Explanation of Tests

For this article, we decided to test several configurations of the Radeon RX 5700 and benchmark each to highlight the performance that can be gained by the changes possible on this card.

• "RX 5700 - Stock": This data point is the unmodified AMD reference card running at everything default, like it would behave after a clean driver install.
• "RX 5700 - Max Pwr & OC": For this test, we increased the "Power Limit" slider in Wattman to its maximum, and also dialed in the highest clocks we could set in Wattman. The problem here is that AMD has implemented limits; the sliders simply end at some point (+20% Power, 1850 MHz GPU, 1.2 V, 930 MHz memory). These limits are considerably lower than on the RX 5700 XT, and every card should be able to run higher than those limits, but there is simply no way to go higher in Wattman.
• "RX 5700 - Flashed to RX 5700 XT": This data point represents the RX 5700 after flashing the XT BIOS. Everything in Wattman is left at default.
• "RX 5700 - Flashed + Max Power Limit": Next, we increased the power limit to the maximum setting (+50%); all clocks are at default. Please note that the XT BIOS has a maximum power limit of +50%, whereas the non-XT BIOS is limited +20%. This is on top of the already increased TDP limits stored in the BIOS (150 W vs 180 W).
• "RX 5700 - Flashed + Max Pwr & OC": This round of benchmarks represents the maximum performance we could achieve in Wattman after the BIOS flash. The power limit is set to its maximum, and the clocks and voltages have been dialed into the highest possible stable setting. We also dropped the voltages a bit to ensure we're not running into the various thermal limits. Otherwise, the card would throttle, dropping performance almost down to RX 5700 pre-flashed levels.

For all tests we used the default fan curve as defined by AMD. What's important here is that both BIOSes are configured to limit fan speed to 2100 RPM—no matter how hot the card gets. This means that noise levels are identical in all tests, with the difference being temperatures (or throttling if temperatures go too high).

Test System

Benchmark scores in other reviews are only comparable when this exact same configuration is used.

• All games and cards are tested with the drivers listed above
• All games are tested using the same game version.
• All games are set to their highest quality setting unless indicated otherwise.
• AA and AF are applied via in-game settings, not via the driver's control panel.
• Before starting measurements, we heat up the card for each test, to ensure a steady state is tested. This ensures that the card won't boost to unrealistically high clocks for only few seconds until it heats up.

Each game is tested at these screen resolutions:

• 1920x1080: Most popular monitor resolution.
• 2560x1440: Intermediary resolution between Full HD and 4K, with reasonable performance requirements.
• 3840x2160: 4K Ultra HD resolution, available on the latest high-end monitors.