Date: May 20, 2001

********************************************
George Gilder's "Storewidth" Conference was held April
10-12, 2001, at the Ritz-Carlton Hotel, Laguna Niguel, CA
********************************************

(to subscribe to our list of some 2500 New Economy mega-power
players--well, their portfolios *used* to be huge, anyway--and receive
future conference reports, click here and give us your contact info)

Dear clients, partners, friends,
and fellow data storage hogs:

Hard to believe -- here I was at my favorite coastal resort/surfing spot, again, about to get the lowdown on another Next Big Thing, they were telling me. Right after the last Big Thing I'd covered, days before in the next state East (see previous report). Amazing.

And you thought things were slowing down.

Okay, so you say it's been a few weeks since this gig went down. Whaddya expect -- free *and* fast? Could a guy please make a living in between getting his brain filled with Next Big Things every so often? Huh? This is work, people -- sorting through where computing and communications is going. No one said it'd be easy when I took this thing on, no siree. Besides, you know what they say now about working in Internet Time: sooo yesterday.

So I took my damn, sweet time. And not much changed in the storage industry in the last few weeks, anyway. Okay, maybe disk drive margins dropped a few more points -- but that's it. All right, 10 more storage-management startups probably got funded, too -- in the province of Alberta alone. But that would definitely be it. I promise. The rest is exactly as you will read it here.

So settle back, buzzhounds, and allow me to enlighten you about what the assembled throngs of storeheads at the beach had to say about their white-hot industry. It's one reporter's attempt at recapping the veritable firehose of content that was "Storewidth 2001"...nonstop, full-on, open spiggot.

Actually, I had my report pretty well summarized some time ago and set it aside. But I kept coming back to my voluminous notes, wanting more -- finding myself fascinated with the topic, the players, the characters, and the coming drama in this whole business -- trying to sort it all out, and generally doing further study of the rapidly evolving storage landscape. (And I was catching lots of media coverage, too, in these ensuing weeks.) After all, we're just in the early stages of this...well, really next big thing.

The largest part of this period was spent in one critical time-warp: I decided to slog through two of George's newsletters on the subject, so I might be able to store and retrieve even more facts and insights for you, Dear Readers. Well, that damn-near froze my personal operating system. Talk about detail! Gilder is amazing -- he (and his crack staff) have this unique ability to bring you to your knees, screaming for mercy, as you wade through their rambling prose, blubbering acronyms and tech specs and all day long and in your dreams -- forgetting your name and where you live.

Hey, George, I didn't want to know THAT much! But, doggone it, you read it all -- every damn word. Kinda makes my reports seem like, well, Fun with Dick and Jane. Nonetheless, for those of you who just haven't had the time to plow through his writings on the subject or travel to his confab, here's an overview (bite-sized and rapid-fire for sanity, I hope) of what transpired at the very excellent "Storewidth 2001" last month, and some of what became apparent thereafter...

------------------------------------------------------------------------------------------

High Drama Following the Money
to IT's Biggest Challenge

Storage Companies Smell Big Margins Upstream;
Even Possible Future Killer App for PC Industry?

by Graeme Thickins
graeme@thickins.com

A juiced-up bunch of mostly technologist-executives in the rapidly changing data storage industry -- with a notable contingent of VCs and private investors mixed in, hanging on every word -- gathered recently in Southern California for a major insider confab/gabfest/debate about the big changes happening in their industry. It was fitting the event was organized by a man who's achieved almost cult-like status as an technology analyst, and whose writings bring nothing less than high drama to new technology breakthroughs.

It became clear at George Gilder's "Storewidth 2001" conference last month that what has loosely been called "the storage industry" is unquestionably becoming a major power center in information technology -- because, simply, it's where the big money is.

To wit, it's now by far the fastest-growing segment in corporate IT spending, and the industry seems downright giddy to be there.

Virtually all the key established players were present at this high-energy/high-buck event, along with many brash upstarts -- threatening, almost taunting, to eat their predecessors' lunch.

It was American ingenuity and technological innovation at its entrepreneurial best -- quite befitting George Gilder's role as the leading cheerleader thereof.

But how techie was the event, you ask? Well, a 10-page list of acronyms was handed out with the program materials -- intended mainly for those newer to the field, one would assume. (Although an acronym contest was held the second evening, which definitely required a cribsheet. The techies were really into it -- you had to be there.) And, in the mostly suited crowd, the two most quotable speakers were a guy in a tie-dyed T-shirt and another with a pony tail--and this isn't even a developer crowd, where such is almost the norm.

All in all, it was a real experience, and a huge amount of content to process. No one complained of not getting their money's worth, nor their fair share of power-schmoozing.

Notably, the event was one of the first to add a pay-per-view webcast (at a reduced amount from the full conference fee), for those who couldn't travel to the site. To illustrate the interest in the topic, the streaming technology firm that handled the webcast (iBeam) said that, to its knowledge, it was the largest-grossing webcast ever -- even before the 30-day archive period that allowed additional viewer registration.

