Gigabyte 990FXA UD7 Motherboard ReviewEvery once in awhile you come across a motherboard that raises the bar on what an enthusiast board should be. It sets the standard for features and performance exceeding all your expectations. This type of board forces every board you see from that point to meet a higher standard in order to impress you. nike air force ireland The Gigabyte 990FXA UD7 is not one of those boards. The CPU was cooled using a home brew water cooling setup consisting of a Swiftech MCR320 triple radiator, 3x120mm fans, Swiftech MCP655 B pump, and a Koolance CPU 350 AT water block with socket AM3 / AM3+ mounting hardware.
We normally don really conduct SLI or Crossfire functionality testing on most motherboards anymore due to the fact that these technologies are very mature at this point and we not seen problems with either of them which could be classified as relevant to a motherboard review. However given the fact that SLI hasn been allowed or licensed to AMD branded chipsets before, we felt that we should at least verify that SLI was indeed working. Going forward this shouldn be necessary, but again we felt the need to confirm SLI functionality on the 990FXA UD7.
I like to say that I was able to get SLI working on the 990FXA UD7 without issues but that wouldn be the truth. It wasn too difficult but after installing the second graphics card, I booted the system and once inside Windows I was prompted to restart after the driver for the card was loaded. http://www.nikeairforceireland.nu After rebooting SLI wasn an available option. I went into the device manager and it showed the second graphics card to be a standard VGA adapter which wasn working. I manually installed the GTX 480 driver which worked. SLI still wasn an option until after I rebooted. The system hung on the reboot but after power cycling it and getting into Windows, I had no more trouble with SLI.
The first image shows the Street Fighter Benchmark run without SLI enabled with variable frame rates and V sync turned off. The middle screenshot shows the benchmark running with the SLI indicators turned on indicating graphics card load balancing. nike air force ireland The final image is the difference in score. Keep in mind that at 1680×1050, the max resolution of my bench monitor, that a pair of GTX 480 in SLI are total overkill for this game. Again the point was to test and show that SLI is indeed working on the 990FX chipset.
The onboard audio MIC IN port was tested using a Logitech Internet Chat Headset. Spoken words were recorded from the Windows Sound Recorder found under the Accessories\Entertainment folder in the start menu within Windows XP. The recording was made with the Microphone Boost option disabled, nike air huarache ireland then enabled. The Microphone Boost option is found within the advanced menu under the microphone section with in the Volume Control Menu.
The recording sample was inaudible with the microphone boost option disabled. It was audible when enabled, but the distortion was extremely pronounced. In fact unusually so even for this type of sound solution. And actually that not too surprising. The Realtek CODECs are pretty lousy when it comes to recording.
To test the capabilities of the on board USB 2.0 connections, we used an ACOMDATA HD060U2FE 72 USB 2.0/FireWire HDD connected first to the USB port. http://www.nikeairhuaracheireland.nu USB 3.0 functionality was tested using a Super Talent USB 3.0 SuperCrypt 32GB Flash Drive. SATA drive tests were performed using Western Digital Caviar Black WD1002FAEX hard drives on all SATA headers. The SATA drives were used for testing in RAID 0 16k block size configurations on all applicable controllers. Testing was also conducted using the same model SATA drives in a stand alone SATA configuration on all applicable controllers. All drive benchmarks were done using the freely available CrystalDiskMark 3.0 program, run with both 50MB and 100MB sized test sets.
The SB950 south bridge is a nice solution on paper. It boasts 6 full SATA 6G ports as opposed to Intel 2 SATA 6G ports and 4 SATA 3G ports. Additionally it offers the ability to separately configure the mode of ports 4 and 5, where as the Intel controllers do not. The mode is set for the whole controller. Unfortunately the AMD south bridge lacks the ability to handle stripe sizes below 64k in most instances. This particular instance allowed for a 256K stripe size which is unusual. The Intel controllers allow for far more versatility in stripe sizes. http://www.nikeairhuaracheireland.nu Unfortunately AMD has often struggled with the performance of their south bridges for some time. Especially when it comes to USB performance. However USB performance in the SB850 / SB950 is pretty comparable to other offerings at this point. Their SATA performance is still not quite where it needs to be though.
