the heatsink so that air can't escape and see if it helps?I'm very interested. It appears that this unit requires a high(er) static pressure to work well.
yyrkoon - Monday, February 26, 2007 - link
These coolers, are all fine and dandy, but what about low profile coolers, that dont weight as much as the Golden Gate bridge ?Me, my personal case, is an Lian Li PC-G50 (silver, if you must know . . .), and the PSU sits right_above about half of the CPU. Silent is great, higher overclocking potential, is better still, but the bigger question is, why arent these companies working on anything that doesnt require special cases, or at the very least, cases that are so huge, you can not put them any place without having large amounts of room.
My case: 15" tall, looks great, is very functional, its just a terrible shame, that no one seems to be making low profile coolers, that would help me eek out a 310mhz + CPU on my ABIT NF-M2 nView + AM2 Opteron 1210 (which I have actually achieved, just wouldnt do much other than BSoD within windows).
Lem - Wednesday, February 28, 2007 - link
Have you checked the Thermalright "SI-type" coolers? I have been happy with a Thermalright SI-120 + 120mm 1200rpm fan (PAPST 4412F/2GLL). I am not much into overclocking though. I only raised the frequency of my X2 3800+ to 2.4GHz because it was a nice round number that my RAM could handle.The newer SI-128 seems to support AM2-socket as well. I prefer that design over these for obvious reasons.
My Lian Li PC-61 is somewhat bigger than your case but I do not think that there would be any problems with PC-G50 either.
BigLan - Tuesday, February 27, 2007 - link
"why arent these companies working on anything that doesnt require special cases"Erm, your case is the special one. These coolers generally fit in standard atx cases, with the regular atx-approved layout (height might be an issue in mid-atx cases.) I think that having the psu above the cpu is ok for micro-atx, but that is still pretty specialized - you might want to look into server coolers (specifically 1-u coolers.) m-atx-style normally requires smaller fans, which have to spin much faster to move the same amount of air, and are very, very noisy (does anyone remember the old 60mm delta screamers?) You're also limited to fin area, which also hurts performance. There are some reasonably good btx coolers which overcome these problems though, but they're often proprietary (built into dell cases etc.)
cujo - Tuesday, February 27, 2007 - link
i love big cases. lots more room to work inside and lots more room for quiet 120mm fans.yyrkoon - Tuesday, February 27, 2007 - link
Good for you. My case also has silent 120mm fans, the system is so quiet in fact, I have to turn something else on, so I can sleep ( I require white noise to sleep ).arswihart - Monday, February 26, 2007 - link
I disagree with using overclocking as a measure of a heatsink's performance. You should focus on noise and temps, it's that simple. I don't know why anyone would even pay attention to that data you are presenting.Wesley Fink - Monday, February 26, 2007 - link
Temperatures at stock speeds are reported on p. 4. However, contrary to your assumption, coolers that perform best at cooling at a stock 2.93GHz are not always the same coolers that enable the highest overclock. Temperature rises as a CPU is overclocked, and coolers that perform well at stock speeds do not always remove heat as well at higher temperatures.Since the OC that can be achieved does not matter to you, the data on p.4 - temperature at stock speed - should give you what you are looking for.
arswihart - Monday, February 26, 2007 - link
Idle and Load temps for stock / OC. No need to look at 50Mhz differences to rank the heatsinks is my point.arswihart - Tuesday, February 27, 2007 - link
To further make my point, you see a 7 C difference between the single and dual fan configs on the Infinity, and only a 70Mhz difference in OC. To me, the 70Mhz difference is negligible, while the 7 C difference is quite striking.Relatively speaking, the difference in temps is much more drastic (12%), and much mnore relevant, than the difference you are seeing in OC (2%).
Do you still want to argue about it?
Sure it makes for interesting reading, so I guess I can't blame you too much, and the audience at Anandtech surely eats it up. I would just tell you I'd rather not have reviews cluttered up with this rather petty data.