Intel Nehalem Mainboard and CPU Picture Debut...

[Omastar]

p3 @ 1ghz needs ~160 seconds

So this A1 rev Nehalem is actually slower, clock for clock, than a 1GHz P3 in Super Pi?

[Blauhung]

So this A1 rev Nehalem is actually slower, clock for clock, than a 1GHz P3 in Super Pi?

Production will probably C-something, I'd be willing to bet that just about every chip (specially a new uArch) starts out very very slow on early stepping.

[Movieman]

Production will probably C-something, I'd be willing to bet that just about every chip (specially a new uArch) starts out very very slow on early stepping.

Everything I know about engineering( which isn't much) is first make it work, then make it work better(faster,less heat,etc)
My guess is that they start at a slow speed to see if the theory actually works in practice,then look at all the individual components to see any potential flaws and then they start to make it go faster.
Just a guess but maybe closer to the truth than we know.

[flippin_waffles]

Everything I know about engineering( which isn't much) is first make it work, then make it work better(faster,less heat,etc)
My guess is that they start at a slow speed to see if the theory actually works in practice,then look at all the individual components to see any potential flaws and then they start to make it go faster.
Just a guess but maybe closer to the truth than we know.

Sounds very familiar.

[Blauhung]

Everything I know about engineering( which isn't much) is first make it work, then make it work better(faster,less heat,etc)
My guess is that they start at a slow speed to see if the theory actually works in practice,then look at all the individual components to see any potential flaws and then they start to make it go faster.
Just a guess but maybe closer to the truth than we know.

exactly

most of the time the further steppings always include various speed path fixes to help ramp up the clocks and stomp out performance hampering bugs.

my job is pretty much the same thing. All the PHD guys make the process work (sometimes it only looks like it works), then they give it to my cohorts and I and we make it work safer, with less defects, and faster.

[Omastar]

Production will probably C-something, I'd be willing to bet that just about every chip (specially a new uArch) starts out very very slow on early stepping.

Gotcha. Didn't know there was that large a variance between early revs and final revs in terms of clockspeed scaling.