United States. Congress. Senate. Committee on the.

The Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) online

. (page 55 of 140)
Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 55 of 140)
Font size
QR-code for this ebook


for computing usage in a few years.

When Control Data sued IBM part of the relief it requested was that
IBM's Components Division be spun out so that the other computer
main-framers had the benefits of what was only then beginning to be
technical superiority in the semiconductor area.

IBjNI's semiconductor expertise must be made available to the other
computer companies in order to assure future competition. Further-
more, the semiconductor test equipment division must be spun out as
well, since it is a necessary adjunct to the use of made-in-IBM compo-
nents.

The next section deals with some technical terminology with regard
to the inability of the semiconductor industry to produce an 8-K or
a 16-K MOS memory whereas IBINI is currently able to do that. And,
then the fact that further on down the road in IBM's FS program —
the future systems computer — that IBM may move to another kind of
technology wliich is called "charge coupled devices" or '*CCD's." While
the other semiconductor houses are working on the CCD's — ^and I have
no doubt that they will be able to produce those in quantity in 1977
or 1978 — there is a possibility which nobody knoAvs and whicli nobody
can answer: That IBM in some of its other aspects of the future sys-
tems will use something known as bubble memories. The semiconductor
companies absolutely will not be able to produce these bubble memories
because they have different properties from all other semiconductors
and, in fact, they are not a semiconductor. The only other company
known to have expertise in producing bubble memories is Bell Tele-
phone Laboratories, and they, also, only produce for in-house use.

I then go through the size of the semiconductor industry which
according to various estimates, and the one cited here happens to be
RCA, should be in terms of U.S. factory sales — close to a $4 billion
industry in 1978. I suggest that if IBM were introduced into the
marketplace as a supplier, initially it would acquire some share of
the computer business equipment niarket which would be the memory
segment of the semiconductor market.



5404

If we assume that the computer semiconductor memory market ex-
pands to $850 million in 1978, and that IBM can capture an un-
realistically high 50 percent of that market, then m 19 < 8 IBM would
have a 22.4 percent share of the total market for computers and busi-
ness equipment. ,. . ^ i tt o j? i.

However, that would result in only 8.3 percent ot total L.b. factory
salesof semiconductors of close to $4 million.

I estimate that IB:M Components has total volume of about $600
million, all of which is in integrated circuits.

IB:^I would therefore be the largest integrated circuit company m
the industry, but Texas Instruments, with sales of— and we have to
estimate this figure also— about $650 million would still be the top
semiconductor in the industry.

Based on some allocations for manufacturing costs, research and
development, selling general and administrative expenses, I have cal-
culated that IBM Components would have an operating margin of
about the same as Texas Instruments, Fairchild, and ]Motorola, but
lower than the newer, smrdler companies in the industry, such as I atel,
J^Iostek, and Advanced M icro-Devices.

In conclusion, I believe that the group of tables which I have
shown in the text shows that IBM would be a viable competitor m
the semiconductor industry, and that the other companies which now
dominate the industry would not be placed at a competitive dis-
advantage.

, Furthermore, the other computer companies would be able to
compete much more effectively if they could use IBM's semiconductor
memory devices which will surpass those of the independent semi-
conductor companies in the 1975-76 time frame.

I have not made any attempt to estimate revenues for IBM's test
equipment operations," but the total market is probably about $150
to $200 million.

The inclusion of revenues from this source could increase IBjM
Components Division revenues by as much as $25 million. However,
TI and Fairchild also build this type of equipment, but we don't
know what their revenues are.

And so we believe that the major companies would probably retain
their ])resent competitive characteristics.

The talile in my statement shows the profitability of something like
IBM Components, Inc., under a number of different assiunptions.

Senator Hart. That is a most interesting and useful analysis both
on computer industry and your lecture on its background. Your
prognosis about the future of components in this industry is
fascinating.

I know staff has a number of questions on the matter in order to
obtain fuller information and data. You suggest some central report-
ing agency with an obligation of confidentiality to put the data
together.

