Cores Duo/Solo: "My God, they're full of flaws"

Of the 34 problems only 14 were labeled red in the chart. 17 are green and either are unlikely to cause noticeable effects or have a software workaround. 2 are listed as yellow which isn't preferred, and 1 isn't actually a flaw at all since it only applies to 64-bit operation which Yonah can't do.

I admit it's not very good for it's image, but with 16 (including yellows) or less than half of the problems without workarounds, it's not as dire as the wording of the article may imply.
 
Granted it has been a while since I checked on erratas in CPU's, but a few years ago, Intel's chips had far fewer than AMD's, and that margin percentage wise favored Intel's more, since Intel does and has had more transistor per chip than AMD. If that is a reference of good/bad chip, Intel wins hands down there.
Good thing they know of the errata's, so they can create workarounds so that they perform as intended anyways, of course this creates a performance hit to correct, but the CPU works.
 
I find this bug to be the most ammusing:

noexecute.JPG


-mpjesse
 
but a few years ago, Intel's chips had far fewer than AMD's,
Not how I recall it, especially since a "few years" put use right around A64/ Prescott time. At that point Intel had more erata, and more critical errors. Even so far back as the barton release, the only reason they had more uncorrected errata was due to the work Intel had done to correct errata on the northwood cores.
 
It may be exuberant optimism (I probably shouldn't steal that term from Bullshitter), but there is a new stepping of Yonah coming out. It's to be released with the T2700 2.33GHz Core Duo and will enable VT support. I would hope Intel decides to solve some of the errata using that stepping.

http://www.theregister.co.uk/2006/01/16/intel_core_duo_vt/