There's really no other way to say this: ASRock's P55M-Pro, Gigabyte's GA-PA55M-UD2 and Foxconn's P55MX all perform almost exactly the same in every single benchmark we ran.

Application tests, game tests, LAN tests, even USB, Firewire, and SATA controller performance was virtually identical across the board. We've known for a while now that motherboards of a given generation all perform the same, but there's honestly no more to it than that.

We've included a wealth of application, gaming and peripheral performance data for you to see for yourself, but honestly, it's not what's important when it comes to picking a micro-ATX P55 motherboard. Features, overclocking potential, support, and price point are what matter here.

The biggest difference between the boards, from a performance standpoint, actually surfaces in power consumption:

Idle Power Consumption

At idle the Gigabyte GA-P55M-UD2 draws the least amount of power while under load, ASRock's P55M-Pro is absolutely stingy. It's not a huge margin, but big enough to crown the Gigabyte board the least power hungry micro-ATX P55 board we've tested from an overall vantage point.  We had to make sure that Enhanced Halt State (C1E) was enabled on the ASRock board along with ACPI Suspend (S3) was set to auto for all power management features to work correctly. If you happen to be using an SSD drive, then the Check Ready Bit option must be disabled on the ASRock BIOS (AMI in general) or the system might not resume correctly from S3. Otherwise, S3 resume worked properly on both boards at stock and overclocked settings.

Application Power Consumption - Cinema 4D R11

 

Application/Gaming Performance

We're presenting all of the application/gaming performance data without commentary because, as we mentioned before - there's no real appreciable performance difference between these three boards. All of the boards were run with the Core i5 750 and we've included the Phenom II X4 965 BE as well as the i7 860, 870 and 920 purely for reference. Please check our Lynnfield launch article to see how well this processor performs against a variety of CPUs.

In many of our tests, the Core i5 750 is the same speed or faster than the Phenom II X4 965 BE. The lack of Hyper Threading prevents it from being a runaway success. In other cases, the Phenom II X4 965 is faster - and by a large degree.

Intel was very careful to disable HT on the 750, without it, there would be no reason to spend the extra money on the Core i7 860. Just as it was with Bloomfield, $284 is the sweet spot for Lynnfield if absolute performance is a requirement. Now for the benchmarks.

Test Setup Multitasking
Comments Locked

55 Comments

View All Comments

  • Sunburn74 - Monday, October 5, 2009 - link

    Thanks.
    Which board? I searched the thoughts section and don't see any mention of sleep :(

    I just know its a huge problem with gigabyte boards, pretty much every p45- and a good number of the x58 boards mysteriosly can't s3 sleep with significant overclocks in place and its something I'm seriously going to explore before my next mobo purchase.
  • Ryun - Monday, October 5, 2009 - link

    I purchased and Asrock 760g, which is a great little motherboard, yet it does not support S3 state (standby mode). I have an email from Asrock's (surprisingly quick-response) tech support saying that none of their boards officially support S3 state and to use their Instant Boot technology instead.

    Did the Asrock motherboard you tested allow you to go into S3 state/standby mode? I really like Asrock's boards but the lack of standby is a deal breaker for me.
  • Gary Key - Monday, October 5, 2009 - link

    S3 is fully supported on the ASRock board. In the power consumption section I did note what needed to be enabled for it work. Also, this was probably in the wrong spot, but in the OC section I briefly mentioned that the board had no problems resuming from S3 with the Bclk set to 215. I can understand why ASRock wants you to use Instant Boot, but S3 operation is just fine, even when overclocked.
  • Ryun - Monday, October 5, 2009 - link

    Ah, looks like I missed that part. Many thanks for pointing that out. Sadly though, I don't have those BIOS options on the 760g board I have but perhaps I can tinker a bit more.
  • n7 - Monday, October 5, 2009 - link

    Just wanted to say a massively huge thank you for testing with 8 GB!

    It's extremely encouraging to see, as the large majority of reviewers do not bother testing with all slots populated.

    Thanx again.
  • vlado08 - Monday, October 5, 2009 - link

    Which board has the quickest Power On Self Test?

    With fast CPU's and SSD I expect fast booting!

    Now I have a Gigabyte board (P965 DS4) and when the Sata is in AHCI mode POST is quite long.

    I hope that in future articles you will include this information.
  • MadMan007 - Monday, October 5, 2009 - link

    S3 is your friend. Really, who boots their computer every time any more?
  • strikeback03 - Monday, October 5, 2009 - link

    I usually do, since I dual boot and won't necessarily know which OS I need the day before.
  • Gary Key - Monday, October 5, 2009 - link

    Cold Boot - Quick Boot turned off in BIOS - AHCI enabled, External Hard Drive attached via IEEE 1394a, LAN attached to our Promise NAS via a Gigabit Switch.

    Time reported is from the time we turn on the board until Win7 has correctly installed the network stack. So this is the full POST and OS is usable process that is being timed.

    ASRock - 44.7 seconds
    Gigabyte - 53.2 seconds

    I have the information since we run this for every board, just did not know if anyone would care to see it. ;)
  • vlado08 - Tuesday, October 6, 2009 - link

    Thanks Gary

    But I was interested in time from pushing the power on switch until the begining of the OS loading. I think that there might be difference between boards depending on their BIOS.
    Time from begining of the OS loading until fully functional OS depends on the computing power ot the CPU and the speed of the HDD (SSD) and not on the design of the board.
    And because you (we) want to distinguish between the boards I thought that this might be one of the criterion.

Log in

Don't have an account? Sign up now