I told ya it was really big.

Some Key Points to Get You Started:

- The term "Storewidth" grew out of the notion of the "hollowing out of the computer" -- which included storage moving out onto the network -- a concept Eric Schmidt, then CTO of Sun, first discussed with Gilder seven years ago. (And Schmidt was here to continue the discussion.)

- The idea is that storewidth and bandwidth feed off each other -- "two virtuous cycles," one speaker said.

- The event tagline got the least respect of all: "The End of Storage" was intended to imply, one would think, the follow-on phrase "as we know it." Yet many speakers felt compelled anyway to point out "It's just the beginning." All in how you look at it, it seems. But what's certain is that the storage industry of old is not the one of today and tomorrow. Whatever you call it, it's fast headed upstream to the bigger margins in managing data.

- More data will be created in the next three years than in all of human history, according to a UC Berkeley study.

- Data stored by ASPs and SSPs (storage server providers) is doubling every four months.

- IT spending is growing at a compound annual growth rate of 11%; spending on PCs and servers is growing only at 6%, while storage is growing at 22% annually -- twice as fast as overall IT spending.

- Another speaker said 64% of IT spending is now in storage -- "and growing" -- also citing a Meta Group report that says stored data will increase 10 times in the next several years.

- Disk drive companies and suppliers to them are increasingly margin-pressed. How do they continue to squeeze out even more manufacturing efficiencies as prices keep dropping? How do they break through physical limits? The conference gave no easy answers to their plight, unless the laws of physics change soon. The obvious concensus is that the margins are well up the foodchain, in storage-management solutions and software.

Notable Insights from "Storewidth 2001":

- The battle continues between two main competing storage management technologies: SAN (storage area networks) and NAS (network attached storage). And executives from the two respective key players thereof presented at the event: EMC (SAN) and Network Appliance (NAS). One fundamental difference between the two technologies, if you care: SAN is block storage, NAS is file storage.

- But perhaps the biggest buzz was a hot, new upstart, which many think is poised to beat the leaders at their own game and thus attracted serious attention with their presentation: BlueArc, makers of the Silicon Server -- the "world's first server for the optical age." It helps that Gilder is a huge fan, having written extensively about the company's technology recently. It also helps that media coverage has been everywhere lately, including another positive piece in the Wall Street Journal on May 7. BlueArc says the real storage bottleneck is server software, so they say they will achieve in hardware (that is, in silicon) what is not possible with software: optical speed, meaning 2 gigabits -- blowing the bottleneck blockage.

- According to longtime Compaq and DEC storage expert Richard Lary (the very quotable tie-dye guy), storage over IP is coming, because it's so much easier to do. "But it will take till 2004-05 for the the crossover to happen," he said, adding that fibrechannel startups "still have time to make money." In the meantime, he said that storage over IP will be used primarily as a way to bridge fibrechannel SANs, and for low-end storage.

- EMC is now saying that the file server is the bottleneck, and it's just announced software for "frictionless" access to information -- for "any source, format, amount of distance, speed, or performance."

- Network Appliance says two key areas of storage it plans to exploit are these: iSCSI, for block storage on network wires, and DAFS (direct-access file system), for file-based storage on direct wires, plus "content-to-edge content management."

- Venture money is pouring into storage technology and storage management startups, fast replacing optical as the hot VC sector -- a fact largely unnoticed in the media frenzy over VCs' problems with failed Internet companies, the market crash, and the IPO dropoff. Many have been busy indeed focusing squarely in this space, and a large contingent of the promising startups were at this event with bells on, along with a strong contingent of those VCs and other investors.

Storage Facts & News to Ponder...and Amaze Your Friends With:

[Attribution noted when speakers provided it.]

- World data will double in the next 2.5 years, one speaker claimed, again citing the recent UC Berkeley study.

- By 2002, 70-80% of server costs will be storage related (Meta Group).

- For every $1 spent on storage, $0.91 is spent on management (Cahners).

- 60% of every dollar spent on storage hardware is underutilized (Gartner).

- Flying heights in disk drives are now 1/10th of the current state-of-the-art 0.13 micron line widths in ICs.

- Storage capacity is now doubling every 9 months, while bandwidth and processing power double about every 18 months.

- How far has disk drive technology advanced? IBM now sells a tiny one-inch "microdrive" that stores a gigabyte of data.

- Investing in companies involved in data storage, access, and management more than *quadrupled* in two years, to $695 million in 2000. [Note: bet on the number growing in 2001!]

- Storage is projected to be a $112 billion market by 2004, with the storage service provider (SSP) portion of that estimated to be $40 billion.

- EMC was the best performing stock of the last decade, up 84,000%.

- EMC says that its average company has about 20 terabytes of data and has a disk drive failure every week.

- Not too many years ago, the entire amount of data stored worldwide was 75 terabytes, but now many companies have that much *each*.

