This site may earn affiliate commissions from the links on this page. Terms of use.

Maingear Shift

When AMD announced it was going to build a 5GHz version of its FX-course processors this by summer, reaction from the enthusiast community was mixed. On the i hand, AMD was simultaneously reaching back to its enthusiast roots and offer a chip significantly faster than anything else in its own product stack. On the other, even at 5GHz, the CPU was going to have a difficult fourth dimension competing with Intel'south Haswell and Ivy Span-E. We recently spent some time with a Maingear Shift equipped with the FX-9590, and decided to put the chip through its paces in a caput-to-head gaming showdown confronting Intel's Core i7-4960X.

We've called to focus on gaming for two reasons. First, gaming was always one of the areas where the original FX processors shone, and it's a question that'southward come up several times in our past Piledriver reviews. Multiple readers have asked for a head-to-head gaming article, and the FX-9590's launch, combined with the R9 290X's debut this month, fabricated information technology an ideal time to examine the question of how well Intel and AMD platforms compare. The second reason is that outside the gaming arena, the FX-9590 admittedly struggles. While information technology runs at a higher clock speed, it still has trouble in lightly-threaded workloads where Haswell and Ivy Bridge can rely on better unmarried-threaded performance and superior scaling across 2-4 cores.

Maingear Shift

Price, availability, and the Intel Ivy Bridge-E

When I began this story, the FX-9590 was merely available from OEMs and priced at $800. Frame pacing driver issues delayed my testing, equally did multiple GPU launches from AMD throughout Oct. The toll gap that's opened up between the Ivy Bridge-Eastward and the FX-9590 means that the $grand Intel CPU is no longer the best betoken of comparison. This is taken into consideration in the terminal assay. On the other hand, the delay immune usa to add functioning figures for the R9 290X, and compare it against the Radeon 7990 that the Shift shipped with by default.

Speaking of the Shift, Maingear's vertical enclosure with a custom Rosso Scuderia paint job is a beautifully designed system that ships with a custom 180mm Maingear Epic CPU libation and as tight a cabling job as yous could enquire for. In the significant corporeality of time nosotros spent with the rig, nosotros had no problems — no lockups, no crashes, no hardware issues of any kind. Obviously when you pay meridian dollar for a boutique system, you expect this kind of custom equipment and attention to detail, but we were well pleased with the system's configuration.

This article is primarily focused on the FX-9590 CPU — for full details on the Shift itself, striking upwardly our PC Magazine review of the Maingear Shift. Our high opinion of the Shift didn't change after spending significantly more time with information technology. We built an Intel Core i7-4960X Ivy Span-East comparison system using 16GB of low-latency Mushkin DDR3-2133, an identical SSD (Samsung 840 Evo), and tested both systems on a Radeon 7990 and Radeon 290X.

Each game was tested at 1920×1080, but we varied the visual settings somewhat depending on the game in question. We'll pause these down equally we get. The goal was to preserve a testing surroundings that wouldn't exist completely GPU-jump, merely to exam games at graphics levels that were representative of how high-end gamers would configure each system.

We've also examined performance in terms of both frame rates and frame latency using the handy open-source tool FRAFS, which analyzes Fraps output. Frame timing is a metric used to examine latency and stutter in a game, and information technology's an expanse where AMD's Radeon cards had a rough time of it early this yr. (Encounter: Later nigh 20 years, GPU benchmarking is moving past frames per second.) AMD rolled out a frame-pacing driver this summer and has been updating it regularly since then, so we can besides await and see what kind of performance difference it makes as compared to the single-GPU Radeon R9 290X. To exercise this, nosotros've included a graph of the worst 1% of frames we tested. This is a lower jump for frame latency.

Here's the rule of thumb when it comes to evaluating frame latency figures. Lower is amend than higher, 16.7ms is the purlieus for 60 fps gaming, and 33.3ms is the 30 fps mark. Merely — and this is key — the differences thing more, the college the figure. The difference between 12ms and 16ms is much less noticeable than the gap between 25ms and 33ms — fifty-fifty though the difference is just 1.33x.  Frame drops below 30 fps are more noticeable than sixty fps, and annihilation that pushes you into the teens or below ten fps is especially egregious.

Next page: Gaming benchmarks

  • 1 of three