Gigabyte GA-F2A85XN-WIFI Mini-ITX Motherboard

The GA-2A85XN-WIFI from Gigabyte offers a nice feature-set for a reasonable price and the best energy effiency I’ve seen out of an FM2 motherboard. Our first in a series of DIY articles on mini-ITX motherboards for small system options.

August 5, 2013 by Lawrence Lee

If you’re looking for a mini-ITX motherboard for an Intel Sandy or Ivy Bridge processor, there are a couple of dozen to choose from at a variety of different price points. If on the otherhand, you’re interested in an AMD model, the selection is pitiful by comparison — just a handful. It’s a bit puzzling why their mini-ITX adoption rate is so low, after all, mini-ITX is the form factor du jour for building home theater PCs and simple small boxes with a low environmental footprint, and Trinity/Richland APUs are well-suited for both.

Of the limited number of boards to choose from, the Gigabyte GA-F2A85XN-WIFI is actually the current premium model, even though its street price is just US$110. I reviewed the GA-H77N-WIFI not long ago and its clear that the two products not only share much of the same DNA but seem to be effectively clones of one another, almost mirror reflections. The two are value packed with usual features like S/PDIF, gigabit ethernet, USB 3.0, and SATA 6 Gbps, but also dual HDMI connectors (no DisplayPort) and a mini PCI-E adapter providing both 802.11n and Bluetooth 4.0 support.

The biggest difference, aside from the socket type, is the H77N-WIFI uses a chipset that lacks overclocking capabilities, something that doesn’t exist on the AMD side. The A85X chipset is AMD’s flagship, comparable to Intel’s more expensive Z77 chipset, and this gives the F2A85XN-WIFI some extra appeal to enthusiasts and the like.

Budget boards aren’t typically loaded with extras and the F2A85XN-WIFI comes with almost
the bare minimum. Included is a driver/utility disc, manuals, I/O shield, two SATA cables, and an external WiFi antennae. The antenna has a plastic exterior, is rotatable, and the base is magnetized to keep it in place. This is a nice step up from the wobbly rubber/plastic antennae included with the H77N-WIFI.

PHYSICAL DETAILS

Given the size of the mini-ITX form factor, layout really isn’t important as long as there are no glaring interference issues. As they typically aren’t installed in tower cases and there’s so little space on the PCB, the ATX layout paradigm gets thrown out the window. While it’s an FM2 motherboard, the F2A85XN-WIFI’s layout is similar to most Intel mini-ITX models with CPU socket is on the bottom half of the board and most of the peripherals and connectors are located on the upper portion.

BIOS/UEFI

The F2A85XN-WIFI features the now standard Gigabyte UEFI DualBIOS with a graphical
user interface and mouse support. It offers all the functionality you would expect from a modern motherboard though some of the settings are scattered in different sub-menus.

Glancing at the options in the BIOS/UEFI, it’s clear that this isn’t a hardcore enthusiasts’ board. The degree of frequency and voltage adjustment is enough for most overclockers but there are no ridiculous high limits aside from the GPU clock. It does have a couple of nice touches that I like to see, the ability to undervolt, and dynamic voltages adjustments which apply both when the APU is idling, on full load, and anywhere in-between.

Mini-ITX systems typically don’t have many fans, so the F2A85XN-WIFI is only equipped with a pair of PWM fan headers. Basic control is available in the BIOS/UEFI. Users can use the predefined “Normal” and “Silent” settings or choose “Manual” to adjust how quickly the fans react with a bit more refinement. After some trial and error, I found that “Normal” was the equivalent of the 1.75 setting while “Silent” corresponded to 1.00.

Boot Performance

To test boot time I optimized the BIOS menu by setting the hard drive and other delays set to minimum, taking care not to disable common functionality like USB support, POST messages, etc. and measured the time it takes to reach the Windows loading screen (I stop here because this is the point where the O/S drive speed becomes a factor).

The boot process on the F2A85XN-WIFI is about average for an FM2 board, hitting the Windows loading screen in under 12 seconds. It’s not particularly quick, but for some users, it will reach this point before the monitor even comes out of standby mode.

