System Builder Marathon, Q1 2013: The Articles
Here are links to each of the four articles in this quarter’s System Builder Marathon (we’ll update them as each story is published). And remember, these systems are all being given away at the end of the marathon.
To enter the giveaway, please fill out this SurveyGizmo form, and be sure to read the complete rules before entering!
Day 1: The $600 Gaming PC
Day 2: The $800 Enthusiast PC
Day 3: The $1,000 Performance PC
Day 4: Performance And Value, Dissected
Day 5: The $1,600 Alternative PC
Introduction
Surprise! You thought our System Builder Marathon ended last week, right? Well, we were busy working on one final follow-up that doubles the price of Don's $800 machine to see if the limitations of this quarter's narrow price band can be overcome at a still-reasonable $1,600.
Most of Tom's Hardware's editors were enthusiasts before they ever dreamed about covering technology in print. We read dozens of competing review sites to gather the most unbiased balance of opinion and test results before finally plopping down our hard-earned dollars on a handful of parts.
Now, a lot of the components we write about show up at our offices before they're even available to everyone else, and sometimes that pile of free stuff pulls us away from the realities of saving up over the course of weeks or months for a value-packed system to replace once-modern technology. Fortunately, regular interaction with our audience helps keep us grounded.
We continue building our own boxes and participating in the comments for each of the stories we write. Because of this, we're pretty good about anticipating the responses we're going to get for each of our System Builder Marathon machines. Although we're not prophets, we have a pretty good bead on the ways of the enthusiasts. We knew that tightly framing this quarter's competition around the expected price/performance pinnacle would leave a lot of you searching for something more.

That anticipation compelled us to order the parts for an upgraded version of our $1,000 PC for a fifth-day bonus build more than a month ago, priced at $1,600.
But wait. We've been sitting on the parts for a month. The Marathon was published last week. What took us so long to get this story done? What did I fail to anticipate? Today's piece is loaded with twists and turns. Grab a cup of coffee or tea, and join us as we put this last System Builder Marathon machine together.
| Q1 2013 $1,600 Alternative PC Components | ||
|---|---|---|
| Processor | Intel Core i5-3570K: 3.4 GHz Base Clock Rate. 3.8 GHz Turbo Boost, Quad-Core, 6 MB Shared L3 Cache | $230 |
| Graphics | 2 x PowerColor PCS+ AX7870 Myst Edition 2GBD5-2DHPPV3E | $520 |
| Motherboard | Gigabyte Z77X-UD4H: LGA 1155, Intel Z77 Express | $170 |
| Memory | Crucial Ballistix Tactical BLT2K4G3D1608ET3LX0: DDR3-1600 C8, 8 GB (2 x 4 GB) | $48 |
| System Drive | Mushkin MKNSSDCR240GB-DX: 240 GB, SATA 6Gb/s SSD | $180 |
| Storage Drive | Western Digital WD1002FAEX: 1 TB, SATA 6Gb/s Hard Drive | $105 |
| Optical | Asus DRW-24B1ST: 24x DVD±R, 12x DVD±R DL | $20 |
| Case | Fractal Design Define R4 w/Window | $120 |
| Power | Corsair HX750: 750 W Modular, ATX12V v2.3, 80 PLUS Gold | $120 |
| CPU Cooler | Noctua NH-D14 | $81 |
| Total Cost | $1,594 | |
The prices in that table were what we paid when the parts were ordered, and a lot of them changed over the last six weeks. For example, the PowerColor card is $20 less, per board. Other prices are up. All told, then, the total cost of buying our machine and replicating the build is within $20 of our original invoice.
It’s makin’ me late…
CPU: Intel Core i5-3570K
Today’s machine is an upgrade to my previous $1,000 build. The one thing I didn't really feel compelled to upgrade, though, was its Core i5-3570K. At least, I didn’t think it needed an upgrade until I read Don's recent Crysis 3 coverage (Crysis 3 Performance, Benchmarked On 16 Graphics Cards). Moving on...

