Community Development Comes to Hardware

We just introduced the first fruits of our collaboration with Fujitsu at a big launch event in New York - and I want to start by offering my congratulations to a unique group of people: those individuals and teams responsible for joint development at Sun, Fujitsu and in the OpenSolaris community. If you believe, as I, that community development is the future of Sun (and our industry), you can point to our announcement as proof - two companies, and a broad open community, worked together to produce a singular product set that presents opportunity for us all.

Together, we're building a line of SPARC/Solaris machines targeting very high scale computing environments. The machines are general purpose, run Solaris without modification, but offer features and scale that were historically the stuff only mainframe customers could love (because no other computers offered them).

The high end of the new family (called M-class, where the M means Mainframe, not Monster, the latter's appealing propensity to eat dinosaurs aside), delivers the industry's most powerful, general purpose computer in a single cabinet, a one teraflop machine (one capable of performing a trillion instructions per second). It looks like the picture at the left (and before you ask, it's targeted at folks who care about the kinds of high performance computational problems that aren't divisible into fleets of smaller machines - although most social networking sites are easily scaled by the addition of more web servers, the same isn't true for large scale airline reservation or ERP systems, or many simulations).

This isn't the only product in the family, there are a variety of other smaller scale systems - which leverage Fujitsu's innovations around mainframe reliability, ours around Solaris and high volume computing, and our joint expertise in building competitive systems and service organizations. And unlike the Rock systems to which I alluded earlier, our M-class systems are shipping today, and designed for conventional workloads.

Below is the note I sent to my counterpart at Fujitsu, expressing how proud we are of the collaboration - and how hopeful we are about continuing that work going forward.

_______________________________________________

Begin forwarded message:

From: Jonathan Schwartz
Date: April 18, 2007 4:18:05 PM PDT
To: kurokawa-san
Cc: citoh, John Fowler
Subject: Congratulations!

Kurokawa-san,

I would like to extend my heartfelt congratulations to you and the entire Sun-Fujitsu SPARC Enterprise product team on the announced shipment of our newest M-class systems. These ground breaking products are a shining example of our collaboration, and the opportunities opened through partnership and cooperation in broadening the SPARC/Solaris ecosystem.

The performance numbers of these new systems are astounding - I understand we can now deliver a full teraflop of computing power in a single cabinet! Additionally, we are achieving spectacular SPARC64 VI performance results - setting world records for SAP performance in enterprise deployments, and Linpack for high performance computing.

These new systems provide breakthrough capabilities for solving massive scale mission critical problems, from financial transaction processing and business intelligence, to simulation, design, and the delivery of enormous volumes of web interactions.

These are truly mainframe systems - that expand the reach of SPARC systems into markets that were previously out of reach. With up to 64 sockets, 2TB of memory, a 368GB/sec backplane, and the addition of instruction retry, memory mirroring and online repair - among many other previously "mainframe only" features - we've unquestionably brought the choice and competitiveness of open systems to an entirely new market segment.

Optimized for Solaris, and entirely binary compatible with existing applications, we're able to offer customers the highest levels of reliability, availability and manageability without high costs, complexity or vendor lock in. By leveraging Sun's expertise in open, partition-based network computing and Fujitsu's experience in mission-critical computing and high performance processor design, we've been able to offer a family of servers that are unrivaled as virtualization/consolidation platforms.

So again, my thanks and congratulations to you and your team - we're honored to work together, and look forward to continuing the development and evolution of the SPARC/Solaris ecosystem.

With warm regards,

Jonathan Schwartz

Chief Executive Officer,
Sun Microsystems, Inc.

Posted on 11:47PM Apr 20, 2007 | Comments[23]

Share this post  | | | | | |

What Brand Means.

I spent a good portion of a weekend a few weeks ago with a customer that was having a quality problem. There's no point in going in to the nature of the customer or the problem, but suffice it to say it was a bad problem, and by far and away the most expensive kind: one that put the customer's brand at risk. For those that deliver service via the network (or free software), brand is all you've got. It's not an asset, it becomes the asset.

The quality problem I mentioned was customer specific, and had a very real impact on tens of thousands of consumers. And lest it go unsaid, I was really proud of our Services team. They managed the unmanageable, they guided everyone through the unexpected (me, even). They understood what Service meant, they understood the customer's brand, and they understood the role Sun played in fulfilling it.

The saying goes, "a brand is a promise." On a personal level, I've always felt that statement was incomplete. A promise is the lowest common denominator of a brand - it's what people expect. Think of your favorite brand, whether search engine or sneaker or coffee shop or free software, and you'll know what I mean - a brand is an expectation. If you experience anything less, you're disappointed. A promise seems like table stakes.

But a brand must go beyond a promise. To me, a brand is a cause - a guiding light. For fulfilling expectations, certainly, as well as dealing with the ill-defined and unexpected. It's what tells your employees how to act when circumstances (and customers) go awry, or well beyond a training course. My first real experience with that was a personal one.

I was married in 1999. It was a small wedding, presided over by family, in a house in San Francisco. It was the best party my wife and I have ever thrown for ourselves. I heartily recommend throwing a good wedding.

