Previous AMD RX 480 Content

• AMD RX 480 8GB Review

• AMD RX 480 8GB vs. 4GB

• AMD RX 480 Hybrid Mod & Overclocking

AMD RX 460, RX 470, & RX 480 Specs

Polaris 10 vs. Polaris 11 Specs & Architecture

When Do Multiple GPUs Do Well? When Don't They?

Multi-GPU configurations come in two primary modes: LDA, or Linked Display Adapter, and MDA, or Multi-Display Adapter. LDA is more traditional and relies upon a bridge, while MDA (exposed best through new APIs) can enable non-same GPUs to be paired together, and uses the PCIe bus for communication between cards.

For the most part, modern games do not yet expose multiple non-same GPUs through their programming or through their APIs – which still see large representation by Dx11. Ashes of the Singularity is among the only games that can actually leverage MDA and see positive results, but it took a year or more of near-negative scaling to Dx12.

Multiple GPUs tend to do well in environments where frames are more independent; that is to say, when the GPU isn't instructed to perform temporal (frame-to-frame) analysis of the graphics it's rendering. Interdependent frames hurt the ability for AFR (Alternate Frame Rendering) to work properly, and can actually cause negative scaling in some instances. That'd be where you want to disable a GPU to make things work better, which is never a good feeling considering it turns into an expensive brick.

Interdependence during frame rendering is mostly caused by post processing effects, or Post FX. Post FX are well represented in Mirror's Edge Catalyst, which makes heavy use of high dynamic range rendering (HDR), bloom, lens flares and motion blur, ambient occlusion, and plenty of other effects (like crepuscular rays, or “godrays”). These are computationally intensive processes that don't play well with multi-GPU configurations and hog cycles, which (one frame to the next) can cause a slowdown on one or both of the GPUs in the system. If one card gets hit more heavily and slows on its frame delivery, we start seeing frame latency gaps that become noticeable and cause “stuttering” or apparent frame drops. This latency between frames is represented by 1% low and 0.1% low metrics in our benchmarks.

Mirror's Edge Catalyst also makes a lot of use of filtration effects and texture filtration, both drawing big on the VRAM pool. This is shown in our 4GB vs. 8GB RX 480 benchmark.

Games that don't rely quite as heavily on Post FX will run better with AFR than those that do. This is why we've had a history of stating that SLI and CrossFire are both still hit and miss, and that the efficacy of such a setup hinges upon the type of game played.

Game Test Methodology

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

The latest AMD drivers (16.6.2 RX 480 press) were used for testing. We used 16.7.1 for GTA V & for power draw. NVidia's 368.69 drivers were used for game (FPS) testing. Game settings were manually controlled for the DUT. All games were run at presets defined in their respective charts. We disable brand-supported technologies in games, like The Witcher 3's HairWorks and HBAO. All other game settings are defined in respective game benchmarks, which we publish separately from GPU reviews. Our test courses, in the event manual testing is executed, are also uploaded within that content. This allows others to replicate our results by studying our bench courses. In AMD Radeon Settings, we disable all AMD "optimization" of graphics settings, e.g. filtration, tessellation, and AA techniques. This is to ensure that games are compared as "apples to apples" graphics output. We leave the application in control of its graphics, rather than the IHV.

Windows 10-64 build 10586 was used for testing.

Each game was tested for 30 seconds in an identical scenario, then repeated three times for parity. Some games have multiple settings or APIs under test, leaving our test matrix to look something like this: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.

For Dx12 and Vulkan API testing, we use built-in benchmark tools and rely upon log generation for our metrics. That data is reported at the engine level.

Video Cards Tested

• AMD RX 480 8GB ($240)
• NVIDIA GTX 1080 Founders Edition ($700)
• NVIDIA GTX 980 Ti Reference ($650)
• NVIDIA GTX 980 Reference ($460)
• NVIDIA GTX 980 2x SLI Reference ($920)
• AMD R9 Fury X 4GB HBM ($630)
• AMD MSI R9 390X 8GB ($460)

Thermal Test Methodology

We strongly believe that our thermal testing methodology is the best on this side of the tech-media industry. We've validated our testing methodology with thermal chambers and have proven near-perfect accuracy of results.