The sequential 50MB read tests show results below that of the Marvell 9172 controller. Normally you expect the Marvell controller to come in last, but again AMD has always struggled with south bridge performance. However USB and IEEE1394 performance is right where it should be. USB 3.0 performance is also solid.nike air huarache ireland I hadn ever seen the Etron EJ168 controller prior to the working with the 990FXA UD7, and luckily it didn give me any problems while performing well. The 50MB write tests showed nearly perfect symmetry between the SB950 and the Marvell 9172 controller. IEEE1394 performance fell a bit below expectations here, but remains consistent with other implementations of the VIA controller. USB 2.0 and 3.0 performance remained excellent.
In the 100MB tests, the sequential read data showed a fair gap between the Marvell 9172 and the SB950 in standalone mode in favor of the Marvell controller. This is of course not what we want to see. Things changed slightly in RAID0 testing as the SB950 picked up a slight gain in performance over the Marvell controller. Unfortunately as usual the SB950 and Marvell 9172 trail behind Intel controllers. http://www.nikeairhuaracheireland.nu Though again USB 2.0, 3.0 and IEEE1394 performance was solid. The 100MB sequential write tests showed better results more in line with the competition and what we prefer to see. The SB950 and Marvell 9172 were very close in standalone mode favoring the SB950 slightly. In RAID 0 testing the Marvell 9172 fell way behind the SB950 controller. IEEE1394 performance was again somewhat disappointing but not out of line. USB 2.0 and 3.0 performance was exactly where it should be.
This data was not easy to collect. Actually it was more time consuming than usual to collect than anything. Specifically there were problems with the Marvell 9172 controller. I was unable to create a RAID array on the eSATA ports or enter the option ROM for the internal ports (both are 9172 controllers with separate integrated adapter configurations) while the SB950 was in RAID mode. If I took the SB950 out of RAID mode and booted from a single AHCI volume on ports 4 or 5 I was OK and the Marvell controller behaved correctly. nike air force ireland Normally I use a RAID array for the C:\ drive as it allows me the ability to do AHCI testing and RAID testing without changing the controller mode. Changing the controller mode to something other than what you setup the OS on can cause BSOD and other issues. In this case I had to have two separate hard drive configurations both with their own OS installation on them. So it was literally almost double the work to setup the testing rig.
Gigabyte has been made aware of the problem.http://www.nikeairforceireland.nu It is unknown when they resolve this problem, or whether or not they even be able to duplicate it. This was with BIOS F3, so going forward I look for newer BIOS revisions if you plan on using the Marvell 9172 RAID features in conjunction with the SB950. I advise making sure this is resolved before purchasing this board.
Hagel Technologies DU Meter software was used with Windows Task Manager to determine the performance levels of the onboard network interface. DU Meter was used to measure bandwidth and transfer speeds, while Windows Task Manager monitored CPU utilization on the test system. For the testing, a 750MB Archive file consisting of several compressed WMA/MP3 files was used for the large file transfer, and 750MB worth of MP3/WMA files were used ranging in sizes from 3 to 30MB was used for the small files transfer test. http://www.nikeairforceireland.nu The test was performed using a plenum rated category 5e crossover cable to bypass any traffic, routing or other transfer issues and possible packet loss or corruption that can be caused by a router/switch or hub. The cables were connected between two test machines, one using the onboard NIC(s) of the board being reviewed and the other is an Intel EXPI9400PT 10/100/1000Mbps PCI Express Gigabit Ethernet adapter installed into a test machine using an Intel D5400XS motherboard.
The Gigabyte 990FXA UD7 features the Realtek 8111E PCI Express network adapter. It is capable of 10/100/1000Mbps speeds. This is the same controller the MSI 990FXA GD80 used and while not as bad, it still gave me some performance issues. Most notably in the small files download test.
The large files download test was a little better than the small files test. The average transfer rate was 40.4MB/s and maxed out at 65.3MB/s. CPU usage was again a modest 3%. http://www.nikeairforceireland.nu Finally in the large files upload test we actually saw excellent performance. The average transfer rate came in at 104MB/s with a peak of 115MB/s. CPU usage climbed slightly to 4%. This test was much more in line with what we should see from this controller.
Still the controller performance is inconsistent and could indicate issues which need resolution. nike air huarache ireland While not as bad as what I saw in the MSI 990FXA GD80 article, the Gigabyte 990FXA UD7 network controller could definitely use improvement.