Read Customer Reviews of Intel's Core i5-3570K
A mere 100 MHz slower than the Core i7-3770K, Intel's Core i5-3570K falls only slightly behind in most benchmarks. A few metrics that would have exploited Hyper-Threading also reflect slightly larger losses. However, overall value favors this less-expensive part.
CPU Cooler: Noctua NH-D14
Noctua’s NH-D14 received last year’s highest award after maintaining its top performance standing for two years in a row. We didn’t really need a cooler this big for a CPU this small, but its reduced noise will improve the experience of this system’s eventual winner.

Read Customer Reviews of Noctua's NH-D14
Excessive cooling can’t be a bad thing, right? We can always hope that this cooler gives our processor the extra few MHz we previously needed to cross the 4.5 GHz barrier.
Memory: Crucial Ballistix Tactical 8 GB (2 x 4 GB) DDR3-1600
Also found in my original build, Crucial’s BLT2K4G3D1608ET3LX0 memory kit is a fallback part. It wasn’t our first choice because, on the day we placed our order, I wasn’t certain of its overclockability. Only after building that $1,000 machine did this memory’s superiority become clear.

Read Customer Reviews of Crucial's BLT2K4G3D1608ET3LXO 8 GB Kit
But superiority wasn’t the reason we altered our original order. Instead, we were forced to give up on higher-voltage parts after suffering memory controller degradation on two E1-stepping Core i5s in a row. Component damage occurred over the period of days, and that delay was a big reason for this article’s late publication. I’ll provide more details throughout today’s build.
It’s keepin’ me waitin’…
Motherboard: Gigabyte Z77X-UD4H
After initially considering the Z77-HD4 for an early project, I had grown dubious of the voltage regulator previously found on my problematic Z77X-D3H-based machine. But when the motherboard we initially picked for this build failed as well, I had to circle back, look closer, and search for issues responsible for killing the parts we were buying.

Read Customer Reviews of Gigabyte's Z77X-UD4H
Instead of the UD3’s tiny five-phase regulator, we instead find a 12-phase design that’s a closer match to Gigabyte's venerable UP5 TH implementation. With questions of capacity out of the way and an advanced feature set worthy of its higher price, we were happy to make the switch. We’ll discuss why our initially-chosen Z77 Extreme4 was later replaced shortly.
Video Card: 2 x PowerColor PCS+ AX7870 Myst Edition
If one Tahiti-LE-based graphics card is good, then two must be better, or so goes the idea behind the GPU performance upgrade in today’s widely-upgraded build.

Read Customer Reviews of PowerColor's PCS+ AX7870 Myst Edition
Axial fans blow GPU heat into our case, and dual GPUs compound the problem of that heat rising into the CPU cooler. Our oversized Noctua heat sink will maintain our overclocking goals in spite of this problem, allowing us to use this pair of top-value cards, even though we'd typically frown on such a thermal configuration.
Power Supply: Corsair HX750
I understand the difference between input and output power, yet still had no idea that our $1,000 PC’s 520 W power supply would prove vastly overkill for a system that topped out around 340 W (calculated output at 85% efficiency).
Those results technically leave enough capacity to power a second graphics card. However, the Antec PSU I was using doesn't have enough connectors to drive my second PowerColor board.

Read Customer Reviews of Corsair's HX750
The good news is that I got excess capacity from Corsair HX750 for free, at least compared to other 80 PLUS Gold-rated units of similar quality. Seasonic’s X650 Gold, for example, was listed at the same price. And smaller high-end units didn’t give me enough auxiliary PCI Express connectors.
Just Chasing After Some Finer Day…
Case: Fractal Design Define R4
We pick our own components, but those selections are subject to availability, just like they are for you in the real world. That explains why every Tahiti LE-based graphics card in our System Builder Marathon comes from PowerColor, and why I went with the windowed version of the award-winning Fractal Design Define R4.

