Heavyweight rematch: Gigabyte X79-UP4 vs. MSI X79A-GD45 Plus

Intel’s ultra-high-end desktop platform got a shot in the arm from Ivy Bridge-E in September. This refresh delivered updated CPU cores, but it didn’t bring any changes to the two-year-old LGA2011 platform. Intel didn’t update the accompanying X79 Express chipset, which is why we didn’t see a wave of new motherboards rolled out with Ivy-E. Asus’ X79-Deluxe was the only fresh face at the time, and neither Gigabyte nor MSI has released anything since.

Part of Ivy-E’s appeal is the fact that the chip is a drop-in replacement for its Sandy Bridge-based predecessor. Existing X79 boards should require no more than a firmware update to work with the latest processors. Gigabyte and MSI both have newish X79 models that we haven’t tested, so we decided to have a little throwdown to see how they compare. In the black and grey trunks, we have the $220 Gigabyte X79-UP4. And, uh, also in black and grey trunks, we have the $250 MSI X79A-GD45 Plus.

These boards have been in the lab for a while now, and I still have trouble telling them apart at a glance. Closer inspection reveals plenty of differences, though.

Gigabyte’s X79-UP4
We begin with Gigabyte’s X79-UP4, which delivers a lot more extras than one might expect from an affordable X79 model.

The UP4 wears its black-and-grey aesthetic well. The circuit board’s matte surface is especially sinister, and I really like the look of the heatsinks. At the very least, the monochrome motif shouldn’t clash with other system components.

Zooming in on the socket gives us a better angle on the seven-phase power circuitry feeding the CPU. Each phase is powered by fancy electrical components from International Rectifier. The board has ferrite-core chokes and extra-beefy copper layers, too. We’d expect nothing less from an enthusiast-oriented motherboard.

As you can see, the socket area is a little crowded. The VRM heatsink and top PCIe x16 slot encroach from the north and south, respectively, while dual banks of DDR3 memory slots flank from the east and west. We can’t check clearances for every hardware combination, but we can convey a few key measurements.

Like on most modern motherboards, the DIMM slots come closest to the socket. Beware of combining taller memory modules with oversized air coolers. Watch out for the PCIe slot, too; it’s all up in the socket’s business. At least the VRM cooler is short enough to stay out of the way.

The socket area is crowded in part because the board’s ATX footprint has limited room for eight DIMM slots. (There are two slots for each of the processor’s quad memory channels.) Gigabyte’s decision to add a seventh expansion slot—one more than on the MSI board—also results in tighter clearances around the socket.

All four of the x16 slots get PCI Express 3.0 connectivity directly from the CPU. The first and last slots have x16 and x8 links, respectively. The middle two share an x16 link that can be split evenly between them or devoted solely to the third slot. Props to Gigabyte for putting enough space between the full-fat x16 slots to provide breathing room for dual-card configs. I’m even more impressed that the X79-UP4 can host four double-wide graphics cards, each one connected to the CPU. This board is officially approved for quad CrossFire and SLI configurations.

The rest of the PCIe slots stem from the X79 platform hub. Although the chip is limited to Gen2 connectivity, the older spec should provide sufficient bandwidth for the x1 slots and auxiliary peripheral controllers.

The X79’s own peripheral payload is relatively weak. There’s no built-in USB 3.0 connectivity, and 6Gbps SATA support is restricted to two of the six ports. Gigabyte provides some relief with a collection of Marvell controllers that adds four internal 6Gbps ports and two external ones. A pair of Fresco Logic controllers handles USB 3.0, providing two ports at the rear plus an internal header for two more.

Four USB 3.0 ports doesn’t sound like a lot in the context of modern Haswell boards, but it’s enough to handle more high-speed peripherals than most folks need to run simultaneously. The X79-UP4 is loaded with USB 2.0 ports for older devices with lower bandwidth requirements. It even sports a combo PS/2 port for the old-school clicky keyboard crowd.