- Network Appliance has one customer with 350 terabytes of storage, adding 3 terabytes a week.

- New "holographic" storage technology (now in development by InPhase Technologies) will let you store the equivalent of 20 CDs each on postage-stamp sized media, and up to 400 CDs on a 5 1/4" size...and let you do things like download a feature-length movie at a video kiosk in two minutes.

- MEMs (micro-electronic memory systems), chips with moving parts, promise access times of less than one microsecond -- an order of magnitude better than 1999 disks.

- Other new alternatives to disk storage, now in development, will allow densities of 4.5 terabytes per square inch, blowing away what had previously been called a physical limit. (One speaker's comment: "We're finding the limits are negotiable.")

- The first WDM (wave division multiplexing) optical home network is coming this year, said an Excite@Home executive, and the lucky market will probably be Dallas.

- BlueArc, which recently raised $72 million in a third round, says its new Silicon Server network-attached storage system is performing in tests 10 times better than EMC, and 5-6 times better than Network Appliance, has 35 times the storage, 100 times the user connections, and five nines

A Sampling of the Best Quotes from Some of Storewidth's Big-Name Speakers:

"At its current growth rate, the costs to store information will
soon exceed the U.S. GNP! So...ah...something must be done."
- Bill Miller, CTO, Storage Networks

"Ten years ago, only 5% of the world's content was created
digitally. Today, that figure is more than 90% -- and it
all has to be stored!"
- Michael Ruettgers, Chairman, EMC

"The disk guys are the real heroes. Never before has so
much brain power been put to work for such low margins!"
- Richard Lary, longtime DEC storage engineer, patent
holder, Compaq fellow (and tie-dye wearer)

"What will it take to master this new world? Infrastructure
that can mask the complexity from the user."
- Michael Ruettgers, Chairman, EMC

"I don't agree with Ruettgers (EMC chairman) that
(startup) BlueArc's not going to succeed."
- Keith Brown, Director, Strategy & Technology,
Network Appliance

"Disk head flying heights have reached the point where
it's now like a 747 flying 1/50 of an inch off the
ground at 1.5 million miles an hour."
- Richard Lary, longtime DEC storage engineer, patent
holder, Compaq fellow (and tie-dye wearer)

"The future requires simplicity. What's needed is to store
and manage information as an application itself. We're trying
to mask the complexity by giving customers high-level tools."
- Michael Ruettgers, Chairman, EMC

"We're a hot industry now! Historically, we've been just
above the power supply. All of a sudden, this is where
it's at! I'm thrilled to see the VC interest -- we're
the only sector getting money."
- Keith Brown, Director, Strategy & Technology,
Network Appliance

"On a cost-per-gigabyte basis, disk has blown away
DRAM and optical, and it will blow away tape, too."
- Richard Lary, longtime DEC storage engineer, patent
holder, Compaq fellow (and tie-dye wearer)

"Storage is in some sense already too cheap and too fast.
But it's too expensive to manage!...Dynamic management
is the back-breaker today. It's so bad that companies
are *overbuying* storage to compensate."
- Richard Lary, longtime DEC storage engineer, patent
holder, Compaq fellow (and tie-dye wearer)

"The Holy Grail is to make storage capacity and performance
management as simple as memory management in servers."
- Richard Lary, longtime DEC storage engineer, patent
holder, Compaq fellow (and tie-dye wearer)

"We're still looking for an architecture we can call a
managed infrastructure....Ask if it's scalable *after*
you know its manageable."
- Jon William Toigo, pony-tailed storage consultant and
author, "The Holy Grail of Data Storage Management"
(a Gilder favorite)

"There are no good vendor solutions yet for fibrechannel
security....Using an SSP (storage service provider) is a
good way to outsource your problem!"
- Richard Lary and Jon William Toigo, respectively
(agreeing)

"I'll cut my hair when a real SAN comes to market! Because
fibrechannel is not a network -- so how can it be a SAN?"
- Jon William Toigo, pony-tailed storage consultant and
author, "The Holy Grail of Data Storage Management"
(a Gilder favorite)

"Fibrechannel has a big 'forklift upgrade' coming...Keep
your eye on Adaptec. They missed fibrechannel, but are
now firmly in bed with Cisco in iSCSI."
- Jon William Toigo, pony-tailed storage consultant and
author, "The Holy Grail of Data Storage Management"
(a Gilder favorite)

"I'm a big believer in storage going to the edge. There's
a huge appetite out there for small servers...Very soon,
you'll even be able to have 2 terabytes at home for a
couple thousand bucks. Napster proved storage at the
edge works!"
- Janpieter Scheerder, EVP, Sun Network Storage Division
(who also led the development of Solaris; now retiring)

"Manageability has to get automated! Where should the
management live? Everywhere! That's why open APIs
and protocols are so important...Software interoperability
standards are needed now."
- Janpieter Scheerder, EVP, Sun Network Storage Division
(who also led the development of Solaris; now retiring)

