Intel gambles with Itanium

But a decade ago, Intel saw that it was nearing a crossroads. Its fingernail-sizechips were based on an aging design, one that would require major revision inorder to fulfill Intel's dream of extending its reach beyond PCs to high-endcomputer servers that power corporate networks and the Internet.

Like a homeowner torn between renovating an old house and building a newone, Intel realized it could spiff up the existing microprocessor line, known asthe x86, with a technical facelift, or it could start on a whole new design. Afterheated internal discussions, Intel decided to plunge into the unknown, forging anambitious partnership with Hewlett-Packard Co. to design an all-new chip.

Tuesday, Intel unveiled the Itanium, the first of a series of processors it hopeswill extend its dominance to the entire computing universe and cement itsposition for decades. The stakes for Intel are high. With prices of its PCprocessors falling steadily, it badly needs to move to higher ground. The fate ofItanium could determine whether the company's dominance slips as the PCwanes in importance--or whether Intel helps shoulder aside big servers fromSun Microsystems and IBM, theway Intel-based PCs once overran rivals such as Apple Computer Inc.

Operation Merced
Developing Itanium, previously known by the code name Merced, has been anintense and unpredictable effort that sometimes teetered on the brink ofdisaster. Time and again, a project team of as many as 500 circuit engineers,chip architects and software wizards found it had underestimated the difficultyof its task, more than once sinking into a quagmire of complexity with noobvious way out.

Like carpenters forced to build new hammers and saws as they went along,Intel's engineers designed and tested new software tools at the same time thatthey were sketching out parts of the tiny chip. The team broke into separategroups, each working on one piece without knowing just how they would fittogether.

"Everything was crazy," says John Crawford, the chip's chief architect. "We were taking risks everywhere. Everything was new. When you do that, you're going to stumble."

As a result of the many setbacks, the first Itanium chip is two years late, an eternity in the world of technology. What would have been a speedy processor if introduced on time in 1999 now will run only half as fast as Intel's next version of the Pentium 4, which still uses the x86 architecture.

For this first Itanium, expectations are low. Many in theindustry think it will be used mainly for testing, with corporate customers waitingfor a second-generation chip code-named McKinley that is due out next year.Microsoft Corp. has only just completed a limited edition of its Windowsoperating system tailored for Itanium and doesn't plan to release a final versionuntil the end of the year. Since the Itanium is aimed at servers and workstations,Intel will continue making Pentium and Celeron chips for PCs.

"This will end up being one of the world's worst investments, I'm afraid,"predicts David House, a former Intel official who is now chief executive ofAllegro Networks, a network-equipment start-up. Mr. House, Intel's chief ofcorporate strategy in the early 1990s, worries that Intel will never get back the$1 billion to $2 billion that analysts estimate Itanium has cost so far. Mr. House,who approved the project at the time despite his own reservations, says thescale of the Merced project "scared the everloving bejesus out of me." Intelsays that his doomsaying is nonsense and that the Itanium family will make itplenty of money.

The Itanium story began in the early 1990s as Intel's designers started to chafeat limitations of the x86 design. The most serious: It processes data in chunks of32 bits--each a one or a zero. For technical reasons, that limits the memory ofa PC or server to four gigabytes.

The limit is still distant for today's PCs, which average 128 megabytes ofmemory. Servers, however, use far more memory because they do so muchheavy lifting, from storing Web pages to running giant company databases.Memory requirements also double every 18 months or so, making the memorylimit a distinct threat to Intel's high-end computing ambitions.

The solution: Intel needed to develop a new processor architecture, orunderlying design, that could handle data in chunks of 64 bits. That would makepossible memory sizes four billion times as large as with the x86.

A new architecture is a complex ecosystem of devices and programs, socreating a new one involves far more than just designing a single chip. Engineersneed to build, generally from scratch, software to help programs run better onthe new processor, sets of "assistant" chips for the processor, and batteries oftesting and verification programs. Little wonder that a group of Intel engineersargued for simply tweaking the x86 architecture to handle 64-bit data.

