that I first recommended NVE to our core
Nanosphere portfolio in July 2003, many of
you are probably wondering what hap-
pened to MRAM? And why should anyone
still own the stock?
NVE took a major hit when Cypress
called it quits on MRAM development. The
two companies had formed a technology
exchange agreement in 2002, in which Cy-
press gained access to NVE's IP and a 17%
equity stake. In turn, Cypress invested $6.2
million in NVE and committed to defend-
ing NVE's IP in patent litigation under
certain circumstances. Under the agree-
ment, Cypress would also manufacture a
minimum of 500 wafers per quarter that
NVE would sell under their own brand.
The deal was never meant to be particu-
larly lucrative for NVE
--
its main purpose
was to prove that NVE's technology
worked. So when Cypress recently
decided to stop pouring money
into MRAM development, it un-
dermined NVE's claims to have
viable MRAM.
"Cypress was a pretty good
advocate of MRAM and they
bailed and said they could spend
their money better elsewhere,"
says Mark Devoss, a senior ana-
lyst at iSuppli. According to De-
voss, the bottom line was that Cy-
press couldn't build a cost effective
single transistor structure. "The hur-
dles were higher than they anticipated," he
says. "Cypress had working parts but they
weren't in a form that Cypress felt could
place it in the right spot to replace existing
products. The whole thing with emerging
technologies is that in order to be success-
ful they have to displace an incumbent, and
the incumbents are all still working, con-
stantly reducing costs and increasing den-
sity. The bar keeps getting raised in terms
of the cost per bit target. And as the other
technologies advance, it gets tougher for a
new technology to replace them."
On a live panel at the Embedded Sys-
tems Conference in San Jose earlier this
month, TJ Rodgers, Founder, President,
CEO and a Director of Cypress said, "I
spent $54 million trying to make magnetic
memories, which is a very fine project but
it didn't work out. I don't think they're
going to happen
--
at least, we couldn't
make them happen." When asked how he
made the decision to abandon MRAM
Rodgers answered, "The decision was made
while sitting with the engineers working on
the MRAM project; there were about 50 of
them. They were the best, the finest R&D
team I've ever worked with in my entire ca-
reer. We had the thing working, we had
sampled it to 8 customers, 4 of the cus-
tomers reported back to us that the RAMs
that they had, they thought they were great.
What they didn't know that I knew was
that if you took 256 RAMs and put them
on a board and took 4 boards so you had
1000 RAMs and ran the RAMs for a week,
one bit would go flaky every single week.
And nobody could stop it from going
flaky
--
I even went back and put error cor-
rection in the RAM to try to fix it and the
flakiness happened to be an oscillation at a
Gigahertz, tunneled literally right through
the error correction. So in this case I
wanted to hear exactly what they were
going to do to solve the problem and the
team walked in to me and said, 'We ain't
gonna solve the problem.' And I said, 'I'll
take your word for it,' and that was the end
of it."
With Cypress unable to get NVE's
MRAM technology up and running, NVE's
only other hope at the moment is
Freescale
--
and things aren't looking so
good there. NVE's original deal with Mo-
torola was signed back in 1995, and though
the agreement was never made public, ana-
lysts at Stephens Inc. speculate it granted
NVE a 1% royalty on sales of Motorola's
MRAM products that utilize NVE's
patents. When Motorola spun off its
MRAM development to Freescale, NVE
saw it as a great opportunity to renegotiate
a better deal. While the companies can't
comment on ongoing negotiations, NVE's
CEO Dan Baker did tell us, "We believe our
agreement with Motorola likely terminated
at the end of 2005 because of the transfer
of the manufacturing capability to
Freescale so Freescale would need its own
license. We had said that we want that li-
cense to be under better terms than the
Motorola agreement. They asked us to as-
sign the Motorola license to Freescale and
we've declined to do that."
It's unclear at the moment whether
that's a good move for NVE. On the one
hand, if Freescale is using its patents, then
NVE could come out with a better deal.
But it's not certain that Freescale is using
their patents at all, and with Cypress no
longer in their corner, NVE can't exactly af-
ford major patent litigation. Perhaps more
Two important changes this month.
First, we make a new Buy addition to
our back page list of stocks, with nano-
silver leader Nucryst Pharmaceuticals
[NCST]. The company has a world-class
partner in Smith & Nephew [SNN], a
strong management team and a host of
promising new applications under de-
velopment. And the April issue also has
us dropping our rating on embattled
longtime Nanosphere company NVE
Corporation [NVEC] following a recent
run-up and a change in its risk/reward
outlook. There's still great potential to
be realized but the company's future has
become cloudier.
The Nanosphere isn't the only
thing getting regenerated. Be sure to
read our interview with Tony Atala,
whose groundbreaking work in tissue
engineering and organ growth might
seem like science fiction. As Arthur C.
Clarke has said, "Any sufficiently ad-
vanced technology is indistinguishable
from magic."
Here's to thinking big about
thinking small...
The Insider
© Copyright 2006 Forbes/Wolfe Nanotech Report
April 2006
·
3
Hear Josh
Wolfe Speak!
At The 5th Annual
NanoBusiness Alliance Conference
May 17-19, 2006
New York Marriot Marquis
Times Square, NY
For more information,
visit: www.nanoevent.com