What role would you see of the SEC ? Would it be possible for the
SEC to do something like that ?

Mrs. Walter-Carlsox. I really feel that's outside the sphere of
SEC responsibilities. Several reporting organizations which now
exist are independent agencies or corporations.



5405

In fact there are, I believe, profitable corporations such as the cen-
tral reporting agency for cars and trucks, wliich is Ward's Auto-
motive Reports.

In the laundry industry there's the Home Laundry Manufacturer's
Association; and the Electronic Industries Association, until very
recently, compiled data on various kinds of semiconductors which
were produced and shipped.

I really think that kind of an organization would work better or
be something with which we are more familiar than having it done
under SEC jurisdiction.

Mr. Nash. Thank you.

You mention Mrs. Carlson, that vou were unable to obtain objective
information from Wall Street during the IBM Telex trial last year
and you spent a considerable amount of your own time going through
the documents.

I would be interested in your perspective of why the lack of avail-
able information from Wall Street about this industry and about
the entire antitrust litigation going on in the industry.

Mi"s. Walter-Carlson. Well, there are two things. There are no
experts on Wall Street on IBM because we really don't know^ what
goes on in IBM, or we did not until we were able to see the docu-
ments which were entered into the public record as part of the IBM-
Telex trial.

The second part of your question or second part of my answer to
your question — why were we not on the institutional side receiving
complete information — I think has to do with the IBM mystique, as
it were, which has been built up over so many yeare. IBM had never
lost a case and it could do no Vrong and I don't think a lot of the Wall
Street analysts were paying a great deal of attention to the odds, if
you want to call it that, as to whether IBM would win or lose. The
assumption was that IBM has never lost before and so surely they
will win and so clearly IBM is a buy and clearly you should own IBM.

There were very, very few people who were willing — and never in
print but verbally— to make the statement that there was a possibility
that IBM could lose the Telex case, but they were hard to find.

Mr, Nash. We have not received many communications in our of-
fice — and it's not necessarily expected that we would — but we have
not received from Wall Street, or institutional investors for that
matter, complaints about lack of information on the computer
industry.

Is there any feeling in Wall Street that they would like more in-
formation, or are they satisfied to go along with what they have ?

Mrs. Walter-Carlson. No. I tliink Wall Street is dissatisfied with
the amount of information available on computer placements, com-
puter installations, and that kind of thing.

I think it reflects in the prices of the other stocks in the computer
industry. The institutional investor with a fiduciary responsibility
does not want to invest in a company Avhere he knows that that com-
pany has a limited market share, but really what is that market share.

I mean, we're told by IDC, as a result of its census figures, that
we think that Burroughs' share is 10 percent, but we don't really know
that.



40-927 O - pt. 7-37



5406

There's only one company who really knows what the market shares
are of itself and everybody else, and that's IBM. And they're not
willing to indicate what their share is. Therefore, the other companies
are certainly not willing to indicate that perhaps their share is even
less than what the IDC census shows. We just don't know that.

Mr. Xash. How does IBM react — or other institutions heavily
invested in IBM — when brokerage firms might ])ut out negative
comments about the stock ? Do you see any problems about that ?

Mrs. Walter-Carlsox. Oh. there could be some problems. I don't
know that they've really occurred, but the AVall Street brokerage firms
right now are in severe financial difficulty, as we know, and I would
think that they would not typically want one of their analysts testify-
ing against IBM, or taking a position against IBM, or recommending
sale of IBM because it would undoubtedly upset some of their major
accounts which pay out a great deal of gross in commission business.

Mr. Nash, "WTien you say "upset a great deal of their accounts,"
what type of accounts do you mean; institution accounts, insurance
companies ?

Mrs. Walter-Carlsox. Yes.

Mr. Nash. In your statement you say :

Based on past experience, prudence dictates that it is unwise to invest in a
company which exists only so long as it does not encroach upon IBM's market
share.

How sure are you that the problems confronted by those companies
are not inefficient management or poor quality products, higher prices,
or other internal problems ?