Read Customer Reviews of Fractal Design's Define R4
Many readers love windows. But my goal was to achieve the optimal balance of quietness and cooling. The window on this chassis eliminates one of the noise-dampening mats we'd normally get in the non-windowed version. Fortunately, quiet fans and optimized vent placement mean that this Fractal Design case is still going to be pleasantly silent.
SSD: Mushkin Chronos Deluxe MKNSSDCR240GB-DX

Read Customer Reviews of Mushkin's 240 GB MKNSSDCR240GB-DX
The 240 GB Chronos Deluxe served as the sole drive in my $1,000 performance-oriented PC. But most folks also need a data drive, too. Our SSD will host the operating system and applications, while documents, music, and movies are stored elsewhere.
Hard Drive: Western Digital WD1002FAEX

Read Customer Reviews of Western Digital's WD1002FAEX Hard Drive
Western Digital’s Caviar Black 1 TB drive was an easy choice for me because it wasn't really mine. We've had a number of readers ask for Western Digital's Black-series drives for holding their user data, and of course we're willing to oblige.
This move became even easier when Western Digital updated its drive configuration from the former 32 MB cache and SATA 3Gb/s interface (FALS) to the current 64 MB, SATA 6Gb/s (FAEX) version.
Optical Drive: Asus DRW-24B1ST

Read Customer Reviews of Asus' DRW-24B1ST
I have low expectations of DVD writers; I only want to believe that on the rare occasion I use mine, it’s going to work. Asus’ DRW-24B1ST gets top ratings from Newegg customers, I’ve never experienced a failed Asus drive, and this unit even gives us the bonus of 24x writes.
We never know about the days to come...
My plan started off simply: starting with the $1,000 PC, I’d add a storage drive, a second graphics card, and the higher-quality case I knew our readers would demand for an extra $600. And, knowing that the graphics cards would vent heat into my case of choice, I even upsized the CPU cooler.

The Noctua NH-D14’s under-socket support plate has a removable foam section that should only be left in place when it's attached to boards that don’t have a plate there already, namely LGA 775. It must be removed before attaching to our LGA 1155 board, or else the excess pressure could warp the socket.

Two notches in the support plate provide clearance for the LGA 1155 mounting bracket's support screws. These must be aligned to the screws, or else the cooler will sit crooked, applying uneven pressure against the CPU socket.
Two cross braces sit above the board, secured to the support plate’s screws via Philips-headed cap nuts. Included plastic spacers must be installed first over each screw to prevent cross-brace contact with motherboard components and/or over-tightening.

The heat sink includes two factory-installed, spring-loaded screws. These must be tightened a little at a time, side-to-side, to assure even contact pressure.
The memory also has to be installed before the fans are clipped into place, since a fan covers the slots. But isn't the memory and motherboard in the picture above wrong? Didn't my parts list include different components?
And tomorrow we might not be together…
A little more than two days passed before my E1-stepping Core i5-3570K suffered a severe loss in memory stability. I spent the rest of the third day trying to figure out why the system was crashing all of the sudden.
It seamed that as soon as Windows finished its updates, the system was no longer stable. I reloaded everything on the Mushkin SSD twice before getting it stable with none of Microsoft's automatic updates. Day four rolled along and I started having trouble with the graphics cards. Uninstalling one of the two boards helped a little, and reloading the graphics driver once again appeared to fix all my stability problems. Until it didn’t.
I pulled the CPU and put it into a known-stable system, only to discover that there were no stable settings for it. Though I’ve never seen a bad CPU right out-of-the-box, this one appeared to be my first. So, I replaced it.
And then the USB 3.0 controller died. Could a bad motherboard have killed our precious processor? Bad boards happen. I had started this build with ASRock’s Z77 Extreme4, so I reached into my pile of Z77 Extreme4s and pulled out a replacement. That one worked fine for about one day, and then started throwing memory errors. The first motherboard was surely bad, but could the CPU’s damage have been caused by a bad setting?
ASRock defaults DDR3-2133 memory to 1.5665 V. Add in the 1% over-voltage we’ve been seeing nearly every board vendor sneak in over the past several product cycles, and we come up to around 1.68 V. I wasn't going to risk any more hardware to pull an exact reading. Instead, I switched to the low-voltage RAM from our previous build, along with its not-fried CPU.
Our memory errors went away, temporarily. But the second CPU was suffering more issues with its memory controller. It had been used with those potentially-problematic voltage settings, so the chances appeared good that its memory controller was just a little less damaged than its predecessor.
A fresh board, a fresh CPU, and fresh RAM later, and all was well. Well, for about one more day. Very minor stability issues started to occur on the third motherboard sample. Then another thought struck: could the CPU cooler be warping boards?
The motherboard was already lying on the bench, so I removed the CPU cooler, loosened its brackets, and then set the CPU cooler atop the CPU. Stability returned. Even though I followed all of the italicized must-dos on the previous page, the Z77 Extreme4 appeared to be flexing too far under the force of Noctua’s installation bracket.
With no Z77 Extreme4s left to continue this fiasco, I moved on to Gigabyte’s similarly-thick Z77X-UD4H. Though the board didn’t feel any stiffer, it also didn’t throw errors after taking on the load of Noctua’s installation kit.
These are the good old days…
Twelve years have passed since the last time I was forced to chase down such a series of catastrophic events, where a board damaged a processor that in turn damaged every board it touched, which in turn would damage every processor it touched. Back then, I had to pay for that troubleshooting myself. This time, I only had to explain why my two-week paycheck hadn't yielded any content.
Switching to a motherboard that held-up under the load of Noctua’s NH-D14 installation kit, and a memory set that wouldn’t damage the apparently-fragile E1-stepping Core i5-3570K’s memory controller, I moved on to re-build the system.

