SPARC history
SPARC history ..
image shows Megabyte reading a scroll - the top 50 SSD articles and recommended SSD blogs
SSD articles ..
pcie  SSDs - click to read article
PCIe SSDs ..

SPARC Product Directory

from the makers of
sparc product directory

News about SPARC systems and related companies

2002, January week 3

SPARC history / SSD market history / memory channel SSDs
news about summary
Dy 4 Systems

Force Computers
Dy 4 Systems Becomes a Business unit of Force Computers

FREMONT, Calif. - Jan. 21, 2002 - Force Computers today announced that Dy 4 Systems has become a business unit of Force Computers, which results from parent company Solectron's December 2001 merger with C-MAC Industries. Dy 4, a leader in ruggedized embedded computing products for the defense and aerospace markets, will maintain its brand identity and focus on providing commercial-off-the-shelf (COTS) products for defense and aerospace.

"With the integration of the Dy 4 and Force product lines, we bring Force's cutting-edge hardware and software know-how from the commercial and telecom spaces to defense and aerospace—which we'll leverage to make our products better and our portfolio even richer," said Tom Quinly, president of Force's Dy 4 business unit. "And existing Dy 4 capabilities in developing DSP solutions are immediately deployable into Force's current OEM business. Overall, our combined strength positions Force to drive the next wave of embedded computing and enables us to offer increasingly competitive solutions over the entire program lifecycle." ...Force Computers profile, high availability SSDs
SSD ad - click for more info
SAN software
SSDs in the cloud
The SSD market is moving into a new phase at the core of which is changed thinking about the role of memory and storage and software.
Where are we heading with memory systems and software?

View from the Hill:

The Emerging Bright Spots in the SPARC market

Last year was the worst year in Sun's 20 year history. The combination of the IT recession, 9/11, and Sun's shooting itself in the foot with technical problems in its cache memory (thereby blowing away its hard won reliability image advantage over Wintel) and the slowness of developing faster SPARC processors could leave you with the false impression that all was doom and gloom in the SPARC systems market.

Not so.

Despite the computer industry having a bad time in 2001 (lowlights included the first ever year to year decline in PC shipments, and the biggest ever losses in the semiconductor market) there were a few bright spots emerging which you may have missed, if you thought Sun's fortunes were typical of the SPARC market as a whole.

  • The rackmount SPARC server segment continued showing double digit year on year revenue growth as customers turned to telco style packaging to reduce cost of ownership. This trend was also reported in the Intel architecture server market, and was a very fast growing segment for Dell Computer. Because rackmount configurations require a high degree of customisation and testing to meet specific user requirements, this is a segment that's not actually dominated by Sun, even though many (but not all) of the companies in this market area do use Sun motherboards. The "pile them high and ship them fast" strategy which worked so well for Sun in the boom dotcom days, doesn't work in this more conservative market segment.
  • The military SPARC market started apprehensively in 2001 as manufacturers and integrators expected to feel the cold winds of the budget cuts promised by the Bush administration finally working their way through to them. But 9/11 changed all that. Most integrators in this market have been reporting publicly (and privately) an upsurge in spending. For the US government national security has replaced tax cuts as the #1 priority, and the underinvestment in security and and defense during the last few years is now seen as a mistake. 2002 and beyond will see high growth rates for companies in this area.
  • The fault tolerant high-availability SPARC server market saw many companies reporting very high double digit revenue growth in 2001. Originally this was a niche segment within the telecom market, but I expect that this kind of system will branch out into the mainstream, and the whole segment could easily see 50% year on year growth in 2002. Although Sun has a toehold, it is really a bit player in this market. Nevertheless Sun benefits from supplying cPCI SPARC motherboards to some of the HA SPARC companies (although Force Computers actually ships more SPARC cPCI cards than Sun.)
  • The SPARC portable market, like the Intel portable market, has not seen the same down downturn experienced on the desktop. While figures are not available, we've seen a new entrant Naturetech coming in to challenge the market leader Tadpole-Cycle. Several other companies also have toeholds in this market, and if the desktop market continues to decline during 2002, we may see more viable alternatives appearing in a portable format.
Outside the SPARC market, other fast growing segments in the enterprise computing market included:-
  • Solid state disks (which many users are turning to, to speed up their SAN access). This segment saw a 500% year on year increase in buyer interest during 2001, and despite falling memory prices, many SSD makers reported high double digit revenue growth during that period. I expect this to accelerate during 2002, as more users become aware of the benefits of using this type of technology.
  • Many companies in the Tape library segment reported double digit revenue growth during 2001 while remaining profitable. Increasing volumes of digital data everywhere make libraries the most cost effective way of archiving network information.
  • The wildcard, and unexpected success story in 2001 was the huge surge in interest in iSCSI. The cutback on IT spending has made this look like an attractive option for mid range users who have not yet wired up with Fibre-channel. It offers the promise of getting some SAN functionality by leveraging existing IP networks. Capability is one thing, but performance is another. But many users will be happy to use the off-site disk synchronization offered by iSCSI and don't really need the faster performance of fibre-channel in their internal networks. There are few real products in this market, and many vendors have been taken by surprise by the speed with which this new concept has been picked up by the user community. If "interest" translates into "buying" then this will be one of the fastest growing product areas in 2002, and could do much to restore the sagging fortunes of Cisco and other IP switch makers.
And what about Sun?