Mrs. Walter-Carlsox. Well, let me give you a personal example of
what I mean here. In 1973, early 1973, on my recommendation we
owned a fairly sizable block of Storage Technology, and Storage
Technology we kncAv to have good management. After all, they had
grown in the face of adversity; they had been able to raise money up
until that time; and they had just announced that they were going
to do an offering to raise additional money. Shortly after that a
rumor got started that IBM was about to announce a new tape drive
which Storage Technology did not have, and in a manner of a couple
of trading hours that day. Storage Technology, which had been trad-
ing at about $17 or $18 a share, went down to $9 or $10. There was
no reason for it because nothing had changed, really, within the com-
pany in terms of what they were going to earn in the current quarter
or in the next quarter. But the fright was that IBM was coming out
with a product which Storage Technology would not be able to produce
and the company would be hurt financially.

As a direct result of that rumor, which didn't turn out to be fact
until about 6 or 7 months later — well, fact in terms of IB^NI actually
introducing that product which Storage Technology could also pro-
duce and which it then did produce — we saw our investment cut in
half in a couple of hours. I could never, at this time, take a recom-
mendation of a Storage Technology, or a California Computer, or one
of those companies, to our investment management committee because
the first thing they would say to me would be remember what hap-
pened when we owned Storage Technology'.

Mr. Nash. The way I read your historical tracing of the semicon-
ductor iudustrv, it seems to me as if the development of the semi-



5407

conductor industry is almost said to be synonymous with the develop^
ment of the computer technology in that semiconductor seems to be
the heart of the computer.

I would like your reaction to that.

Mrs. Walter-Carlsox. This has become much more true with the
advent of the integrated circuit than it was while transistoi'S were in
existence.

First of all, there just aren't that many different ways that a
transistor can be manufactured because it's small-scale integration
and it's typically a single unit. So when transistors were being used
in all computers it really didn't matter because Honeywell or Sperry —
or even GE or RCA at that time — were still in the business, and could
call Fairchild or Texas Instruments or Motorola or National Semi-
conductor and say send me 1 million transistors and that's what
they got. They were basically configured in the same manner and the
computers came out and they kind of all looked the same.

Well, when you became involved in integrated circuitry, which is
when IBM began to bring its components division strictly in-house,
then the semiconductor manufacturer, who is supplying the heart
of the machine, really becomes in part the designer of your system.
However, when you're going on the outside, obviously there are non-
disclosure agreements which are signed so that the information doesn't
get traded between competitors.

But, you are at a disadvantage because you say to your supplier,
"Well, now, you're telling me that you're going to be able to manu-
facture this memory device and you will absolutelv ship it to me in
September 197-3." This, in fact, happened, and then the semiconductor
industry found out that they couldn't make 4-K devices and IB]\I
found out tliat they could not make 4-K devices. But IBM, because
their components were internal, were able to substitute a package,
a module, of four 1-K devices on a single substrate until thev were
able to produce the 1— K. It did the same thing except the 4— K ulti-
mately will be cheaper than four 1-K's. They could do that because
thev didn't have any design in timelag because they had always
designed around the 4-K.

On the other hand the other computer manufacturers couldn't do
that because they were waiting for tlie 4-K, and mavbe they had left
only the amount of space for that package in their machine and then
they couldn't get one 4-K. Thev had to take four 1-K parts and their
equipment was delayed because they couldn't redesign the entire box
to contain the extra components.

Mr. Nasti. Would it be fair to sav that significant advances in com-
puter industrv technology have been made by semiconductor com-
panies rather than computer companies themselves ?

Mrs. Walter-Caklsox. Absolutely. The semiconductor companies
have been extraordinarily responsive to their customers needs. They
have worked di recti v with the customer. They've tried to design
around the problem. The semiconductor manufacturers, in fact, really
attempted to fcet th.e — well, not tic main-frame computer houses be-
cause it was too late in their design cycle to shift — the minicomputer
manufacturers to go to MOS memories loner before IBM introduced
the up<rraded 158 and 168 equipment with MOS memories. Those little
minicomputer companies just were afraid to do it because they couldn't



5408

take the chance on the viability of their company by introducing a
new technology.