"Nobody wants to do backup, but everybody wants data
recovery."
- Janpieter Scheerder, EVP, Sun Network Storage Division
(who also led the development of Solaris; now retiring)

"File servers will move out of the OS. The only way to
solve the problem is out at the edge."
- Janpieter Scheerder, EVP, Sun Network Storage Division
(who also led the development of Solaris; now retiring)

"My philosophy is information is only as true as your
ability to know where it came from....The storage industry
is very loose about what is 'data' vs. 'information' --
that must change."
- Janpieter Scheerder, EVP, Sun Network Storage Division
(who also led the development of Solaris; now retiring)

"I was the lone voice in 1993 talking about storage!"
- Mark Leslie, Chairman, Veritas

"CIOs aren't really interested in 'storage management'.
Data availability is what lights them up."
- Mark Leslie, Chairman, Veritas

"Our new appraoch is to replace the 'backup' paradigm.
It's a '70s style, batch process that consumes IT
resources and is obsolete as soon as it's done. We
say integrate it into the application production drive
train -- backup each block in real time."
- Mark Leslie, Chairman, Veritas

"Our growth is 50%, and our clustering business is
growing at 200% a year. We had a great fourth quarter.
This is a $10 billion opportunity!"
- Mark Leslie, Chairman, Veritas

"The eBay outage caused the loss of a billion dollars
that day in their market cap, which they never recovered."
- David Wright, CEO, Legato

"There are no tools to manage SANs today. But only
5% of enterprise storage is in SANs now -- 95% is
still direct-attached. The game hasn't even started! ...
Over the 5-year life of a SAN, 90% of the costs are
in management. We're still about 18 months away from
a full-fledged solution."
- David Wright, CEO, Legato

"With current virtualization, 2 terabytes takes about
100 hours to restore -- which is not acceptable. So,
ask vendors about restore, not just backup!"
- Jon William Toigo, pony-tailed storage consultant and
author, "The Holy Grail of Data Storage Management"
(a Gilder favorite)

"You never eliminate the bottleneck, only move it.
You're chasing your tail."
- Mike Wendling, Architect, Excite@Home

"Move the bottleneck out onto the fibre!"
- George Gilder (then moderator)

"This whole thing is about data sharing. That's what
SANs are battling with now -- trying to pull control
from the OS! And now the application software vendors
are getting in."
- Jon William Toigo, pony-tailed storage consultant and
author, "The Holy Grail of Data Storage Management"
(a Gilder favorite)

"We're thinking about how to build persistent data
across the web, out to the edge. But it all must migrate
back -- that's very important."
- Mark Leslie, Chairman, Veritas

"I'm starting to lose my enthusiasm for SANs and clustered
storage. It's not a panacea. Point solutions for specific
apps may be the way to go."
- Jon William Toigo, pony-tailed storage consultant and
author, "The Holy Grail of Data Storage Management"
(a Gilder favorite)

"This SAN thing will take a long time."
- David Wright, CEO, Legato

Words of Wisdom from the Guy Who Started It All:

Eric Schmidt first proposed the notion to George Gilder of the "hollowing out of the computer," with storage going out onto the network, some seven years ago. Then CTO of Sun Microsystems, which has long held that "The Network Is the Computer," Schmidt and Milo Medin, CTO of Excite@Home, planted the seeds of interest with Gilder for the "storewidth" premise and his subsequent conference series. Schmidt now serves as chairman of Novell and Google, but is also involved in other emerging, development-stage companies.

What follows are some key quotes from Schmidt's very well received keynote, kind of a rambling look at everything tech...and a highlight of the event's second day:

"The next killer app on the 'Net is content -- voice, video, rich content. That's where I'm now focused."

"Our industry tends to over-estimate what we'll accomplish in one year, but *under*-estimate what we'll achieve in ten."

"Scaling the 'Net is the defining challenge of science."

"So, what's happened lately? The tortoise caught the hare! The fiber connection is faster than the computer network."

"What's next? Fiber directly to the CPU. And also better fiber...We'll end up with nets that look a lot like MANs (metro area networks). 'Sphere' networks -- campus like."

"We've been talking about IPv6 for some six years now, but it's not been widely adopted. I like what The Economist said: 'Four billion addresses, for four billion people, for four billion galaxies -- that should last a while.' "

"Current economics is forcing centralized nets. The capital markets funded all the growth, ahead of the curve. But centralized will win -- the ones that own the fiber. Who's that? The telcos."

"Instant messaging has too many protocols. They can't talk to each other -- it's ridiculous!"

"802.11b and Bluetooth -- one's gonna happen...and 3G is not going well so far."

"In ten years, you'll get the entire contents of Google on a disk drive in your pocket." [Note to disk drive companies: he thinks you'll still exist then! Problem is, they'll probably sell for a dime a dozen.]

"The storage companies have a problem -- they're not good enough. We'll end up with a hybrid solution." [Note: implying neither SAN nor NAS will win.]

