Reading the Times - about Intel

The really important point everyone missed is that Intel's announcement was about manufacturing, but signalled a strategic reversal: they're getting back into the gigahertz game.
Written by Paul Murphy, Contributor on

As regular readers know, I'm both fascinated and frustrated by the extent to which the main stream press misinforms our bosses -the non technical people who ultimately make most of the strategic technology decisions in the organizations we work for.

The media reaction to last week's announcements from Intel make the point - or, more precisely, that fact that every single one of the stories I looked at on this both missed the point and bought into Intel's pixie dust - even those triggered by IBM's apparently off topic rebuttal- makes the point.

Consider, for example, John Markoff's January 27/07 New York Times story: "Intel Says Chips Will Run Faster, Using Less Power.

Here's the opening paragraph:


Intel, the world's largest chip maker, has overhauled the basic building block of the information age, paving the way for a new generation of faster and more energy-efficient processors.

In reality Intel doesn't even place in the top five of the world's chip makers by volume, and, more importantly, Intel spun the announcement as about processor design, but it wasn't about design at all, it was about the application to manufacturing of research published in 2004 by people working mainly at the University of Texas in Austin.

To see both the reality and the spin Markoff fell for, look carefully at the wording in Intel's own press release:


In one of the biggest advancements in fundamental transistor design, Intel Corporation today revealed that it is using two dramatically new materials to build the insulating walls and switching gates of its 45 nanometer (nm) transistors.


Intel is the first to implement an innovative combination of new materials that drastically reduces transistor leakage and increases performance in its 45nm process technology.

It's easy to see how that kind of thing would mislead the uninformed - but notice that it actually makes only two unambiguous claims: that Intel is "using" the new transistor technology, and that Intel is the first "to implement" the new technologies.

Markoff's second paragraph isn't any more insightful than his first:


Company researchers said the advance represented the most significant change in the materials used to manufacture silicon chips since Intel pioneered the modern integrated-circuit transistor more than four decades ago.

Intel's people may have said this, but Intel didn't pioneer the micro-processor -or the "modern integrated-circuit transistor", whatever you want to imagine that to be.

The first sentence of paragraph three is more subtly wrong, but just as grating:


The microprocessor chips, which Intel plans to begin making in the second half of this year, are designed for computers but they could also have applications in consumer devices.

I know what the writer means, and so do you - but "microprocessor chip" is a pleonasm (the opposite of an oxymoron; two words or phrases that are mutually redundant); Intel's press release to the contrary the first 45nm, "penryn", CPU isn't expected until 2008, and the distinction drawn between "computers" and "consumer devices" is simultaneously reasonable enough and illustrative of appalling cluelessness in a technical reporter.

Now none of that would matter, except for the fact that the original language, as used in IBM's press release, made sense: "The achievement is expected to have widespread impact, leading to improvements in electronic systems of all kinds, from computers to consumer electronics."

And no, I don't think that's a coincidence. Compare, for example, this bit from the story:


The Intel announcement is new evidence that the chip maker is maintaining the pace of Moore's Law, the technology axiom that states that the number of transistors on a chip doubles roughly every two years, giving rise to a constant escalation of computing power at lower costs.

first to this from Intel's press release:


It also ensures Moore's Law, a high-tech industry axiom that transistor counts double about every two years, thrives well into the next decade.

and then to wording from IBM's:


As a result, the use of this material could allow the industry to continue on the path defined by "Moore's Law", the chip industry axiom that predicts a doubling of the number of transistors on a chip every 12-18 months, thereby allowing chip performance and function to increase as well.

Notice that IBM's statement is correct while Intel's press release is brilliantly manipulative, but no more technically correct than the reporter's rather unsporting emendation of IBM's wording to hype Intel.

What's oddest about this, however, is that IBM's press release responds directly to what they expected Intel to announce - a manufacturing advance. Thus when people like Markoff mis-interpret the announcement as about research fundamentals, the IBM response appears unfocused - leading to further mis-interpretions, this time of IBM's position:


Word of the announcement, which is planned for Monday, touched off a war of duelling statements as I.B.M. rushed to announce that it was on the verge of a similar advance.

I.B.M. executives said their company was planning to introduce a comparable type of transistor in the first quarter of 2008.

Many industry analysts say that Intel retains a six-month to nine-month lead over the rest of the industry, but I.B.M. executives disputed the claim and said the two companies were focused on different markets in the computing industry.

The I.B.M. technology has been developed in partnership with Advanced Micro Devices, Intel's main rival. Modern microprocessor and memory chips are created from an interconnected fabric of hundreds of millions and even billions of the tiny switches that process the ones and zeros that are the foundation of digital computing.

In reality IBM's schedule is actually about the same as, or marginally ahead of, Intel's; both companies were among the co-sponsors of the underlying materials research via the SRC/Sematech consortium; both are currently moving from theory to manufacturing; and both plan to ramp up manufacturing late this year with marketable product in early 2008.

In other words, the bit about IBM "being on the verge of a similar advance" continues the mis-understanding of what Intel really announced while the claims attributed to "many [unidentified] industry analysts" are at best wrong and at worst imaginary.

And then, there's this:


Today transistors, for example, are made with systems that can create wires and other features that are finer than the resolving power of a single wavelength of light.

The article is actually quite long (about 1200 words), and that means the author found room for quite a few other mistakes - including the fundamental one made in every report I looked at:


Many executives in the industry say that Intel is still recovering from a strategic wrong turn it made when the company pushed its chips to extremely high clock speeds -the ability of a processor to calculate more quickly. That obsession with speed at any cost left the company behind its competitors in shifting to low-power alternatives.

Now, in reality, Intel concentrated on selling megahertz, not processing performance - a 3.0Ghz dual core "woodcrest" is a blinding 18% faster than its 2Ghz counterpart - more comparisons here - but the really important point everyone missed is that Intel's announcement was about manufacturing, but signalled a strategic reversal: they're getting back into the gigahertz game.

That's what IBM's doing because that's what this technology is aimed at; but not a single reporter, not even the ones who mentioned IBM's announced intent to release the PPC980 Power6 derivative at something close to 5Ghz and then drive it to 6.0 or higher as manufacturing switches to the new materials, seems to have understood the issue.



Editorial standards