Gigabyte earns two gold stars for populating the cluster with both common connector types for digital S/PDIF audio output. Bypassing the onboard DAC is the best way to get good sound out of integrated motherboard audio. Unfortunately, digital audio output is limited to stereo playback and surround-sound content with pre-encoded tracks. Music and movies should work great, but multi-channel game audio can’t be encoded in real time. The drivers for the Realtek audio codec at least offer some virtual surround mojo that fakes multi-channel output for stereo devices.

The cushioned I/O shield pictured above is pretty awesome—there are no tiny slivers of metal to slice your fingers or get caught up in the ports. Little touches like this can make the building process much easier. Too bad Gigabyte made the front-panel connectors unnecessarily difficult to use.

The front-panel pins are nicely walled off, but there’s no external block to simplify the wiring process. Each connector must be attached individually, which can be difficult to do inside a fully-loaded system. MSI and others employ an elegant solution that adds just pennies to the cost of the motherboard. Speaking of which, let’s see what MSI’s X79A-GD45 Plus has in store…

MSI’s X79A-GD45 Plus
MSI takes a slightly different approach with the X79A-GD45 Plus, right down to the whole back-and-grey look.

The circuit board has a glossy finish, and there are hints of brown around the traces. This isn’t a pure black like on the Gigabyte board. At least MSI sticks to one shade of grey, though. The overall theme looks a little more coordinated as a result.

Predictably, the GD45 is populated with exotic electrical components. Driver MOSFETS? Check. Super-ferrite-core chokes? Uhuh. Solid-state capacitors? Yep, including special Hi-c flavors. Hi-c denotes high conductivity, and my childhood makes me suspect that high-fructose corn syrup is the secret ingredient.

Nine power phases supply juice to the CPU—two more than on the Gigabyte board. MSI and Gigabyte use different electrical components, so I wouldn’t read too much into the phase counts alone. We’ll get a sense of each board’s overall power efficiency a little later in the review.

The GD45’s DIMM slots are just a smidgen closer to the socket than on the Gigabyte board. The VRM heatsink is a little farther away, though, and so is the top PCI Express x16 slot.

To avoid potential clearance issues, we use a closed-loop water cooler on our X79 test rig. Modern CPU blocks are nice and compact, and their retention mechanisms screw right into the metal bracket surrounding the LGA2011 socket.

If you want to run four-way SLI or CrossFire setups, the X79A-GD45 Plus comes up short. Those black x16 slots are limited to a single lane of Gen2 connectivity each. The spacing wouldn’t work for four double-wide graphics cards, either, but three-way configs will be just fine. The first two grey x16 slots have 16 Gen3 PCIe lanes each, while the third slot has eight lanes. There’s enough room between the first two slots for a pair of triple-wide cards—or a slot’s worth of airspace for dual-double-wide setups.

The X79A-GD45 Plus is entirely devoid of old-school PCI slots. I see no reason to complain, probably because I can’t think of a single PCI device that I couldn’t easily go without in a modern system.

A four-pin Molex connector supplies extra power to the graphics card slots, and the SATA ports sit next it. The four 3Gbps ports are on the left side of the cluster, while the two 6Gbps ones are on the right. MSI doesn’t equip the GD45 with auxiliary Serial ATA controllers, but it uses a couple of NEC chips to fuel four USB 3.0 ports. Those ports are split evenly between an internal header and the rear cluster.

The external I/O panel ticks a lot of the right boxes. The Gigabit Ethernet jack is powered by an Intel chip, there’s a switch to clear the CMOS without cracking open the case, and digital audio output is available in two S/PDIF formats. Unfortunately, the onboard audio suffers from a familiar handicap: it can’t encode surround-sound game audio for digital output. MSI does bolster the Realtek codec with Creative’s Sound Blaster Cinema software, which provides speaker virtualization, dialog enhancement, and smart volume control, among other functions. I’d rather have DTS or Dolby Digital Live support for real-time multi-channel encoding, though.