And for you Napster fans out there...or biggots, for that matter:

"My focus now is rich media and distributed directories."

"Napster is a distributed directory with enduser-distributed content. That's the next major architecture above the substrate I'm now describing."

"In his book 'Life After TV' in '92, George Gilder described what was coming as 'A new worldwide web of glass and light.' That's the earliest major reference I've found accurately predicting what's happening." [Note: borrowing a phrase my nephew uses to describe his first-born: George is a little advanced.]

"We had enough problems when we went from 50 to 500 channels. What happens when we go to 50,000? Will it be regulated?"

"The biggest mistake of voice over IP was calling it telephony. Because now it's regulated."

"Here's my prediction for the next 120 years: for the next 20, we'll have a pretty good time. Then, overcapacity will cause consolidation. The analogy will be like the railroad business -- exactly. Read the history! The standards were set by the Federal Government -- Lincoln. We're going to be part of a political process."

"Censorship -- we'll see the issue over and over again. People want information and will get it!"

"Here's a picture of the next Internet: the rise of 'always on'. Voice, video, and data...quality will be expected, and will happen."

And Schmidt's closing summary remark:

"Information sharing and communication are the past, present, and future of the Internet."

The audience Q&A after Schmidt's talk provided a few more of his insights:

"Instant messaging is just an app -- and a bad one! One that reaches only a portion of the people inherently is."

"How we upgrade the Internet is a huge technological discussion, and a big problem. We're getting to the point where India needs more address allocation -- and China, too. There's a possibility IPv6 will come via the proliferation of wireless...Some say do it via tunneling, but I think the only way it will happen is with two protocol stacks -- two Internets...IPv6 is only ten years late."

In answer to the question, "What are the Napster implications for network architecture? What happens now?"...this was Schmidt's response:

"Pure P2P (peer-to-peer technology) has fundamental scaling problems -- I haven't yet seen them solved. But I think the way to go is 'P2P server optional'. If the server is present, then you need something like Akamai -- and the lawyers go after the peers... But P2P, the people want this!"

In answer to the question "A key tool of the Web has been search, but isn't it less effective now?"... Schmidt had these comments:

"It's still number one, with email number two. Google has grown very rapidly...This business will scale with accuracy and more timely hits....And some in the future will have to be paid for."

"The underlying growth of the Internet is not slowing! Cell calls are cheaper, everything's on sale! The consumer wins!"

Quotes from Some of the Key Up-and-Comers Present at the Event:

"If you can dream it, you can implement it in silicon."
- Dr. Geoff Barrall, SVP Engineering & CTO, BlueArc

What about EMC saying the server should be eliminated?
"Only for large files and streaming."
- Dr. Geoff Barrall, SVP Engineering & CTO, BlueArc

"NOTHING costs more than information management."
- Thomas Isakovich, Founder & CEO, TrueSAN Networks

"Our goal is building a ubiquitous utility of storage.
The Internet *is* storage -- plug, play, and forget.
There is no market bigger!"
- Steve Mattioli, CEO, Yotta Yotta

"The server used to be able to produce more data than the
pipe could carry. But that changed between 1998 and 2000 --
by an order of magnitude!...The standard server supports
hundreds of users. Ours supports tens of thousands, and
has up to 250 terabytes of storage each."
- Dr. Geoff Barrall, SVP Engineering & CTO, BlueArc

"Traditional server architecture was built for computation
and cannot achieve gigabit throughput! Our objective was
a server that works at optical speeds: 2 gigabits. So we
built it from the ground up."
- Dr. Geoff Barrall, SVP Engineering & CTO, BlueArc

"With all the storage approaches out there--proprietary
SANs, semi-proprietary SANs, the proprietary "VI"
NAS protocol, and hybrids thereof--it's starting to look
like the Unix wars or yore. On the same timeframe as
peace in the Middle East."
- Steve Higgins, Director, Software Technology,
Procom Technology

"The bottleneck in most systems is disks that are down.
Our approach, adaptive network storage, is to push it
out to the network."
- Steve Higgins, Director, Software Technology,
Procom Technology

"200,000 simultaneous streams can saturate the entire Internet."
- Simon Khalaf, President, Volera

"The Metafabric Architecture we're introducing today uses
software that brings SAN and NAS together...Some companies
are buying both, when they shouldn't have to!"
- Thomas Isakovich, Founder & CEO, TrueSAN Networks

"I/O is the bottleneck, so that's where we're focused
as an IP storage company. Expect Gigabit Ethernet and
IP to be dominant in I/O in the next few years...The
problem with fibrechannel today is that the switches
don't talk to each other--it's not scalable beyond a few
switches...Many solutions today are being designed
for data centers, not companies."
- Aamer Latif, CEO, Nishan Systems

"The Internet is going to a ubiquitous pipe -- intelligent
devices, PDAs & pervasive computing, P2P, instant
messaging, and many-to-many interactions, which is
what the 'Net is best for!"
- Scott Davis, VP & CTO, Mangosoft

