the SSD Buyers Guide - click to see article
SSD buyers guide
these are the SSD companies you have to consider on any buyers shortlist - this quarterly series predicts future market winners by tracking search volume of millions of SSD buyers
top SSD oems ...
SSD myths - write endurance
SSD endurance ..

SPARC Product Directory

from the makers of
sparc search
How Sun Got Burned... What should Sun do now?
October 9 - 2001:- This is an occasional column by Zsolt Kerekes
See also:- Why did EMC's revenue implode?
Looking Back on Sun's Cache Memory Problem
more articles, storage news, market research
October 5, 2001 the messages start coming out from Sun Microsystems.

Quarterly revenue will be 38% below the number for last year, 4,000 jobs will be shed to reduce costs and get back into profitability. We've been hearing messages like this all year from other computer computer companies. But Sun's revenue decline is worse than most of its leading competitors. And the climb back up the hill will be harder. Did it have to be so bad?

My article in April this year Why Did Sun's Revenue Growth Hit a Brick Wall? showed that Sun gained more than most other companies from the dotcom boom and lost more tha most in the dotcom bust, and Sun's communications talk about the effects of the recession, about the impact from the September 11th atrocity. All are true, but not the whole truth. The root of Sun's problems today extend back years. To reclaim its leadership role Sun will have to address many internally created problems. This article is my analysis of what has gone wrong and what needs to change.

/ Branding / Public Relations / Advertising / Resellers / Storage / IHVs / End users /
Zsolt Kerekes - Publisher
Zsolt Kerekes has been editor and publisher of the SPARC Product Directory since 1992. Zsolt is also editor of STORAGEsearch.
A market research based action plan and guide for 4,000 nervous Sun employees. Don't miss it. "Goodbye Sun! Hello New Employer!" In a recession? I'm deadly serious. If you know someone who may be affected, tell them to take a look.
Branding The problem:

Sun has become disassociated from what should be its strongest brand:- SPARC.

IHVs and VARs in the storage market sometimes ask me about my other publication. "What's that SPARC thing about?" they ask. "The Sun market" I reply. "Oh. We sell in the Sun market, it's important to us, but I've never heard of SPARC."

When was the last time you saw an ad from Sun with the folowing kind of headline?

Sun's new SPARCmainframe! Uses good old fashioned solid SPARC technology, still scaling processor architecture after 14 years.

Sun started winding down its branding programs for SPARC in 1995, when it launched its first 64 bit Ultra 1, and 2 workstations. Before then all workstations and servers had been given names with a SPARC prefix, such as SPARCstation, SPARCserver etc. The Ultra was a weak brand. Ultra what? Other Ultra brands in the computer market include the better known UltraSCSI. So that was a tactical mistake. But in 1996 Sun went even further and decided that the JavaStation was the hot new product of the future.

When I went to a Sun partners event the following year the word SPARC was not even mentioned. "This is a company which is trying to break away from the one thing which has made it successful" I remember thinking.

I parodied Sun's current relation to the SPARC brand in my spoof article about Sun's shutdown in July Diary of a Workaholic Sun Partners Program Manager . I make no apology for quoting this extract below.

"Next, take a look at another historic program, called "SPARC-compatible." When I first saw it in the research notes, I looked up "SPARC" in a dictionary, but unlike "Catalyst", it wasn't there. Guessed it might be foreign, and tried a German online dictionary. No luck there either. Asked researcher. She told me it was a pre-Solaris, pre-Java, pre-Jiro legal term, like a copyright symbol, and never used in marketing communications in any font size larger than 6pt in print or 20 pixels high on the web. I looked at the program details and was horrified to find that it included a lot of our competitors in the Solaris Compatible and Sun Ready markets. As including direct competitors is one of the no-no's for the new program, decided not to read any more about it."

What should be done now?
  • rename all new Sun products, even the Sun Fire mainframe to new names with a SPARC suffix
  • start an advertising campaign which leads on why SPARC has been so successful in the past and why it will continue to be successful in the future
  • increase funding in the branding organisation SPARC International, Inc so that it can advertise and promote the brand more widely

Selected Reader Comments:-

A reader from a large user site which I have verified, but who wishes to remain anonymous made the following observations.

"I'm a diehard sun fan. Been a sysadmin on Solaris for 8 years, but I disagree with Sun pushing SPARC as brand.

Reliability? We've replaced over 1000 UltraSPARC 2 400mhz cpus because of the ecache issue. Ecache issues alone cost us millions of dollars. Will the U3 fix that? maybe, but we've already had several DOA U3 chips on new U3 boxes. That does not look good from a management standpoint.

