Sun Microsystems is a pioneering tech company that is having trouble getting any respect.
A Forbes article on Thursday notes that the company's market cap has dropped below $3 billion: "The company has become so toxic that no one dares to swallow it."
As Sun CEO Jonathan Schwartz likes to say, the Forbes writers "over-rotate." But Sun has fallen further and harder on Wall Street than its main competitors over the last few years and months. Schwartz has bravely pushed Sun down the path of open source and created demand for its hardware and service via free software, but the big payoff has been slow in materializing. Add in the crumbling economy, and Sun has no choice but to take cost out of its business model.
From a stock market perspective, Sun has fallen further than its competitors.
(Credit: Yahoo)This morning, Sun revealed that it is taking the headcount reduction route to profitability, letting go of 15 percent to 18 percent (up to 6,000 employees) of its global workforce and taking a charge of $500 million to $600 million over the next year. The headcount reduction will reduce annual expenses by $700 million to $800 million.
The economic reality is that 2009 isn't going to be a good year for the tech industry. Sun is facing reality with the cuts. Other tech companies will follow with headcount reductions too. This week, IDC cut its 2009 growth rate for spending on tech by enterprise companies worldwide from 5.9 percent to 2.6 percent. The U.S. growth rate for next year was revised from 4.2 percent to 0.9 percent.
Sun CEO Jonathan Schwartz
(Credit: Dan Farber/CNET News)In the Forbes article, various analysts who cover Sun suggest ways, in addition to headcount reduction, that the company could become more profitable. Among the suggestions: selling the Sparc microprocessor business to Fujitsu, spinning out the Java language group, dropping the low-end hardware business, and selling more customized servers to cloud computing providers.
In an e-mail response Thursday night to my query about the Forbes article--and just hours prior to announcing the layoffs--Schwartz gave his take on the substance of the Forbes piece:
Various analysts have told me our revenue was $299 million last quarter (it was $2.9 billion), that we should lay off 50,000 employees (that would be more than 100% of our employees), that no "real" companies use open source (I guess Google and GE don't count), that we're losing customers in droves (we gained customers last quarter), that we're losing cash (we generated more than $150m last quarter), that Niagara/SPARC is a niche (it was a billion dollar a year business, growing 80% last quarter), that we're losing share on x86 (our biggest competitor was down 18% last quarter, but we grew more than 4%), and that we lost $1.7 billion in cash last quarter (no - we impaired a goodwill asset, just like CNET's parent company, CBS, wrote down $14 billion - it's an accounting change).
So, I'm a tad skeptical of folks looking for sensational column inches... we're very comfortable we're on the right path. We had more than 1,000 requests for our new ZFS-based Storage platforms just a day after launch. And we're deluged with requests from big customers wanting to talk about open source adoption as a vehicle to reduce proprietary licensing fees.
But with even larger companies pre-announcing 15% revenue declines, it's evident the whole industry's got some challenges. I understand everyone's worried, but sensationalism belongs on grocery store checkout counters, not in the business press.
Schwartz is waiting for the world to change, to move to more of a cloud computing model where Sun can power millions of data centers with its hardware, software, and services. This model requires that Sun get more than a fair share of the market compared with competitors like IBM, Hewlett-Packard, Dell and eventually Google. Open-source, free software is Sun's disruptive element. Schwartz maintains that free software brings the marginal cost to acquire a customer to zero and helps drive revenue.
"The majority is going to buy hardware (to run the free software), and not just from Sun," Schwartz said earlier this year.
If Sun cannot intercept enough of the enormous demand for its hardware and services in the coming cloud era, no amount of headcount reduction will earn Sun the respect it craves.
Padmasree Warrior, Cisco CTO
(Credit: Dan Farber/CNET News)SAN FRANCISCO--The cloud was omnipresent at the Web 2.0 Summit as industry executives discussed the migration from the client to millions of virtualized servers as the information pipe.
