January 5, 2009 Timothy Prickett Morgan
Welcome to 2009, everybody. For the first time in I can’t remember how long, I actually took much of the vacation that was my due, and I spent lots of time with friends and family eating and drinking and horsing around. I feel human again, even if it is in relation to what still remains a troubling economy and a difficult world. Guild Companies, which publishes the Four Hundred stack of newsletters, made it through 2008, and provided there are a whole lot less shocks this year, it looks like we will be here to serve the AS/400 community in 2009.
This is a relief to all of us who create and produce the content for the newsletters, as well as who sell our advertising services and produce this content–without which, I will emphasize, no AS/400 newsletters would be possible. From my perspective, it was difficult enough in 2008 for me to shut down our non-AS/400 newsletters and go get a second job over at the The Register to make ends meet. I have been writing an AS/400 newsletter for so long, I don’t know who I am without doing it. I am grateful that I have two jobs, and I am going to keep repeating that to myself as I carry my workloads in 2009. Best I can figure, it could be a hell of a lot worse. I could have no jobs at all and I could have had to fire my employees.
As 2008 was ending, I said that I was going to shut down a lot of thinking and worrying about 2009, just to clear my head and to stop revving so high. (My adrenal glands needed a rest; that is for sure.) I hope you all did the same, too. We are going to need to think clearly in 2009, and it will be important to minimize mistakes and to keep our eyes on what’s happening and what we can do to help. Great tumult provides great opportunity, or so I have been led to believe. The key, I think, is to remain flexible and to keep an open mind.
This is a very odd time in the history of the AS/400 and successor product lines, and I must confess, I am wrestling with this a bit. With the convergence of the iSeries and pSeries lines last year into a truly single Power Systems lineup, with the separation of SMB marketing and enterprise marketing in the Business Systems and Enterprise Systems divisions, and with the retirement of key systems people (like Bill Zeitler and Frank Soltis), I am not only not sure what is reasonable to ask IBM to do in support of the i product set, but also unclear who to ask. I have argued for so long for IBM to give the AS/400 and its progeny a fair shake when it comes to technology and pricing that now that it is here, I am not sure what we need next. Ditto for support of new languages such as PHP or new databases like MySQL. We asked for it, and we got it. The convergence of the product lines, which really only finished last fall, has brought price parity to the lines.
While all of this has been necessary for the continued health of the i platform, it is by no means sufficient. There are hundreds of thousands of customers in the world who by and large still have RPG applications hitting DB2/400 databases running on an i operating system. And IBM needs to get the i platform competitive with Windows and Linux boxes using the latest multicore X64 processors in a virtualized environment. As I showed last year, the entry Power 520s running the i software stack are absolutely competitive with Windows, but the Power 550 and Power 570 machines were increasingly overpriced compared to Windows alternatives. In the coming weeks, since IBM doubled up the core counts in the Power Systems lineup back in October, I will be taking a look at how the Power Systems boxes compete with X64 iron, particularly machines based on Advanced Micro Devices‘ new “Shanghai” quad-core Opteron machines and the forthcoming “Nehalem” Opteron-alike processor due from Intel around the end of March or so. IBM cannot let X64 server makers open up too much of a price/performance gap compared to Power Systems, whether they are running i, AIX, or Linux.
For now, IBM seems content to work on its “Blue Business” server appliances running i and Linux, and probably at some point AIX, to take on the SMB space in an AS/400-on-steroids style. There is some expectation that IBM will this year deliver a Power6+ kicker to the current Power6 processors, but that may or may not happen. It better happen, if Big Blue knows what is good for it. IBM is expected to get eight-core Power7 chips out the door in 2010, and that is too long to wait, particularly with Intel and AMD cranking out the best chips they have ever delivered this year and maybe even Sun Microsystems getting its “Rock” 16-core UltraSparc-RK processors released in the second half of 2009. There is a chance that budget pressures will compel Sun to spike the Rock chips and just move entirely over to Fujitsu‘s Sparc64 clones, which are more of a standard SMP-style processor than the Rocks. It is hard to say, and a lot will depend on what Sun’s investors are pressuring the company to do to cut costs and boost sales.
It is going to be an interesting year, this 2009. That is for sure. And we’ll all be right here, keeping you up to speed on what IBM does with the i platform and how this compares to the other technology out there in the field.