search for: lowball

Hi, I have two ten-page websites that I need produced quickly. I can pay a reasonable rate; this isn''t an attempt to lowball--the priority is speed and quality. I have all of the Photoshop files you will need to understand the work you will be producing. More parameters: - Both websites should have identical html, but different stylesheets. (like CSS Zen Garden) - Both websites should be produced in well-separated XHTM...

...on the ability to finish each operation on those short vectors relatively quickly rather than padding everything to a full vector register. For example, if a loop with a trip count of 20 is vectorized on a machine with 64 elements per vector (that's 64b elements in a 4096b register, so this is lowballing it!), using only masks and not the vector length register makes your vector unit do about three times more work than it would have to if you set the vector length register to 20. That keeps the register file and functional units busy for no good reason. Some microarchitectures take on the burden...

...tdistr(z$efficiency, 'normal') q = 55:95 + 2.5 lines(q, dnorm(q, z.fit$estimate['mean'], z.fit$estimate['sd']), col='red') legend('topleft', legend=c('density', 'fitted'), col=c('blue', 'red'), lwd=1, inset=0.05) # Factor out lowball values. z.f = z[which(z$efficiency >= 65),] by(z.f, z.f$method, summary)

...on the ability to finish each operation on those short vectors relatively quickly rather than padding everything to a full vector register. For example, if a loop with a trip count of 20 is vectorized on a machine with 64 elements per vector (that's 64b elements in a 4096b register, so this is lowballing it!), using only masks and not the vector length register makes your vector unit do about three times more work than it would have to if you set the vector length register to 20. That keeps the register file and functional units busy for no good reason. Some microarchitectures take on the burden...

Philip Reames <listmail at philipreames.com> writes: > Question 1 - Why do we need separate mask and lengths? Can't the > length be easily folded into the mask operand? > > e.g. newmask = (<4 x i1>)((i4)%y & (1 << %L -1)) > and then pattern matched in the backend if needed I'm a little concerned about how difficult it will be to maintain enough