"There is a lot of hype. We think about the cloud as the next evolution in computing," said Cisco Chief Technology Officer Padmasree Warrior. "It's a way of abstracting the services and applications from the physical resources and using a more on-demand layer."
Warrior believes that cloud computing will evolve from private and stand-alone clouds to hybrid clouds, which allow movement of applications and services between clouds, and finally to a federated "inter-cloud."
"We will have to move to an 'inter-cloud,' with federation for application information to move around. It's not much different from the way the Internet evolved. It will take us a few years to get there. We have to think about security and load balancing and peering," she said. "Flexibility and speed at which you can develop and deploy applications are the basic advantages that will drive this transformation."
Warrior laid out the cloud-computing stack as having four layers: IT foundation, flexible infrastructure, platform as a service, and applications (software as a service). Paul Maritz, CEO of VMware, noted that platforms as a service, such as Salesforce.com or Microsoft's forthcoming Azure, present a challenge for developers. "Developers have to make big bets and choose a platform. I think there is room to see other sources of technology that are more open and standardized," he said.
Paul Maritz, Kevin Lynch, Dave Giroaurd, and Marc Benioff at the Web 2.0 Summit panel on cloud computing.
(Credit: Dan Farber/CNET News)Adobe CTO Kevin Lynch, acknowledged that compatibility at the cloud platform layer is a problem. "The level of lock-in in the cloud in terms of applications running and data aggregation is at a risky juncture right now in terms of continuity," he said.
Dave Giroaurd, president of Google Enterprise, brought up the potential legal tangles of moving intellectual property between clouds. "It's an unclear area of the law as to who owns what," he said. Salesforce.com CEO Marc Benioff touted integration between his Force.com platform and Google and Facebook as an example the way cloud services can be mashed up.
Eventually, cloud computing providers at all of the layers will become more open and adhere to standards that allow for federation and movement between clouds. Maritz sees cloud computing as helping to drive the information economy and stimulate new information marketplaces.
"The challenge is how to selectively and securely make information available so other parties can add value," he said. But just having the data and applications in the cloud won't be enough to create dynamic information marketplaces. "We need to find new data representations and ways to annotate data," Maritz said. Indeed, cloud computing won't be very compelling without what is variously called Web 3.0 or the Semantic Web.
SAN FRANCISCO--After a decade as the chief evangelist in the wilderness of software as a service, which has morphed into cloud computing, Salesforce.com founder and CEO Marc Benioff is having a more challenging time coming up with groundbreaking industry disruptions. But that isn't stopping him from enthusiastically preaching the cloud computing gospel.
Benioff and company have built a $1 billion business and gradually expanded a CRM application, run like Google runs search in the cloud, into a platform that greatly reduces the friction involved in business software development and delivery. With Microsoft recently entering the cloud-computing platform arena with Azure and practically every vendor staking a claim to the cloud, Benioff's vision has been legitimized and turned into the next big thing. But that just makes Benioff try harder. Salesforce.com is no longer the underdog, but Benioff is relentlessly touting his "no software" theme and irreverently characterizing Microsoft, SAP, and Oracle as dinosaurs.
Salesforce.com CEO Marc Benioff
(Credit: Dan Farber)At the Dreamforce annual customer conference here Monday morning, before a crowd of 10,000 adherents, a supercharged Benioff came out on stage, seeking to maintain his crown as the Pied Piper of cloud-based business software. "There has never been a better time for cloud computing and for Salesforce.com," he said. He was likely referring to the troubled economy, which makes cloud-based software services an attractive alternative to traditional software business models.
The news of the day is an evolution of the Force.com, the company's development platform for building and running business applications in the cloud. Force.com sites will allow customers to run their Web applications on Force.com, and takes care of the domain, URL, and RSS management. In effect, the new service further consolidates Salesforce.com's hold on a company's data and public Web presence. The company also announced Force.com for Amazon Web Services, which allows applications to be built between the Amazon and Force.com clouds.