Note that neither of these X79 boards has FireWire ports, a common omission that may irk folks with older video cameras. There’s no Thunderbolt connectivity, either, but a platform limitation is probably to blame for that. We haven’t seen any X79 boards with Thunderbolt.

Although we don’t mind missing out on Thunderbolt, we wouldn’t want to be without the little connector blocks pictured above. The blocks allow the bulk of the front-panel wiring to be constructed outside of the case, where there’s plenty of space to fiddle with the tiny, individual connections. The blocks then plug into the motherboard, making the assembly process much easier. Connector blocks are included for the front-panel cluster and for one of the internal USB 2.0 headers. If only the GD45 came with a padded I/O shield like the Gigabyte board, as well.

Tweaking options
PC enthusiasts demand tuning options, and Gigabyte and MSI both deliver. Their boards are loaded with much of the same tweaking and overclocking functionality overall, but the packaging is very different.

Regrettably, Gigabyte and MSI both do a little extra-curricular tuning behind the user’s back. Intel’s Turbo mechanism defines a peak multiplier for single-core loads and lower multipliers as the number of active cores increases. On the Core i7-4960X, for example, the peak single-core multiplier is 40X, while the peak all-core multiplier is 37X. Those multipliers are observed by the firmware defaults. However, adjusting the memory frequency on either board invokes special Turbo modes that apply the maximum single-core multiplier regardless of how many cores are active. That’s overclocking, according to Intel, and it technically voids the CPU warranty.

Although we fully endorse wanton warranty voiding, users should be the only ones who decide when and how to overclock their systems. At least MSI’s “enhanced Turbo” mode can be disabled with a single switch. Gigabyte’s implementation can only be reversed by setting the load-based Turbo multipliers manually. The firmware displays the correct values, but it’s maddening that those values are presented as the multipliers being used by “auto” config. Ugh. The things motherboard makers will do to score a little higher in benchmarks.

After using the next-gen firmware interface on Gigabyte’s Haswell motherboards, the X79-UP4 feels like a big step back. Its UEFI can be traced to the first X79 boards, and the Ivy-E update actually strips out Gigabyte’s mouse-centric “3D BIOS” GUI. All that remains is an old-school layout with basic mouse support and slightly jumpy cursor tracking.

Navigating the firmware with the keyboard is easy, but there’s an odd quirk on that front. Pressing numlock causes all keyboard input to be ignored for several seconds. Otherwise, the interface feels fast and responsive. Most values can be keyed in directly, and the organization generally makes sense. Gigabyte should really consolidate the voltage controls on a single page, though. It’s annoying to have to switch between separate pages for CPU, memory, and chipset voltages, especially when they could all fit easily onto one page.

The firmware is loaded with overclocking options, and it has perks like fast boot settings, configuration profiles, and an integrated flashing tool. Those features are all pretty common nowadays. Good firmware-level fan controls are rarer, and Gigabyte’s need some work. Temperature-based fan speed control is available for the system and CPU fans, but the selection of PWM/°C values is difficult to understand and fairly limiting. Those values describe the slope of the fan response curve.

The fan controls in Gigabyte’s EasyTune software are much more straightforward. This Windows-based tuning utility has a collection of overclocking options, too, but the interface is several years old and very much looks the part. The completely revamped EasyTune app designed for Gigabyte’s 8-series motherboards hasn’t been ported back to the X79-UP4.

Like its Gigabyte counterpart, the X79A-GD45 Plus has a solid array of firmware- and software-based tuning options. MSI hasn’t made any major changes to the UEFI, which has a few mouse-friendly elements and impeccably smooth cursor tracking. The firmware is short on exotic extras, but it’s easy to navigate, and everything seems to work properly.

Don’t get too excited about the Live Update feature teased by the Utilities tab; the firmware can’t update itself automatically over the Internet. Live Update requires Winki, a Linux-based OS that comes bundled with the board. The OS offers basic functionality like a web client, photo viewer, and instant messaging app. Users have to install Winki themselves, though, and the distro isn’t nearly as slick as free alternatives like Ubuntu.

