Updated Authoritative German magazine c't is reporting that there could be a bug in Intel's latest flagship desktop chip, the 733MHz 0.18 micron Pentium III. But Intel says that the problem is down to the defect in the i820 chipset the magazine used to test the Coppermine chip. In the latest printed edition of the magazine, which reviews the Coppermine processor, the magazine reports compiler optimisation errors using both Intel and Microsoft software. The magazine apparently also reports hardware errors, possibly related to the brand new L1/L2 cache in the top end chip. The SPEC suite test computed false results about every 20th run. But whether there's a bug or not, the entire question may be totally academic. The largest distributors of Intel chips in the world don't have stocks of the newest part. Ingram and Tech Data don't even list the processors yet, while Hallmark and Keylink, although they list them on their dealer lists, say they don't have any stock. An Intel UK representative said: "We're expecting a very strong demand this quarter and we'll meet all backlogs. The supply line is tight, we're not denying that." A representative from Intel Germany claimed that the problem was due to the i820 (Caminogate) chipset. "I'm very confident about that," he said. He said that the software errors did not occur when 666MHz and 600MHz were used on the same motherboard, and that the problem did not occur with competitive, third-party products. Intel refers to problems with its processors as errata, not bugs, except in the famous case of the Pentium bug, where the cock-up was disingenuously referred to as a flaw. Whether this reported problem is down to the i820 chipset or the 733MHz processor, errata are fairly common in recently released processors. ®