In addition, the company announced Force.com for Facebook, which allows developers to use the Facebook APIs within Force.com applications and tap into Facebook social graph data via the Facebook Platform and Facebook Connect. The combination will lead to social CRM and social sales, Benioff said. "Facebook has over 300,000 pages run by businesses," said Facebook COO Sheryl Sandberg during the Dreamforce keynote. "By coming together with Force.com we are about to unleash enterprise apps on our network," she said.
Steve Fisher, senior vice president of the Salesforce.com platform, showed a Force.com recruiting application running within Facebook. The connection with Salesforce.com could also be another source of revenue for Facebook, beyond advertising.
Cloud computing is becoming mainstream and Benioff is trying to ensure that he is upstream from the competition. He may not remain in that position, but he will continue to push the industry a whole deeper into the cloud.
Logitech announced Tuesday that it has paid about $30 million in cash to acquire SightSpeed, a maker of cross-platform video chat and calling software. It's a logical acquisition for Logitech, which includes video cameras among its large portfolio of peripherals for PCs.
Berkeley, Calif.-based SightSpeed was founded in 2001 and has 25 employees. SightSpeed's video communications solution is based on the session initiation protocol (SIP) and offers 30 frame-per-second video and an integrated instant messaging service. SightSpeed supports 640 x 480 video windows at any speeds above 1.5 Mbps, multiparty conference calls and in-call file sharing. A free version is available and pricing for the SightSpeed Plus for consumers is $9.95 per month or $99.95 per year. Pricing for the more full-featured Business edition starts at $19.95 per user per month or $189.95 per year. The acquisition is expected to close next month.
Guest post: Christopher Lochhead, the retired chief marketing officer at Scient and Mercury, offers some turnaround strategies (learned the hard way) for weathering the economic storm.
Economic downturns require extraordinary leadership. They require brutal honesty. They require action. If your market and company are truly in trouble, here are some turnaround strategies (learned the hard way) to weather the storm so you can live to fight another day.
1. There Is No Such Thing As One Bad Quarter
When your markets get weak and/or you really screw up, fixing it will take a lot longer than you think it will. Pray for spring, but get ready for a long, cold winter.
2. Get The Facts Yourself
People don't like to deliver bad news. As an executive, your job is to get to the heart of the problem fast. You (not someone who works for you) need to figure out how bad your problem is.
How bad is the sales forecast? How late is the next release of the product? What is the cash burn rate? How many critical projects are broken? You must drill into the "whys" to make sure you understand the facts and the causes of the problems. The key is to ask "why?" five times.
Why is the project late (you will get an answer)? Then ask why that is the case (you will get another answer), then ask why that is the case (you will get yet a deeper answer), and so on.
Once you've asked why five times, end every conversation with the most powerful question you can ask, "Is there anything else?" Before my grandmother was heading into surgery to fix a broken hip, I asked her doctor that question. He told me something he had not wanted to tell us--that there was a 25 percent chance she would die during the operation. People don't like to deliver bad news. Real leaders get the real facts so they can take real action.
3. Get 2 Top 10 Lists Fast
Get the smartest, most courageous people in the company together this weekend (no more than 10 as big groups do stupid things) to brainstorm about the top 10 ways to drive revenue and the top 10 ways to cut costs. Here are a few ideas to get you started.
Drive revenue:
- Assign every big deal in the pipeline to an executive and make the execs and the salespeople accountable for closing the deals
- Give customers a new incentive to buy this quarter
- Focus on your core markets and ignore the rest
- Announce a competitive replacement program (provide an incentive for customers of your competition to switch)
Cut cost:
- Do a lay-off
- Pull out of under-performing markets or geographies
- Sell under-performing assets or business units
- Stop all stupid travel, off-sites and trade shows (anyone at AIG from there?)