Users can tweak the CPU cooling by changing the minimum fan speed and setting the target CPU temperature. They can also enable “auto” profiles for two of the four system fans, but there’s no granularity beyond that. The system fans otherwise spin at a static, user-defined rate between 50 and 100% of full tilt.

If the screenshot above looks familiar, that’s because MSI’s ClickBIOS II Windows software has a nearly identical interface to the firmware. The shared layout should make navigation a breeze for newbies. However, ClickBIOS II takes forever to load—12.5 seconds according to my stopwatch. Applying changes also takes longer than it should, and reboots are often required, which destroys the convenience of making adjustments from Windows.

MSI’s Control Center is more responsive, but it still takes about seven seconds to load. The software’s multiplier options didn’t work for us even with OS-level multiplier control enabled in the firmware. Control Center scores points for its graphical fan speed profiles, though. The app can also be used to disable the onboard LEDs for stealthier operation.

The neatest thing about Control Center may be the remote management server, which interacts with MSI’s Command Center apps for Android and iOS. Remote overclocking via smartphone is gimmicky, I’ll admit, but remote system monitoring is a really good idea.

Overclocking
We took our Core i7-4960X up to 4.7GHz on the Asus X79 Deluxe. That board’s auto-tuning mechanism managed to overclock the CPU to 4.6GHz with little more than the click of a button, too. Let’s see how the X79-UP4 and X79A-GD45 Plus compare.

Gigabyte got the first kick at the can. We started with EasyTune’s auto-overclocking mechanism, which offers three QuickBoost levels with clock speeds based on testing done in Gigabyte’s labs. For our Core i7-4760X, the most aggressive QuickBoost profile targeted 4.78GHz… and failed miserably. There was no video signal when the system rebooted, and we had to clear the CMOS manually to get things up and running again.

The second QuickBoost setting aimed for 4.53GHz, and it failed, too. The system rebooted without issue, but each attempt to load Windows produced a BSOD. At least the most conservative QuickBoost profile worked without issue. It turned the CPU up to 4.28GHz using a 42X multiplier and a 102MHz base clock, and that combo was stable under load.

Interestingly, the other QuickBoost profiles cranked the base clock to 103-104MHz. I suspect pushing that frequency caused the instability we encountered with those configs. Outside of officially supported gear ratios (100, 125, 167, and 250MHz for Ivy-E), the last few generations of Intel CPUs have offered very little base clock headroom.

Most auto-overclocking schemes rely on pre-baked profiles, but we prefer routines that slowly increase clock speeds and test for stability along the way. That’s how we do things when overclocking manually, and the iterative approach yielded much better results on the UP4.

Through multiplier tweaking alone, we got our CPU up to 4.4GHz. The board supplied extra juice as necessary, but the automatic CPU voltage option was too conservative at 4.5GHz. That speed required a manual voltage bump to keep BSODs at bay under load. 4.6GHz wasn’t stable no matter how many voltage and power settings we tweaked, though. Our load test, which combines AIDA64’s CPU stress test with the Unigine Nature graphics benchmark, consistently crashed the system.

Next up: overclocking on the X79A-GD45 Plus. Again, we let the built-in intelligence have the first crack at the CPU. Pressing the OC Genie button on the circuit board loads up a pre-defined profile based on the system’s CPU. For our Core i7-4960X, OC Genie picked a 43X multiplier and stuck with the default 100MHz base clock, yielding a 4.3GHz peak CPU frequency. This config was stable under load, so we started turning the screws manually.

Like on the Gigabyte board, the automatic voltage tuning worked through 4.4GHz. Higher speeds required manual adjustments. We also had to turn up the offset for CPU voltage droop. Those adjustments worked with the CPU running as fast as 4.6GHz. At 4.7GHz, however, the system consistently spit out BSOD errors under load. After playing with various voltage and power settings—with no success—we called it a day.

