| Title: | The Digital way of working |
| Moderator: | QUARK::LIONEL ON |
| Created: | Fri Feb 14 1986 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 5321 |
| Total number of notes: | 139771 |
NO MARGIN OF ERROR FOR DIGITAL --- OR ANYONE ELSE
by Jack Fegreus
� Digital News and Review
May 9, 1994
Used with permission to circulate freely within Digital
Last week, on the heels of the wretched third-quarter financial
results for Digital Equipment Corp., I sat down for an hour with CEO
Robert Palmer. The interview embraced numerous topics, ranging from
Digital's survival in a tide of red ink to the company' involvement in
what may well be the first wave in the late David Stone's "information
appliance revolution."
Digital has begun a fundamental transformation,, and while its
tactics may at times seem restrained, unsure, or even wrong-headed, the
underlying strategy remains clear and rock solid.
Every company in the computer industry --- whether systems giants
Digital and IBM, software monoliths Microsoft and Novell, or specialty
companies dedicated to a particular niche like MTI and Quark --- faces
similar problems in coming to grips with the essential discontinuity in
computer price/performance metrics created by the RISC microprocessor.
Many of those companies seem to have already weathered the storm. But
as Little Buttercup pointed out in Gilbert and Sullivan's "H.M.S.
Pinafore", "Things are seldom what they seem."
Far from being a trailing-edge indicator of the disruption caused
by the introduction of RISC computers, Digital's ongoing troubles are
much more likely a leading-edge indicator of the struggles that are yet
to come for many computer giants. As the price/performance of hardware
heads south faster than a college student on spring break, intense
downward pressures will be brought to bear on all facets of the
computer market. Expect the road-kill count to increase as we
accelerate down the information autobahn.
Probably the greatest myth to implode will be the belief that the
desktop PC and the companies that specialize in that market are all at
the head of of the trend toward lower price/performance and somehow
insulated from its potentially devastating effects. Nothing could be
further from the truth. While those in the desktop market entered it
at much lower price points and amidst much fiercer competitive forces,
they entered a market whose rate of technology change --- from '386 to
'486, or from '030 to '040 --- was no more precipitous than that of
going from CVAX to NVAX.
Certainly the PowerPC chip in the Macintosh market, and perhaps
Alpha, Mips, et al., in the Windows NT market, will alter the desktop
landscape significantly. Exponentially cheaper MIPS brings greater
demands for software functionality, which in turn calls for greater
storage for larger programs and even larger data files, which therefore
puts greater stress on networks, and so on and so on. And because the
cost of the CPU traditionally sets our expectations for all of our IT
costs, no company is safe. In the Darwinian world, there is no
Switzerland.
The net result is incredible pressure on profit margins. Digital
was not alone in dealing with the dichotomy of rising quarterly
revenues and falling quarterly profits of the computer market. Yes,
the company is losing an astonishing $1 million a day, but that's a far
cry from the even more astonishing $3 million a day that it was losing
when Palmer took over in October 1992. For Digital and every other
company in the computer market, the challenge is to rapidly make the
transformation from being revenue-driven to becoming margin-driven.
For Digital, the easiest --- and the bloodiest --- part of this
strategy is mostly over: wringing out excess capacity. The next
phase, to standardize as much of the product line as possible, is well
under way. Just look at the DEC 2100 to see where the future lies.
Remove the Alpha CPUs from that system and what remains could have come
from any vendor, from AST to Zenith. Even the last phase of the plan
has already begun, as Digital spins-off those product lines that don't
directly help to define it as the "open client/server" company.
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 3091.1 | MSBCS::BROWN_L | Fri May 20 1994 14:14 | 4 | ||
Assuming 91 (or 92) days in a quarter, and $183m loss last quarter,
that works out to almost exactly $3m/day (not $1m/day), so Palmer
has returned it to the same rate as Ken. kb
| |||||
| 3091.2 | Check both peoples math! | MIMS::SANDERS_J | Fri May 20 1994 14:20 | 8 | |
re .1
183/91 = 2.01
183/92 = 2.01
NOT 3.0
| |||||
| 3091.3 | MSBCS::BROWN_L | Fri May 20 1994 14:28 | 2 | ||
embarrassment... yes, $2b a day.
| |||||
| 3091.4 | OKFINE::KENAH | Every old sock meets an old shoe... | Fri May 20 1994 15:08 | 1 | |
Two gigadollars a day works out to about $23+ per second... | |||||
| 3091.5 | NOTIME::SACKS | Gerald Sacks ZKO2-3/N30 DTN:381-2085 | Fri May 20 1994 15:25 | 3 | |
> embarrassment... yes, $2b a day. More embarrassment: it's $2m, not $2b. | |||||
| 3091.6 | ok, nevermind ;-) | MSBCS::BROWN_L | Fri May 20 1994 15:27 | 1 | |
| 3091.7 | SMURF::STRANGE | Steve Strange - USG | Fri May 20 1994 16:23 | 6 | |
re: .5
Hey, when you're talking millions, what's three orders of magnitude?
