Reason instead of obesity: Gigabyte RX Vega64 Gaming OC in test

If there were enough Vega chips and thus a wide range of cards, the Gigabyte RX Vega64 in this form would probably be the ideal entry-level card, rather than a top-of-the-line model. But it really must always be the demonstration of what is technically feasible... Board layout As with the RX Vega56 Gaming OC, Gigabyte is extremely different from AMD's reference layout, which should also make the use of existing full-cover water coolers impossible. Also at the popular Raijintek Morpheus, the end of G... We have to put it in the first place that we actually had to double-benchmark all the games. At the last moment, Gigabyte incorporated AMD's new kernel from the source BIOS, which accommodates the custom models with a customized power table. D... Power consumption at different loads The power consumption in the gaming loop is at the measured approx. 261 watts in Balanced Mode about 20 watts below what could be measured at the reference in the default BIOS. This is all the more astonishing because... Overclocking and undervolting The conventional overclocking by means of an even higher power limit and more clock counteracts the current cooling concept, because the rather quiet cooler gets along quite well with what it has to dismount ex works. More on the other hand... Cooling system and backplate Of course, the generated waste heat is directly related to the recorded power, for which the cooling solution is responsible for optimum dissipation. And this is precisely where we are faced with various compromises, which are... Summary For the Gigabyte RX Vega56 Gaming OC we have already written a very detailed conclusion, which applies unreservedly to the bigger sister in the same forum. Butter-and-bread instead of lighthouse project and cost optimization for ...

Overclocking and undervolting

The conventional overclocking by means of an even higher power limit and more clock counteracts the current cooling concept, because the rather quiet cooler gets along quite well with what it has to dismount ex works. More, on the other hand, is already quite loud. In our articles "Overclocked and boiled: AMD Radeon RX Vega64 water-cooled" and "AMD Radeon RX Vega64 without temperature limit with interesting findings" has already been explained extensively.

On the other hand, a healthy sub-voltage can achieve much better results. Especially the use of the program OverdriveNTool can produce real miracles here. More or less, because as always it depends on the respective chip quality (and the driver version). However, since these very individual results cannot be flattened, each user would have to test it for himself. Whether the effort is worth it, however, is up to everyone. A good guide for this can be found, e.g. in this tutorial.

Temperatures and clock rates

Why we limit ourselves to the output of the values reported as GPU temperature is because the telemetry of our test sample also uses this value. What it is all about with the sometimes significantly higher hotspot temperature, you can read here again: "AMD Radeon RX Vega64 without temperature limit with interesting findings". We were able to log values up to 15 degrees higher on the map, but they seemed a bit borderline in their height if you take advantage of all the possibilities of the map.

We now tabulate the achieved start and end values for temperatures and GPU clock (boost), but also point out that these clock rates can be much higher in loops with significantly lower load. In some scenes of Wolfenstein 2 (indoor), the map reaches 1.5 GHz, only to break in the open terrain again.

	Initial	Final value
Open Benchtable
GPU Temperatures	42 °C	74-75 °C
GPU clock	1449 MHz	1352 MHz
Ambient temperature	22 °C	22 °C
Closed Case
GPU Temperatures	43 °C	74-75 °C
GPU clock	1433 MHz	1338 MHz
Air temperature in the housing	24°C	47°C

Overview Graphs: Temperatures vs. Clock

For better illustrations now again the respective courses considering our timeline of a total of 15 minutes each for the warm-up time.

We see that the clock in the gaming loop is around approx. 100 MHz higher than the reference. This increase of an average of just under 10 percent in this very demanding game is bought with a 40 watt higher power consumption, which is already 18%. However, gaming performance only increases by a maximum of 5-7%, which is no longer in any meaningful relationship. The Torture Loop looks similar:

Thermal analysis of the back of the board

Finally, we consider the thermal analysis of the respective load states. In order to remain as practical as possible, we have pierced the backplate at the relevant locations for the IR measurements and cut out the pads accordingly. Since these are very small diameters, this does not change the performance.

Gaming

We can see that the card is already reaching its limits in the gaming loop with Witcher 3. As long as you operate it vertically and in an open structure, everything still seems to be in the green area, but that can change quickly if you install it. Interestingly, the balancing between the individual voltage converters is a little more optimally solved than with the Gigabyte RX Vega56 Gaming OC.

When used in closed housings, we then measure higher temperatures for the voltage converters by up to 2°C. Here it is revenge once again that the fans only take into account the GPU temperature. With only a little more bums for the fans you would get back to about. 3-4°C cooler, without the ears falling off. Gigabyte has responded to our hints and has given the latest BIOS a one-to-one 100 rpm higher fan speed.

Stress

The stress test is slightly lower in terms of power consumption compared to normal gaming, but some areas, such as the voltage converters, still get a tick warmer. In the end, this is also due to the more constant loads, which are generally more difficult to cool down.

Even in the closed housing, this time it is only up to one or two degrees more at the neuralgic points, nothing more. In sum, however, the voltage converters remain even a tick cooler in all four scenarios than the almost identical Vega56, which, however, operates far above its sweet spot.