.OP
.PF "''Company Confidential       May 31, 1985''"

.DF
From: Rob Horning                            Date: May 31,1985

  To: Mark Forsyth                        Subject: Static RAM's
      Gerry Gassman (Cupertino)
      Gordon Goodrich (Cupertino)
      Ed Holland (Cupertino)
      Leith Johnson
      David Means (Cupertino)
      Craig Robinson
      Russ Sparks
      Darius Tanksalvala
      John Wheeler

  cc: Denny Georg
      Peter Rosenbladt (Cupertino)
.DE


I resently visited the major Japanize SRAM vendors with the purpose of 
finding out what they are planning for high speed (25 ns) SRAM's and to
let them know that we have needs for the parts.
This is a summary of what they are planning for the next year or so.

Some of the manufactures are moving in the direction of higher density
as opposed to higher speed.  We let them know that we were more interested
in high speed than high density.  Two of the manufactures are considering
changing there priorities in order to provide 25 ns SRAM's sooner.
Most of them have the technology or will
have the technology within a year to build 25 nS 16K bit SRAM's.

As expected the most common high speed part was the 4K by 4 part.  The next
most common 25 ns part was the 16K x 1.  Hitachi has a 25 ns 16K x 4 that
they will sample this year.  Toshiba is the only vendor planning a 25 ns
2K x 8 part and it is not due to be sampled until the middle of next year.

There were no plans for faster than 25 ns TTL compatible SRAM's.

Below is a summary of where each of the vendors stand as far as high speed
static RAM's.  This information is confidential and should not be given
to the SRAM vendors.

.H 1 " Hitachi"

Hitachi is a computer manufacturer and they have had internal need for high
speed static RAM's.  I think this is one reason why they are  farther along than
the other Japanize manufactures.

There are two divisions at Hitachi developing high speed static RAM's.  There
is the division that does the ECL parts that is using Bi-CMOS to get lower
power high speed parts.  This division is developing a 25 ns 16K x 4 and
64K x 1 part.  They have already provided HP with a few engineering samples
of the 64K x 1 part and will have engineering samples of the 16K x 4 in
4Q85.  They plan to have qualification samples in the first quarter of next
year and production parts about 4 months later.

The other division that has high speed static RAM's has been doing high
density CMOS SRAM's.  They currently have a 45 ns 4K x 4 part designed with
2.5u design rules and are planning a 25 ns 4K x 4 part designed with 1.3u
design rules to be sampled in 4Q85.
They currently have a 16K x 1 35 ns part (2u design rules)
but do not have plans to make this 25 ns.

The two divisions have some charter overlap in the high speed SRAM area.
It does not seem likely that this will continue.  It appears that the
ECL division has better products planned and are farther along in the
development of these products.

.H 1 "Toshiba"

Toshiba currently has 35 ns 4K x 4 and a 35 ns 2K x 8 parts that use a
2.0u process.  They will sample 25 ns parts in the first half of next year
that use a 1.5u process.

The Toshiba parts are NMOS as opposed to CMOS or mixed parts like most
vendors are doing.  They estimate that the maximum power will be about
.5 watts (about the same as fast CMOS parts).

.H 1 "NEC"

NEC is just starting to sample a 16K x 1 35 ns part and will sample a
4K x 4 35 ns part in 1Q86.  There current plans are to have 25 ns parts
in mid 87.  They have been focusing there attention on getting higher
density parts in the 45 to 70 ns range.  They are considering changing
there priorities in order to get the 25 ns parts out sooner.  We
encouraged them to do this.

.H 1 "Fujitsu"

They are now sampling 16K x 1 and 4K x 4 35 ns parts (2 u process).  They do
not have plans to do any 25 ns parts at the time.  They are considering
developing 25 ns parts and will get back to us in mid June to let us
know what they decided.

.H 1 "Mitsubishi"

Mitsubishi will be in production in 3Q85 with a 16K x 1 and a 4K x 4 35 ns
part.  All of there future plans are higher density parts.  They have no
plans for faster parts and did not indicate that they would consider
changing their plans.