By the early 1990s, however, new processors such as Digital EquipmentCorp.'s Alpha chip were already showing greatly improved performance withnew designs. Mr. Crawford led a faction that said Intel had to push forwardwith a new architecture or risk trailing competitors "from day one." Technicalarguments raged until his group eventually convinced Albert Yu, then generalmanager for microprocessors, that an aggressive approach was best.

Across Silicon Valley, Hewlett-Packard also had some decisions to make. Apower in high-end computing, H-P had long built its own microprocessors butreserved them for its own workstations and servers. Recently, engineers in itscorporate labs had designed an advanced processor architecture calledPA-WideWord, which promised blazing speeds by letting the chip performseveral calculations at once in what is known as parallel processing.

But the costs of chip-making were soaring to the point that a single newmanufacturing facility could run more than $1 billion. Despite its tradition ofgoing it alone, H-P decided it needed a partner that could share the financialburden and help sell the chip to other computer makers, possibly making it anindustry standard.

H-P approached Intel, which was intrigued. Not only did H-P's team includeseveral chip-design luminaries, but Intel also noticed that H-P had a big headstart. "When we saw WideWord, we saw a lot of things we had only beenlooking at doing, already in their full glory," Mr. Crawford says.

At a preliminary technical exchange, says WideWord architect Rajiv Gupta, "Ilooked Albert Yu in the eyes and showed him we could run circles aroundPowerPC [an IBM processor], that we could kill PowerPC, that we could killthe x86. Albert, he's like a big Buddha. He just smiles and nods."

Lawyers worked out some ground rules, and by early 1994, technicaldiscussions had begun in earnest. They took place at an out-of-the-way H-Psales office, with no documents allowed out of the room. At each day's end,material was stored in a double-locked filing cabinet, with one key held byIntel's Mr. Crawford and one by H-P's Mr. Gupta. "The idea was that if wedidn't do a deal, we would take the filing cabinet to the parking lot and blow itup," Mr. Crawford jokes.

In June 1994, the companies announced a partnership, with Intel leading thedesign of a 64-bit processor that would use many of H-P's ideas. An exuberantMr. Yu incautiously declared: "If I were competitors, I'd be really worried. Ifyou think you have a future, you don't."

Culture Clash
But the task was just beginning. H-P officials, accustomed to consensus-drivenmanagement, were jarred by an Intel culture of "constructive confrontation" thatembraces argument and interruptions. For their part, Intel officials wonderedwhy H-P people wouldn't stand up for their ideas. After some meetings, saysH-P designer John Wheeler, "the H-P folks were exhausted, while the Intelguys would be slapping us on the back saying, 'This is the best meeting we'veever had.' "

Managers eventually drew up a 75-page guide to cooperation that explained,among other things, how to interpret the behavior of both sides. Engineersquickly dubbed it the "owner's manual."

Some disputes about technical issues dragged on. H-P engineers argued thatcertain arithmetic functions known as floating-point operations could be handledoutside the chip by software, saving chip space. Intel wanted the functionsdesigned into the chip, so they'd be faster. As the two sides arm-wrestled, Intelsuddenly found itself facing a flap over a flaw in its Pentium chip, one that justhappened to involve a hardware fault in the floating-point unit. "That just sort ofended the discussion," says H-P mathematician Alan Karp. The team went withthe software method.

Once H-P and Intel had hammered out the design basics, the developmentteam expanded rapidly. But with the Internet boom just starting, they werehard-pressed to find experienced engineers. Intel ended up hiring many recentcollege graduates. "I was one person signed up to design this giant thing thathad no idea what it was," says Nadeem Firasta, who joined Intel in 1995.

Managing such a diverse, intense and fast-growing team took a toll onmanagers. The first to head the project, Avtar Saini, lasted only a year beforerequesting a reassignment. He now runs Intel's operations in India.

Too Big
His successor, a laconic Missourian named Gary Thomas, joined the project in1995--just in time to see it hit a devastating roadblock.

