Archive 1997–2024

DFI UT P35-T2R: Tweakers Rejoice!

Test Setup

DFI UT P35-T2R
Standard Benchmark Testbed
Processor Intel Core 2 Quad Q6600
Quad Core, 2.4GHz, 8MB Unified Cache, 9x Multiplier, 1066FSB
CPU Voltage 1.2500V
Cooling Tuniq Tower 120
Power Supply OCZ ProXStream 1000W
Memory OCZ Flex PC2-9600 (4x1GB)
Memory Settings 4-4-4-12 (DDR2-1066)
Video Cards MSI 8800GTX
Video Drivers NVIDIA 163.71
Hard Drive Western Digital 7200RPM 750GB SATA 3/Gbps 16MB Buffer
Optical Drives Plextor PX-B900A, Toshiba SD-H802A
Case CoolerMaster Stacker 830 Evo, Lin Li PC75
BIOS DFI 9/13
Operating System Windows Vista Ultimate 64-bit
.

With such a vast array of overclocking BIOS functions at our disposal, a decision was made to stick with a single operating system environment. Microsoft XP SP2 was chosen due to its maturity, taking away some of the time burden looking for possible software level conflicts that may exist on a newer OS (though we ventured to our standard Vista test suite for the non-overclocking benchmarks). This does not mean to say that the board cannot work with Vista at present when overclocking. We are just trying to eliminate possible downtime on non motherboard/BIOS related issues and also considered the fact that a vast majority of overclocking enthusiasts have not switched to Vista (yet).

The test setup utilized the supplied Transpiper, set up to exit our Lian Li PC75 case via the space available by not inserting the supplied I/O backplate. The fin section of the Transpiper heatsink was aligned with the exhaust fan of the OCZ 1000w ProXStream PSU, though at times we felt this might not have been the best solution during extended overclocking sessions because of the already high temperature of the PSU exhaust. Another reason this setup may not be ideal when benchmarking is that the Transpiper solution is partially handicapped by an adjustable construction. We found most of the heat that was transferred into the single heatpipe never quite made it past the elbow joint connecting to the dual pipe heatsink. The biggest gain in temperature reduction was realized by simply pointing a fan at the PWM heatsink itself.

We could not find any gains in using the supplied CPU to PWM heat transfer plate; CPU temps with the plate installed were considerably higher at full load (~10C). This was partially due to the increased thermal resistance of an extra surface, which decreases thermal transfer potential from the most critical junction of a cooling solution - directly over the CPU die. We also need to consider the concave or bi-concave bases of some modern coolers and waterblocks and indeed the IHS itself (unlapped). The heat transfer plate would need to be quite flexible to ensure good contact when under pressure.

We used our QX6800 G0 ES CPU, concentrating on the 8x and 9x multipliers that both our air- and water-cooled setups can take advantage of currently. The CPU itself is a particularly high temperature sample of a G0 stepping; load temps were around 15C higher at equivalent voltages in comparison to some of the better Q6600 G0 examples we have in the labs. We later lapped the IHS flat (as flat as possible) which resulted in a seven degree drop in load temperatures, but that still ends up around 8C higher than our best Q6600 temperature ranges.

Both high resolution 1920x1200 and 1280x1024 resolution game benchmark tests were run to find performance benefits for gamers, should such advancements through BIOS tweaking exist. We utilize new drive images on each board in order to minimize any potential driver conflicts. Our 3DMark results are generated utilizing the standard benchmark resolution for each program. We run each benchmark five times, throw out the two low and high scores, and report the remaining score.

We limited our standard Vista test suite to a few benchmark results to indicate the general performance of this board. After all, a P35 is a P35 for the most part at stock settings, with most P35 boards scoring within 1%~2% of each other. It really does not matter if you utilize a $90 abit or Gigabyte P35 board or the $200+ versions from ASUS or DFI as stock performance will not vary much in most cases.

As such, we decided to take a different look at this board and focus on overclocking. This board was designed for the PC Autobahn and not your typical school zone road or office park avenue. Our overclocking tests will only show results on this board for now, but we will expand coverage to include direct comparisons as we review other boards that are designed for the overclocking enthusiast market sector.

Originally published at https://www.anandtech.com/show/2354 (Wayback Machine).