Scalability? Yeah you can stick 72 cpus in a box and scale. And the new E15k looks good for scaling. But the price/performance for the U3 just sucks compared to other chips out there. If I'm paying a premium, I expect the type of performance the Power4 is showing not a Pentium Pro. Sun needs to step to the plate and find the numbers somewhere. They need to invest in the chip and get the peformance up or drop it and move on. It's absymal, if Intel can get the scalability, SPARC is dead.

Before Sun can push SPARC they better improve the above two issues. More and more Sun shops are looking towards Intel and other solutions for performance and reliability." ...Editor Verified Anonymous Reader

Editor again - I had similar comments about the Ecache reliability problem from a major Sun reseller a few weeks ago, but forgot to mention it because everyone has problems from time to time with their chips. Remember the Pentium floating point bug a few years back? Also I have a vested interest in the success of the SPARC platform, which is why I'm keen that Sun should take some positive action.

Reader Eliezer Ramm, who comments below, works in a web services / ASP company.

I totally related to your article "how Sun got burned" I told someone that dealing with Sun is like the abused wife who keeps on coming back to her husband.

Sun needs to do one thing in these trying times. Promote Sun!! Not Java, or the other stupid projects. They should learn from Corel who had one great company and burnt all their cash on stupid projects (Word Perfect, Linux, Games) and died. Sun needs to do one thing:- SPARC and Solaris. Stop buying stupid companies !!!!!!!!!!

Sun should be evangelizing Solaris. Big Admin should read like LinuxToday !! Next time they feel like blowing a few million on something stupid they should spend it on free certification programs or free tech support programs. Why don't realize that the happier I am as a customer the more likely I am to purchase more equipment.

I called Sun in the USA with a problem, however the guy answering the phone couldn't verify my right to tech support and wouldn't pass me on. So I said " here you guys are, taking a week off to save your company, yet me, who has 10 Sun boxes you won't help ?" Don't they realize that satisfied customers = more purchases?

What about Solaris x86? Take some of the money that you spend on stupidness and invest it in Solaris x86. People who use x86 will go out and buy SPARCs when they can. Those who use Linux will continue to do so on Intel boxes. I am not saying compete with Linux but x86 should be just as good. Heck ! It comes from the same source tree as Solaris.

Help us the users sell your stuff. Case in point. I was at a seminar given buy Sun professional services here in Israel. The guy was giving a pitch how to build an environment for an e-business. 3 web servers, 2 app servers a database w/ storage. I asked the guy if he had any benchmarks to help gage how many hits each server can handle. He pointed me to some sites like and specbench. I told him those sites show Sun in last place !! His response "Sun is more concerned with placing boxes than individual performance" go tell that to your boss !!!!!!

If Sun is the best prove it !!!!!!!!!!!!!! I know you can.

Sun is beautiful because it has only focused on Solaris during its' history. This is wonderful compared to IBM/SGI and their unix/nt/linux phases.

Sorry for the rant but I have been meaning to get this off my chest for a long time.

Lets organize the thoughts and comments your article generates and send them to scott, ed and bill.

Bring back SPARC/Solaris to the people who use it , know it, and love it!


Public Relations The problem:

During most of the last year Sun's press releases about market share and its business growth have been disengenuous, and deliberately contrived to create the misleading impression that things have been very much better than they actually were. Examples abound in our news archive where you will see my own comments and warnings. Now the chickens have come home to roost, and many Sun users (if they haven't been seeing the news on this site) will be very surprised and shaken by the apparently sudden turnaround.

What should be done now?

Sun's PR department always prefer putting out positive messages rather than negative ones. But as the company needs to start a recovery plan, they should quote a few less optimistic statistics from the pre recessionary days and admit that Sun is doing worse than the IT market at large. That's already started this week.
Advertising The problem:

For years Sun's advertising has been aimed at showing that Sun can waste money on advertising in the (badly targeted) consumer media just as well as the worst of the dotcoms. The message, if there was any message in most of that advertising, has been saying to me (and to many of Sun's partners)

"We at Sun are so successful that we have money to burn, and this stupid ad, in a media which isn't even a computer media, proves it."

Sun has historically shown a deep aversion to advertising in independently managed publications which are aligned to the Sun, SPARC or Solaris platform, unless it can exert some form of control. In many cases, such as the sites, it has tried to close those independent publications down.

What should be done now?

Spend less on print advertising. Spend more advertising on the web. Sun marketers don't have to like every article they read in publications like this one. The important fact is that independent publications like this do serve a useful purpose and are used by lots of readers who are trying to find their way around a very complex market. Advertising is a way to communicate to those people and expand the overall SPARC compatible market.
Resellers The problem:

Sun had a very successful relationship with VARs in the 1980's and 90's. Sun's proud boast was that aimed to sell 80% of Sun product through resellers. Since the recession started, it has seemed that Sun would like to change the ratio around the other way. The practise of shifting systems via eBay and bypassing the VAR channel and taking more business direct, started long before the IT recession. Sun has been enviously eyeing Dell Computer's business model and working to disentangle itself from the hundreds of companies which propelled Sun from an unknown to a major server platform provider. That alienates many VARs (I know, I talk to more of them than Sun does.)

