We now have specifics on what AMD will launch to contend from Nvidia, from an abnormal source. Newegg publishes its very own graphics and computing guides, together with a guide to which GPU you really should acquire. Included in that guide, at the very least briefly? Specs for AMD’s impending Radeon RX 6700 XT, 6800 XT, and 6900 XT GPUs.
Definitely these specs really should be viewed as preliminary, but they line up with other rumors we’ve listened to about the household. Hat-tip to Videocardz for getting the information and facts:
There are some oddities in the chart that are tricky to describe. The Radeon 6700 XT is a marked downgrade from the 5700 XT, with significantly less clock, significantly less RAM, significantly less memory bandwidth, and — admittedly — a incredibly nice TDP of 150W. We can consider that some of RDNA2’s effectiveness-for each-watt uplift will do the job in AMD’s favor below, but the smaller sized RAM buffer is a very little surprising. AMD may well have offset the clock fall and memory bandwidth reduction with more cache and bigger efficiency, but there is no way to offset minimized VRAM.
It’s probable that the 6700 XT will be offered in a 6GB and 12GB flavor, equivalent to GPUs like the RX 570. A different leak — this one particular from Reddit — provides some far better clock knowledge courtesy of Apple’s macOS 11 beta.
We really do not know if these numbers are precise both, but they imply a Navi 21a and 21b. 21a has a greatest frequency of 2050MHz, while 22b has a top frequency of 2200MHz. Socket energy limitations are provided as 200W and 238W respectively. This implies it requires 1.19x more energy to push 1.07x more effectiveness on the chip. The truth that the Apple chart refers to socket energy could mean this is only the energy attract for the chip alone, not the relaxation of the card. If we believe that, the TDP listings from Newegg and Apple are a bit closer to each individual other.
Various months ago, I wrote that if AMD could enhance its core clocks up to 2.2GHz (equal to the PS5), it could support the company square off with the RTX 3080. The GPU could pick up 1.22x effectiveness from the more clock (assuming a 1.85GHz – 2.2GHz jump). We know effectiveness really should nearly double from the core count enhance, while the jump won’t be properly linear. Specified that the RTX 3080 is about 2x speedier than the 5700 XT in many assessments, AMD may well be ready to land a competitive GPU for Ampere in conditions of raw effectiveness, while certainly energy use and thermals will be their very own matters. It’s also early to know if the early complications we’re listening to about with GeForce RTX 3080 shipments will influence comparisons when RDNA2 launches or not.
We know that Nvidia is proclaiming a 1.9x effectiveness/watt enhancement about Turing, when compared to AMD’s 1.5x enhanced effectiveness for each watt about RDNA. It’s heading to be incredibly interesting to see how the two corporations evaluate from each individual other when playing cards fall in the not-so-distant foreseeable future.