Although, the prospects for many SPARC companies look good for 2002, Sun, like most large corporations will be slower than more nimble medium sized competitors to react to changing customer needs. It demonstrated its inability to do that in the fast growing network storage market in recent years, where despite its best efforts it has lost significant market share. It may be 2003 before Sun again reports quarterly revenues as good as its historic highs.

In 2001 Sun proved what many in the semiconductor market have been saying for years, that Sun's strength lies in writing C code, operating systems and compilers, and not in chip design. Unless Sun can cure the constipation in its chip design area to get faster products out the door, it will lose more enterprise server market share to the faster Intel chips.

This may be the best opportunity ever for Fujitsu to show its chip making prowess. The 2GHz SPARC chips are what the market is waiting for.
SSD ad - click for more info
Sun Microsystems Sun Microsystems Reports Second Quarter Results

SANTA CLARA, Calif. - January 18, 2002 - Sun Microsystems, Inc. (NASDAQ: SUNW) reported results today for the second quarter of fiscal year 2002 which ended December 30, 2001. Revenues for the second quarter were $3.108 billion, a sequential increase of 9% compared with revenues reported for the first quarter of fiscal year 2002, and a decrease of 39% compared with the same period a year ago. Net loss for the second quarter was $99 million (excluding losses on Sun's equity investment portfolio, restructuring charges, in-process research and development and related tax effects).

For the first six months of fiscal year 2002, Sun reported revenues of $5.969 billion, a decrease of 41% compared with the same period a year ago. Net loss for this period was $257 million (excluding losses on Sun's equity investment portfolio, restructuring charges, in-process research and development, and related tax effects).

Sun's Chairman and CEO, Scott McNealy stated, "Sun's vision, strategy, and execution continue to provide focus and strength in this difficult economic environment. We're innovators, pure and simple; always have been, always will be. Sun is a low cost provider by concentrating on one chip technology, the SPARC architecture, and one operating environment, the Solaris platform. This means we can efficiently leverage R&D investments across products ranging from multimillion dollar mainframe replacement servers right down to sub-$1,000 entry-level servers." ...Sun Microsystems profile

Editor's comments:- no surprises here for SPD readers, the year on year revenue dropoff is in line with previous results, and is due to the special circumstances which made last year a particularly difficult year for Sun. IBM did much better, and may have gained from Sun's woes see STORAGEsearch news today for the details.
Red Hat Linux vs. Unix: An IDC Study Comparing Total Cost of Ownership (TCO)

January 16, 2002 - Sun's competitors in the Solaris X86 space have lost no time establishing where they stand on the Intel versus SPARC (and other RISC processors) issue. Red Hat, which you may remember dumped SPARC, from its support list a few years ago, seems to have a better crystal ball than most. They commissioned research company IDC to do a study at the back end of last year, and if you're interested, you can sign up on their web site to get the resulting white paper, "The Role of Linux in Reducing the Cost of Enterprise Computing".