4. Horde Cash
In March of 2000 as the tech bubble was getting ready to burst, my accountant, the legendary Greg Finely, called me and yelled, "Horde cash!" It's good advice in bad times. Meet with your CFO and finance team to figure out how to optimize the cash, and never forget the sage words of the Coen brothers, "Where's the money, Lebowski?"
5. Tear Off The Band-Aid Once
Take all the pain in one big shot. Cut deeper and make bigger changes than you think you need to. The more quarterly forecasts you miss and layoffs you do, the harder it is to recover. Once you know that your company is in trouble, assume it's in worse shape than you realize. Because it is.
If you are going to miss quarterly numbers and forecasts, miss them once. If you have to do a reorg, do it once. And if you are forced to do a layoff, do it once.
6. Fire Executives
It is stunning how many companies do a layoff without firing any executives. You can't layoff 20 percent of the company without letting go of some of the executives (who are at least partially responsible, if not completely responsible for the problem). And don't just demote them, or move them to some other job. Fire them.
7. Chop The Dead Wood
Every company has people on the team who are "C" players. Rather than doing an across-the-board 10 percent cut, make sure that the people you are cutting are the worst performers in your company. Your "A" and "B" players will appreciate the fact that you did the right cutting. This may sound harsh, but no one wants the "C" players around anyway.
8. Tell The Truth
Some executives think that lying, misleading, and otherwise obfuscating will "soften" the blow in bad times. Wrong. Lying never works. It sounds obvious, but companies and executives do it all the time. It can land you in jail or ruin your career (trust me--I've seen this happen to well-meaning but misguided execs). People hate delivering bad news, so they tell a "white lie," which they often rationalize as somehow doing good for others.
Be honest and direct about the facts. Brutally honest. Be honest with your stakeholders. If you are laying off 25 percent of your people, then say that's what you are doing. Don't say, "We are laying off 15 percent and expect some additional headcount reductions through normal attrition."
9. Communicate Clearly and Powerfully
The truth will never be as bad as the rumors will become. "No comment" will increase the untruths and gossip. It will also unleash the venom of the people you used to be forthright with. The press will attack harder, and your employees' distrust will grow deeper. Both will undermine your efforts with customers and drive your stock price even further down. It doesn't matter how much it hurts. You must over-communicate.
When you're in trouble, get clear about what you are going to say before you open your mouth. Rambling or trying to make 16 points will make you look confused, defensive, or stupid. Then get clear on three--and only three-- key messages to deliver: the facts as you know them, the actions you're taking now, and how your actions today position you for future success. Write these messages down, and practice saying them.
10. Sign A Pact In Blood
In November of 2005 Mercury's board of directors fired our CEO, CFO, and general counsel because of a stock-option accounting problem. Our stock tanked, our competitors attacked, and our employees were scared. The key executives in the company agreed to stick together come hell or high water (and boy, did we go through hell and high water, but that's another story).
We didn't wavier. And neither should you. Nine months later, we had settled the accounting problem, turned the company around, produced some of our best quarters ever, did an acquisition, and ultimately Hewlett-Packard bought us for a significant premium. That turned out to be a big win for shareholders, customers, our people and HP.
11. Drive It Like You Stole It
Legendary teams execute their turnaround plans like it is the last thing they will ever do. Take action. Bust your butt. Get on planes and meet with all of you key customers. Rally your teams in town hall meetings in all of your key offices. Refine your strategy. Focus your efforts. Get your people focused on results. Meet with your top investors to tell them how and why your turn around will work. Get help from some wicked advisers. Recruit new talent to the company. Sell, sell, sell, and lead, lead, lead.
Leading a company through a turnaround is arguably the hardest thing to do in business. If you actually do it and pull through, it will become the most rewarding thing you have ever done in business. Good luck and knock 'em alive.
After 20 years in business and being the marketing chief at three public companies, Christopher Lochhead retired at 38. Now, he serves on a few boards and is a part-time strategy adviser. Every year he gives a handful of speeches, and from time to time writes something. Check out www.lochhead.com.
