InvestorsHub Logo
Followers 5
Posts 807
Boards Moderated 0
Alias Born 08/24/2013

Re: Dmcq post# 139241

Thursday, 02/05/2015 9:59:36 AM

Thursday, February 05, 2015 9:59:36 AM

Post# of 151689

I was assuming 0.5W maximum for the A72 at 2.5Ghz on a 16nm FinFET process like The Register said, working back from that I estimate less than a watt per core at 28nm at 1.8Ghz but it certainly requires some fudge factors.


Well, I doubt we'll see any quadcore A72 designs on 28nm at all. Are there even A57 quadcores at 28nm in the market?

Regarding power, that is what I also read. ARM seems to be basically comparing an old, power hungry processor (A15) on an old process (28nm) with a new processor (A72) on a new process (16nm). Marketing, as you said (Intel couldn't do it any better). ARM didn't give much detail about the figures as far as I understood.

Regarding your power estimations: Well, A72 delivers more performance per clock. That needs to come from something so I expect it, just like any other ARM core in the past, to consume more power clock for clock (and on the same process) than its predecessor. No magic pixie dust for ARM either.

Volume:
Day Range:
Bid:
Ask:
Last Trade Time:
Total Trades:
  • 1D
  • 1M
  • 3M
  • 6M
  • 1Y
  • 5Y
Recent INTC News