Powers Unfiltered

Attention .NET Developers of the greater Seattle area (I hear there are some up there…).

Digipede Evangelist Kim Greenlee will give a presentation on Concurrent Software Development at the .NET Developers Association meeting on Monday, December 11.  The meeting is on the Microsoft campus – in Building 40, the Steptoe Room (#1450).  She blogs about it here; be sure to bring this meeting announcement with you, or Microsoft security may stop you.

 

 

 

On Wednesday, December 6, Microsoft, Digipede, and other Microsoft HPC partners will host a half-day informational event describing how to achieve faster “time to insight” from computationally intensive financial applications. 

Nathan Trueblood and I will be there (with one or more Digipede customers and a bunch of our friends from Microsoft) to help make this a great event for anyone in financial services who needs more computing horsepower for their Windows applications.

I’ll present information on how the Digipede Network integrates with and adds value to Microsoft’s entire technology stack, inclulding the new Windows Compute Cluster Server 2003 (CCS), Visual Studio 2005, .NET, Excel 2007, SharePoint 2007, and more.  We’ll run through some real-world examples of how our financial services customers use the Microsoft / Digipede solution to make more money by dramatically improving application performance and scalability.

One of Digipede’s clients will also be a featured speaker at this event.

Date:  Wednesday, December 6, 2006
Time:  Welcome, 8:30; event 9:00 AM to 2:00 PM
Location:  1290 Avenue of the Americas – 6th floor, New York (that’s between 51st and 52nd street in the Axa Equitable building)

Registration and other details:  http://msevents.microsoft.com/cui/eventdetail.aspx?culture=en-US&eventid=1032306637&x=15&y=14

We hope to see you there!

Big Webcast today — Hear UBS Director of IT Eric Kristoff, Microsoft Industry Architect Stevan Vidich, and yours truly expound on the wonders of grid computing in the financial services industry.  Go here:

http://www.xtalks.com/gridcomputing.ashx

 

I found this old draft from the end of the Supercomputing conference (almost two weeks ago).  Sorry to be slow, but here you go.

I’m NOT an old-time supercomputer guy, and I therefore have NOT been to every SCxy show (as the Supercomputing show refers to its various annual incarnations).  Indeed, my first show was SC03, in Phoenix, so I can’t bring the perspective of those who have attended since SC88.  However, the changes just in the past four years are dramatic.

In SC03 and SC04, Microsoft’s efforts in supercomputing were absent and ridiculed, respectively.  By SC05, with the beta of Compute Cluster Server out (and the show on Microsoft’s home turf in Seattle), reactions turned more to curiousity.  This year, with Compute Cluster Server shipping, reactions varied from respect to fear to hope (not to mention residual ridicule), but no one could claim Microsoft was absent.  There were dozens Microsoft employees — many from the HPC team, but more than a few others as well.

There were Microsoft clusters in the Microsoft booth (obviously), but also in the HP booth (where Dan and I showed off the Digipede Network running on an 8-node cluster), and in booths of several other hardware vendors.  So — slowly but surely — Microsoft is definitely moving the dial, gaining visibility and partner presence for its new HPC offering.

Software was a different story.  While there were a few application vendors in the Microsoft booth with traditional high-performance applications (e.g. Ansys/Fluent, BioTeam, Mathworks, and Wolfram), I saw no other software vendors running Windows versions of software in their own booths (except us, in the HP booth).

More important than the HPC crowd at Supercomputing (by far, at least for us) will be the traction Microsoft and its partners gain with Compute Cluster Server in enterprise markets like financial services.  We’re working like mad to help with that.  More on that another day.

I had another of my favorite type of interactions with a stranger from Microsoft (or a contractor thereof) this morning.  “Rachel from Microsoft” called me to tell me of many exciting opportuntities surrounding the launch of Vista, Office 2007, and Exchange Server 2007.  We get such calls every week or two, which is fine — sometimes we learn of good opportunities for training, or opportunities to participate in Microsoft events of one sort or another, and so on. 

