Cooling, Noise, and Power
My last round of testing is also one of the most important when you have already decided which GPU you want. Once you know you want to go with an RTX 2070 SUPER you then have to decide which of the MANY models will fit you best. Size and aesthetic issues aside, the best way for cards to set themselves apart is with overclocks or with better or quieter cooling. Having already tested two other RTX 2070 SUPERs this is a great situation where we can see how they all compare.
First, though I wanted to check out the power usage. For this, I ran two tests. On both, I monitored the power usage of our test system using a Kill-A-Watt. To put things under load I first used 3DMark to replicate a gaming load on both the CPU and GPU. Here the test bench with the Gaming X pulled 383 watts which was the highest of the three 2070 SUPERs. In fact it was 9 watts more than the Gaming OC with its slightly higher clock speed which makes me think MSI has gone with a safer voltage for the overclock. The GPU only AIDA64 stress test showed similar results with the Gaming X pulling 262 watts over 253 for the Gaming OC. This also puts the power that the two Gigabyte 2060 SUPERs pulled in perspective as well.
For noise testing, I did a few different tests. Both I used our decibel meter 18 inches away from the bench. I tested for noise with fixed fan speeds set at 50% and at 100% fan speed. These two results give an idea of the noise range that you can expect. Most cards don’t run anywhere close to 100% fan speed but I like to know how noisy a card can get. In this case the Gaming X was quiet at 54.2 decibels and right with the Gigabyte Gaming OC model, both well ahead of the Founders Edition. The Gaming X was quieter than the Gaming OC for the 50% result though. The reason the Gaming X was quieter can be seen in the Fan RPM graph that shows the two huge fans don’t have to run as fast, running at 3671 RPM at full speed. The last test I ran is also the one that looks at real-world performance. This is where I get the Gaming X up to temperature using AIDA64 Stress Test and let that sit for over a half-hour until the card has leveled off and I measure what the fan noise is at. The Gaming X was all the way at the bottom of the chart, again right with the Gaming OC with both at 34.4 decibels.
All of the fans and all of the work designing a heatsink come down to one thing, temperatures. To check this out I ran two tests. Using AIDA64 Stress Test to load up the GPU I let the video card get warmed up and level off over a half hour. One test was with the stock fan profile to see out of box performance and the second was run with 100% fan speed to just get an idea of the max cooling the cooler is capable of. With the stock fan profile, the Gaming X peaked at 64 degrees which was the warmest of the three cards. Even the Founders Edition was two degrees lower. In the 100% fan speed test however the Gaming X swung the other way completely coming in 2 degrees lower than the Gaming OC and being at the bottom of the chart with the exception of the water-cooled Vega 64. Clearly MSI has the card tuned a little for noise but not with a significant temperature cost.
While doing the stock fan testing I also pulled out our Flir to look at the Gaming X thermals. The temperatures don’t have to be too accurate here, I’m, mostly just looking to see if there are any concerning hotspots. The back of the Gaming X was mostly hot at the center behind the GPU itself but you can see up on the top edge where the backplate leaves some PCB exposed and you can see it is hotter and that the backplate isn’t pulling all of that heat out. That was visible also on the top edge photo. The fan side you can see mostly even cooling behind the fans. Down under the card you can see the hotspot where the air coming out of the bottom of the card does still get trapped up against the motherboard.