The New York Times
November 16, 1998
Depth of an Old Microsoft-Intel Rift Is Disclosed in Court
In the End, Microsoft Reconciled With Intel
By STEVE LOHR
The Microsoft trial is not merely a legal prizefight. With its parade of witnesses and mountains of documents, it has opened a window to the inner workings of Microsoft Corp. and how it deals with the rest of the computer industry, raw material for a digital-age "Dynasty."
Last week's episode: Family Feud.
The story told in a federal courtroom in Washington was how, from the spring of 1995 to the summer of 1996, an extraordinary rift developed between Microsoft and Intel Corp. -- partners in the most lucrative business relationship in modern capitalism -- and how it was patched up.
At the time, people in the computer industry knew about the Microsoft-Intel split over multimedia software developed by Intel, called "native signal processing," or NSP.
But the depth of the rift, its significance and the personal animosities involved were not really apparent until last week with the testimony of an executive of Intel and internal e-mail and documents from both companies that were submitted as evidence.
The crack in the partnership was repaired personally by the chief executives of the two companies, Bill Gates of Microsoft and Andrew Grove of Intel, over lengthy dinners, which included shouting matches and frequent e-mail dialogues.
In the end, Intel backed down from a technology initiative that Grove once described in an e-mail to Gates as "key to the growth of the PC industry."
The antitrust issue in the case is: Why did that happen? The government argues that the Intel episode demonstrates how Microsoft unfairly uses its market muscle to bend the industry to its will -- in this case even Intel.
Nothing so sinister, Microsoft scoffs. Intel, Microsoft says, was eventually persuaded to stop pushing a misguided technology effort that could have hobbled the rollout of new personal computers running Windows 95, the operating system Microsoft released in August 1995. It was a decision that assuredly helped Microsoft, the company admits, but it was also in Intel's enlightened self-interest because the giant chip maker's main business is selling new computers powered by Intel microprocessors, and the increased demands of each new iteration of Windows sends buyers scurrying to buy more powerful computers.
To be sure, Microsoft and Intel often debate which technology standards to champion. But the debate over NSP became a more public and a more significant issue than most. Intel's senior executives had declared at industry conferences that NSP could help encourage more households to buy personal computers because improving audio and video performance would make "the PC sing and dance."
And as a software effort, NSP was clearly seen by Microsoft as an unwelcome foray into its domain. "The main issue was that it was crucial software that Microsoft did not control," said Richard Doherty, president of Envisioneering, a research firm in Seaford, N.Y.
The e-mail submitted in court sheds light on the ambivalence in the Microsoft-Intel relationship, painting an environment in which they circle each other warily. There is mutual distrust and personal sniping on both sides. Yet ultimately, the benefits to both sides of the sometimes bumpy "Wintel" partnership led them back to a pragmatic accommodation.
By April 1995, Microsoft had awakened to the potential significance of Intel's NSP effort. After a meeting with Microsoft executives, Gerald Holzhammer, an Intel executive, wrote in an e-mail to his colleagues on April 13, 1995, that Microsoft was "upset with us being in 'their' operating system space -- no surprise there."
By May, Gates decided to protest directly to Grove. In an e-mail on May 25, 1995, Gates called NSP "a significant problem here that is on a path to get a lot worse." Clearly upset by software being developed at the Intel Architecture Labs in Hillsboro, Ore., he complained to Grove, "I don't understand why Intel funds a group that is against Windows 95." Later, he added, "In software you have a group that won't allow us to lead and has all the prestige and profits of Intel to drive them forward."
Commenting on Gates' e-mail, Paul Maritz, a senior Microsoft executive, noted on the same day that there was "a strong faction within Intel" that wanted to use NSP as a step toward expanding Intel's intellectual property claims in the industry. "We have to be very careful about this," Maritz wrote in his e-mail to Gates and others, "given our recent discussions about needing balance to Intel in the market" -- an apparent reference to Microsoft's work with other chip makers to keep Intel from becoming more powerful.
On June 8, 1995, Grove made his appeal for NSP to Gates. Terming it "key to the growth of the PC industry," Grove added that while Microsoft and Intel had "differences of opinion on the detailed approach," he was confident that the two companies could resolve them.
In a philosophical digression, Grove wrote, "I think Microsoft and Intel have special responsibilities toward the PC industry."
Later, he added: "If you believe that what we are trying to do is good for the industry, please help us (plural us: Intel and Microsoft and whoever else needs to participate) achieve it. Merely being critical of what we have undertaken is not helpful."
On June 22, 1995, Maritz wrote to Ron Whittier, an Intel executive, to inform Intel of Microsoft's stance on NSP in no uncertain terms. Microsoft, he wrote, had told personal computer makers that "we have major issues with NSP and our advice to them is NOT to use it."
On July 7, 1995, Gates sent an e-mail to a group of Microsoft executives briefing them on a three-hour dinner with Grove two days earlier in San Jose, Calif. "Fundamentally," Gates began, "Intel and Microsoft have not been doing a very good job of working with each other in most areas."
"The main problem between us right now is NSP," Gates wrote. "We are trying to convince them to basically not ship NSP."
Not known for being awed by the intelligence of others, Gates included an aside about Grove: "Although Andy is super-smart in software-related areas and in some aspects of the PC market, it is amazing what he doesn't know."
By August 1995, when Windows 95 was released, Microsoft's pressure on Intel and personal computer makers to step back from NSP had succeeded, Steven McGeady, an Intel vice president testified last week. By then, McGeady said, he believed Intel's NSP program had been "shot in the head."
In an Oct. 18, 1995, e-mail, Gates told several executives that Intel felt that after Microsoft's "NSP chill" personal computer makers would not do anything to fine-tune their machines to exploit Intel's new multimedia chip, the MMX, for Windows 95 "unless we say it's OK." Gates added, "This is good news," because it meant the computer makers "are listening to us."
The regular meetings between Gates and Grove continued. While conflicts persisted, especially over what Microsoft viewed as Intel's support for its rivals in the Internet software business, like Netscape Communications Corp. and Sun Microsystems Inc., the Microsoft-Intel relationship clearly was on the mend after the NSP issue was settled.
On June 9, 1996, Gates wrote an e-mail to his senior executives, briefing them on a two and a half hour meeting he had had two days earlier with Grove. "Our relationship with Intel is advancing on many fronts," Gates began in a lengthy e-mail sent at 10:26 p.m. on a Sunday.
The strengthening of the Microsoft-Intel relationship was a message Gates was eager to communicate to a wider audience. He and Grove had recently given a chummy performance for Fortune magazine. He passed this along in his e-mail to his colleagues: "The Fortune magazine coming out in a week features a joint interview with Andy and I where we are very friendly to each other."