The silicon on that card that I will use is free of defect, and there is currently only speculation as to whether there's any defective silicon. (unless you're some sort of engineer who's qualified to examine the functionality of microprocessors and has found the nV40s to have defect?)
Your "argument" about the 6800s is total speculation and supposition rbV5. You can't just say that the hardware is broken until all the facts are in. It was originally thought the 5800Ultras had inferior AF image quality due to hardware limitations, then 3-4 months later driver revisions improved AF to the point some sites said it was better than ATIs.
You have to remember rbV5 that nVidia isn't just rehashing the exact same core for four calendar years in a row like ATI is? [if this is unclear to any of you: I bought my 9700P in October 2002, ATI is now "introducing" the same core in 2005! (X850)] When you have years to work on drivers and core design, rbV5, you have some time to improve on things. The nV40 is a whole new core design that has been on the market less than half a year. nVidia has actually released a new core the last two years running, each of which is a more advanced design in some ways than ATIs core design they bought from ArtX.
My Rev. 1 9700Pro had all kinds of wacky problems, as did everyone else's, necessitating Rev.2. (wouldn't run on many motherboards, rolling wavey lines)
My rev.1 8500 retail couldn't run stable at default speeds, many others had the same problem.
We don't need to go into the MAXX, as much as I loved it, it had issues.
So you see, when ATI either comes up with something new (8500) or buys it (9700), they have problems too.
