No datacenter needed now, just run it on your computer. Anything with 12gb+ of vram can run it comfortably at low quants. If they can come out with a good MOE model it will crush the competition. At this current rate of progress, China will out-compete foreign AI development within a year or two.
And it does it by defaulting to xhigh reasoning in the most benchmaxxed manner, so that it oneshots your queries alright, just after wasting tens of kilotokens of doubt per every turn. 12GB VRAM can probably fit it replying to “say hi”.
Still impressive, just not a leap you’re selling here.