"Hey, folks -- keep making storage cheap and difficult to
manage, Because I'm in the storage software business, and
the margins are great!"
- Steve Murphy, CEO, Fujitsu Storage
Technology Corporation (Softek)

"It's amazing how many organizations don't see they're
overprovisioned and low yield -- which is why we have
a product that does visualization of their SANs."
- Steve Murphy, CEO, Fujitsu Storage
Technology Corporation (Softek)

"Storage prices are plummeting, yet it's still
the biggest part of the CIO's budget."
- Steve Murphy, CEO, Fujitsu Storage
Technology Corporation (Softek)

"This virtualization thing (backup & restore)
is boring. But boring like a bank!"
- Steve Murphy, CEO, Fujitsu Storage
Technology Corporation (Softek)

"A host of storage service providers is being created,
almost weekly...To be in the game, you have to eliminate
storage management hassles, since it's not corporate IT's
core competency, and free up IT resources they can apply
to their core business objectives. And you must reduce
the storage procurement cycle."
- Jaff Ascah, CEO, Storage Alliance

"Leveraging the abundance of bandwidth ain't cheap...
We saw plenty of SSP hype in 2000, but there are few
survivors...We'll have 40 sites by June, for several
big, strong customers."
- Bill Johnson, CTO, Managed Storage International

"The object is to reduce the complexities of management.
But this is the end of the current SSP model. The
telcos and co-los will own it....'utility' services."
- Matthew Westover, CEO, Storability

"I found 29 SSPs when I was planning this conference!"
- Mary Collins (moderator), Technology Analyst,
Gilder Publishing

"We've raised $150 million to date, and our market is
the Global 250, whose storewidth needs are well defined."
- Jon Oltsik, VP Marketing, GiantLoop
(a "storewidth implementer")

"We're just 21 months old, have raised $290 million in
equity, have no debt, and we're now in 20 cities...Ours
is the first nationwide network of managed optical nets."
- Ron Young, Cofounder & CMO, Yipes Communications
(another "storewidth implementer")

"We're saying good-bye to ATM and Sonet, and hello to
IP, GigE, and DWDM (dense wave division multiplexing)...
GigE's growing 3 times faster than Moore's Law, and Yipes
16 times! ...IP over GigE yields huge savings...To have
disruptive technology, you must have disruptive pricing...
We provide double the bandwidth at half the cost of a T1."
- Ron Young, Cofounder & CMO, Yipes Communications

"Ethernet will win! Not fibrechannel. GigE and IP will
be the network for all enterprises! ...We have lots of
confidence we're *all* going to buy more bandwidth --
as George Gilder always says."
- Ron Young, Cofounder & CMO, Yipes Communications

One VC's Perspective on Storage:

At lunch on the second day, Barry Eggers, general partner of Lightspeed Venture Partners, gave an overview of this hot sector. [Note: Lightspeed's biggest claim-to-fame to date? An early investment in Brocade.]

Here are some of his comments:

"Storage networking will be the 'optical' market of 2001 and 2002. There will be over-investment, with many hopping on the bandwagon."

"Startups trying to compete with EMC face a tough battle when the gloves come off. And a pretty conservative attitude from customers."

"In looking through the crystal ball haze, I think every good market needs consolidators -- and EMC, Brocade, and Cisco are the three main ones here. But will storage service providers survive? Which ones? Only the well capitalized."

"There are many positives. The growth is happening. But the market is getting crowded."

"Religion and technology don't mix." [Implying don't place your bet on one or the other of the currently competing storage-management approaches winning.]

"At Cisco (where he once worked), we were always told: 'The only thing to be religious about is customers'." [That is, knowing your market is what matters, not focusing on some industry-insider standard or approach.]

In closing, Mr. Eggers offered up his email address -- so, startups, fire away! It's eggers@lightspeedvp.com

Even *More* Lively Quotes from "Storewidth 2001":

"Yahoo's Broadcast.com did 100,000 hours of streaming
video in a recent month. They're probably the largest
consumer of bandwidth there is."
- Jay Adelson, CTO, Equinix

"American employees have wasted 2.5 billion hours
waiting for downloads. At $20 per hour, that's
fifty billion bucks!"
- Simon Khalaf, President, Volera

"And 35% of Internet content is delivered with
8-second latency! We have a basic plumbing problem
here, and an expectation problem..."
- Simon Khalaf, President, Volera

"TV quality is 300 kilobits per second. It ain't
gonna happen today! Blockbuster is still the best option."
- Simon Khalaf, President, Volera

"Our solution is three-tier. We keep arguing here
about edgewidth or storewidth -- we need both! What's
needed is content delivery like the delivery of goods
to a supermarket -- then companies pay for prominent
display of those goods (and it fans out to consumers
from there)."
- Simon Khalaf, President, Volera

"Let's take the analog of TV and learn from it! Preload
the content into servers and deliver it locally!"
- Simon Khalaf, President, Volera

"With distributed computing, the hares (PCs) multiply
fast -- Napster taught us this. Now, the last-mile
problem relating to TV quality on the Internet can be
solved with this technology!"

