Here is an update about one of our potential gorillas that most have given up on...
tsrec.com
#4. Rambus NASDAQ: RMBS
Initial Recommendation - 12/98:
A major problem in the configuration of personal computers deals with the
number of commands that can be sent from the microprocessor of the computer
to its RAM at any given time. The problem's so great that Intel predicts
that without new technology, an increase in microprocessor speeds will soon
be inconsequential in increasing the overall speed of a PC. Not only does
Rambus, an intellectual property company, have the "new technology" to deal
with this problem, they have a virtual monopoly on it through a series of US
Patents. Like ARM Holdings, Rambus doesn't involve itself in the low-margin
business of producing chips; all they have to do is license their
technology, and then sit back and collect the profits. Customers of their
products include Intel, Motorola, and Samsung, with Dell, Compaq, and IBM
planning to buy next year. This company has an expected 5-year growth rate
of 75%, and given that Intel has stated that they expect Rambus technology
to be embedded into all of their microprocessors within a year, that number
could be even higher. Also, given their track record, there's always the
chance that Rambus may come up with a new breakthrough sometime soon. Even
if all they do is just improve on past innovations, I'd be very surprised
not to see them as one of the top 25 high-tech companies of the early 21st
century.
An Update - 11/19/99:
To say that Rambus hasn't had a smooth year is like saying that Mike Tyson
isn't a good role model. They've had numerous chip delays holding back the
release of RDRAM, the DRAM chips based on Rambus' technology. First it was
problems with RDRAM itself, and then it was incompatibility issues with
Intel's 820 chipset, the first motherboard chipset meant for desktops that
supports RDRAM. Finally, on November 15, months later than expected,
RDRAM-based PCs make their debut, but not before Intel had announced chipset
support for a competing DRAM technology known as PC133. So why do I still
recommend Rambus? Because, quite simply, they still have by far, the best
DRAM technology out there. Winbench tests showed that RDRAM-based PCs blow
away computers with competing types of DRAM when it comes to applications
that require large amounts of data to be transferred from the PC's DRAM to
its microprocessor in a short amount of time, applications such as Adobe
Photoshop, AutoCAD, 3D gaming, streaming video, and most important of all,
multitasking. Furthermore, RDRAM's uses aren't limited to PCs. Compatibility
with Intel's 840 chipset is allowing it to make strong inroads into the
server and workstation market, and laptop proliferation shouldn't be too far
away. Sony's licensed it for its Play Station II, and in the future, RDRAM
has a good chance of being used in digital set-top boxes, PDAs, smart
phones, switches and routers, digital cameras, and in any other device that
requires data to be quickly transferred to and from its DRAM module. One
complaint about RDRAM has been the fact that it's more expensive than
competing technologies, costing 25-30% more. While this may sound like a
lot, it isn't much when considering its minimal effect on the overall cost
of an entire PC, server, laptop, or workstation; and given the
price/performance benefits derived when compared to those reached by paying
for a faster CPU, it's not a bad deal to say the least. Also, RDRAM costs
are expected to come down over time, further marginalizing this issue. DRAM
sales are expected to top $70 billion in 2003. RDRAM has a good chance of
taking over 50% of this market by then, and Rambus gets a 1.5% royalty for
everychip using their technology. Given the low costs of running a
royalty-based business, profit margins of over 80% are very likely. As for
how much Rambus gets in potential profits if these estimates hold true, I'll
let you do the math for yourselves, and if you do take the time to do it,
I'm sure you'll like the numbers that you'll get.
|