Looking ahead, as systems become more complex, and the server becomes a smaller part of the cost of each system, the role of VARs is going to expand, not diminish.

What should be done now?

Sun should just admit that it was wrong. It needs VARs. Now more than ever. Sun should give VARs money to educate them into better marketers, so they can do a better job of understanding the emerging markets, and the selling opportunities for SPARC systems.
Storage The problem:

I identified the problems of Sun's failure to grasp the importance of the emerging storage market early in 2000, and wrote about it extensively. (See - Will Sun Succeed in the STORAGE market?) Since then, Sun has been on a buying spree of storage companies. But it was mostly too little, too late.

To understand the scale of Sun's monumental failure in storage, particularly if you are a shareholder in SUNW, try to imagine what the shares would be worth if you added half of EMC's revenue to Sun's during the last year. That's a simplification, but it gives you a true picture of the opportunities lost. EMC has its own problems now, which I write about in STORAGEsearch, but the storage market is still growing.

What should be done now?

The list of acquired storage companies keeps growing, but there are still more than 500 manufacturers and around a 100 software companies listed on STORAGEsearch, including recent storage startups. Should Sun buy another company now? I don't think so. Maybe the least worst strategy now is to resell other storage products, which Sun did before, and with the recent HDS deal, Sun continues to do now.
IHVs The problem:

The SPARC market has always needed Independent Hardware Vendors (IHVs) to introduce the latest technologies such as iSCSI, InfiniBand, NAS, or whatever it is at the time, but the barriers to entry for IHVs who want to introduce new products into the Sun market are too high... Because the Wintel market is larger and easier to access, there's a risk that Sun users will not always get the best products first.

IHVs often tell me that it's very difficult understanding the Sun market, and selling into it. That's unlike the Wintel market which is very well served with a wide range of independent information sources and publications. New IHVs usually go to Sun to find out how they can market their products to the installed Sun base and get reliable marketing information. This is a problem with the IHVs rather than Sun, who will happily sign them up to marketing program and sell them some advertising on Sun's own web site. Then they're left on their own. Unless they successfully identify that "SPARC thing" with the Sun market, and end up safely here.

What should be done now?

The real solution is that the independent vendors in the Sun market need to improve their market research about publications and information channels in the SPARC market, and spend more money advertising on them. My own list of Sun, Solaris & SPARC portals, and Linux portals is far from complete, but it's a good start. Sun should refer vendors to external independent publications, which are aligned with the SPARC platform. Instead of just one publication they should have a list. That will help their partners better without costing Sun anything.

Sun does actually have links to this publication, but in some very obscure pages in a massive web site. I'm suggesting they have a SPARC publications list, including their own publications, linked from their home page.

Another helpful suggestion for IHVs is to consult our advertising page because this publication is the #1 source of independent information about SPARC compatible vendors.
End users The problem:

Without end users who believe that Sun's products and the company itself are going in the right direction, there would be no future. In recent years Sun has become arrogant, disconnected and unapproachable to many people who would regard themselves as Sun users or evangelists. Worse still, many of Sun's communications about where it's going and what it thinks is important have been confused with a jumble of hard to understand inconsistent and irrelevant marketing slogans.

"Is Sun going into the wireless business?" I've been asked. "Does anyone know what this Jiro stuff is?"

Any company, as large as Sun does inevitably get involved in many strands of product and technology development, but the main communication channels from Sun have been blurred by a fog of hype which at best has been confusing, and at worst, has been insulting the intelligence of its users.

Sun is not going to control the desktop anytime in this parallel universe. JavaStations and their descendants do not a great company make. Hyping every crazy idea which proclaims the end of the Wintel platform, is just nuts. Ignoring the reality of heterogeneous systems has been folly.

The Solution

Sun should admit that its users often know more about the realities of computing that Sun. Sun's strengths in the past are the key to its recovery and success in the future...
  • SPARC architecture - scalable from portables to mainframes and bulletproof reliability. Let's hear more about that. In letters more than 20 pixels high.
  • Open collaborative marketing with VARs, IHVs and others who can help to improve the platform. Sun doesn't need another partner program. It needs to say "Sorry guys, we realise we do need you." ...and mean it.
Sun is not alone. It didn't get where it is today without friends. Its friends and partners and allies are waiting for it to come back down off that cloud and participate in making everything work better.
SSD market history
SPARC market history
How will hard drives fare in an SSD world?
SPARC Product Directory SPARC computers SBus & PCI cards
SPARC manufacturers USASun/SPARC Resellers in the USA UKSun/SPARC Resellers in the UK

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.