The first comparative results are already there and as such, it is not that they were surprised by the performance of the RX-6700, but precisely because of its supposed consumption. And we say assumed because what is filtered does not belong to the series or the reference model, as everyone likes, but to one customized by Sapphire, which logically leads to overclocking. What conclusions can we draw from this? 3DMark?
After the controversy over its name, where Sapphire itself had just displayed 6700, while AMD naturally launched it as the RX 6700, now comes the controversy over the data to give a GPU some play which could sneak in as a very interesting model.
The RX 6700 is competent enough, but with the latest driver
AMD and NVIDIA overlay the performance of their graphics cards as master samples so that the performance, if leaked, will look different than it actually is. We see it again in this particular case, since the 6700 as such, name given by Sapphire when leaking the specifications, made in 3DMark 9,427 stitches with the original power settings and 10,304 points at full power as AMD’s TBP allows.
Being the same card and changing its name to the real one given by AMD, the RX 6700 with the new 22.5.2 driver achieves a bit better performance with 11,166 stitches with the original power mode and nothing less than 11,414 stitches with said driver and the maximum power configuration in TBP. These data give us some percentages of +18% and +11% compared to those mentioned above, so there is undoubtedly plenty to choose from in AMD.
Where is this new graphics card compared to its sisters? Well, given that the RX 6700 XT is a 11% higher in Shaders, this GPU achieves a 15% higher score for only 4% more consumption. The problem is that the comparison is not fair as such, because the RX 6700 GPU is from Sapphire and that means more frequencies and consumption, so the distance should be greater.
The RX 6650 XT is perhaps the most affected
Logic tells us that AMD has segmented the range very well and has thought of everything. All the more so if we take into account that the RX 6650 XT has a consumption of 180 wattsin front of the 175 watts as her older sister says (in theory, of course).
With that in mind, leaked data reveals that the RX 6700 falls between a 7% and 10% faster, but considering that the frequency is 45 MHz above to the production model, these figures should in fact be lower.
The problem and the funny thing about it is that the RX 6700 is listed with a TBP of 175 watts, while the RX 6650 XT is stock 180 watts. Is a superior GPU more efficient, but with less consumption? Well, it could be, but it looks like this RX 6700 won’t be a range between the two XT versions, but more beating its lower sibling than the upper, although in all honesty the consumption is really good.
It remains to know the price to be able to determine if we are facing a GPU that, if it arrives with a large stock, could be a bestseller as soon as the prices drop to levels below the MSRP.