It was not until IBM did announce the upgraded 158 and 168 that
there was a run on the semiconductor industry for MOS memories.
They didn't have enough plant and capacity in place and they couldn't
ship as quickly as their customers would want.

Mr. Xash. That's the point you made, I believe, when you indi-
cated that in August of 1972 IBM announced upgraded vereions of
certain machines using upgraded memories.

You referred to IBM's technology and offering new but obsolete
machines at a cost of $300,000. My question relating to that is, and
I just want to bring out for the record to make sure I read it properly,
is this one of the examples you had in mind when you indicated IBM
selling outmoded technology- or better technology existing but not
being accepted in the marketplace because it is not vet utilized by
IBM?

Mrs. Walter-Carlson. Yes. I think, as I also pointed out in the
text of my testimony. IBM was really very late in developing an ex-
pertise in the semiconductor technology.

In fact, they used very, very simplistic devices when there were
many, many more complicated things available.

We think — and we don't know this because nobody really knows a
lot of the answers to what IBM does — that when IBM introduced the
370, that they really planned to use all MOS memories.

But MOS is a different kind of processing technique from bipolar,
which is what IBM had used in the past. Many of the new companies
Avhich started to produce MOS devices never made it because it's tough
and it's different to do. We think that the reason why IBM held off
until August of 1972 and then announced an upgraded version of the
machine which had only been out for 18 months was because they
simply could not produce the devices. It was only when they could
produce the devices in quantity that they served themselves by intro-
ducing the new machines, to the detriment of the user.

Mr. Xash. In reviewing whether an industry structure is good or
bad many people ask the effect on the user. Is the user paying higher
or lower prices than otherwise ? Do you have any suggestions to us as
to how we could attempt to quantify the costs to users from the type
of technological hold-back vou have been referring to ?

Mrs. Walter-Carlsox. I'm really not sure that I can answer that
question. I don't mean to avoid it but I'd have to go around it a little
bit because I can't answer it directly.

When a semiconductor part is introduced for the first time it is ex-
tremely expensive. It is not cost-competitive with whatever has been
in the field which is being produced in great quantity because the com-
panies are not yet down the learning curve and they don't have their
processing techniques in place and so on.

A 4— K memory right now has an average selling price of $14. 1 think
that within 12 to 18 months that average selling price will be some-
where between $3 and $4, and I can't come any closer than that.

The semiconductor companies have to maintain those high prices
until they start making a profit. That delays the main-frame com-
panies from designing those parts because they're too expensive, and
that raises the cost of their equipment.



5409

IBM, on the other hand, doesn't have any of those problems. They
don't really care, initially, whetlier the part costs $14 or $4 because
they can change the pricing within the various aspects of the system so
that they're getting the overall profit margin that they want. When
the cost of the components comes down to a justifiable level they can re-
duce the price of the memory, because that's where the components are
going.

But because of th.e economics iuA^olved. that equipment has already
been shipped and is already in place and has been generating revenue
over that period of time for IB]M as a total unit.

Mr. Nash. You seem to be saying that as to the future the market
control will be predicated upon component technology' and utilization
of the IBM Component Division.

Am I on the right track ?

]\frs. Walter-Carlson. Yes.

Mr. Xash. We woidcl be interested in receiving for the record your
recommendations as to what you believe the critical steps might be to
create a viable competitive computer industry in the 1970's. 1980's, and
so on.

INIrs. Walter-Carlson. Well. I don't really think that I have any
better answer than anybody else in that regard. I do feel that, as my
paper shows, the expertise in the components area, which is very recent
for IBM — I meaii, after all, we're talking about August of 1972. so
that's only occurred within the last 2 years — has got to be made avail-
able to the rest of the industry; and I think that will help.

