Archive Home arrow Reviews: arrow Motherboards arrow X79 Express Motherboard Performance Comparison
X79 Express Motherboard Performance Comparison E-mail
Reviews - Featured Reviews: Motherboards
Written by David Ramsey   
Tuesday, 22 November 2011
Table of Contents: Page Index
X79 Express Motherboard Performance Comparison
X79 Express Chipset
Motherboard Feature Comparison
Motherboard Testing Methodology
Overclocking Results
SiSoft Sandra Memory Bandwidth
AIDA64 Extreme Edition Tests
PCMark Vantage Tests
CINEBENCH R11.5 Benchmarks
CPU-Dependent 3D Gaming
PassMark PerformanceTest
Media Encoding Benchmarks
SPECviewperf 11 Tests
SPECapc Lightwave
Blender and POV-Ray
Final Thoughts and Conclusion

Overclocking Results

Overclocking LGA 1155 Sandy Bridge CPUs can be frustrating: since most system clocks are derived from the base clock, overclocking by increasing BLCK is all but impossible, as more than a few MHz increase will render your system too unstable. Multiplier overclocking-- that is, specifying how high the CPU's "turbo" multiplier should go under load-- works well, but you must have a "K" series CPU and the correct motherboard to fully utilize this mechanism.

The Sandy Bridge Extreme CPUs and the X79 chipset remove these restrictions. Still, turbo multiplier overclocking has the advantage of being fast and easy to do, without having to worry about the rest of your system. Also, this allows the processor to clock down to stock speeds at idle, reducing power usage and heat output; raising the BCLK keeps things overclocked all the time.

When you overclock a Sandy Bridge (regular or Extreme) CPU by increasing the turbo multiplier, you can do it on a per-core or all-cores basis. The former allows higher overclocks when fewer cores are used; the latter runs all cores at the same speed. I prefer to see the highest speed at which I can run all cores under load, so I chose the all-cores method.

The overclocking results varied on each motherboard, and probably don't represent the best possible results, as we were limited by our schedule as well as the fact that the only available cooler was Intel's High Performance Liquid Cooling System RTS2011LC. The Asetek-sourced cooler uses a standard thickness 120mm radiator with a single LED-lit fan, which Intel rates at 74CFM. The glowing blue-lit fan is pretty and very quiet (frankly I think the CFM rating is optimistic), but I needed more for the overclocking runs, so I replaced the stock fan with two high-speed Delta AFC1212D fans, rated at 113CFM airflow each, in a push-pull arrangement. While this setup is very loud it does move quite a lot of air through the radiator!

The Intel DX79SI board was saddled with a beta BIOS (of which you're reminded with a large BETA on the display every time you boot the board), and unfortunately I wasn't able to achieve any reasonable overclock with the board (I don't think 200MHz counts as "reasonable"). This was disappointing since the DX79SI has an auto-overclocking feature Intel calls the "DX79SI Overclocking Assistant", which seems much more elaborate than the various auto-overclocking features built into other boards. I'll be interested to see how well this feature works when a stable BIOS is released.

I was able to achieve significant overclocks on both ASUS motherboards. On the ASUS P9X79 Deluxe, I got to 4.6GHz under load:

p9x79_overclock.jpg

On the ASUS Sabertooth X79 TUF, I made it to 4.8GHz:

sabertooth_overclock.jpg

The ASUS Motherboards have detailed power phase controls that control VDROOP and allow you to specify current usage as well as voltage. All I needed to do to reach these overclock was set most of the power phase controls to "high" or "extreme", and then set the multiplier. While these overclocks were stable throughout my benchmark runs, the very high voltage (1.52V) resulted in CPU temperatures that exceeded 80 degrees Centigrade in a few runs, which is higher than you'd want to maintain for any length of time. I think with a better cooler and more time I could possibly break 5GHz and get the voltage down a little, too. Surprisingly, CPU temperatures were significantly higher on the P9X79 Deluxe, which is why the overclock is a little lower on that board. I think with more time to experiment with the power controls I could have lowered the temperature and brought the overclocking up to the level of the Sabertooth.



 

Comments 

 
# light and confortableCHARLEYAM 2011-11-23 08:58
SWEET WHAT YOU GOING TO DO WITH THE ONES YOUR BENCH TESTING WHEN YOUR DONE??? I WOULD LIKE TO RIDE THIS MACHINE TO ITS LIMITS AND BACK. REALLY GREAT CHIPSET AND SECOND GENERATION PROCESSORS ARE OUT OF THIS WORLD INTEL GREAT JOB GUYS AND MOTHERBOARD COMPANIES KEEP THE X79 CHIPSET ON THE MBS SO WE CAN TAKE ADVANTAGE OF THESE NEW CHIPS
HAPPY HOLIDAYS DUDES DUDETTES AND ALL OTHER LIFE FORMS ON THIS PLANET...PEACE OUT BYE BYE
Report Comment
 
 
# RE: light and confortableDavid Ramsey 2011-11-23 09:01
I'll probably use the X79 Sabertooth for upgrading my personal Intel machine, and keep the Intel board for future testing (if they ever release a new BIOS, anyway!)
Report Comment
 
 
# RE: X79 Express Motherboard Performance Comparisonodiebugs 2011-11-30 23:10
I see Asus still using the FAILED garbage Marvell 9128. Can't wait to see if they lie about the pci-e 3.0 like they did with the TRUE Sata 3, which is Sata 2. Good idea blowing heat onto the usb and sata port.
Report Comment
 
 
# RE: RE: X79 Express Motherboard Performance ComparisonDavid Ramsey 2011-12-01 09:03
Well, bear in mind that the PCI-E 3.0 is a function of the Intel X79 Express chipset. The actual status of this feature seems a little fuzzy right now: it hasn't been "officially" certified and nobody can test it yet since there are no PCI-E 3.0 cards to test it against.
Report Comment
 

Comments have been disabled by the administrator.

Search Benchmark Reviews
QNAP Network Storage Servers

Follow Benchmark Reviews on FacebookReceive Tweets from Benchmark Reviews on Twitter