:-)
Steve
| |||||
| 3091.8 | Read the article, don't spell check it :) | NZOMIS::DUKE | Sun May 22 1994 19:28 | 33 | |
Morning,
This piece really make more sense than most I have seen lately. It very
clear that with PCI and Risc, component plug and play is moving from the
desk top to the departmental machine and within a short time to the large
back ends.
The result will have to be "margin" run businesses. The only major
current ones in our industry are component vendors.
In some areas Digital is well placed, a sinlge machine, using PCI and
Risc with multiple operating systems. In the areas I work I can't see
Digital ever again selling 7000 or 10000's.
Within the cabinet of the future will we see Mips processors, PowerPc
etc in the 2100 ? I can't see why not really. Technically this could be
hard but why not ?
It would be nice selling really, buy my cabinet and from me buy
whatever conponents you need. Components would of course include
processor, disk, memory, operating system, data base, applications etc.
As a supplier we could then provide the best in class to our customers
all the time rather than just "Digital". In addition in this time of
customer choice there would be far greater flexability to work with
previously non Digital accounts.
I have noticed recently that with our local PC group leading in major
accounts that we could not previously get thru the door with that is
easier and actually they are keen to work with us having once had the
experiance.
| |||||
| 3091.9 | We are on the right track ! | RTOEU::KPLUSZYNSKI | Tue May 24 1994 03:31 | 24 | |
Customers want to get two simple things when they buy a computer:
- Buy the (currently) best product for the best price
- Get a reliable solution that supports their business
Modularity of a product like Sable is important to lead in
price/performance, because it can use the best products on the market.
Providing a reliable solution to the customer requires us, the computer
vendor, to take over the task of integrating the components, do the
quality assurance and support the usage of the product.
And by "components" I mean all parts of a solution, starting from
hardware and operating systems, including middleware, all the way up to
applications, consulting, project services and support. Basically
anything a customer needs and wants to buy.
We certainly will not develop, manufacture or provide all components
ourselves. But we are the ones to manage and integrate our partners and
suppliers in such a way, that the products are created that the market
wants to buy from us. And that is the essential job of Marketing.
Klaus
| |||||
| 3091.10 | RISC is not for granted | POLAR::MOKHTAR | Mon May 30 1994 14:29 | 4 | |
i think we should not take it for granted that RISC ( Alpha,PowerPC,
Mips ) will replace intel's 80x86 design. I think it is possible
but we need to be more aggresive with Alpha to do it.
| |||||
| 3091.11 | LGP30::FLEISCHER | without vision the people perish (DTN 223-8576, MSO2-2/A2, IM&T) | Mon May 30 1994 18:26 | 13 | |
re Note 3091.10 by POLAR::MOKHTAR:
> -< RISC is not for granted >-
Intel seems to be planing on a transitional architecture, one
that perhaps favors easy emulation or partial emulation of
the classic architecture.
I'm sure that they will make every technological -- and
marketing -- effort to make sticking with "Intel Inside" the
apparent easiest path.
Bob
| |||||
| 3091.13 | Pile Em High, Sell Em Cheap. | LARVAE::TREVENNOR_A | A child of init | Fri Jun 10 1994 03:19 | 24 |
RE: .-3
Yes. The whole industry is going shrink-wrapped and commodity.
Inevitable the day that Unix, TCP/IP, DOS, Windows became the
dominant technologies that everyone had to integrate with.
There will still be high-value custom projects, but there will be
intense competitions for those.
In general the companies that stay around will be:
1) Lean and mean. We are painfully staggering towards this, Ihope we
dont trip and fall.
2) A "famous" company. That is, a household name - something that
Digital definitely cannot aspire to be (try the "name 3 computers
companies" test on someone you meet at a party). The "DEC" (what
95% of our customers call us) versus "Digital" (what we pompously
insist on calling ourselves) debate has a bearing here.
3) Able to make/sell price/perf leading products with confidence and
knowledge. We can (and are) doing most of this.
Alan T.
| |||||