But my favorite part of this call was when Rachel informed me of a new “single source for all information regarding…” and I didn’t even hear the rest I was laughing so hard.  I believe I now have over a dozen humans at Microsoft who are my “single point of contact” for X, and a dozen or more URLs (past and present) that are my “single source for all information” about Y. 

It’s tempting to blame Microsoft for this, but they’re just trying to be responsive to a chorus of lazy partners.

I hear Microsoft partners whine (and yes, I’ve done it myself) about how big and confusing (and confused!) Microsoft is, and Microsoft hears this and thinks “well, what can we do about it?”  Throughout the organization, well-meaning individuals and teams think “Ah ha!  We’ll just give our poor partners a Single Point of Contact to help them navigate the many options we offer, and that will help them out.  That will make us more accessible.  That will make us more partner-friendly.  That will make them happy!” 

Come on.  Digipede is a fairly small company, yet I bet we have over 200 non-trivial personal relationships with Microsoft.  While the idea of a single point of contact may be appealing to some people (hell, if Microsoft wants to give me a full-time administrative assistant who works on the Redmond campus, who am I to argue?), it’s also unrealistic — and inefficient.  

Far better is to talk to people you know and trust, meet more people, figure out which ones have interests in common, see whom you can help and who can help you, and repeat.  In the rest of the world, this is just called “networking,” and it’s not viewed as some painful burden — it’s a way to build worthwhile relationships, in business and in life.  If you’re not good at it — don’t take it out on Microsoft.

So Microsoft (are you listening, Allison Watson?), don’t give us any more single points of contact.  First, you can’t — you’ve proven that over and over.  Second, we don’t want them.  Give us good networking tools, so we can discover the people and resources that can improve our relationship — to your benefit and ours.  Microsoft is brilliant at this in person (e.g. at events like the Microsoft Worldwide Partner Conference, or better yet at Channel Builder events), and inexplicably inept at this online (e.g. at partners.microsoft.com, or worse yet at the online Channel Builder).  And yes, if anyone’s listening, I’d be happy to make (many) suggestions.

 

The annual Supercomputing conference (SC06) is in Tampa, Florida next week.  Actually, it starts this weekend and runs for a ridiculous seven days, but Dan and I will be there Monday night through Thursday afternoon — which will be plenty.  SC is a “must attend” show in the high performance computing / grid computing / supercomputing / cluster computing / scalable distributed computing world (this industry needs lots of names for very, very similar concepts so that experts can argue about the subtle differences).

Like many events in this niche, SC suffers from a severe identity crisis that provides for a pretty amusing culture clash.  You can find booth babes and conference swag next to pony-tailed UNIX bigots next to high-powered corporate sales dudes next to uppity startups next to dazed computer science professors.  Nevertheless, the show is growing by leaps and bounds — this will only be my fourth one, but since SC03, the floorspace and attendance have increased dramatically.  SC05 was too big for the venue in Seattle, and SC06 will be bigger.

I will probably spend most of my time (when not in meetings with customers and partners) on the exhibit floor.  Dan and I will be in the HP booth a lot — their HPC and grid teams both like what we’re doing, and they’ve provided some space for us.  Much less hassle than a booth of our own!  Dan and I will also be looking for opportunities to execute the “insta-booth” maneuver we mastered (out of necessity) at the overflowing SC05 — so wherever the nearest Starbucks line is, watch for spontaneous Digipede demonstrations.

If you’re going to be there, let me know.

Congratulations to Mike Gunderloy on the one thousandth issue of The Daily Grind.  There is no better place for your daily dose of .NET developer news.  Great work, Mike — Keep it up!

Emil Sit has an excellent post on his recent experience with SunGrid, charitably titled “Observations on SunGrid Customer Care.”  He begins with:

I  haven’t used the SunGrid this week. In fact, no one has: there was a four day outage from last Saturday morning through this morning.