Power consumption
While testing the Gigabyte X79-UP4, we ran into an odd issue with our Ivy Bridge-E CPU. The system hard-locked after prolonged idling but was perfectly stable under load. Weird, huh? After much head scratching, trial, and error, we traced the problem to the CPU’s C3, C6, and C7 sleep states. Disabling those sleep states in the firmware resolved the locking issues completely.

At first, we suspected the motherboard. The same CPU works just fine in the three other X79 motherboards we’ve plugged it into. But a replacement UP4 exhibited the same behavior, so we swapped the CPU for a fresh one from Intel. Problem solved.

Since we had already completed the bulk of our testing with the old CPU and its problematic sleep states turned off, we spot-checked the results with the new chip and all its power-saving mojo enabled. The two configs perform pretty much identically. There is a notable difference in power consumption, though. The results with the asterisk come from the new CPU.

We measured power draw at the wall socket with our test system at idle, then playing a 1080p YouTube video, and finally under a full load combining Cinebench rendering with the Unigine Valley demo.

With all the CPU’s sleep states enabled, the Gigabyte board draws nearly 10W less at idle and almost 5W less during video playback. The savings only amount to a drop in the bucket on a typical monthly utility bill, but keep in mind that lower power consumption also means less heat for one’s cooling system to expel.

We should note that Intel custom tunes the voltage ranges for Ivy Bridge-E CPUs. Our second sample runs at slightly higher voltages than the first one, but the differences are very small: 36 mV at idle and 12 mV under load.

But I digress. The real story here is the MSI X79A-GD45 Plus, which has the lowest power draw of the bunch. To be fair, it also has the fewest auxiliary peripheral chips. The others all have additional SATA controllers, and the Asus X79-Deluxe features built-in wireless connectivity and extra USB 3.0 chips, too. Peripheral controllers don’t consume that much power, though. Asus’ X79 boards have a history of higher power consumption, and that trend continues with Ivy Bridge-E.

Memory bandwidth
All the boards were tested in the same system using Corsair DIMMs running at 1866MHz with identical 9-10-9-27-2T timings.

Productivity performance

Kraken JavaScript performance
We tested the latest Kraken release, version 1.1, in Chrome 27.

TrueCrypt disk encryption
TrueCrypt’s AES algorithm benefits from acceleration via Intel’s AES-NI instructions, which are supported by our Ivy Bridge and Haswell CPUs. We’ve also included results for another algorithm, Twofish, that isn’t accelerated via dedicated instructions.

7-Zip file compression and decompression
The figures below were extracted from 7-Zip’s built-in benchmark.

Video encoding

x264 video encoding
We’ve devised a new x264 test, which involves one of the latest builds of the encoder with AVX2 support. To test, we encoded a one-minute, 1080p .m2ts video using the following options:

–profile high –preset medium –crf 18 –video-filter resize:1280,720 –force-cfr

The source video was obtained from a repository of stock videos on this website. We used the Samsung Earth from Above clip.

Gaming

DiRT Showdown
We busted out our Inside the second methods to testing gaming performance. While we aren’t showing all of our fancy latency graphs, we have included results for FPS and the 99th percentile frame time.

So, yeah. Motherboards have little impact on application and gaming performance. Some boards are a little faster or slower here and there, but the differences are slim at best.

Peripheral performance

Serial ATA performance
Our Serial ATA and USB tests were run on a Samsung 830 Series 256GB SSD. To ensure peak performance, the drive was secure-erased before each batch of tests.

With our second CPU installed, the Gigabyte X79-UP4 runs a little slower in the sequential read speed test. Changing the chip and sleep states does’t impact the board’s performance in the other tests, though.

The boards exhibit similar SATA performance overall, at least with the native controller. The performance of the auxiliary SATA chips varies quite a bit more, and the UP4’s Marvell implementation is particularly slow.