Microsoft Chairman Bill Gates appeared on the NBC Nightly News Wednesday speaking with Tom Brokaw about the current economic crisis. Gates wasn't concerned about the state of the U.S. economy in the long run. Historical data would support his longer-term view, but that won't make the current disarray and uncertainty about the economy any less scary for investors riding the daily, nausea-inducing roller coaster.
Brokaw observed that Gates seemed to be cool, or not terribly worried, about the U.S. financial crisis. "The U.S. economy in the long run is going to do very, very well. There are some interesting and meaningful decisions to be made in the next weeks," Gates said. He didn't get into the details about those decisions facing Congress, but legislators and the business community are likely seeking his advice. His good friend Warren Buffett of Berkshire Hathaway is investing around $5 billion in Goldman Sachs, providing a confidence boost to the market.
As for continued technology innovation in light of the economic upheaval, Gates said, "In terms of inventing new medicines or improving software, or new ways of doing things, the level of investment will stay very high." That said, conservation of capital will be in the minds of VCs and start-ups until the economy rights itself.
Earlier in the day, Gates told Bloomberg that problems with the U.S. economy would likely reduce government support for combating the world's problems, such as poverty and disease. "There are the rich-world economies and the developing-world economies and, while the degree to which they are linked is not well understood, when one suffers it can't be good for the other. Rich-world budgets may not have room for increased generosity.''
As Wall Street struggles to redefine itself, Cisco is busy this week introducing its latest wave of collaboration products to compete with Microsoft, IBM and Oracle.
"It's a major launch for us, including a comprehensive update to our Unified Communications platform, a new collaboration client for WebEx and Telepresence Expert on Demand," said Rick McConnell, vice president and general manager of Cisco's Unified Communications division.
Cisco's Unified Communications release 7.0 adds support for Windows Mobile, in addition to Symbian and Blackberry. iPhone support is in the works, McConnell said, but Android support is not planned at this point. It also interoperates with collaboration suites from competitors such as Microsoft and IBM. McConnell said that the software suite no longer requires a router for connections from a remote location, such as a home.
WebEx, the Web conferencing service that Cisco acquired in March 2007 for $3.2 billion, has been turned into a Web applications platform. WebEx Connect includes e-mail, calendaring, team spaces, bookmarking, and document sharing, and integrates with the Unified Communications portfolio. Like Salesforce.com, Facebook and other Web platforms, WebEx will allow third parties to build applications (widgets) that integrate with the platform and sell them through a marketplace. Cisco plans to add soft phone, video phone, speed dialing, and federated presence widgets.
PostPath and Jabber, two recent Cisco acquisitions, will be build into WebEx Connect for e-mail and instant messaging in the coming months, McConnell said. WebEx will be priced in the sub-$10 range per seat per month.
Cisco has also found a way to make its high-end, $300,000 TelePresence videoconferencing product more attractive to large corporations with multiple locations. Telepresence Expert on Demand provides integration with the Cisco Unified Contact Center, dealing with scenarios in which an expert can be summoned to a session. For example, a bank branch could call upon an expert in financial planning to meet with a customer via Telepresence.
"With the Telepresence system installed, you hit speed dial, which integrates with the Contact Center and rounds up an expert to address questions," McConnell said.
Most of Cisco's $40 billion in revenue comes from selling infrastructure to power networks and data centers. Don Proctor, senior vice president of Cisco's software group, called collaboration the "next phase of the Internet" and a $34 billion market opportunity. It's clear that Cisco hopes to bundle its communications hardware and software, creating collaborative infrastructure in a box, as it goes after more share of market and mind with its growing product portfolio.
PC assembly at Dell's Parmer North 1 facility in Austin, Texas.
(Credit: Courtesy of Dell)During the conference call reporting on second quarter financial results, Dell's new CFO Brian Gladden said several times that the company has "more work to be done," to improve profitability and decrease costs. He wasn't kidding. Over the last year Dell has cut headcount by more than 8,000, and now The Wall Street Journal is reporting that the company is planning to radically alter its production line by selling off its factories to contract manufacturers.