TEST METHODOLOGY

Test Setup:

• AMD A8-5600K
  processor – 3.6 GHz, 32nm, 100W, integrated Radeon HD 7560D graphics
• Scythe Kabuto
  CPU cooler – stock fan at 800 RPM
• Kingston HyperX LoVo 4GB memory – 2x2GB, DDR3-1600, 9-9-9-24
• Western
  Digital Scorpio Blue notebook hard drive – 500GB, 5400RPM,
  8MB cache
• Asus
  BC-08B1ST Blu-ray drive
• Seasonic
  SS-400ET ATX power supply
• Microsoft
  Windows 7 operating system – Ultimate, 64-bit

In-Case Test Setup:

• AMD A10-6700
  processor – 3.7 GHz, 32nm, 65W, integrated Radeon HD 8670D graphics
• Noctua NH-L9a
  CPU cooler
• Kingston HyperX LoVo 4GB memory – 2x2GB, DDR3-1600, 9-9-9-24
• Corsair Force GS solid-state drive – 240GB
• FSP Group FSP300-60GHS SFX power supply – 300W
• SilverStone Sugo SG05 case – mini-ITX
• Microsoft
  Windows 7 operating system – Ultimate, 64-bit

Measurement and Analysis Tools

• CPU-Z
  to monitor CPU frequency and voltage.
• CPUBurn
  CPU stress software.
• Prime95
  CPU stress software.
• FurMark
  GPU stress software.
• Cyberlink
  PowerDVD 13 to play H.264 video.
• Mozilla
  Firefox and Adobe
  Flash Player to play Flash video.
• CrystalDiskMark benchmarking tool for storage devices.
• SpeedFan
  to monitor system temperatures and fan speeds.
• Seasonic
  Power Angel AC power meter, used to measure the power consumption
  of the system.
• Infrared Thermometer to measure heatsink temperatures.
• Digital Multimeters to measure voltage drop.
• Icy Dock MB981US32-1S eSATA/USB 3.0 dock to test storage subsystems.
• Kingston HyperX 3K 120GB solid state drive to test storage subsystems.

Video Test Clip

1080p | 24fps | ~22 mbps

H.264/MKV 1080p: A custom 1080p H.264 encoded clip inside an Matroska container.

 

1080p | 24fps | ~2.3 mbps

Flash 1080p: The Dark Knight Rises Official Trailer #3, a YouTube HD trailer in 1080p.

 

1080p | 24fps | ~33mbps

H.264 Blu-ray: A short section (chapter 4) of the Blu-ray version of Disturbia, the motion picture, played directly off the Blu-ray disc. It is encoded with H.264/AVC.

 

Estimating DC Power

The following power efficiency figures were obtained for the
Seasonic SS-400ET used in my test system:

Testing Procedures

If available, the latest motherboard BIOS is installed prior to testing. Certain services/features
like Indexing, Superfetch, System Restore, and Windows Defender are disabled
to prevent them from causing spikes in CPU/HDD usage. I also make note if energy
saving features like Cool’n’Quiet/SpeedStep or S3 suspend-to-RAM do not function
properly. If a WiFi adapter is present, it is disabled unless the system lacks wired ethernet.

Our main test procedure is designed to determine the overall system power consumption
at various states (measured using a Seasonic Power Angel). To stress the CPU, I use either Prime95 (large FFTs setting) or CPUBurn depending on which produces higher system power consumption. After 10~15 minutes of load (when temperatures stabilize), I also measure the hottest points on the external heatsinks using an infrared thermometer. To stress the IGP, I use FurMark, an OpenGL benchmarking and stability testing utility.

Finally, storage subsystems are tested briefly using CrystalDiskMark (1000 MB of 0x00 fill test data) and a Kingston HyperX 3K 120GB solid state drive. For USB and eSATA I use an external eSATA/USB 3.0 dock to connect the drive.

TEST RESULTS

Power Consumption (A8-5600K)

Unfortunately it’s difficult to ascertain exactly how much of the energy draw
is generated by the processor alone, as the amount of power pulled from the
AUX12V/EPS12V connector depends on how board power regulation has been implemented.
On this board, the CPU is driven by a 6 power phase design but surprisingly the proportion of power derived from the +12V line was very similar to the 4-phase F2A75M-D3H.

The F2A85XN-WIFI’s energy efficiency was similar to previously tested models when playing H.264 video but the power consumption during Flash video rendering was about 5W lower. The idle draw was also very low, tied for the lowest I’ve encountered on an AMD board.

As the load increases, the efficiency of the F2A85XN-WIFI only got better. During all the load tests, the average reading on the power meter was 8~12W lower than the closest performing model, the ASUS F2A85-M Pro. Typically mini-ITX models have stripped down power regulation due to size limitations, giving them better idle than load power efficiency. This makes the F2A85XN’s results ever more impressive.

Cooling

To test the board’s cooling, the CPU was stressed for ~15 minutes with Prime95. Temperatures of the boards’ chipset heatsinks were recorded using a spot thermometer. The highest temperatures were taken for comparison.

