AMD & MSI Radeon R9 390X Specs

The R9 390X hosts familiar architecture. At the center of it all is AMD's existing Hawaii arch, which debuted on the 290X, albeit with a slight update. Everything I'm about to say has already been covered in our R9 390 & R9 380 reviews.

The R9 390X's version of Hawaii has been dubbed “Grenada,” though the core architecture is identical to Hawaii. Changes of note include AMD's power (and, thus, thermal) optimizations and marginal frequency increase. As with other R300 series video cards, the R9 390X sees a reference clock increase of +50MHz, bringing its stock option to 1050MHz from 1000MHz on the 290X. The clockrate increase impacts other specifications of the card; for instance, texture fill-rate is the product of clockrate and TMUs, resulting in a gain to 184.8GT/s from 176GT/s (reference). Read more about texture fill-rate in our drill-down article.

AMD also increased the memory clock frequency to 1500MHz base, which boosted memory bandwidth to 384GB/s maximum throughput. The R9 290X sat at 320GB/s reference.

It is worth noting that, despite AMD's 4GB HBM solution in Fiji and the Fury X, the R9 390X sticks with 8GB of GDDR5 memory. The R9 290X saw a refresh to 8GB at one point, but has shipped in both 4GB and 8GB models; the 390X is strictly available in its 8GB flavor. Because this is a single-GPU card, that's a true 8GB allocated to a single GPU.

MSI R9 390X Differences

There are a few obvious (but key) differences between the reference and MSI Gaming specs. The most immediate and obvious is the cooling solution. MSI's R9 390X utilizes the company's Twin Frozr V cooler, equipped with two large-diameter fans and idle fan passive cooling abilities. Every company has its own cooler: Sapphire deploys the Vapor, Dual-, and Tri-X coolers; EVGA uses ACX; ASUS uses Strix. For the most part, high-end cooler deployments by these brands doesn't make that massive an impact on the user's experience – a few degrees between them, at best – but is still a key differentiating factor.

Twin Frozr V uses a “Torx” fan, a push fan that reduces turbulence and decreases friction, allowing for more fluid distribution of air. This also decreases noise levels, something all the listed manufacturers have attempted this generation of nVidia and AMD products.

As with nVidia's Maxwell cards, modern AMD coolers allow for passive GPU cooling when an insignificant amount of load is applied. During non-gaming workloads, the Twin Frozr fans will spin-down and relinquish cooling entirely to the (massive) heatsink on the card, which is more than capable of keeping the GPU cool during times of low demand. As may seem obvious, the card is dead silent when passively cooled.

That's the biggest note for this video card – its heatsink is huge. We measured the card at nearly 50mm wide and 110mm high, making it taller than the expansion slot and wider than the usual dual-slot config. The card takes about 2.2 expansion slot spaces. Keep this in mind for dual-card configurations – you'll want a gap between your PCI-e slots. The unit is also prone to sag when mounted vertically, as would be done in most cases, and can dip down toward the far end of the card.

MSI's biggest point of bragging is its cooler, but the company also notes the presence of an LED backlight for the “MSI Gaming” dragon logo. The light can be configured for five different modes through software.

High-end super ferrite chokes are utilized in the VRM design, which operate at 35C lower than reference and, MSI suggests, offer a +30% increased current capacity. Super-efficient capacitors are used across the card (Hi-C Caps) that are rated for 93% efficiency. The use of high-quality VRM and electrical components assists in thermal footprint reduction.

Test Methodology

We tested using our updated 2015 Single-GPU test bench, detailed in the table below. Our thanks to supporting hardware vendors for supplying some of the test components.

The latest AMD Catalyst drivers (15.7.1) were used for testing. NVidia's 355.6 drivers were used for testing. Game settings were manually controlled for the DUT. All games were run at 'ultra' presets, with the exception of The Witcher 3, where we disabled HairWorks completely, disabled AA, and left SSAO on. GRID: Autosport saw custom settings with all lighting enabled. GTA V used two types of settings: Those with Advanced Graphics ("AG") on and those with them off, acting as a VRAM stress test.

Each game was tested for 30 seconds in an identical scenario, then repeated three times for parity.

Average FPS, 1% low, and 0.1% low times are measured. We do not measure maximum or minimum FPS results as we consider these numbers to be pure outliers. Instead, we take an average of the lowest 1% of results (1% low) to show real-world, noticeable dips; we then take an average of the lowest 0.1% of results for severe spikes. Anti-Aliasing was disabled in all tests except GRID: Autosport, which looks significantly better with its default 4xMSAA. HairWorks was disabled where prevalent. Manufacturer-specific technologies were used when present (CHS, PCSS).

Overclocking was performed incrementally using MSI Afterburner and AMD's OverDrive. Parity of overclocks was checked using GPU-Z. Overclocks were applied and tested for five minutes at a time and, if the test passed, would be incremented to the next step. Once a failure was provoked or instability found -- either through flickering / artifacts or through a driver failure -- we stepped-down the OC and ran a 30-minute endurance test using 3DMark's FireStrike Extreme on loop (GFX test 2). 

Thermals and power draw were both measured using our secondary test bench, which we reserve for this purpose. The bench uses the below components. Thermals are measured using AIDA64. We execute an in-house automated script to ensure identical start and end times for the test. 3DMark FireStrike Extreme is executed on loop for 25 minutes and logged. Parity is checked with GPU-Z.

Continue to Page 2 for the gaming benchmarks!