Motherboard selection is mercifully less intimidating than picking a laptop for gaming. With boards, we can establish a set of criteria and narrow down the selection immediately to something more manageable; lower prices than other components also make selection somewhat easier to mentally justify. Our criteria for motherboard selection typically includes consideration of socket type, form factor, ability to overclock, and chipset
We've previously published chipset guides for both AMD's latest chipsets and Intel's Haswell chipsets, each of which shows the differentiating features between various inter-platform options. This buyer's guide looks at the best gaming motherboards for Intel's Haswell and Devil's Canyon processors, then AMD's FM2+ platform. AM3+ is not considered in this guide, given its age and our decision to abandon the platform in PC build guides. We've also opted to exclude X99 motherboards from this guide, given the added complexity and entirely different architecture.
We'll start with tables, then cover the things to look for in a motherboard, and then move on to our selections for this season.
Since the creation of gaming laptops, their performance has generally been substantially lower than their desktop counterparts. Somewhat excitingly, this has been changing in recent years. The release of nVidia’s Maxwell-based mobile GPUs introduced laptops that are inching closer to the performance of their desktop brethren.
While gaming laptops generally provide worse performance-per-dollar and customization options when compared to self-built PCs, their advantages in mobility are unrivaled for traveling gamers.
Marketing speak comes in fad-like waves of bombastic claims that are often not founded in reality (or are yet untested in reality).
When it comes to gold-plated USB cables – something we've griped about in the past – that's exactly what they are: Marketing speak. The advertised advantages are generally stated in the form of “reduced latency” or “improved signal quality” (direct quotes from marketing materials), often citing greater conductivity of gold as the catalyst for these claims.
GPU overclocking changed with the release of Maxwell's updated architecture. The key aspects remain the same: Increase the clock-rate, play with voltage, increase the memory clock, and observe thermals; new advancements include power target percent and its tie to TDP. We recently showed the gains yielded from high overclocks on the GTX 980 in relation to Zotac's GTX 980 Extreme and the reference card and, in some instances, the OC produced better performance than stock SLI pairing.
This GTX 980 overclocking tutorial will walk through how to overclock nVidia's Maxwell architecture, explain power target %, voltage, memory clock, and more.
A large portion of the enthusiast market caters to overclocking: Intel sells unlocked CPUs at a premium, Asus, Asrock, and MSI all market their motherboards as “overclocking ready” and guaranteed to OC higher, and GPU manufacturers market their graphics cards as having better cooling, a bigger VRM, and binned chips. All this, and monitor overclocking is rarely -- if ever -- advertised or discussed. This may change with the recent popularity of Korean off-brand monitors, like the QNIX 2710.
Most of our readers (and staff) are avid PC builders, generally opting to select and install components from one of our DIY PC guides. There are entire companies devoted to custom PC builds, though, and they often build and ship hundreds of custom gaming PCs each day; that's a huge number, considering the relative size of the “gaming PC” market compared to biz-client sales. Out of curiosity, we toured a few of these SIs (system integrators) to observe the process and learn about the automation involved in system building.
We previously toured CyberPowerPC, where we looked at a high-end PC build with custom NZXT components. Today, we're looking at iBUYPOWER's warehouse and assembly line, where you'll see a wall of thousands of dollars of video cards, conveyer belts moving rigs from one bench to another, and even packing tape automation. Yes. A robot that does nothing but tape boxes.
Most of the tech industry’s major players are located somewhere in California – a state that has, in our experience, proven to be very large and very saturated with horrifyingly bad drivers. It also happens to be saturated with leading technology innovators and game development companies; the hardware split is pretty even between SoCal (Orange County, Fountain Valley, LA, Industry) and NorCal (home to Silicon Valley). Game developers mostly hang-out in San Francisco and San Jose.
We’ve previously toured both regions, with some of our best content focusing on nVidia’s silicon failure analysis lab (San Jose) and Kingston’s automated RAM/SSD manufacturing line. Following Game24 and the GTX 980 launch, we returned to the Los Angeles area for more. In our most recent California trip, we visited NZXT, HyperX, CyberPower, and iBUYPOWER to see their assembly lines and warehouses.
Retailers and manufacturers are always happy to give consumers purchasing options: Spend an extra $30 and get buying insurance, another $50 and you get an extended warranty, spend untold thousands on a car to add Bluetooth, and in the case of video cards, an extra $20 and you get a “faster” card in the form of a pre-OC or “SuperClock.”
We’ve explained overclocking as it pertains to GPUs in the past, but never looked specifically at pre-overclocked or SuperClocked cards. The realistic intent of higher-clocked GPUs is to enable users who are either too busy/lazy to overclock, would prefer to have an expert do it for them, or who are legitimately unaware of or afraid of overclocking. Some of the high-end overclocking cards are binned-out with hotter chips (chips that can overclock higher), but not all SuperClocked and pre-overclocked cards are like this. Many of the available options are just overclocked versions of the stock card.
SSD benchmarks generally include two fundamental file I/O tests: Sequential and 4K random R/W. At a very top-level, sequential tests consist of large, individual files transfers (think: media files), which is more indicative of media consumption and large file rendering / compilation. 4K random tests employ thousands of files approximating 4KB in size each, generally producing results that are more indicative of what a user might experience in a Windows or application-heavy environment.
Theoretically, this would also be the test to which gamers should pay the most attention. A "pure gaming" environment (not using professional work applications) will be almost entirely exposed to small, random I/O requests generated within the host OS, games, and core applications. A particularly piratical gamer -- or just someone consuming large movie and audio files with great regularity -- would also find use in monitoring sequential I/O in benchmarks.
This article looks at a few things: What types of I/O requests do games spawn most heavily and what will make for the best gaming SSDs with this in mind? There are a few caveats here that we'll go through in a moment -- namely exactly how "noticeable" various SSDs will be in games when it comes to performance. We used tracing software to analyze input / output operations while playing five recent AAA titles and ended up with surprisingly varying results.
UPDATE: Clarified several instances of "file" vs. "I/O" usage.
This test was spawned out of a general lack of equipment in the GN lab. We've got a few monitors available for testing, but of the three best units (120Hz displays), only one natively operates at 1920x1080; the others -- fabled unicorns among monitors -- run at 2048x1152 and 1920x1200.
The 1920x1080 120Hz display isn't always available for our game GPU benchmarks, making it desirable to use one of the larger displays at a lower-than-native resolution (for consistent / comparable testing). In effort of honest benchmarking, we decided to double-check an existing suspicion that forcing lower-than-native resolutions would not negatively impact FPS or produce synthetic artifacts that do not exist at native resolutions.
The hypothesis says "nope, should be identical performance other than visual scaling." Let's see if running a monitor at a non-native resolution will negatively impact testing with artifacts or lower FPS.
We moderate comments on a ~24~48 hour cycle. There will be some delay after submitting a comment.