@Idontcare
Fair enough. Do you know if the jaguar cores are hand drawn?
Could they too see 15-30% gains from simply being redrawn by a automated program?
Does Intel do the same thing? how long have they had it if so?
if not, why not?
Considering that the Bobcat core was basically 100% done by synthesis (it wasn't really 100%, but it was a large portion), I expect Jaguar was as well. But I can't confirm that.
Synthesis isn't an "all or none" proposition. You can have a varying admixture of both synthesis and hand-layout circuits.
Don't know how much synthesis goes on in Intel's chips, but I recall they used synthesis in portions of the Prescott P4.
....but one of the big flaws as stated by certain engineers in the press about bulldozer was synthetic design?
That particular engineer had questionable expertise in both the areas of synthesis design as well as bulldozer itself, having left AMD many years before bulldozer was brought to market.
I think what people got there was an outdated but educated opinion on the state of things five years prior, which made it seem kinda like a compelling story (it was believable) but at the same time there were parts of it that just failed standard sanity checks (like the fact the chip clocked crazy high, not a hallmark of traditional synthesis which would suggest a "game changer" in synthesis had transpired in the creation of bulldozer itself).
Automated processes - because it was cheaper - not more effective.
Using more of that - seems dubious if you want to maximize whatever process your on?
Remember your
project management triangle.
Automated processes enable the scope and the schedule to do things that would not otherwise be feasible within the given cost envelope.
Had bulldozer (or bobcat) not heavily relied on synthesis for computer-determined circuit optimizations then the respective development schedules would have needed to be even lengthier or the scope (complexity and performance of the cores) would have had to have been dialed way back.
As enthusiasts it is easy to get the cart-before-the-horse with synthesis and see it as a way to turn out sub-par designs. But it really is the opposite, it is an enabler. Were it not for synthesis the designs would have been even more sub-par.
We only think of the existing synthesis designs as being sub-par because (1) self-proclaimed experts tell us to, and (2) we forget that not everyone gets to have Intel-like R&D budgets.
The trivial solution is the one in which we define the optimal product sans all fiscal restrictions during development. "Bulldozer would have been teh awesome if only it had been 100% hand-designed!"
Yawn, not interesting. AMD didn't have another billion dollars to throw at bulldozer's development. And if they didn't reach for synthesis to get the job done then the final product would have been even more derpdozerish.
In a practical world, in a budget-constrained and time-constrained world, synthesis enables the scope of your project to reach degrees of complexity that would not otherwise be attainable if you limited your design team to hand-design methods of the 90's.
Everybody must evolve in how they do their job, synthesis is the future.
edit: Just saw/read CTho9305's post, his +1