They claim the report will show a 45-80 percent lower TCO for Linux on the Intel architecture over RISC/Unix environments. I never sign up for anything, so I haven't read it. But Red Hat and othe Linux makers would seem to be the main beneficiaries of Sun's recent decision to pull the plug on Solaris X86.

If you care about Sun's decision, and believe they will take the slightest notice of what users tell them, then Save-Solaris-x86.ORG is a site where you can voice your concern in an open letter to Sun. If there's a real market opportunity out there, and enough Solaris X86 demand, maybe some future ex-Sun techies will do a startup to produce a Linux which can run Solaris binaries... They'll need good lawyers though.

There's actually no good technical reason why a SPARC/Solaris server should cost any more than an Intel/Linux one. A smart move for Sun would be to issue a press release with a pricing roadmap to reassure users.

See also:- Linux portals, ...Red Hat profile

PS - now Sun will have to pay another company to write a report which says that SPARC has a lower TCO etc, or maybe they'll just write their own like they seemed to do with storage... It shouldn't be too hard. Since SPARC system pricing is totally within Sun's control, they can just tweak the prices of the models in the report to make the result come out right... I doubt if anyone believes these kind of comparisons or acts on them. They just stir up hot air in the press on a slow news day.
Solaris on x86 Solaris on Intel Has Been Axed

January 16, 2002 - according to an article in ZDNet today, Sun has terminated plans to do a production release of Solaris 9 on Intel architecure processors.

See also:- Surviving the Solaris x86 Wars

Editor's comments:- this move, designed to save Sun development and marketing resources in the short term, will start a surge of users away from Solaris onto Linux. It cuts off one of the business development options for Sun, which I outlined in a recent article:- that of becoming a reseller for Intel servers running Solaris.

It's a high risk strategy for Sun which will increase their profitability, but also reduces the available market which Sun can address, because some companies are never going to buy SPARC.

So the great experiment has ended...

This makes it even more critical for Sun to produce new SPARC chips and re-establish good relations with other OEMs who use them. Many moons ago, Microsoft had a similar experiment, with a product called Xenix, which was a version of Unix running on Intel. I can't remember when they dumped it, but that didn't do their business any harm. One way to view this move is that Sun is saving costs by cancelling its Intel insurance policy. That's OK as long as there's no fire. Now Sun will make or break on the SPARC platform. That's a good thing if they get it right. But they can't afford another chip related fiasco. The next thing they should dump (or transfer to Open Source) is Java.
GNP Computers
. "The Product of the Year award is meant not only to honor the outstanding teams who developed and manufactured the products and services on our winning list, but also to educate our readers on the vast array of offerings in this market. Choosing the winners was a difficult process, and the editorial team spent many hours studying the applications we received from vendors, as well as looking through materials we collected on our own over the past year," said Internet Telephony's editorial director, Greg Galitzine.
GNP Computers Receives Internet Telephony's 2001 Product of the Year Award

MONROVIA, Calif. - January 15, 2002 - GNP Computers received Internet Telephony's 2001 Product of the Year award for Continuant Cluster Suite, its applications development framework and operating environment for high-availability telecommunications applications.

Unveiled in June 2001, GNP's Continuant Cluster Suite was recognized as an innovative solution because of its ability to deliver significantly more than 99.999 percent availability in carrier-grade telecommunications platforms based on Commercial off-the-shelf (COTS) technology. In addition, it provides a complete applications development environment that includes documentation, embedded intelligence and professional services for optimizing the application architecture, implementation and execution.

Continuant Cluster Suite was developed for applications such as softswitches, gateways and controllers, intelligent voice response, wireless, Voice-over Internet Protocol (VoIP) and other critical systems that require fewer than five minutes of downtime per year. ... Internet Telephony 2001 Products of the Year article, ...GNP Computers profile

Editor's note SBS Technologies also won an award for their CP61016 platform

current news
Marketing Views STORAGEsearch SPARC Product Directory ACSL - the publisher

SPARC(R) is a registered trademark of SPARC International, Inc. SPARC PRODUCT DIRECTORY(SM) is a service mark of SPARC International, Inc used under license by ACSL. Products using the SPARC trademarks are based on an architecture developed by Sun Microsystems, Inc.