Conducting thermal tests requires careful measurement of temperatures in the surrounding environment. We control for ambient by constantly measuring temperatures with K-Type thermocouples and infrared readers. We then produce charts using a Delta T(emperature) over Ambient value. This value subtracts the thermo-logged ambient value from the measured diode temperatures, producing a delta report of thermals. AIDA64 is used for logging thermals of silicon components, including the GPU diode. We additionally log core utilization and frequencies to ensure all components are firing as expected. Voltage levels are measured in addition to fan speeds, frequencies, and thermals. GPU-Z is deployed for redundancy and validation against AIDA64.

All open bench fans are configured to their maximum speed and connected straight to the PSU. This ensures minimal variance when testing, as automatically controlled fan speeds will reduce reliability of benchmarking. The CPU fan is set to use a custom fan curve that was devised in-house after a series of testing. We use a custom-built open air bench that mounts the CPU radiator out of the way of the airflow channels influencing the GPU, so the CPU heat is dumped where it will have no measurable impact on GPU temperatures.

We use an AMPROBE multi-diode thermocouple reader to log ambient actively. This ambient measurement is used to monitor fluctuations and is subtracted from absolute GPU diode readings to produce a delta value. For these tests, we configured the thermocouple reader's logging interval to 1s, matching the logging interval of GPU-Z and AIDA64. Data is calculated using a custom, in-house spreadsheet and software solution.

Endurance tests are conducted for new architectures or devices of particular interest, like the GTX 1080, R9 Fury X, or GTX 980 Ti Hybrid from EVGA. These endurance tests report temperature versus frequency (sometimes versus FPS), providing a look at how cards interact in real-world gaming scenarios over extended periods of time. Because benchmarks do not inherently burn-in a card for a reasonable play period, we use this test method as a net to isolate and discover issues of thermal throttling or frequency tolerance to temperature.

Our test starts with a two-minute idle period to gauge non-gaming performance. A script automatically triggers the beginning of a GPU-intensive benchmark running MSI Kombustor – Titan Lakes for 1080s. Because we use an in-house script, we are able to perfectly execute and align our tests between passes.

Noise Testing Methodology

Our noise testing methodology is new and still being revised, but has been kept consistent across all tests contained herein. We test noise in a real-world environment and do not presently use an anechoic chamber. The results align with what consumers will encounter in their own rooms.

We use a REED logging dB meter mounted to a tripod, whose mic is positioned 20” from the face of the GPU (mounted in an open bench). The REED meter is approximately 6” above the bench. All open bench fans are disabled. The Kraken X41 CPU cooling fan is configured to its “silent” mode, minimizing its noise output to be effectively imperceptible.

A noise floor measurement is taken prior to each test's execution to determine ambient without any systems running in the room. We then take an idle measurement (GPU & CPU at idle). Our noise floor has a fluctuation of approximately +/-0.6dB.

Noise levels are logarithmic, and are therefore not as simple to perform delta calculations as thermals or framerates. Noise percent differences are calculated using dB=20*log(V2/V1) (where V is amplitude). You cannot perform a simple percent difference calculation to determine the delta. For an example, a 10dB range (50dB vs. 40dB) is not equal to a 22% delta.

After the noise floor is determined, we log idle fan dB, 50% speed dB, and 100% speed dB (configured in Afterburner). We also measure auto fan dB at an identical stepping for every test; we do this by running Kombustor for exactly 5 minutes prior to beginning dB logging, which is useful for fans which use two push fans. Some dual-push fan cards will only trigger the second fan if the VRM is under load.

Power Testing Methodology

Power consumption is measured at the system level. You can read a full power consumption guide and watt requirements here. When reading power consumption charts, do not read them as a GPU-specific requirements – this is a system-level power draw.

Power draw is measured during a FireStrike Extreme - GFX2 run. We are currently rebuilding our power benchmark.

For regular readers: We have begun the process of migrating off of our Z97 test platform for power. This means that the new power chart is scarcer, but on the same platform as everything else.

Continue to the next page for thermal benchmarks, noise, and power!