It will help in two ways. It will reduce IBM's in-house lock on
secrecy, which none of the other houses have, and it will also permit
the other houses to have superior components available to them, be-
cause with that technological lock, along with the capital lock, and
the software lock, it pretty well means that nobody gets into the bus-
iness: and an awful lot of people who are in the business probably
don't got to stav verv much longer.

I think another thing that has to happen is that you somehow or
other have to set up a GMAC of sorts with IBM's incredible amount
of cash. I think in something wliich they call "securities held for re-
payment of long-term debt" they have $3.8 billion in cash, in com-
parison with $26 million for Sperry Rand.

I mean it's just incredible. That's fjot to be made available some-
how or other to not the other main-frame companies but to the in-
dustry as a GMAC and I think one of the other things that you have
to do is you have to say IBM for ,r number of years — and whether
that's 4 years or 5 years or however long — may not lease; they must
only sell their equipment.

That, then, permits the other companies who have been trying to
build a lease base on which they have been losing money to also com-
pete fairly, because then they can go out and they can say we only
want to sell you equipment' because that's all that IBM does and
that is now the standard in this industrv rather than a risk-lease
basis.

Mr. Xash. Thank you very much.

Senator Hart. Mr. Granfield ?

Mr. Granfield. I'll be brief as possible. I find your statement ex-
tremely interesting ; and also let me extend a greeting to you.



5410

I spent the last 6 years at UCLA and you make me feel somewhat
nostalgic for those beautiful surroundings which you find in your
city.

I find your statement interesting because most of the witnesses
today have given evidence that IBM is indeed a sluggisli monopolist
because they lag in the introduction of certain technological inno-
vation.

This morning I tried to shed some light on the whole area of
technology, technology not only in the technical but in a cost sense
and that may explain some of the technological innovative behavior
of IBM vis-a-vis its competitors.

But this afternoon you tell me that IBM is even sneakier than
that, that what they have done while they have "lagged" traditionally
in the introduction of technological innovations is suddenly achieved
such a technical superiority in terms of semiconductors that we
shouldn't break IBM up or force it to divest itself of one of its
branches because it's a sluggish monopolist and denies its customers
technological innovations but because it's going to out-innovate its
competitors.

So I'm becoming a little mystified as to exactlv what is the implica-
tion of somehow affecting the structure of IBM.

Some witnesses tell us because they lag behind. You come and tell
us, "Well, they have perhaps traditionally lagged behind; suddenly
now they're going to leap so far ahead that this is whv we must curb
IBM."

So I find your testimony quite interesting because suddenly the
sluggish monopolist is going to lead the pack and that's wliy they
are a threat to the competitive liability of the computer industry.

The witnesses to date have told us the great threat they pose to the
customers because of their technological rapidity.

Perhaps they listened to these witnesses. I don't laiow. But I cer-
tainly am very uneasy with whatever path is or Avhatever welfare
costs that preserving IBM is going to impose upon its customers or
upon the computer public as a whole because you tell me in your
testimony that because they have taken the technological role internal
to the firm they may be able to out-innovate firms whose primary
purpose is the production of semiconductors.

This seems to me a direct benefit of the IBINI structure, that they
would be able to specialize in the production of semiconductoi'S for
information processing for computers.

Could you shed some light on my mystification?

Mrs, Walter- Carlsox. Well, that was a multifaceted statement,
and I'm not sure that I know precisely what your question was.

Mr. Graxfiei.d. Really what I'm trying to home in on is, "Is IBM
bad because it doesn't innovate, or is it bad because it does innovate,
or is it bad at all?''

Mrs. Walter-Carlson. Well, first of all we're talking about two
different time frames. We're talking about what hapjXMied before the
Justice Department and the other people sued IBM on the basis of
antitrust.

I think it is a fact that since the Justice Department suit was insti-
gated IBM has become far more predatory in its acts than it had been
in the past, so there is a rhetorical question.



5411

If the sleeping giant was, in the past, able to lead the industry and
get along with what it had and continue to reap the benefits of its
huge revenue base and not need to innovate, that's one facet.

If, on the other hand, you are under attack because you have not



Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 55 of 140)