Directory  Reviews Cooling Reviews |
| Coolermaster HyperTX S775 Heatsink Review |
|
|
|
Page 4 of 5 Installation Removing the easily accessible clip on the top of the cooler allows the shroud to lift from the fin stack. It slides off with ease. All that is left is the naked tower, almost like a wet cat. When using the stock thermal pad, installation is a cinch. The tower simply rests on the CPU in any of four orientations. Locking of each plastic retention mount follows, and they provide a satisfying clicking sound when they are fully seated. All four of the retention pieces depress with little effort. Application of pulling and horizontal forces resulted in one of the plastic mounts popping out of the motherboard on the first attempt. The stand did not slide far enough into the motherboard before the pole spread the plastic clips. Reseating the HSF again, while making sure each mount was flush with the board, fixed this issue. The average user might overlook the stress test. Failure to do so may result in HSF popping loose, unnoticed, while transporting to a LAN party.    Performance: All that is left is to power up the system. To make sure our seating job went well, the first stop was the bios temperature monitoring section. Upon boot, the first temperatures were 46.6 C, or 116 F, and rose to 47.7 C, or 118 F. This seemed a bit excessive, but stable for idle temps. We proceeded to allow the computer to boot into windows regardless of the unusually high thermal readings. The resultant temperatures directly into windows were 52.7 C, or 127 F. These temperatures leveled off at 48.8 C, or 120 F. To confirm how high the temperature would go, the software Hot CPU Tester Pro 4.3 was used. Running the torture test for about thirty minutes resulted in temperatures as high as 60 C, or 140 F. We definitely assumed that something must have gone wrong. After a shutdown, some arctic silver 3 replaced the regular thermal pad. Repetitive reseats of HSF revealed more problems with the simple mounts. With the mounting poles in a "used" state, they did not fit into the mounting holes without running into too much resistance. This resulted in the mounts prematurely setting themselves. Having to reset one of the retention mounts also ruined the thermal paste spread. The CPU had to be cleaned and the thermal paste reapplied. The use of two sets of hands allowed us to maintain the seal with the CPU while seating the retention mounts. Failure to maintain the seal properly will result in the thermal paste on one side. Proper seating of the HSF with the arctic silver resulted in 45 C, or 113 F, while booting into the bios. After an hour, the bios temperatures leveled off at 46.6 C, or 116 F. This was still significantly higher than normal temperatures most editors experience. However, it was an improvement often associated with applying arctic silver 3, and it was stable. The computer spilled onto the desktop at 47.8 C, or 118 F. Because the HSF was correctly installed, this was very puzzling. It was still returning abnormally high readings. On the brighter side, the temperatures dropped from load very fast. Often the temps would drop about 5.5 C, or 10 F, per second and would reach +2 F of idle within 5 seconds. It became obvious that we had a hotter than normal CPU, or Abit's temperature monitoring was reporting faulty, but consistent, data. We decided to install the stock HSF to allow for context with this hot CPU. The CPU was a toasty 55 C, or 131 F, and this reading was straight from the bios. That means that the HSF from Cooler Master provided a 10 C, or 18 F, improvement over stock cooling efficiency at standard clocks. We reseated the stock heat sink another three times to ensure we could reproduce the results consistently. When we moved back to the Hyper TX, the earlier results were replicated. It was time to stress test this "optimal" setup. We ran the Hot CPU Tester Pro diagnostic for a little over an hour. The load temps were 56 C, or 133 F. This was far enough for us at the stock rate. For over-clocking potential, we started with the highest post possible, followed by the highest boot possible, and finally, the highest stable long term possible. We wanted to turn this 160-dollar processor into a 250-dollar rendition for the price of the Hyper TX. We stepped the processor from a 272 MHz FSB to 426 MHz before the computer would no longer post. In the bios, the 2.982-gigahertz dual core idled at 51C, or 123.8 F. After an hour of stress testing, the load temperatures settled at 60 C, or 140 F. The on-board audio stopped functioning at these speeds. With the voltages at their stock ratings and the temperatures so high, we decided against raising it. It would not be productive for a long term over-clock. Unfortunately, the most stable windows bootable speed we obtained was 400 Mhz, or 2.7979-gigahertz. Effectively, we went from a 3.72-gigahertz processor to a 5.6-gigahertz processor. That puts this e6300 right up there with the e6700 in performance, but a slower bus speed of only 800 MHz (4-4-4-12-21). When we used 1066 ram from super talent, we were able to achieve a 1000 MHz FSB with a 4:5 clock ratio and timings of (4-4-4-11-16). With the corsair ram, we were able to come within 1000 points of the e6700 on all of the CPU benchmarks within Everest Ultimate 2006. With the Super Talent ram, we were able to come very close to every mark the e6700 made. The AW9D was probably the limiting factor at this moment, with a "limitation" of DDR2 800 speeds quoted for the motherboard. It was time to stress test the system and see what temperature we reach. From idle of 49 C, or 120.2 F, we ran Hot CPU Tester Pro for a little over 4 hours. The final load temp for the system was 59 C, or 138.2 F. The same speeds with the stock cooler idled at 58.8 C, or 138 F and reached over 69 C, or 156 F. This particular core is one of the hotter renditions that Intel has available. Because this is a new test system, we were only able to compare this review with respect to stock cooling abilities.   
|
|||||||||||||||||||||||||||||||||||||||||||||||||