Fractal Design’s trays support 2.5” drives by direct mounting, and 3.5” drives through grommets and included shoulder screws. The shoulder screws prevent grommet crushing, and the grommets prevent the transmission of vibration from mechanical drives to the drive cage.

We probably don’t need to re-shoot the CPU cooler installation, since the same method was used a second time on the Gigabyte motherboard. The big difference is that Gigabyte supplies only one CPU fan header, so we had to include the NH-D14’s fan power splitter in our installation.

Finally, a working system that doesn’t slowly die after a couple of days!
Dare I overclock the system that I spent days repeatedly breaking at stock settings? I had to think long and hard about it, but of course I wasn't going to leave this thing at its default performance level.
My goal this time was to repeat the previous $1,000 build’s clock rates, and I credit the good airflow of Fractal Design’s case for getting PowerColor's graphics cards to the same 1,200 MHz core clock and GDDR5-6400. The CPU wasn’t as friendly, unfortunately.


I eventually got the processor to 4.40 GHz, and I didn’t encounter the voltage regulator limits of our previous Gigabyte-based endeavor, but I still had to go past my target 1.280 V to achieve stability.


The 1.305 V core setting returns 1.30 volts, while the “Extreme” Load-Line Calibration setting is needed to maintain that voltage level. Voltage actually climbs slightly to 1.308 V when the system remains loaded for long periods of time, but we have a powerful CPU cooler to deal with its extra heat.


The memory still needed 1.60 V to reach DDR3-2133 at its default CAS 9-9-9-24 timings. We used “XMP” mode for the system’s stock configuration, with Profile 2 providing DDR3-1600 CAS 8-8-8-24 at 1.50 V.