Despite its undersized chipset heatsink, the surface temperature of the FCH cooler peaked at just 31°C above ambient on full CPU load which is comparable to previously tested FM2 boards. The VRM heatsink on the otherhand took quite a beating, heating up to 46°C above ambient (close to 70°C in the lab).

Note that our test heatsink is of the downblowing variety but it’s a large model with the fan sitting high up, far away from the PCB and set to a relatively low speed. A slimmer model with a fan that sits closer to the socket would presumably do much better.

Fan Control

Using SpeedFan to track the CPU and System temperatures and fan speeds, I connected 2000 RPM PWM fans to the CPU_FAN and SYS_FAN headers, set both to “Silent” in the UEFI/BIOS (the equivalent of the manual 1.00 PWM value/°C setting) and put the system on full CPU load.

The CPU_FAN reacted fairly linearly to the CPU temperature but it seemed that a fail-safe activated at about 65°C, kicking it into high gear. Surprisingly I found that the SYS_FAN header responded to the system (FCH chipset) temperature which is a truly odd choice for a mini-ITX board.

Given the proximity of everything on the board, any fan that can affect the CPU temperature will also affect the FCH. Designing them to react to different sensors doesn’t make any sense. This feature is far more useful in an ATX tower where perhaps a front or side fan can blow over the chipset. In a mini-ITX case, everything is jammed tightly together.

Gigabyte’s antiquated EasyTune utility makes an appearance once again and its fan control capabilities are almost as limited as in the UEFI/BIOS. The CPU and System fan can be set to a minimum speed of 10% and 20% respectively and the maximum speed cannot be altered from 100%. If you decide to let EasyTune handle your fans, you’ll need to set the maximum temperature fairly high in order to keep things running quietly, even if you have quality cooling.

If you prefer to use SpeedFan, both headers are fully controllable via PWM once the application is set up properly (find the “IT8728F” chip in the Advanced menu and change PWM 2-3 mode from “SmartGuardian” to “Software controlled”). The same CPU and System (chipset) temperature sensors found in EasyTune are also available, reported as “Temp3” and “Temp1” respectively.

Storage Subsystem Performance

To test storage subsystems I used CrystalDiskMark, the 1000 MB setting with 0x00 fill test data, and a Kingston HyperX 3K 120GB solid state drive (compressible data produces the best possible speeds out SandForce drives). The drive was connected using an Icy Dock external dock which supports eSATA and USB 3.0 (limited to 3 Gbps and 5 Gbps respectively).

SATA 6 Gbps

All A75/A85 chipset motherboards are equipped with AMD’s own native SATA 6 Gbps controller and I’ve run this test many times with similar results so I didn’t bother doing it again on this particular model. Intel’s solution is superior across the board, especially with the smaller 512K block size, which means operations dealing with small files.

It’s notable however that the difference is only above ~400 MB/s which is beyond the capability of most SSDs. The popular SandForce drives like the one I used can only achieve these numbers using highly compressible data. All things being equal, its unlikely any difference in performance would be noticed.

USB 3.0

USB 3.0 can purportedly produce data rates of up to 4 Gbps (512 MB/s) but in practice I’ve never seen anything come close. When connected internally, the test drive is capable of 500 MB/s when working with 512K block sizes. The A85X’s controller is a only slightly slower than solutions from Intel and VIA but they’re all roughly in the same ballpark.

This lack of performance isn’t a huge deal if you’re working with external hard drives and flash drives, but SSDs and other high-throughput devices would probably be better off using Thunderbolt as an interface.

WiFi

For the WiFi test, I sent a large file transfer (700MB) to a GBLAN-connected machine. I should note that the 802.11n router servicing our lab is not the greatest — it has never been able to obtain a throughput higher than 50 mbps. Still the relative difference should be noteworthy. For this comparison, I pitted the F2A85XN-WIFI’s Atheros NIC against a USB adapter form ASUS, and a PCI Express controller from Intel.

The ASUS USB-N13 got completely trounced in this test, not even cracking the 10 mbps mark. I wouldn’t describe either the Intel or Atheros adapters as particularly fast either but they were miles ahead of the USB adapter. The onboard WiFi trailed the Intel NIC by 6.4 mbps but that’s still enough for many forms of high definition video.

The file transfer results mirrored what I saw with regard to signal strength. The Centrino reported the greatest signal strength of the three, particularly for further away off-site networks, followed by the Atheros, and then the ASUS. I should also note that USB-N13 only picked up three SSIDs, while the other two detected five (each spotted a different fifth network so I excluded it from the chart). The USB adapter’s poor results are probably in part due to the lack of an external antenna.

