I've never before had GN's lab so fully equipped. It's an exciting period of growth for us. For the first time in the site's seven-year history, it feels appropriate to slap the “lab” label on our multi-room testing setup. Following weeks of intensive cleaning and organization efforts, we've now got shelving units installed (all connected to ground and ESD compliant) that house motherboards, video cards, CPUs, and more – much bought out of pocket – and a complex network of systems.
The brain of the network is our rendering rig, which sits opposite my main production system. The rendering rig is used exclusively to render and edit videos for GamersNexus, with the primary user being GN's Keegan Gallick. The system hosts RAID HDDs that are utilized for all of our test data, video media, and photo media.
We put together a quick video showcasing the rig:
Our publication schedule is constantly littered with research-intensive articles. The workload is split – Michael Kerns heads-up feature posts like keyboard round-ups, Keegan Gallick assists with sustaining news posts and videos, and Patrick Lathan handles social media and some posts. This is done intentionally, as I'm normally pulling my (abundant) hair out trying to devise new testing methodology for upcoming products. It's fun, though, and we do our best to split resources well enough that the site is sustained during times of intense research.
This week's slow-down on major content posts has allowed for some behind-the-scenes work on a new benchmarking platform – an X99 rig using an i7-5930K and DDR4 memory for SLI and CrossFire benchmarking. System integrator iBUYPOWER had some spare X-rev lab samples left over, so with their support and our own investment in other hardware, the site will soon be able to produce SLI tests without concern of throttling due to lane and processing limitations of the 4790K CPU.
We've been hard at work with some behind-the-scenes efforts lately. As the site has continued to grow, we've found more room in the budget to make upgrades in necessary areas to allow continued growth; we've also allocated some funding to review hardware, camera / convention equipment, and staffing power.
If we were “on the map” before our GTA V and Witcher coverage, we've become a much more noticeable – but still medium-sized – dot on that map. GTA V's launch saw our day-1 publication of a GPU benchmark, followed rapidly by texture comparisons, a heavily-trafficked graphics optimization guide, and a CPU benchmark. All totaled, these items helped contribute to the site's first time exceeding 1,000,000 pageviews in a one-month period. Huge news for us, as we've traditionally rested in the 300-500k pageviews per month range. Much of that persisted through The Witcher 3, offering the same types of content (GPU benchmark, graphics optimization guide, more) with a slightly smaller inbound traffic metric.
As we ramp into GDC and PAX East, we're using the gap in review time to overhaul our testing methodology and test platforms. Yesterday's post revealed our open air GPU testing station, a direction that'll drastically improve our efficiency when testing multiple graphics configurations. Today, we're looking at the new case review test bench. The site has grown substantially in the past two years; we'll no longer be using the same bench for testing all components, and will now use individual systems for testing each component. This will eliminate chance of test error, improve efficiency, and allow each of our writers to specialize in an area.
GN's Staff Writer & Social Media Manager, Patrick Lathan, will be handling most ATX and micro-ATX case reviews going forward. As such, I dropped off a load of parts for Patrick's new test bench, which will be put to immediate use with NZXT's S340. Following his review of the S340, we'll look at Be Quiet's Silent Base 800.
We try to rotate our official test bench hardware every few years to maintain a relevant, but realistic configuration. Ideally, we use mid-range components that more accurately reflect what a consumer would build – an i5-3570K or i5-4690K, for instance, rather than a $340 i7-4790K. This platform is applied to all of our game benchmarks, which aim to report real-world framerates of newly-launched games.
In a recent attempt to increase testing efficiency, we opted to remove the tried-and-died NZXT Phantom 820 (the case is shot from years of abuse) and migrate to an HPTX top-deck tech station, or “test bench,” as we call it. The unit is fully open-air, exposing expansion cards and video cards for easier access and swapping when performing tests. There are some test concerns by opting for this route, but we've addressed those in our methodology (and will briefly mention them below).
As we've always updated our readers on the site's methodologies and proceedings, I thought I'd put together a video showcasing the new test bench. You can find that below.
As part of our ongoing effort to define major component specifications, we've now published our “PSU Dictionary” to about 80% completion. Additional entries will be listed throughout the week's end. The PSU dictionary details critical terminology to selecting a quality power supply, including 80 Plus certifications, Rails (PSU), electrical protections, voltage ripple, and (soon) coil whine.
The first PAX South sees the showcasing of several indie titles, a few mid-sized titles (Grey Goo, Dreadnaught), and the occasional hardware vendor. We figured a quick sales recap could commemorate the first major gaming convention in the southern US.
On December 31, 2013, I promised that we'd have a new website with a focus on compatibility for large screens and mobile devices. I also promised that we'd have it “sometime soon in 2014.” It's been one year exactly. That's soon-ish. Hey – at least I stuck to my story.
Just a quick note: Following-up on the previous post mentioning our impending upgrades, there is anticipated downtime for the site at around 12AM EST 12/31. The duration of the downtime is yet undetermined. We will be migrating and upgrading the website, so it could take the better part of the day. We'll attempt to be finished by noon EST 12/31, though.
Take it easy!
We moderate comments on a ~24~48 hour cycle. There will be some delay after submitting a comment.