| Test Hardware Configurations | ||
|---|---|---|
| $1,600 Alternative PC | Q1 2013 $1,000 PC | |
| Processor (Overclock) | Intel Core i5-3570K 3.40 GHz, Four Physical Cores O/C to 4.40 GHz, 1.30 V | Intel Core i5-3570K 3.40 GHz, Four Physical Cores O/C to 4.40 GHz, 1.28 V |
| Graphics (Overclock) | 2 x PowerColor 2GBD5-2DHPPV3E: 975 MHz GPU, GDDR5-6000, CrossFire O/C to 1,200 MHz GDDR5-6400 | PowerColor 2GBD5-2DHPPV3E: 975 MHz GPU, GDDR5-6000 O/C to 1,200 MHz GDDR5-6400 |
| Memory (Overclock) | 8 GB Crucial DDR3-1600 CAS 8-8-8-24, O/C at 1.50 V to DDR3-2133 CL 9-9-9-24 | 8 GB Crucial DDR3-1600 CAS 8-8-8-24, O/C at 1.50 V to DDR3-2133 CL 9-9-9-24 |
| Motherboard (Overclock) | Gigabyte GA-Z77X-UD4H: LGA 1155, Intel Z77 Express Stock 100 MHz BCLK | ASRock Z77 Extreme4: LGA 1155, Intel Z77 Express Stock 100 MHz BCLK |
| Optical | Asus DRW-24B1ST: 24x DVD±R | Lite-On iHAS124: 24x DVD±R |
| Case | Fractal Design Define R4 | Rosewill Redbone U3 |
| CPU Cooler | Noctua NH-D14 | Cooler Master Hyper 212 Plus |
| Hard Drive | Mushkin Chronos Deluxe DX 240 GB, SATA 6Gb/s SSD | Mushkin Chronos Deluxe DX 240 GB, SATA 6Gb/s SSD |
| Power | Corsair HX750: 750 W Modular, ATX12V v2.3, 80 PLUS Gold | Antec Neo Eco 520C: ATX12V v2.3, 80 PLUS |
| Software | ||
| OS | Microsoft Windows 8 Pro x64 | |
| Graphics | AMD Catalyst 13.2 Beta 6 | AMD Catalyst 13.1 |
| Chipset | Intel INF 9.3.0.1026 | Intel INF 9.3.0.1026 |
The same software installation that gave me good results with a single card returned OpenCL errors with two cards. An update to Catalyst 13.2 Beta 6 solved those errors.
| 3D Game Benchamarks | |
|---|---|
| Battlefield 3 | Campaign Mode, "Going Hunting" 90-Second Fraps Test Set 1: Medium Quality Defaults (No AA, 4x AF) Test Set 2: Ultra Quality Defaults (4x AA, 16x AF) |
| F1 2012 | Version 1.2, Direct X 11, Built-in Benchmark Test Set 1: High Quality, No AA Test Set 2: Ultra Quality, 8x MSAA |
| Elder Scrolls V: Skyrim | Update 1.5.26, Celedon Aethirborn Level 6, 25-Second Fraps Test Set 1: DX11, High Details No AA, 8x AF, FXAA enabled Test Set 2: DX11, Ultra Details, 8x AA, 16x AF, FXAA enabled |
| Far Cry 3 | V. 1.04, DirectX 11, 50-second Fraps "Amanaki Outpost" Test Set 1: High Quality, No AA, Standard ATC., SSAO Test Set 2: Ultra Quality, 4x MSAA, Enhanced ATC, HDAO |
| Adobe Creative Suite | |
| Adobe After Effects CS6 | Version 11.0.0.378 x64: Create Video which includes three Streams, 210 Frames, Render Multiple Frames Simultaneosly |
| Adobe Photoshop CS6 | Version 13 x64: Filter 15.7 MB TIF Image: Radial Blur, Shape Blur, Median, Polar Coordinates |
| Adobe Premeire Pro CS6 | Version 6.0.0.0, 6.61 GB MXF Project to H.264 to H.264 Blu-ray, Output 1920x1080, Maximum Quality |
| Audio/Video Encoding | |
| iTunes | Version 10.4.1.10 x64: Audio CD (Terminator II SE), 53 minutes, default AAC format |
| Lame MP3 | Version 3.98.3: Audio CD "Terminator II SE", 53 min, convert WAV to MP3 audio format, Command: -b 160 --nores (160 Kb/s) |
| HandBrake CLI | Version: 0.98: Video from Canon Eos 7D (1920x1080, 25 FPS) 1 Minutes 22 Seconds Audio: PCM-S16, 48,000 Hz, Two-Channel, to Video: AVC1 Audio: AAC (High Profile) |
| TotalCode Studio 2.5 | Version: 2.5.0.10677: MPEG-2 to H.264, MainConcept H.264/AVC Codec, 28 sec HDTV 1920x1080 (MPEG-2), Audio: MPEG-2 (44.1 kHz, Two-Channel, 16-Bit, 224 Kb/s), Codec: H.264 Pro, Mode: PAL 50i (25 FPS), Profile: H.264 BD HDMV |
| Productivity | |
| ABBYY FineReader | Version 10.0.102.95: Read PDF save to Doc, Source: Political Economy (J. Broadhurst 1842) 111 Pages |
| Adobe Acrobat X | Version 10.0.0.396: Print PDF from 115-Page PowerPoint, 128-bit RC4 Encryption |
| Autodesk 3ds Max 2012 | Version 14.0 x64: Space Flyby Mentalray, 248 Frames, 1440x1080 |
| Blender | Version: 2.64a, Cycles Engine, Syntax blender -b thg.blend -f 1, 1920x1080, 8x Anti-Aliasing, Render THG.blend frame 1 |
| Visual Studio 2010 | Version 10.0, Compile Google Chrome, Scripted |
| File Compression | |
| WinZip | Version 17.0 Pro: THG-Workload (1.3 GB) to ZIP, command line switches "-a -ez -p -r" |
| WinRAR | Version 4.2: THG-Workload (1.3 GB) to RAR, command line switches "winrar a -r -m3" |
| 7-Zip | Version 9.28: THG-Workload (1.3 GB) to .7z, command line switches "a -t7z -r -m0=LZMA2 -mx=5" |
| Synthetic Benchmarks and Settings | |
| 3DMark 11 | Version: 1.0.1.0, Benchmark Only |
| PCMark 7 | Version: 1.0.4 x64, System, Productivity, Hard Disk Drive benchmarks |
| SiSoftware Sandra 2011 | Version Version 2013.01.19.11, CPU Test = CPU Arithmetic / Cryptography, Memory Test = Bandwidth Benchmark |
3DMark gets a 45% boost from the $1,600 PC’s second GPU, but its scores get even more interesting when we overclock. Tuning our machines boosts the $1,000 build by a mere 20%, but pushes the $1,600 PC even higher, giving us more bang from the same overclock.
A CPU bottleneck limiting the second GPU might help explain that disparity; processor overclocking would help open up that bottleneck. On the other hand, all GPUs are overclocked by a greater percentage than the CPUs. We’re probably seeing benefits from a combination of factors, including a faster CPU configuration and the $1,600 PC's beta graphics driver, which was added to solve an issue with OpenCL benchmarks in CrossFire.