Energy Efficiency vs. Intel

Aside from the processor, our Intel and AMD mini-ITX test components are identical and therefore ideal for a direct comparison. The two chips used also have similar power envelopes. The Core i5-2500K and A8-5600K are rated for 95W and 100W respectively.

The two platforms are roughly in the same ballpark when it comes to power consumption. If you discount Blu-ray playback, the AMD configuration performed slightly better under light loads. Under heavy loads, the Intel combination had some compelling victories when running real-life, non-synthetic applications.

In-System Test

To give our readers an idea of what to expect with regards to heat and noise, we installed the F2A85XN-WIFI into an average sized mini-ITX case for, the SilverStone Sugo SG05, and conducted some thermal and acoustic tests. The case’s stock FSP 300W SFX power supply and 120 mm fan are both relatively quiet. To keep noise down further we used a Corsair Force GS SSD, a newer, more energy efficient APU, the 65W A10-6700, and a high quality low profile cooler, the Noctua NH-L9a.

Building a quiet and cool running mini-ITX system is obviously a bigger challenge than a spacious microATX/ATX configuration but with the right components, it’s not particularly difficult. Our system was fairly quiet regardless of what actions it was performing. At idle and during H.264 video playback, I cranked the CPU fan speed to as low as SpeedFan would allow; the resulting noise level was a respectable 20 dBA@1m and all the internals were operating at comfortable temperatures. It could have been much lower if we had slowed down the system fan but as the board can only control PWM models, I felt it was fair to leave it as is. For my load tests, I adjusted the CPU fan speed to 1,650 RPM in an attempt to keep the temperature under 70°C. Video encoding and Prime95 bumped the output to 23 dBA but the noise level was certainly bearable and the pitch/tone was not unpleasant. I did not perform our standard CPU + GPU load test as our A10-6700 chip throttles the CPU to keep the APU within its designated power envelope causing similar power and thermal measurements as the CPU load test alone.

The peak power consumption was 106W AC but assuming 80% efficiency, the DC power draw was about 85W. This system could easily have been powered by a silent DC-DC power supply with a 120W brick.

Gigabyte will be sending us an equivalent Intel mini-ITX motherboard in the near future so we can make a direct comparison of environmental properties between the two platforms.

FINAL THOUGHTS

The Gigabyte GA-F2A85XN-WIFI is very similar to the GA-H77N-WIFI I reviewed a couple of week previous. Much of what I said about the H77N-WIFI applies here as well. It’s nice energy efficient board for basic computing, hooking up to a TV, or for a small server (a proper server requires additional SATA ports). However, for gamers and enthusiasts, the F2A85XN-WIFI isn’t a compromise as it has a VRM cooler and overclocking capabilities. This extra versatility really makes the Z77N-WIFI, a more expensive Z77 chipset board, its true analog on the Intel side. The F2A85XN-WIFI also has a better designed and more attractive WiFi antenna.

Even discounting these differences, I found the F2A85XN-WIFI to be more impressive based on how well it stacked up against other FM2 boards. Of the three FM2 boards we’ve reviewed, the F2A85XN-WIFI absolutely dominated them in load energy efficiency. I’m talking about a difference of about 10W, which is huge when the total consumption is only 75~100W.

Like most budget boards, there are a couple of places that need improvement. You can’t expect the fan control to be exemplary on a budget mini-ITX model but limiting control to PWM fans is frustrating when almost every fan included with a case is of the 3-pin variety. In addition, if you let Gigabyte handle the fan speeds, the system fan reacts to the chipset rather than the CPU temperature, something that makes absolute no sense for this form factor. To top it all off, the accompanying software is antiquated and restrictive — use SpeedFan instead, it’s just better in every conceivable way. Also, both the chipset and VRM heatsink are undersized, though for the chipset its more forgivable as the FCH on FM2 boards doesn’t seem to heat up as much as the PCH on Intel motherboards.

Despite these issues, if you’re looking for a small AMD motherboard for a SFF system, the F2A85XN-WIFI fits just about any niche you have in mind, and does so with first class energy efficiency, at least for an AMD model.

* * *

Gigabyte GA-F2A85XN-WIFI
is Recommended by SPCR

Articles of Related Interest
Gigabyte GA-H77N-WIFI Mini-ITX Motherboard
Intel DQ77KB: A Low Power LGA1155 Motherboard
Gigabyte GA-F2A75M-D3H Budget FM2 Motherboard
ASUS F2A85-M Pro microATX FM2 Motherboard
ASUS P8Z77-V Pro: Xpert Fan Control for LGA1155

Intel DZ77GA-70K Z77 Motherboard: Waiting for Ivy Bridge

* * *

Discuss
this article in the SPCR forums.