"We have a financial problem on the 'Net (showing a slide
with a complex pattern of content providers). Who's paying
for what? We don't know who's screwing who here!"
- Simon Khalaf, President, Volera

"Hold onto TV for a while -- the 'Net is still under
construction. But we'll have TV quality by 2003."
- Simon Khalaf, President, Volera

"Broadband is still a supply-contrained world --
people want it more than we can get it to them."
- Mike Wendling, Architect, Excite@Home

"AtHome's original core architectural principle
was moving data to the edge."
- Mike Wendling, Architect, Excite@Home

"Having a broadband connection to the 'Net makes
latency *more* of an issue, not less."
- Mike Wendling, Architect, Excite@Home

"Current CDNs (content distribution networks) are
useful -- Akamai, iBeam, etc -- but they're not really
at the edge, though they are effective at getting around
interconnection bottlenecks....There's a physics problem
here: an OC3 can handle about 450 simultaneous streams,
an OC12 about 1800. But, with 3 million subscribers,
29% of whom have multiple PCs, and growing, it would
take multiple OC48s! So the answer, again, is edge
caching...and huge increases in storage capacity help
enable it. We're in a unique position to deliver a
national footprint of that."
- Mike Wendling, Architect, Excite@Home

"In the future of 'Net appliances and set-top
boxes, we'll see edge caching extending into the
home, feeding households servers -- TiVo-style
and PC-like -- pushing as far out and as close to
the user as possible."
- Mike Wendling, Architect, Excite@Home

"Why are so many smart people working on this
and not realizing? The edge is in my PC, not on your
CDN! With Napster, latency is not a problem! But,
then, latency isn't an issue when you think of how
long it takes to find something you like on TV!"
- a particularly vocal audience questioner

In response to the audience question, "How do
you cache dynamic web pages?"...

"Only 60% of pages on the web are now cacheable.
But the challenge is to work with major vendors
like Vignette and Broadvision, which create pages
that are not cacheable. Solutions still have to be
found...perhaps from companies like InfraSearch
(a P2P startup recently acquired by Sun)."
- Mike Wendling, Architect, Excite@Home

-----------------

Some Must-Know Storewidth Acronyms (even a few not on the conference list):

- SAN: Storage Area Network
- NAS: Network-Attached Storage
- DAS: Direct-Attached Storage (now declining in usage)
- DAFS: Direct-Access File System
- MAN: Metropolitan Area Network
- WAFL: Write Anywhere File Layout
- WDM: Wave Division Multiplexing
- SSP: Storage Service Provider
- iSCSI: SCSI over IP
- SCSI: oh, come on--you know "scuzzy"!
- FPGA: Field-Programmable Gate Arrays
(the chip technology used by hot startup BlueArc in its Silicon Server)

-----------------

Funky Company Names to Remember (with dead-serious business plans):

- Yotta Yotta
From the word yottabyte: a trillion terabytes...think
distributed network storage across an optical network.
- Big Bangwidth
Another Canadian upstart present...check out Edmonton--
another hot place for storewidth startups.
- Yipes
As in "That's fast!" (their tagline)...think GigE and
IP...attracting VC money like flies to honey, which
now gets them to a game-over position, they say.

And Now for Something REALLY Different:
A Totally New User-Interface Paradigm...
and Even a Cure for the PC Industry Blues?

An Interview with David Gelernter

Just before lunch on day two of the conference, we saw a videotaped interview of noted Yale researcher David Gelernter, who is also Chief Scientist of startup Mirror Worlds Technologies. The interview was conducted by Richard Vigilante, publisher of The Gilder Report and editor & publisher of The American Spectator.

"We want to make the Internet and the computer disappear," Gelernter opened up. "The things that will make a difference to us will be 'information structures'....coherent and simple. There may be a hundred billion of these..."

"Like our Lifestream technology. I want my entire electronic life available to me in something I understand," he said. "It's like a journal -- everything that comes into my life. Each of these objects gets popped onto the end of my stream -- instantaneously searchable and readily browsable."

"Everything is on the stream," Gelernter said. "The software handles it. I don't even have to know how to search it...It's on the verge of coming to market. We're in beta now...Everything's completely indexed, including images -- you just sweep with the cursor and get a summary of documents, fast...In designing it, we found that, when you add the act of human browsing to the mix, the search function doesn't have to be so powerful -- which is less software overhead."

Q: What is a stream?

A: "A heterogeneous electronic library. It moves as a stream in nature moves. A static structure won't work."

Q: What's driving this?