PCMark also benefits from the $1,600 PC’s CrossFire configuration. Though the alternative motherboard’s firmware might also have been responsible for a boost in SSD performance, storage scores are scarcely affected.
The $1,600 PC’s CrossFire configuration bumps up against Battlefield 3’s 200 FPS limit at 1680x1050 and 1920x1080. Naturally, though, you wouldn't buy this much graphics horsepower unless a single 2560x1600 display or three 1920x1080 screens were your goal. The 200 FPS limit certainly hurts the alternative system’s overall performance advantage, but a separate chart will show off its greater-than-90% gain at 5760x1080.
Cranking the detail levels up to Ultra quality, we find that just one Tahiti LE-based card isn't capable of achieving playable frame rates at 5760x1080 or, thanks to Battlefield 3’s high frame rate variability, 2560x1600.


Leaning hard on CrossFire yields playable Far Cry 3 frame rates at 5760x1080, but only at our lower test settings. Dialing up to the Ultra quality preset drops the $1,600 PC’s playability to 2560x1600, though you might be able to get similar performance at 4800x900.
The $1,000 PC is primarily limited to 1920x1080, High quality details, and no extra graphics features in this title.


F1 2012’s proven memory bottleneck appears to give way to CPU overhead in the CrossFire-equipped $1,600 PC, where the extra card monopolizes additional CPU resources. The expensive system shines at 5760x1080, but the cheaper single-card solution proves itself more than adequate at the same settings.
An upgrade to stereo-capable displays would appear to be the only reason to splash out for an extra card in this title.


The $1,600 PC’s extra graphics power pushes the system's CPU limits in Skyrim at anything less than Eyefinity-inspired resolutions. The single-card $1,000 PC remains adequate, if barely, through 5760x1080.


