POST OF THE DAY
Rambus
Solomon Hiding Failed Prediction

Format for Printing

Format for printing

Request Reprints

Reuse/Reprint

By ptnewell
June 20, 2001

Posts selected for this feature rarely stand alone. They are usually a part of an ongoing thread, and are out of context when presented here. The material should be read in that light. How are these posts selected? Click here to find out and nominate a post yourself!

As a herd, essentially without deviation, Wall Street analysts predicted far too high of prices for RDRAM, and far too low production and market share for it (essentially every major firm with analysts does cover RMBS, at least to the extent of issuing advisories, although usually unwilling to issue official recommendations). Although Jonathan Joseph has been among those consistently wrong, he is nothing deterred. In fact, he is rashly spinning failure as victory.

There is no doubt that the first quarter of P4 sales fell below Intel's hopes. However not nearly so far as a comparison of the 800,000 Q1 shipped versus the 20 million hoped for the year would imply to the naive. Intel was only capable of producing modest numbers at fairly high cost in the early going. Moreover, because of the lackluster performance on legacy software, the price/value was not attractive to many. Intel always predicted the bulk of the sales to occur later in the year, especially as the 0.13 Micron version (much cheaper to produce) came available, as did higher speed grades.

However the P4 performs quite well with recompiled software (SSE2 enabled), which has been gradually accumulating. Even more crucially, as yields dramatically improved, Intel slashed prices in late April -- greatly improving the value proposition. And of course RDRAM prices have been falling.

The early results are phenomenal. Asutek saw a four-fold jump in P4 MBs shipped in May. Gigabyte saw similar results in May versus April:

P4 market share triples after Intel's price cuts (June 4, 2001)
"In the Q3 time frame, Intel's Pentium IV will drive the market. They are being very aggressive with price to get it into the market and the Pentium IV's share has been getting bigger, almost triple or four times what it was a month ago."

To be honest, I do not know whether this huge upsurge, seen by so many companies, in P4 sales is a two-month wonder, or will last and grow. But neither does Joseph, and neither do you. With the introduction of a 4-layer i850 board, and continuing drop in RDRAM and P4 prices, and the introduction of higher speed grades, it is reasonable to hope that P4/RDRAM systems will have more than a temporary upsurge.

It is discreditable for Joseph to quote from the Q1'01 only, when production was difficult and limited, before the price cuts and before software which used the P4 came available -- and to conclude that he was right in arguing that RDRAM was too expensive for the mass market. Indeed, since RDRAM prices are ALREADY below what he and other analysts thought SDRAM prices would be at this point, it is already impossible for him to be correct in his original premise (i.e., RDRAM had permanently high costs which would forever keep it out of the mainstream).

Although I here take one particular analyst to task, in fact the groupthink has never been higher than on the memory question. They seem to pretty much all have been making similar claims -- including the logically impossible elements of the argument (RDRAM prices would always be much higher than most people would pay, although it is now already lower than most people were paying for PC133 last year).

Worst of all is the complete focus on Q1, including a complete avoidance of the P4 upswing in Q2 already reported by numerous MB makers. If Intel keeps upping the speed grades, and prices for RDRAM systems continues to fall at anything like the present rate, the May upsurge is very likely to grow, not abate.