AMD Emergent is really also the story of Intel Divergent. In the past 3-5 years, AMD has been able to gain consistent technological advantages over the overwhelming market leader Intel in the race for computer chip supremacy. Now this is no mean deed. Intel dominates the CPU chip market despite the entry and fierce competition of IBM, Motorola, Sun, and bevy of Japanese, Taiwanese, Korean and European chipmaking stalwarts. Only AMD has been able to “chip away” at that dominating position. Remarkable feat.
And it is not a case of Intel not having a plan. In fact it has been simple, elegant. Just as Gaulia est divisa in tres partes, so also decreed the Caesars of Intel about the chip market. So Intel bought the rights to the ARM processor for the low end, its own I86 for the mid-range and with the help of HP, Itanium would bring on a new era of super-mainframe computing. Meanwhile lowly AMD had a different vision. The low-end market would inexorably upgrade to the x86 architecture, the mid range would need low power and multi-core processing, but the real prize would be to bring the x86 architecture to full mainframe capabilities. Preserve all that huge software investment in x86 as much as possible while moving x86 to mainframe capabilities.
And look how well AMD has succeeded with some great engineering:
1) established their x86 64bit architecture, not Intels, as the market standard and beat Intel to market by 2-3 years;
2) also established their bus infrastructure as a better solution and still lead Intel by 1-2 years;
3) beat Intel to dual-core and now multi-core processors delivering working product ahead of Intel by 1-2 years again;
4) established a low-power priority which AMD again appears to be able to beat Intel on delivery plus by percent savings;
5) lead the charge into hardware virtualization, and the dies will be cast this year as to who has engineered best;
With a clear plan to take the x86 architecture into mainframe computing while incorporating benefits to other markets with co-processing and low power, AMD has emerged as a viable technology innovator. Now it must also win at the other side of the chip-> fabrication ops with its ability to deliver cost effective fab-plants. This is an arena where Intel has matched their onetime stock-holders/financiers, IBM.
But the key point is that AMD has succeeded because they took Intels x86 customers where they really wanted to go – upgrading up to mainframe capabilities with the x86 architecture largely preserved and doing so with the minimal disturbance to their customers enormous cumulative x86 software investments. And now AMD is capable of going to the next level, see what Infoworlds Tom Yager has to say about that (Made for Opteron). But in general there is a lesson here.
The Lesson
The lesson is obviously on the other side of the hardware fence, the other big monopoly which is software in transition too .
Microsoft with Vista is about to try take its Windows customer base to a place they may not unanimously want to go. Vista is a superpoweruser operating system which will have the shaky beginnings of a security citadel. WinPF will supply a 3D animation interface on top of the current Windows XP GUI that will start to provide some interchangeable, but proprietary Web or desktop elements but at the price of doubled hardware resource requirements. WinFS will supply a databased architecture for file storage and management including new metadata, encryption, compression, and searchability but at the price of substantial learning curve and proprietary technologies that if adopted will be charged for at Software Assuredly monopoly prices. WinCF will supply a Web and messaging framework that is at the cusp of being standard but also moving to highly proprietary and again subject to future monopoly pricing. All of this is being delivered in a new security and reliability hardcore coating. WinWF is Microsofts attempt to use its desktop monopoly to garner dominant entree into the workflow and BPM field where it is both pioneer (Biztalk and Infopath) and follower (Autonomy, BEA, Cognos, IBM/Lotus, Hyperion, Hummingbird, Microstrategy, Vignette, etc, etc – info integration is into full bloom).
In effect Microsoft is getting around to dotting all the “i”s and crossing all the “-ties” (think first reliability, then scalability plus availability and now security and interoperability)that OS/2 had in their targets back in 1993-4 but the market went for GUI, glitz and most apps. Unix has been able to deliver the “-ties” but at such a premium price that Windows Server, despite it shortcomings, has been able to make inroads along with Linux in the server market. However, that very same Linux and Open Source presents a problem to Redmond. Microsoft has a formidable competitor that is able to both maintain and innovate as fast (or as some would argue even faster) than Redmond.
But perhaps the biggest distinction is that commercial Open Source (the Redhats, Novells, MySQLs, etc) are dedicated to service, training and support (that is where they make their money) in a way that Microsoft with its huge and nasty EULA is weakest. EULA allows Microsoft to be mercurial. When Microsoft wants to be good, they can be very solicitous and helpful. But EULA also allows Microsoft to be very verrry bad – “we know whats best for you” bad. Just ask Win 9x, WinNT and now Win2000 users. Or VB6 developers. Or a long line of Office users who have seen premature obsolesence and inability to exchange data between versions of Office as expeditiously as they would like. Or a whole raft of Software Assurance customers. Or a whole host of ex-software partners who find their utility, games, small business, project management plus now BI and security markets eaten alive by Microsoft leaving customers with much smaller choices of apps in these segments on the Windows platform.
The bottom line is this – in service, support and making sure Microsoft software works the way the customer wanted and expected, EULA gives pre-eminent power to Redmond. In contrast, Open Source makes their money by being dedicated to service, support, and making Open Source software work the way the customer wants. Otherwise – there is a lot of competition in the Open Source arena and also plenty of SaaL-Software as a License commercial vendors to vie for the business.
Now Bill has declared that Microsoft has rededicated itself to being a Service oriented company. But he did not tell you the time or price. The time when Redmond becomes Service oriented is related to the price – new Service will come as soon as you, the customer, are willing to pay for it – EULA, monopoly, and Software Assuredly, Redmond will determine that price. Meanwhile the AMD of software, Open Source, offers very compelling alternatives.
(c)JBSurveyer 2006