The Tom’s Hardware team normally groups application benchmarks by type, but only OpenCL-enhanced Photoshop stood out as a test that might potentially benefit from the $1,600 PC’s second graphics card. We opened up a GPU monitoring program and watched the needle jump between zero and 14% utilization throughout this test.
Visual Studio 2010 also saw moderate performance gains on the new overclocked system, but those gains can only be linked to storage performance and/or chance. This long bench typically shows up to one-minute variance between consecutive runs.


In my search for noticeable changes in performance, I grouped the non-impacted benchmarks together. If you’d like to see exactly how close these results are, simply click on a photo to expand it.

The $1,600 PC’s second GPU does a great job of spinning down as part of AMD's ZeroCore technology suite. We attribute most of the power differences in non-gaming scenarios to the motherboard and our slightly higher CPU voltage.
When it's subjected to load, the second graphics card adds more than 200 W.

Manual fan adjustments keep the overclocked system’s GPU cooler, though the $1,600 PC’s second GPU pushes the temperature of its first GPU up by 9° Celsius.

Non-overclocked CPU temperature differences are far larger, though that’s the fault of a 90°-max fan slope. A low 18° ambient temp produces the big 70° delta.

Our newest build isn't only designed for gaming; it's a solid general-purpose PC that happens to be an excellent gamer, too. Graphics are simply the easiest place for us to get a return on our investment. Today's $1,600 setup is over 40% more powerful in games, both at stock and overclocked settings. CPU and DRAM bottlenecks at lower resolutions mean that our high-end performance chart (on the next page) will be far more important.

The only place we’ve seen the $1,600 PC pay (in performance) for its power increase is during games, and then only at 5760x1080. When we put all benchmarks on a chart of averages, its efficiency looks fairly poor.
The good news for our pricier build is that overclocking yields larger increases in performance than in power. Average efficiency actually goes up for this particular overclocked configuration.
Today’s $1,600 build started out as an upgrade to my original $1,000 Performance PC, and should have contained only the components needed to make that build perform better. At the end of the day, however, I had to spend a little more on a motherboard that didn't help my performance, but rather addressed a couple of show-stopping flaws. Those issues were revealed when I paired an E1-stepping CPU with 1.65 V RAM, and then added a very heavy cooler with a very high-tension mounting kit.
Because I had problems with both the CPU memory controller and motherboard flex, I can’t be completely sure that Noctua's NH-D14 isn’t responsible for the whole mess. I only know from experience that the big cooler had a negative impact on at least three of my ASRock Z77 Extreme4 motherboards. And I’m not willing to talk about what might have happened to my fourth motherboard sample in my mad rush to find the problem.
I also know that Intel explicitly states that 1.50 V plus or minus 5% is the limit for Ivy Bridge-based processors, while at the same time validating the use of 1.65 V memory. I further know that the Z77 Extreme4 automatically sets 1.665 V for DDR3-2133, that ASRock’s set voltage levels are slightly lower than its actual voltage levels, and that nobody has given me proof of the E1-stepping Core i5’s ability to cope with voltage levels approaching 1.7 V.
In the end, I was forced to either give up the big CPU cooler or try a different motherboard model. In the end, I was forced to scale back to low-voltage RAM. And, in the end, I was left with ten days, two processors, and four motherboards wasted.

All of that work, and I still don’t see any increase in overall value when it comes to the performance-per-dollar charts. I do like that the system is quieter, that its case is more durable, that its power supply has more connectors, and that the extra hard drive provides room for all my old media files.
I also like that it can game at 5760x1080. You see, the $1,000 PC, for all of its performance and overall value hype, simply couldn’t be called a consistent triple-display performer. Most games needed a second GPU to produce smooth 5760x1080 frame rates, and some game settings even required the second card simply to reach 2560x1600.

The only thing that could make the $1,000 PC a universal gaming box is if we limited it to 1920x1080. Real quality-seekers will find real gaming quality in the $1,600 build, which also has a better case, motherboard, CPU cooler, and more storage capacity. Even though we paid money to get those non-gaming improvements, the second card still pushes this system’s gaming value far beyond that of our original $1,000 build.