USB performance

Without SuperSpeed support in the chipset, X79 motherboards have to rely on third-party chips for USB 3.0. The MSI X79A-GD45 Plus’ NEC controllers deliver the highest transfer rates overall, and in the random I/O tests, they thoroughly outclass the Gigabyte board’s Fresco Logic controllers. That said, I’m not sure how much random I/O performance matters for an external storage interface typically tasked with sequential transfers.

Ethernet performance

All the boards deliver roughly equivalent networking performance. Move along.

Analog audio signal quality
RightMark Audio Analyzer grades analog signal quality on a scale between “very poor” and “excellent.” We’ve translated those values to a numerical scale that starts at low of one and peaks at six. Higher values are better. Our results were gathered with the systems under a combined CPU, GPU, and USB load.

The X79A-GD45 Plus and X79-UP4 score identically in RMAA’s signal quality test. In fact, all of the boards have the same overall score.

If you’re really concerned with analog output quality, we suggest a discrete sound card. Even budget models have noticeably higher output quality than the best integrated solutions.

Detailed specifications
Unless you’re into nerdy technical specifications and detailed test procedures, the rest of this page is a little dry. Feel free to skip ahead to the conclusion on the following page.

If you are into specs, then you’ll probably enjoy the following comparison table.

That’s a little mind numbing, even for me. But anyone who digs that level of detail might also appreciate shots of the hardware that made up our test systems.

Betcha can’t tell which board is which without looking a second time.

Our testing methods
We used the following system configurations for testing.

Thanks to Intel, Corsair, Samsung, and Asus for providing the hardware used in our test systems. And thanks to Gigabyte and MSI for providing their boards for review.

We used the following versions of our test applications:

• 7-Zip 9.20 64-bit
• TrueCrypt 7.1a
• Chrome 27.0.1453.94
• Kraken JavaScript Benchmark 1.1
• x264 r2334
• DiRT Showdown demo
• CrystalDiskMark 3.0.2f
• FRAPS 3.5.9
• NTttcp
• RightMark Audio Analyzer 6.2.5
• Cinebench 11.529
• Unigine Valley 1.0

Some further notes on our test methods:

• All testing was conducted with motherboard power-saving options enabled. These features can sometimes lead to slightly slower performance, particularly in peripheral tests that don’t cause the CPU to kick into high gear. We’d rather get a sense of motherboard performance with real-world configurations, though; we’re not as interested in comparing contrived setups with popular features disabled.
• DiRT Showdown was tested with ultra detail settings, 4X MSAA, and a 1920×1200 display resolution. We used Fraps to log a 60-second snippet of gameplay from the demo’s first race. To offset the fact that our gameplay sequence can’t be repeated exactly, we ran this test five times on each system.
• Power consumption was measured at the wall socket for the complete system, sans monitor and speakers, using a Watts Up Pro power meter. Our video playback load used this 1080p YouTube trailer for the movie Looper. The full-load test combined AIDA64’s CPU stress test with the Unigine Valley DirectX 11 demo running in a 1280×720 window.
• The Force GT 120GB SSD was used as the system drive for all tests. The Samsung 830 Series 256GB was connected as secondary storage to test Serial ATA and USB performance, the latter through a USAP-compatible Thermaltake BlacX 5G docking station. The Samsung SSD was secure-erased before each test that involved it. The Corsair drive was also wiped before we loaded our system image.
• Ethernet performance was tested using a remote rig based on an Asus P8P67 Deluxe motherboard with an Intel 82579 Gigabit Ethernet controller. A single Cat 6 Ethernet cable connected that system to each motherboard.
• Analog audio signal quality was tested using RMAA’s “loopback” test, which pipes front-channel output through the board’s line input. We tested while the system was loaded with Cinebench’s multithreaded rendering test, the Unigine Valley benchmark, and a CrystalDiskMark 4KB random I/O test running on the Samsung SSD attached via USB 3.0.

The tests and methods we employ are usually publicly available and reproducible. All tests were run at least three times, and we reported the median of those results. If you have questions about our methods, hit our forums to talk with us about them.