A: "For one thing, Palms and cell phones, anything with small screens. People need access to a stream, and they can get a good peep at it through even a small screen....The 'desktop'--first via the Mac, then Windows--used a metaphor it really couldn't get away with. A desk is really much bigger, and that metaphor can't really handle someone's life....What will happen is people will vacuum all this stuff out of their directories and put it into their stream, because it will be faster to get at."

Q: Is there a consumer market?

A: "I've got to be able to get at my stream through any computer anywhere. We're failing to get use out of all the information we have. Like digital pictures--where do you put them all? How do you easily access them?"

Q: Okay, we're creating all this stored data, but it's just too big without having access to it -- the natural way, the way we want. So, what happens to storage?

A: "It becomes a higher level, more sophisticated service. We buy capacity and a faster response time to our stream. A provider sells a box, for maybe a year -- a 'Streambox'."

Q: Why not just out on the 'Net?

A: "It could be, but it's a big opportunity for the box sellers. They're so cheap, and local storage is so cheap...There's a utility industry here. Companies will have a service you can subscribe to -- give me my stream, and I'll tune it in."

Q: But if it all has to be the same standard, do we get a Microsoft problem? What about diversity?

A: "I can't see how a single company can do it. And people don't care about the underlying technology."

Q: You're creating the next fundamental interface?

A: "Yes, if it works out. The 'desktop' is dead -- and the whole idea if an operating system is obsolete."

Q: Won't you need to know about my browser?

A: "The browser is obsolete, too. It's so easy, and simple to use. It's an ambitious prediction. I don't feel comfortable making an economic prediction -- it could be our company, it could be others."

Q: How did this all begin?

A: "It was a research project at Yale, then it became commercialized -- which has been a big transition, to form Mirror Worlds Technologies. It's simple and powerful. My goals were esthestic. I'm a writer and a painter, not a technology person. This is like spies in the enemy's camp."

Closing Thoughts and Other Observations About "Storewidth 2001":

- In their attempt to give the folks their money's worth, I think Gilder overdid the content at this one. Too many sessions, which cut too deeply into schmooze time -- equally important for many, if not most, attendees. And, please, kill the evening sessions! Coming back into the ballroom after the cocktail party and dinner to hear *yet more* is too, too much. The brain needs a rest.

- And speaking of too many presentations, too many of them were sales pitches, rather than commentary on issues, and educational or informative in nature. [A constant conference caveat.]

- Finally, there were so many program changes (times switched, speaker substitutions, etc) that it was almost impossible to keep it all straight. Despite the fact that an addendum was handed out for each day, it required major cross-reference and flipping back and forth to try to make sense of it all. I say, just give me one agenda sheet that's correct and complete for that day, the morning of -- and can the addendi!

- Last but not least, the weather report: it was sunny (but cool), the surf was great, and the sunsets were drop-dead gorgeous -- as usual. [If you've never been to the Ritz, you owe yourself at least dropping in sometime for a sunset cocktail.]

Well, that's about all my personal storage management system can handle for now, friends.

The write was fun. Hope the read was worth it.

Now go forth and multiply the file through LANs, MANs, SANs, and WANs everywhere.

your ever-vigilant, stream-of-event-consciousness,
store-and-forward conference reporter,
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Graeme Thickins, Founder/Editor/Analyst
GT&A Strategic Marketing Inc.
*Twin Cities *LA *SF *Anywhere
Voice: 952/944-1672
Fax: 952/944-1673
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
http://www.gtamarketing.com
"Strategies to Win in the Knowledge Economy(tm)"

(a web site and email list for senior executives and power
players in the Internet Economy, including VCs, angels,
analysts, investment bankers, and related professionals,
now 2500+ strong and growing!)

A Little Background:

My conference newsletter began humbly in 1997 as just a
sideline to my consulting business, strictly for friends and
clients -- based on my long career in tech and obsessive
search for what's next. It's grown, essentially one-at-a-time,
to an email list now 2500 strong and counting, made up of
an amazing bunch of smart people, most of whom I know or
have met personally -- and am continually connecting with one
another (which I guess makes me a community or relationship
guy, as well as a writer). The list includes many big-name (you
know who you are) CEOs and senior Internet and technology
executives, VCs, founders, angels, analysts, investment
bankers, and related movers-and-shakers -- all involved in
building great tech companies....including the often-unsung
marketing and business development professionals.

So, it's about staying out in front -- but with a decided
long-term view, grounded in the undying belief that
technology is a major force for good in our personal
lives and in our society.

I hope you like my reports and, as always,
your thoughts and feedback are welcomed.

---------------

See our other recent conference reports at our home page.

Or at this page: www.gtamarketing.com/NewsletterInfo.html

Also see some of our coverage posted periodically at:
Conferenza.com ... the premier source of event intelligence.

---------------

Watch for our upcoming conference reports, which we
hope will include The Industry Standard's flagship event,
"The Internet Summit," July 23-25 in Carlsbad, CA.

If you have suggestions about other tech events you'd like to
see covered, please let me know, would ya? Thanks.