Over the weekend, I tested out some of the best heatsinks available now. This may not include all of the best but I picked out what I considered were some of the best contenders at reasonalbe prices(except for "The Core"...too expensive). This comparison is far from ideal but I just wanted to do some quick testing to find out which one cools the best under full load. I probably should have recorded idle temps as well but I was too far into the testing when I realized this. I realize there are factors such as ambient temperature, what is best way to measure temp, how accurate are motherboard readings, etc that I may not have taken into account but the results here were obtained under pretty consistent conditions.
The CPU is a is PIII 1GHZ @ 1.2GHZ with 2V. I realize the voltage is on the high side but I'll take my chances. Motherboard is an ASUS CUSL2 in an Inwin Full Tower Case with 2 Delta 92mm fans blowing in, one 2CoolPC blowing air up into the case, and a Delta 92mm and Top Motor 80mm fan on the back for exhaust. Video card is CL Geforce 2 GTS and memory is Mushkin Rev3 running at 2-2-2-5/7. Temperatures were taken using the Asus Probe v2.12.07. As I said earlier, this may not be the ideal way to measure temperature but it is the simplest way.
I realize a lot of people have switched over to Athlons so I can only guess that you would get similar results using overclocked Athlons. The one advantage that the Athlon has over the PIII is the larger core. Although PIII's give off less heat at a given speed vs an Athlon, the smaller core of the PIII can make it a little bit hard to keep cool at higher speeds. This probably explains why Intel is switching to an integrated heat spreader (IHS) on some of the newer cD0 PIII 1GHZ flip chips. I have not yet seen any PIII's with the IHS but I would imagine they'll start showing up in the next month or so. Also, the IHS should also make it easier to mount heatsinks since the cooler has a larger area to rest on.
The motherboard temperature stayed at 71F throughout much of the testing but sometimes the Delta 38 seems to have heated the temperature up to 73F with all of the air it was blowing(blowing hot air off the heatsink that is). Prime95 was used to stress the CPU. It usually took anywhere from 10-15 minutes for temperature to stabilzie. I would wait for Asus Probe to have a flat line(no changes) all the way across the graph. A thin and fresh layer of Ultra Silver compound(from ColdCPU.com) was applied to the CPU core right before each heatsink was mounted. Also, the heatsink and CPU core were cleaned with rubbing alchol before each mounting.
Here are the results in no particular order:
Cooler----------------------------------------Max Temp(°F)---MB Temp(°F)
CAK 38 with Delta---------------------------------107-----------73
CAK 38 with YS-Tech-----------------------------111-----------73
ThermoEngine w/Delta----------------------------102-----------71
ThermoEngine w/YS Tech------------------------107-----------71
WBK38 w/Sanyo Denki(80mm/36cfm/29dB)-113-----------71
WBK38 with YS-Tech----------------------------109-----------71
WBK38 with Delta---------------------------------105-----------73
The Core with Delta-------------------------------109-----------71
The Core with YS-Tech--------------------------113-----------71
Alpha 6035 with Delta/sucking------------------100-----------71
Alpha 6035 with Delta/blowing------------------104-----------71
Alpha 6035 with YS-Tech/sucking--------------105-----------71
Alpha 6035 with YS-Tech/blowing--------------104-----------71
Alpha PEP66T with YS-Tech/sucking----------102-----------71
Alpha PEP66T with YS-Tech/blowin-----------102-----------71
Alpha PEP66T with Delta/sucking---------------96-----------71
Alpha PEP66T with Delta/blowing---------------98-----------71
Delta = 60x25mm, 38cfm/46dB
YS Tech = 60x25mm, 26cfm/34dB
PEP66 was modified to mount the fan on top. Also, a thin piece of cardboard was taped over the open side of the shroud to create a better seal.
The CPU is a is PIII 1GHZ @ 1.2GHZ with 2V. I realize the voltage is on the high side but I'll take my chances. Motherboard is an ASUS CUSL2 in an Inwin Full Tower Case with 2 Delta 92mm fans blowing in, one 2CoolPC blowing air up into the case, and a Delta 92mm and Top Motor 80mm fan on the back for exhaust. Video card is CL Geforce 2 GTS and memory is Mushkin Rev3 running at 2-2-2-5/7. Temperatures were taken using the Asus Probe v2.12.07. As I said earlier, this may not be the ideal way to measure temperature but it is the simplest way.
I realize a lot of people have switched over to Athlons so I can only guess that you would get similar results using overclocked Athlons. The one advantage that the Athlon has over the PIII is the larger core. Although PIII's give off less heat at a given speed vs an Athlon, the smaller core of the PIII can make it a little bit hard to keep cool at higher speeds. This probably explains why Intel is switching to an integrated heat spreader (IHS) on some of the newer cD0 PIII 1GHZ flip chips. I have not yet seen any PIII's with the IHS but I would imagine they'll start showing up in the next month or so. Also, the IHS should also make it easier to mount heatsinks since the cooler has a larger area to rest on.
The motherboard temperature stayed at 71F throughout much of the testing but sometimes the Delta 38 seems to have heated the temperature up to 73F with all of the air it was blowing(blowing hot air off the heatsink that is). Prime95 was used to stress the CPU. It usually took anywhere from 10-15 minutes for temperature to stabilzie. I would wait for Asus Probe to have a flat line(no changes) all the way across the graph. A thin and fresh layer of Ultra Silver compound(from ColdCPU.com) was applied to the CPU core right before each heatsink was mounted. Also, the heatsink and CPU core were cleaned with rubbing alchol before each mounting.
Here are the results in no particular order:
Cooler----------------------------------------Max Temp(°F)---MB Temp(°F)
CAK 38 with Delta---------------------------------107-----------73
CAK 38 with YS-Tech-----------------------------111-----------73
ThermoEngine w/Delta----------------------------102-----------71
ThermoEngine w/YS Tech------------------------107-----------71
WBK38 w/Sanyo Denki(80mm/36cfm/29dB)-113-----------71
WBK38 with YS-Tech----------------------------109-----------71
WBK38 with Delta---------------------------------105-----------73
The Core with Delta-------------------------------109-----------71
The Core with YS-Tech--------------------------113-----------71
Alpha 6035 with Delta/sucking------------------100-----------71
Alpha 6035 with Delta/blowing------------------104-----------71
Alpha 6035 with YS-Tech/sucking--------------105-----------71
Alpha 6035 with YS-Tech/blowing--------------104-----------71
Alpha PEP66T with YS-Tech/sucking----------102-----------71
Alpha PEP66T with YS-Tech/blowin-----------102-----------71
Alpha PEP66T with Delta/sucking---------------96-----------71
Alpha PEP66T with Delta/blowing---------------98-----------71
Delta = 60x25mm, 38cfm/46dB
YS Tech = 60x25mm, 26cfm/34dB
PEP66 was modified to mount the fan on top. Also, a thin piece of cardboard was taped over the open side of the shroud to create a better seal.
Facebook
Twitter