For years, I've been meaning to do a blog post about the Benchmarking Sweatshop. Thing is, every time there's an opening in my schedule, my lab is a complete mess. "I'll take pictures the next time I clean up in here," I say to myself. And I do clean up, occasionally, but that only makes a small dent in the chaos of hardware that lines every free surface and the pile of boxes spilling onto the floor. Without an attached warehouse, it's difficult to keep up with the flow of new hotness delivered sometimes several times a week.
Screw it. The Benchmarking Sweatshop is never a pristine, clutter-free environment; showing it as such would be disingenuous at best. Besides, it would take hours to really clear the decks in here. There really isn't anywhere for all the boxes to go, since the laundry room that doubles as TR storage is still in the midst of repairs due to a blown water main. One wall in my living room is already lined with motherboard and PSU boxes displaced by the flooding, and a shipping container in the driveway is brimming with even more refugee hardware.
Really, I don't have it that bad. My current lab is larger than previous iterations of the Benchmarking Sweatshop, which got its name from a particularly toasty duplex I lived in years ago. The place was really quite nice, but the black tar roof and lack of air conditioning were a toxic combination during the height of summer. It didn't help that I was running a desktop, multiple monitors, and several test systems from the Prescott era, all of which conspired to raise temperatures past 35°C. Good thing I have a flexible dress code. Pants are optional, at least until FedEx knocks on the door.
Remarkably, much of the furniture in my lab has been with me since I left university residence about a decade ago. The configuration has never been completely overhauled, but it's been subtly refined in each move. Let's start with mission control, which has received the most attention.
The crown jewel of my workstation is the six-screen monitor array hooked up to my desktop and test systems. That's a whole other blog post, so I won't go into much detail. The bottom two screens on the right are connected to my desktop, while the remainder are linked to test systems on an adjacent shelving unit. In a pinch, I can switch my desktop to the top-mounted displays and enjoy a three-screen config for surround graphics, ahem, testing.
Despite the mix of papers, receipts, sunglasses, thumb drives, and other widgets that clutter my desk, there's still room for a trio of keyboards and mice. The lower combo on the left is attached to an ancient Avocent KVM switch, while the Corsair duo on the tray above makes a direct connection to one of my test rigs. The speakers perched above the monitors, plus the headset hanging on the wall to the right, are connected to my main PC. Test systems share the headphones on the left. One of these days, I'm going to spring for a receiver and revamp the audio system completely.
Perhaps the most important detail in the shot above is the chair, which cradles my butt for entirely too many hours in a given week. I've had this Nightingale 6200 CXO for probably close to eight years now, and it's showing no signs of wear despite heavy use. The chair is extremely comfortable, whether I'm leaning forward and typing intently or reclining back to wait for a benchmark run to finish. My only complaint is the so-called "soft" casters, which are supposed to keep floors scratch-free but inevitably get impregnated with little bits of dirt that have left a faint halo on the faux-wood flooring beneath my feet.
To the left of my workstation sits The Rack. This Ikea Ivar shelving unit has hosted my test systems for what feels like forever, and I've grown quite fond of its ability to stack four rigs within a relatively small footprint. The testing we do often requires benching multiple systems concurrently, and this seems to be the most space-efficient approach outside purpose-built test stations. I do have to be careful not to have the shelves too close together, though. The Rack's only airflow is generated by each system's CPU fans, GPU coolers, and PSU. Occasionally, collection of auxiliary 120-mm spinners will be deployed to troubleshoot problems that might be heat-related.
Obviously, running four test rigs alongside six screens, a desktop, and a handful of other systems requires quite a lot of cabling. I'm content with the modest jungle as long as it stays out of my way, which it does thanks to the liberal use of zip ties. I've even taken to color-coding the ends of various cables using smaller ties, so that I know which plugs go where. The residue left behind by adhesive labels always bugged me when I had to re-tag things after a move and subsequent layout tweak.
Getting to The Rack involves negotiating the pile of boxes and hardware slowly eating away at the only free floor space in my lab. Those boxes on the floor are from just the last week. Admittedly, though, the towers over on the right have been sitting there for months. Those are older systems waiting to be broken down for parts. If they weren't there, you'd be able to see The Beast, our custom PSU tester, which is tucked away in a half-height Ivar shelving unit.
Apart from my chair and a filing cabinet, the Benchmarking Sweatshop is made up exclusively of Ikea furniture. It's cheap. Also, the Ivar's adjustable shelves and multiple size options are perfect for the mix of items I need to store in the lab and anywhere else there's room for a few rows of motherboard boxes.
Atop the units pictured above is what can only be described as my junk drawer. Everything from swimming goggles to power adapters to loose change litters the top shelves, plus a few props for product shots, a bundle of business cards, and yet more boxes. There's a small TV, too, but it's rarely on unless I'm benching into the evening. 99% of those optical discs haven't moved in years. It's probably time to put them in the garage in a box labeled "dead media."
Our tour continues at the back corner of the lab, which hosts another Ivar shelving unit sitting on a Jerker desk that's been raised on a stack of 2x4s. I build all my test systems here, always while standing, so it helps to have a higher work surface. If you think this station looks busy now, you should see it in the midst of a motherboard round-up.
All the little things tend to live in this corner: DDR memory, solid-state drives, CPU coolers, fans, and drawers filled with the different screws needed to put together a PC. There's also a generous supply of isopropyl alcohol used to scrub off thermal paste, several screwdrivers, and a roll of duct tape—of course.
The last stop on our panoramic journey through the Benchmarking Sweatshop is its makeshift photo studio. Cyril has a nicer setup, but these hardware-store work lights seem to do the trick. It helps that the walls and ceiling are painted a neutral white. I can bounce the light off them without worrying about diffusing umbrellas
Like Cyril, I use a big sheet of white paper as the backdrop for product shots. Several microfiber cloths are at hand to buff away fingerprints and smudges, and an old brush designed to clean graphics cooling fans does a good job of banishing dust particles. The camera, which I used to take these pictures, is a Canon Rebel T2i. For product shots, it's usually paired with a tripod and a wired remote shutter. Having my PC serve as the control station for the photo studio doesn't really work, since it's difficult to walk between the two when the lights are in place. A combined 2000W of lighting gets really hot. It's best not to get too close, but the path between box mountain and the lab's gargantuan beanbag chair is pretty narrow.
The beanbag is necessary for the security system I have in place to protect all this hardware.
Vargas, a two-year-old Vizsla, is the final fixture in the lab. He can most often be found sleeping on the beanbag, one eye on the entrance. Truth be told, he's a lousy guard dog, more likely to lick a burglar than to bare his teeth. Thankfully, he's agile enough to avoid the stacks of hardware and webs of cabling that permeate the lab. Only occasionally do I have to interrupt his snoring for noise testing.