Cooling, Noise, and Power

For my last batch of testing, I like to focus on the more card and cooler specific benchmarks. So I take a look at things like power usage, cooling performance, and noise. All issues that very a lot from manufacture to manufacture, even when the cards are the same GPU. I start with the power usage. To test this, I use a Kill-A-Watt to see our test bench’s overall wattage and I run through Valley Benchmark and we use the highest peak wattage pulled in a realistic benchmark (aka not a power hog like furmark). The RX 470 at its core is a 120-watt TDP card just like the GTX 1060, but with this being an overclocked card it’s hard to say where things will land. So the XFX card actually ended up pulling even more than I would have ever expected. The GTX 1060, for example, pulled 188 watts with the same TDP and the reference RX 480 pulled 242. Well the overclocked RX 470 pulled 239, being just three watts behind the stock 8GB RX 480. I don’t normally mind a little extra power draw, but it does make me question the decision to stick with the 6-pin power connection.

Next, I took a look at overall noise output of the XFX Double Dissipation cooler on the Black Edition. Or if you like initialisms I tested the DD cooler on the XFX RX 470 RS BE. I test using a basic decibel meter sitting 4 inches away from the fans. These results, being so close and on an open air test bench, aren’t going to show you what you will experience with the card in a case, but they do help compare from card to card. Anyhow, I test at 100% fan speed and again at 50% fan speed. 50% fan speed is a better representation of what you will experience in normal use. The XFX, in fact, would run at 54% in my thermal testing. So with the fans turned all the way up the XFX was a little on the loud side but at least came in below the triple fan Gigabyte card. The 50% fan speed results were similar but the XFX was a touch louder here as well.

For the last test, I want to see what you will experience for cooling performance and then I test the max cooling performance of the XFX RX 470 as well. To do this I loop in Valley Benchmark until the card levels off to a peak temperature with the fan settings untouched and then again with the fans at 100% fan speed. At stock settings the XFX came in a touch higher than the 1070’s with the HUGE coolers like the Gigabyte and MSI cards but below the reference RX 480 by a large margin. When I turned the speeds up the XFX came in the middle of the pack. The reference RX 480 cooler was only a few degrees away. This tells us that XFX is running a more aggressive cooling profile than the reference design but the overall cooling power of the cooler is only slightly better. Given the fans having gaps with no heatsink under them at all on the cooler, I’m not all that surprised. That said the results are still good, but I will be curious to see how they hold up compared to other manufacturers.