I got answers from computer programmers(talkchess) A0 is about 800 times hardware advantage than SF. In term of elo there is 50 elo on doubling of hardware. A0 beat SF with 100 elo margin . (400+ elo=2^9=512 times hardware advantage) .
If A0 and SF were played in equal conditions, SF is estimated to be 400 elo stronger.
On my Macbook Pro, SF 8 chooses 8. Qe1 after about 30 seconds of thinking (once it reaches depth 25).
Also, comparing hardware isn't really reliable, as the two entities work differently and have different processing needs. The hardware given to Stockfish in the match was more than sufficient. If there are any qualms, it's the one-minute-per-move setting that both players were restricted to, which the SF team says weakens SF (though I'm sure the one-minute-per-move setting also weakened AlphaZero).
The missing Opening Book also hampered SF, which needed to find its own opening moves—though that was pretty much the point of the exercise: to see how two engines (or rather, an engine vs. an AI) play when human knowledge is not a factor. SF's opening book is based on human play. Taking away SF's opening book makes a lot of sense, in that regard.
I analysed game 1 of SF vs Alpha Zero recently. At move 8, SF played Qe1??. I analysed with full strength SF8, (and also SF 9). at multiple depth. SF never considered such a weak move. Most probably SF has serious handicap. Even 1000-1200 rated people wont consider such a weak move. ( At certain level of handicap, SF will play such weak move)
Can anyone tell me, what handicap level of stockfish played such a weak move?