Hi re Dataquest and their RDRAM predictions. They've
been wrong in the past, why do you trust them now?
December 7, 1998
Targeted at the massive PC main-memory market, Direct
Rambus solutions will represent 5% of the total DRAM market
by next year [i.e. 1999], 30% by 2000, and about 60%
by 2001, according to Dataquest.
techweb.com
June 14, 1999:
According to Dataquest Inc., San Jose, Direct RDRAM, as
tracked on a per-Mbyte basis, will account for just 3% of
all DRAM shipped this year [i.e. 1999], but will jump
to half of all DRAM shipments in 2001.
techweb.com
Dec 13, 1999:
Rambus is likely to penetrate only 16% to 20% of the
market in 2000, according to Dataquest Inc., San Jose-or
about half that, according to more conservative
estimates.
techweb.com
RDRAM percentages for various year, by Dataquest, by quote
date of prediction:
% 12/98 12/99 Actual
1999 5 NA 3% (probably less, this is a pre-Camino est.)
2000 30 16-20
2001 60 NA
The actual figures for 1999 were 3%, I believe, so Dataquest
overestimated RDRAM production for that year by 66%. Their
figures for 2000 have already been revised down by half.
The 16% estimate is from late '99, and I predict that it
too, will have to be revised downwards.
In fact, according to you, they've already revised it
downwards:
For 2000, though, we see a more gentle progression.
RDRAMs will penetrate to reach 13 percent of the year's
market over the course of the year, the bulk of this number
shipping toward the end of the year. #reply-13465725
Given that Dataquest has never done anything but
overestimate RDRAM usage, why are you trusting their figures
in the future? More conservative estimators are saying 8 to
10% for 1999, they are probably closer than Dataquest. You
have to take into account the fact that only a fraction of
Samsung's production is in RDRAM, and Samsung (the major
producer of RDRAM) is only a fraction of the DRAM industry.
-- Carl
P.S. Surprised you found my suggestion to short AMZN in
3Q99. Pretty good trade, by the way.
Expert design engineers are out there designing stuff for
clients that don't want their name splashed all over the
internet. Dataquest experts need to have their name on as
much paper as possible. Note that Dataquest wasn't
predicting reliability and manufacturing problems for RDRAM
before the Camino introduction, I was.
I don't mean to run down Dataquest in the above. It's just
that they are at too high a level to appreciate certain
facts that are more evident to the workers in the trenches. |