Dell has four factories in the U.S. and six outside the country. Competitors such as Hewlett-Packard have shifted some of their assembly work to more efficient contract manufacturers to lower production costs and increase operating margins. The Journal noted that Dell may not have an easy time ridding itself of its factories, however:
Dell could face several obstacles to selling its plants. Contract manufacturers may be hesitant to buy factories in places with high labor costs, like the U.S., said one person with knowledge of the talks. And some facilities could be encumbered by agreements with local governments. Dell's North Carolina plant, for example, received several million dollars of state and local tax incentives that are contingent on the factory meeting certain employment and local-investment goals by 2015.
Google is well known as a one-trick pony.
Almost all of the company's revenue comes from its search engine, which last quarter accounted for more than $5 billion. New initiatives, such as the Chrome browser, Google Gears, and Google Friend Connect, are focused on building a mostly open-source Internet operating system out of Google technology in order to funnel more user data and targeted advertising opportunities into the Googleplex financial engine.
It's easy to draw parallels to Microsoft, which gradually built the dominant 20th century operating system and applications platform. Bill Gates and company realized that attracting developers to the Windows platform was key. Google is following that advice with its open-source projects and allowing its mad scientists to try to remake the early 21st century software world and take on Microsoft.
Microsoft has led the way with productivity software, gaining a more than 90 percent share of market with Microsoft Office. Google is hoping to replicate Microsoft's office suite success with Google Apps. It's far less feature-rich than Microsoft Office, but Google Apps Premier edition is far cheaper at $50 per user per year.
For some companies, Google Apps is "good enough," and its cloud-based, collaborative core is an advantage--no Microsoft SharePoint server required. Even with a few enterprise wins, Google Apps is a puny business. According to a Fortune article, Google brought in about $4 million with its Google Apps business in 2007, compared with $12.2 billion for Microsoft Office. Google Apps is a profitable business, according to Matthew Glotzbach, enterprise product management director at Google.
Since early this year Google has been touting 500,000 active business customers, primarily small businesses, using at least one of the Google Apps, and more than 10 million active users. In addition, thousands of universities, with more than one million active users, are using Google Apps, the company said. So far, Google's biggest wins are Valeo, a leading automotive suppliers, with 32,000 users, and the District of Columbia, with 38,000 employees.
However, the vast majority of Google Apps users are not paying customers. The company maintains that "hundreds of thousands" of users are paying the $50 annual fee. The $50 per-user-per-year Premier Edition offers several features lacking in the free Standard Edition, including Postini messaging security, APIs for integrating Google Apps with IT infrastructure, 24x7 support, 99.9 percent uptime guarantee for e-mail, Google Video and 25GB of storage per account.
At this point, Google is underplaying the number of Google Apps business customers. The company has been saying that it is adding 3,000 businesses a day, which amounts to over 1 million per year. The reality today is that Google has more than a million Apps business customers. In addition, the Apps suite continues to fill out, most recently with Google Video.
It took Microsoft years to build a base of applications and developer ecosystem for Windows and Office. Google faces the same uphill climb for Apps and its fledgling Web operating system. The company hopes to ride on the backs of the younger generation that has grown up on the Web and identify with the Google brand. As the Google generation moves into positions of purchase authority within businesses, Google is betting that those decision makers will shun Microsoft, especially as Apps product features improve. Of course, the resilient and relentless Microsoft will respond to Google's challenge when it is more than a $4 million or even $20 million blip.
SAN FRANCISCO--Standing 52 stories in the air at the upscale Carnelian Room in the Bank of America building here, executives from Dell, Facebook, and Salesforce.com discussed the meaning and use of the latest technology buzzword, cloud computing.
The sky was blue and cloudless, but it didn't adversely impact the atmosphere of what turned out to be a Dell marketing event. It was pitched as an announcement about a partnership that involves "the next generation of cloud computing."
