Beau,
Well, I was kind of ignoring those levels, since it's effectively impossible for anyone but the majors to use, and they do still have the other issues.
Which brings up an interesting point, did anyone declare this 'impossible' yet? Is that what's holding things up? Finishing the prop2 in 180nm with 16 cogs and 512k and 64 analog pins, impossible. There, that should do it.
I wonder if 64 smart pins are going to make 16 cogs look like total overkill.
It's a bit hard to know right at this point, but we should know more after the next images come out.
There seem to be several overlapping ways to do things, between cogs, streamers, interrupts and smart pins. Plenty of room for implementation preference.
P2 may be extravagant but the P8x32A was also extravagant with its video generators in every cog, and its 16 counters. We're used to being spoilt.
Which brings up an interesting point, did anyone declare this 'impossible' yet? Is that what's holding things up? Finishing the prop2 in 180nm with 16 cogs and 512k and 64 analog pins, impossible. There, that should do it.
re:We need to see a graph of cost vs process technology. 350nm masks are currently $35k, while 45nm masks are about $1M, or ~30x cost for ~60x the density. I wonder what 14nm masks would cost.
Thanks! Chip. It's always interesting to get new insights into how decisions are made and what they are based on.
Comments
Well, I was kind of ignoring those levels, since it's effectively impossible for anyone but the majors to use, and they do still have the other issues.
Which brings up an interesting point, did anyone declare this 'impossible' yet? Is that what's holding things up? Finishing the prop2 in 180nm with 16 cogs and 512k and 64 analog pins, impossible. There, that should do it.
It's a bit hard to know right at this point, but we should know more after the next images come out.
There seem to be several overlapping ways to do things, between cogs, streamers, interrupts and smart pins. Plenty of room for implementation preference.
P2 may be extravagant but the P8x32A was also extravagant with its video generators in every cog, and its 16 counters. We're used to being spoilt.
re:We need to see a graph of cost vs process technology. 350nm masks are currently $35k, while 45nm masks are about $1M, or ~30x cost for ~60x the density. I wonder what 14nm masks would cost.
Thanks! Chip. It's always interesting to get new insights into how decisions are made and what they are based on.