After the ceremony, we gathered up our friends, and drove over to the hotel where we'd booked a suite. We had it all planned out, an evening with our friends, a view of the Golden Gate Bridge, a night in the city before leaving town. They even had a package deal for honeymooners.

We arrived at the hotel, mingled with our friends for a while (I in a tux, my wife still holding the bouquet she forgot to toss), and walked up to the reservation desk after a half hour or so. I said "Hi, we're here to check in, last name is Schwartz." Bear in mind, this is long before my name had any value in San Francisco (unless you were a total geek).

The guy at the front desk crinkled his nose, looked at his computer, then looked back at me - "sorry, we're full up." I didn't know what to say, I figured it was a friend playing a prank. So I said, "You're hilarious. But I have a reservation. We have your honeymoon suite, we just got married." He said, "Nope, sorry, no more rooms, looks like we over booked, really apologize." And after taking a few minutes to gather my sanity (and lower my blood pressure), I asked if he could help us find another room or other accomodations. He said, "Nope, but good luck, if you come back in a week or so we'll definitely have space."

Uh, right.

So we went to another hotel. Just down the road, we ended up marching in at 10:30 at night, in full wedding garb, still. The restaurant was closed, as was the bar. Everything was. And I went up to the clerk at the front desk and asked if they had any room. He was kind enough to notice the wedding dress, and asked if we'd been at a wedding - I explained yes, I'd just gotten married, and he said he'd take care of everything.

He opened the lobby, turned on all the lights, and recruited a couple employees to reopen the bar, and make us feel at home. They served us until the wee hours of the morning, put up with our noisy reunion, then put my wife and me in a beautiful room, and found other rooms for our friends. They managed to put a handwritten note in our room, "Congratulations," it said, next to a bottle of champagne. I don't remember what they charged us - I remember feeling like it was nowhere near their going rate. The following morning, fresh faces at the checkout desk somehow knew to offer their best wishes.

Needless to say, I will now go out of my way to stay in their hotels. I recommend them to my friends. I am a huge fan. Even when their brand breaks their promise, I dutifully fill out my room survey to help them improve. I want them to win, I'm an evangelist. It's not a promise, it's a cause. They are the white hats, I'm on a mission to see them succeed.

The other guys? The other hotel? I never think of them. I don't bad mouth them, life's too short for that. I just don't care. I simply reciprocate the attitude of that clerk seven years ago. Careless indifference.

What's a brand?

It's not a logo, an ad campaign or a money back guarantee. At minimum, it's a promise that helps to define those items. Beyond that, it's a cause that gives definition to the ill-defined, that tells you how to deal with the unexpected or the uncomfortable. It's what motivates you to hire that fellow at the front desk, and to foster his instinct to feel, "Eureka, I found an opportunity to build an evangelist!"

That's not about money or resources or training or contracts. It's a cause. One your employees - and more critically, your customers - willingly join.

Posted on 07:35AM Apr 17, 2007 | Comments[44]

Share this post  | | | | | |

Rock Arrived

We just got the first silicon for our Rock systems - you have no idea how cool it is to have this arrive on my desk in a plastic sandwich bag (don't worry, with packaging like that, they don't send me the ones we put in saleable systems):

That's the front, this is the back:

The chips are running billions of instructions already (not quite at Hello, World, but not far away). Being a professor of the obvious, my first question was - how many pins are on the back? (Answer: nearly 50 on a side, or 2395 pins). Care to guess how many are used for signal (or functionality) vs. power/ground?

Answer... 812 are used for 'signal' (or real work), and 1514 are used for power/ground (the remainder are unconnected). Another interesting stat... for those fond of in-memory databases, Rock will allow us to build coherent systems supporting 256 terabytes in a single software domain (plain vanilla OpenSolaris, no less). That is an awful lot of RAM in a single system (and given the cost of memory nowadays, you'd want to post an armed guard next to that machine).

Rock is 16 cores - we haven't said how many threads per core. Nor have we said why this chip heralds the golden age of effortless parallel programming, or how it brings fault tolerance to the masses. But stay tuned, I think we're planning on talking up both in the next few weeks.

The good news about minting - I didn't say fabricating, we leave that to others - but the good news about minting your own silicon is you get to tell the transistors what to do.

Posted on 10:31AM Apr 10, 2007 | Comments[58]

Share this post  | | | | | |

Neptune Away!

Just in case you missed it, our Microelectronics team is off to a good start - we just signed a deal with Marvell Technology to license them the core intellectual property in our Neptune ASIC. They'll take it from here, add their own innovations, and turn the results into derivative products for the open market. Just the kind of partnership we love, and exactly how we want to build our business - bringing core innovations to the broadest markets possible, not just to other businesses within Sun.

For years we were called proprietary - a moniker that did more damage to Sun than any market downturn. And frankly, we've spent years recovering. But at this point, my hope is we've completely turned that slur on its head, that we've come to define open - more open than any other vendor, more open than open itself. From silicon to systems, software to storage and services.

Where open translates to "open to opportunity."

Posted on 07:50PM Apr 04, 2007 | Comments[14]

Share this post  | | | | | |