Gigabyte GA-X79-UD5

[cadaveca]

To read this review go to: http://www.techpowerup.com/reviews/Gigabyte/X79-UD5/

[Delta6326]

Wow, thats a great board! Would be great to get. 300w!

[cadaveca]

That 300 W...that's all that I am able to test under conditions that normal users might see(ie. watercooling). It might be capable of more...I'm not sure. If you run LinX or Intelburntest @ 5 Ghz, 300W is no problem.

Not that you should be running LinX or IBT 24/7......most usage, even overclocked, will be far below that.

[Bundy]

Great, thanks. I know you have put a lot into these reviews and the result is greatly appreciated.

[VulkanBros]

Why excactly is the USB 3.0 performance so underwhelming on X79 boards??

[fullinfusion]

As always great review Dave

[cadaveca]

Originally Posted by VulkanBros

Why excactly is the USB 3.0 performance so underwhelming on X79 boards??

The performance is based on the drive used for testing, too, so it could just be that. The lowest result and highest result from over the last year shows about 8 MB of difference using the same drive, with the same data on it.

That said, the ECS X79 board gave the best result I've seen so far. The ASUS was a bit underwhelming. The Gigabyte board is like right between the other two. That said, there's only about .75% difference between the three X79 Express products. Minor.


X79 Express introduces this USB SuperSpeed stuff...but they all seem to use far different controllers for what seem to be the same technology...I'm not sure who designed USB SuperSpeed directly, as it seems to be like JEDEC for memory, where several OEMs work together to develop the technology, and becuase of the many different controllers, obviously results will vary.

Of course, part of USB 3.0 interface is the driver. This is critical in USB 3.0 SuperSpeed...ultimately, I think it's the drivers that make or break a result. As far as I know, USB 3.0 doesn't even work in Windows XP properly.

The ECS board used a Texas Instruments controller, this Gigabyte board uses Fresco controllers, and the ASUS was a weird mix that I couldn't quite figure out, as there was 3 controllers, capable of 8 ports, but only six ports active.

Anyway, it's clear the results vary so much because of the different controllers. Whether it's the controller itself, the driver, or some weird hardware issue, I am not 100% sure, but I did expect the TI controller to be the fastest, and it was. I'm not sure what conclusions you want to draw from al lfo that, but USB 3.0 performance is not something I consider the fault of the board maker, as they are not responsible for driver development. They merely buy the parts.

[btarunr]

I wish they stuck to ALC889 CODEC. It's the closest onboard audio gets to a sound card. Those IR controllers provide rock-solid LLC, my Biostar TPower I55 had them. Nice review.

[cadaveca]

Originally Posted by btarunr

I wish they stuck to ALC889 CODEC. It's the closest onboard audio gets to a sound card. Those IR controllers provide rock-solid LLC, my Biostar TPower I55 had them. Nice review.

I think the audio resutls are affected by the Dolby software. I'm using this board now, and the audio is OK, but there is just something about how it sounds that kinda relates to the results that RMAA gave. Could be the faked 3D positioning. Either way, I don't really like it...but it's not too bad...I could use my Xonar D2X, but it's not worth it.

LLC is a bit funky on this board. There's some noticible droop, but it's easily compensated for by adjusting the voltage accordingly.

The new BIOS let it OV as far as it did..before F7, I could not get over 4.5 GHz to even boot. Now I can get into the OS @ over 5 GHz, but I'm left with just too much heat to be able to test with fully. These CPUs NEED real watercooling to really clock.

[Sihastru]

Nice review Dave. I've beem eyeing this board since the first day I saw it in the ES form. I was a little worried about the recent "motherboard up in flames" reports, nice to see they fixed it with a simple BIOS update.

There is one thing I would like to know though. Is there any "coil" noise coming from this board? My current X58A-UD7 does have some amount and sometimes I wish it didn't. It doesn't drive me nuts, but it doesn't make me feel good either.

[cadaveca]

I haven't noticed coil noise except from a single product, A Biostar FM1 board, over the past year. Most are using chokes and analogue/digital VRM designs that are not susceptible to such issues unless really pushing things.

I'll run LinX @ 450 W, if it's possible, to try to suss out issues like that. I tend to spend a fair bit of time checking VRM behavior already, just because it's one of the few real ways that boards can differ, and I'm the only reviewer regularily reporting power consumption via 8-pin EPS(others report full system consumption), so you can be sure that if I've not mentioned it, it should not be an issue. At the same time, I do expect some designs to make audible noises under extreme loads...just the nature of the beast.

The problem with the BIOS and VRM is pretty well documented at this point. I did not have anything remotely close to the issues that one user did, under the same conditions, although I do have a slightly different product. However, at the same time, the UD3 has a larger VRM cooler than the UD5, so I'm kinda ignoring that whole thing at this point. Seems like there was a real issue, based on Gigabyte's response to the problem, but I don't think it's really something most users would encounter...and that the steps Gigabyte took were merely a safety precaution. I'd love for my board to pop and for Gigabyte to replace my CPU, too. But no such luck.

[Milky]

In the pictures of the motherboard there is 'PCI EXPRESS 3.0' printed on the PCB but in the review it says 'three PCIe 2.0 x16 slots' ??? Am I missing something?

Realy nice review otherwise

[cadaveca]

Originally Posted by Milky

In the pictures of the motherboard there is 'PCI EXPRESS 3.0' printed on the PCB but in the review it says 'three PCIe 2.0 x16 slots' ??? Am I missing something?

Realy nice review otherwise

My bad, that's a typo. The board supports whatever the CPU supports, as the PCIe x16 slots are directly connected to the CPU without any bridges or switches between the two.

The PCIe x1 slots are 2.0 though, run off of the X79 Express PCH.

[freakshow]

i want this board so freakin bad!!!

[AthlonX2]

Finally got around to reading x79 reviews,figured id stop by and read my buddy daves review. just one question though why are you testing sata6 performance with a hard drive instead of an ssd? seems a bit odd to me but good review anyways .

[cadaveca]

I test with more than just a mechanical drive, but only mechanical numbers are reported as at the time testing was started, more people where using mechanical.

Now that time has progressed, and pricing is much more affordable, updated drive controller testing with SSD's both in SATA 3 Gb/s and SATA 6 Gb/s will be included in future reviews. I have a Sandforce-based SATA 3 Gb/s and Crucial M4 for SATA 6 Gb/s. At this point in time there is no nagging differences between using either format of drive, but due to requests the test rig has been updated.

At first I was loath to do so due to nagging firmware issues that were present in nearly every drive. Since most of those have been sorted, and SATA 6 Gb/s SSDs are more reliable, I have no problems using them. Breaking the $100 price barrier for 60 GB drives was part of that too.

If you have a request for testing with a specific product, donations of such hardware are accepted.

[Jez2k]

I was torn between getting this or the UD3, but seeing that the screws are nowhere near the MOSFETs on the UD5, as they are on the UD3, I feel more inclined to buy the UD5.

According to the manual, the board is XP compatible. However, I thought the UEFI BIOS uses GUID instead of MBR (Master Boot Record). Isn't MBR needed to install XP? Futhermore, I wish to install XP 32-bit (far too many of my games would be lost if I move to 7 or 8). Is this poissible?