His post is quite illuminating about Sun, customer care, individuals involved in support services, communication philosophy, and related issues.  His conclusion:

As an idea, the SunGrid is a fast and easy way to get parallelism and performance flexibly. But Sun has to continue to improve the user interface (e.g., beyond the clever hack for job monitoring suggested to me by a Sun engineer) and relability of their infrastructure. Unless they do, people without CPU grants are going to start looking at alternatives like using Amazon’s hosted EC2 or running their own DigiPede.

Emil is, sadly, more than right.  Sun has to improve a lot of things, not just the UI, and not just reliability, because most people ARE ALREADY looking at (and using) alternatives (and — thanks for mentioning us!).  The problem isn’t UI — the problem is getting way, way ahead of the market, technology, and Sun’s own skill set.

Let me confess that I spent 20 years in or near the electric utility industry, and I know  A LOT about utilities.  SunGrid is not a utility.  The “utility computing” analogy bothers me in general, and I don’t have nearly the time or energy to break it down fully, but the standards, reliability, infrastructure, training, and commitment are not there (anywhere in the IT industry, including Sun) to run a utility.    

I almost wrote “not there YET” in that last sentence, but I remain unconvinced that a utility model will ever (or should ever) apply very well to computing.  Let’s be clear — I think outsourced computing in several forms can be viable, and that there are things you may want to compute on someone else’s computing infrastructure.  That doesn’t make it a utility.  And that’s probably OK.

The electric utility industry is boring.  Yes, I still have many fine friends there, and they are not, mostly, boring, but the industry as a whole does best when it changes slowly or not at all — and with good reason.  Appliances built in 1910 still run just fine when plugged into a socket — a socket whose voltage and frequency standards have not changed in a century.  Inventions since Edison, Tesla and Westinghouse have been largely incremental — in generation, transmission, and distribution.  Ways to USE electricity have changed radically over time, but even these are constrained (dramatically) by unchanging standards for voltage, frequency, and a handful of other key parameters. 

The tech industry has very few of the constraints (or benefits) of standards like these.  (Again, I don’t have time to elaborate fully (stay tuned for future posts), but if anyone wants to debate me on whether a Web Services standard compares to the 60-Hertz standard, I warn you — you’re going to lose.)  The tech industry delights in overthrowing standards, and as a result has made some pretty phenomenal progress — but as a result, use and generation of computing power have remained very closely coupled. 

Is this good or bad?  The jury is out (more on this another day).  But is it fact?  Yes.  The tech culture is not the utility culture (again, watch out if you want to debate me on this).  So even as we all root for SunGrid (even me), the deck is stacked against them. 

 

I have to admit, I was pulling for the Mets in the NLCS — but now that it’s the Cardinals and Tigers, some long-unaccessed memories of the 1968 World Series come flooding back.  (OK, OK, I’m old.)  I was eight years old and just entering my formative years as a baseball fan; I can’t remember a single World Series event of any kind before that, but I remember this one well. 

My grandparents had just moved in with us at our house in New York, and my grandfather was a fanatical baseball fan.  He had pitched a little semi-pro ball when he was 17 or 18, and had been a sportswriter on several newspapers in the Midwest after that — he knew baseball.  My grandfather was a National League fan, and I inherited that from him; we were rooting for the Cardinals as Bob Gibson and Lou Brock and Curt Flood and the rest took on Denny McLain, Mickey Lolich, Al Kaline and company.  I sat and watched transfixed as the ’68 Series went seven games, with more twists and turns of plot and momentum than I can remember —  but the Tigers won that one. 

Like Robert Anderson, I’ll be rooting for the Cardinals this time — I’m a LaRussa fan from his tenure in Oakland as well — but mostly, I hope the series is as entertaining as the last time these two teams met.

Greg Nawrocki has another excellent article today, featured in GridToday (under the unlikely title Building the Perfect ‘Grid Sandwich’). He reviews various issues related to grid adoption:  Hardware? Check. Network? Check. Middleware? Many choices.  Eventually he orbits back to his favorite topic and mine — applications.