You might recall that Dell is the company that owns the URL Cloudcomputing.com, and made a failed attempt to trademark the phrase. Earlier this month, the United States Patent and Trademark Office rejected the company's application. Dell marketing head Andy Rhodes wasn't willing to comment on whether Dell would appeal the USPTO decision.
Despite the cloudless sky, the speakers offered genuine insights into cloud computing, an umbrella term for "hyperscale" computing that covers everything from delivering compute services like a power utility delivers electricity, to simply hosting applications off-premises (see also software-as-a-service and on-demand computing).
(Credit:
Dell)
Event host Forrest Norrod, vice president and general manager of data center solutions at Dell, defined cloud computing as an economic enabler for applications, not just for single applications but for platforms-as-a-service, such as Salesforce.com. He emphasized the economies of scale advantage that cloud computing has over client/server and previous generations of infrastructure deployment.
Forrest Norrod, Dell's cloud computing chief
(Credit: Dan Farber)Dell is currently a cloud computing arms supplier to companies such as Facebook and Salesforce.com. "Dell is focused on early adopters and large customers, about 50 worldwide, to provide optimized servers, storage, and data center infrastructure," he said. "Cloud computing is still an emerging market, with standards across the framework and software stack still emerging. We are trying to promote an ecosystem to build the software stack on top of the infrastructure. You will gradually and judiciously see us add capabilities up and down the stack.
Norrod pointed to recent Dell acquisitions--Message One, Silverback Technologies, and Everdream--as examples of Dell's focus on software, not just the hardware piece. Increasingly, both Dell and HP are building out their software stacks to compete with Sun and IBM for providing highly automated data centers running commodity hardware optimized for cloud computing.
Jonathan Heiliger, Facebook's vice president of technical operations, had some praise for Dell. "Dell is doing the most aggressive things possible to optimize for cloud computing," he said. "We think Dell is perhaps the furthest along and we see them as a thought leader." Facebook has more than 10,000 servers, Heiliger said, and it's safe to assume any of them come from Dell.
He noted the price of hardware is not the biggest issue. Vendors can even sell hardware at a loss or at a fixed margin cost to get the initial business. "What we have seen in the landscape is that most server providers are trying to provide Lexus quality products at a Toyota price. We are looking for Scion products at a Scion price," Heiliger explained. "(Vendors) have to be creative around power and airflow optimization. The cost of operating the hardware is key; you have to take down the operating cost, not just the server cost."
For Heiliger that means bare-bones servers. "We don't need fancy graphics chips and PCI cards," he said. We need one USB port and optimized power and airflow. Give me one CPU, a little memory and one power supply. If it fails, I don't care. We are solving the redundancy problem in software." Blade servers are not ideal, he said, because of the higher cost and proprietary lock-in that come with the lack of a standard chassis.
Check out the video interview I conducted with Heiliger about managing infrastructure hypergrowth as Facebook adds 250,000 users per day.
Claus Moldt, vice president of technical operations at Salesforce.com, offered similar comments to the previous speakers. The company is phasing out Sun equipment and standardizing on Dell servers (Dell is a customer of Salesforce.com). Salesforce.com has two data centers in the U.S. and one due to go online in Singapore later this year. Moldt said his biggest challenge is capacity planning, making sure that as customer usage patterns change, the Salesforce infrastructure can adapt instantly.
Dell is betting big on cloud computing to boost its enterprise footprint. At this point, Dell doesn't have plans to build its own cloud to provide hosting for external applications, Norrod said. But, there may come a time when being an arms supplier won't be enough for Dell to be competitive. In addition, selling bare-bones servers can't be much of a high margin business, which is why Dell is moving more into software and services. Norrod said Dell's cloud computing efforts have been a large component of Dell's recent market share growth. Dell's second quarter earnings due tomorrow should give a more precise indication of the impact of the cloud on company's business.






