Letter A reader has pointed out that we missed quite a few other things that Intel got wrong in 1999. And he adds that there are other factors, earlier in the decade, which ought to be taken into account when judging the company's performance. The author, who wishes to remain sub rosa, said: "It's not too late to insert a comment about Merced too. How about Intel converting the "processor of the next millenium" into a test chip. HP's interest and support for Merced is teetering on the edge. "Other '99 screw ups include the six month delay of AGP4x - the graphics chip makers were ready in the spring. "ATA 66 was introduced on low end systems six months before making it to high end systems. "Intel killed the 752 graphics chip and Intel's entire graphics business just months or weeks prior to the planned release of the chip. It then killed Real3D in the wake of this disaster. "And what about the BX chip set shortages, which are still continuing? "It's questionable whether Intel wasted both its own time and the industry's time on the NGIO bus wars, but resistance was futile against PCI-X. There are probably others. Intel actually screws up a lot. "There are a number of blunders which came earlier in the decade too, including the Covington, the i860, the Intel-VLSI Polar Chip set for handhelds, the acquisition of Chips & Techologies when it had a declining mkt share in the portable graphics market." (email name and address supplied). Want to take up arms in Intel's defence? Post a message in The Register forum. ® Related Story 1999: Annus Horribilis for Chipzilla