NVIDIA GeForce4 Ti 4400/4600 Roundup - April 2002

Cooling Performance

In order to test cooling performance and efficiency we used a Radio Shack IR Thermometer and took the temperatures of the cards while running a 10 minute loop of the Unreal Performance Test 2002. The GPU temperature was recorded from directly behind the GPU on the PCB and the memory temperature was recorded from behind the same memory chip on each board, once again on the back of the PCB. The highest sustainable temperature was recorded during the 10 minute period.

GPU Temperatures - Stock Clockspeed Temperature in Degrees Fahrenheit (Lower is Better) ASUS V8460 Ultra Ti 4600 MSI GeForce4 Ti 4600 Gainward GeForce4 Ultra 750XP Visiontek Xtasy GeForce4 Ti 4600 PNY GeForce4 Ti 4600 NVIDIA Reference Ti 4600 eVGA e-GeForce4 Ti 4600 Chaintech GeForce4 Ti 4600 MSI GeForce4 Ti 4400 ABIT Siluro GF4 Ti 4400 Visiontek Xtasy GeForce4 Ti 4400 ASUS V8440 Ti 4400 146.5 147.5 152.5 155.0 156.5 159.5 161.0 163.5 143.5 151.5 153.5 155.0 | 0 | 33 | 65 | 98 | 131 | 164 | 196

It's always good to see that it's not always the most elaborate cooling solutions that end up doing the best. If you recall, the most bells and whistles belonged to the eVGA Ti 4600 yet it was one of the worst in terms of cooling performance.

Memory Temperatures - Stock Clockspeed Temperature in Degrees Fahrenheit (Lower is Better) MSI GeForce4 Ti 4600 ASUS V8460 Ultra Ti 4600 Gainward GeForce4 Ultra 750XP eVGA e-GeForce4 Ti 4600 Visiontek Xtasy GeForce4 Ti 4600 PNY GeForce4 Ti 4600 NVIDIA Reference Ti 4600 Chaintech GeForce4 Ti 4600 MSI GeForce4 Ti 4400 Visiontek Xtasy GeForce4 Ti 4400 ASUS V8440 Ti 4400 ABIT Siluro GF4 Ti 4400 117.5 119.0 120.5 123.0 123.0 126.5 128.0 130.0 115.0 119.5 120.0 120.5 | 0 | 26 | 52 | 78 | 104 | 130 | 156

Once again we see that the best cooled memory is that which doesn't have a poorly designed heatsink on it. The two coolest Ti 4600 and Ti 4400 cards both fail to use any heatsinks on their memory yet they perform at the very top of the pack.