Anybody tuning in to BBC television during July 1953 (and there can't have been many in those days) were in for a treat. They were going to be kept on the edge of their seats for weeks by a thriller called The Quatermass Experiment.
In a similar way, observers of the computer industry have been on the edge of their seats for the past few months over Intel's rumoured 64bit extensions to the x86 architecture.
Anticipation reached fever pitch as the date approached for Intel's developer forum in San Francisco, traditionally the venue for unveiling hot new technology. What was the chipmaker doing with its "Clackamas" experiment, the latest codename for the fabled hybrid 64bit project?
Well, we all know the answer now. After years of secrecy the company has finally come clean, and Clackamas Technology will finally see the light of day. Some observers are already saying that this is an unprecedented vindication of AMD's strategy for chips that extend the x86 architecture to 64bits while keeping compatibility with all current 32bit software.
With hindsight it's easy to say Intel got it badly wrong and is now rushing to patch the hole in its product plans. The big question is, is it too late? If the reaction of some of the big vendors such as IBM and HP is anything to go by, the answer appears to be no, which must be fairly galling for AMD after all its hard work in building up demand for so-called 64bit extended systems.
The next question is: how will Intel try to market the new processors?
Will Intel salespeople be shouting from the rooftops about the 64bit extensions in Nocona, the next version of the Xeon, or will they be trying the softly, softly approach, pointing out that it has this capability, but it's only there for those awkward customers who really, really want it?
When The Quatermass Experiment was released for the big screen by a then little-known outfit called Hammer Films, the US version was titled The Creeping Unknown, which also seems apt, given the way Intel has kept end-users in the dark over the very existence of Clackamas, not to mention the real shape of its 64bit roadmap.
How well will the new chips perform? At clock speeds over 3GHz, I would hazard a guess that these extended 64bit chips will probably perform better than the current Itanium chips - which could spell further trouble for Intel's high-end processor family.
The final episode of The Quatermass Experiment was titled State Of Emergency, which sounds about right for Intel's Itanium team right now. Corporate IT buyers, attracted by Opteron but wary of buying AMD technology, will soon get exactly what they want from Intel's Nocona. They will be able to run current 32bit software, then upgrade easily to 64bit applications when these become available.
As a side effect, Intel's move to 64bit on what is essentially a Pentium 4 core, could mean that adoption of 64bit desktops will also move faster than it would otherwise have done.
The sequel to The Quatermass Experiment was Quatermass II, which was itself followed by the cult classic Quatermass and The Pit. If Intel finds that customers are far more interested in chips with Clackamas than any other 64bit technology, the Itanium processor could soon be looking over the abyss into a pit as well.







reader comments