The chip's architects had divided functions into separate modules, like lettingteams of subcontractors design the rooms of a house. In mid-1996, Mr.Thomas slotted the modules together for the first time in what the team calledthe floor plan. Bad news: The floor plan was larger than anyone had expected,far too big to fit on a die of silicon that Intel could manufacture economically.

Mr. Thomas maintains that his only emotion was a realization that "it was timeto get to work." Mr. Crawford, the chip architect, is less restrained. "We hadblown out the walls," he says. "This was a lot worse than anything I'd seenbefore."

The team found itself sweating through a "die diet" as it worked feverishly toslim down bloated functions and subsystems. Mr. Crawford, who called himselfthe "chief liposuctionist," searched for cuts that wouldn't hurt performance toomuch. Internal memory was scaled back, as was a module that maintainedcompatibility with x86 chips.

Downsizing
But individual modules, initially only rough designs, kept growing larger as theywere refined. After months of struggle, senior Intel managers realized that theycould solve the size problem only with a radical step: a new manufacturingprocess that would let engineers shrink every wire and transistor. The changewould drop the chip's tiniest dimensions to 0.18 microns--or millionths of ameter--from 0.25 microns, making each module smaller.

The switch to a new fabrication process appeared likely to solve most of theMerced project's problems, at the cost of a few months of delay. But theproject team soon found itself in a fresh predicament as they worked to tune upthe movement of signals across the chip.

In a well-functioning chip, signals flit from module to module in a precisely timedchoreography, with the speed of the chip as a whole determined by the slowestsignals. Merced engineers started looking for those slowpokes and found waysto speed them up via slight changes to the chip design. Soon, however, itbecame clear that many of these changes were disrupting the chip's delicatesignal ballet, forcing engineers of other modules to rework their designs as well.

The team found itself in a nightmarish world where a change to one modulewould ripple through the work of several hundred other people, leaving moreproblems in its wake. If engineers couldn't balance their signals, the onlysolution would be to slow down the entire chip--unacceptable for what wassupposed to be a groundbreaking design. By mid-1998, the problem hadgrown so serious that Intel announced the chip would be delayed at least sixmonths beyond its planned late-1999 launch.

Not long after, Mr. Thomas decided to pack it in. "I was tired, and probablyclose to burnout," he says. He moved to another post within Intel.

New Boss
By coincidence, an Intel engineering manager from Israel named Gadi Singerwas visiting the U.S., planning to spend a week at headquarters in Santa Claraand then take his family for a trip to the Rockies. Mr. Yu caught up with him thenight he landed and implored him to take over Merced. Mr. Singer made aquick check with his family, agreed, and started at once, not even taking time topack up his apartment in Israel.

An intense, thickly bearded man given to sketching diagrams on any nearbywhiteboard, Mr. Singer was a natural choice, having previously overseen thefinal stages of the Pentium design. His first priority was team morale, which hadslumped badly as the signal-timing problem kept the design from "converging."He stepped up training programs for engineers to give them a break from theirobsessive focus on design issues and started posting a list of "Merced babies"born to team members. One night he ran out to Toys "R" Us, returning withbags of Nerf toys to help people blow off steam.

Mr. Singer also shuffled managers' duties and demanded new informationsystems that helped engineers see more quickly the ripple effects of designchanges. Slowly, Merced began to converge.

The design was finally completed on July 4, 1999. The next month Intelproduced the first physical chip, which worked right off the assembly line. Upto that point, engineers had to run test computations on a simulator, but with areal chip in hand they could begin testing in earnest.

AMD's Plans
Itanium's challenges aren't over. While Intel had made sure the new chip wouldrun older applications made for the x86 line, it is likely to do so more slowlythan equivalent x86 chips. Intel's remedy has been to spend hundreds ofmillions of dollars encouraging software developers to tailor applicationsspecifically to the new architecture, which they're doing.

In addition to H-P, Compaq Computer Corp., Dell Computer Corp. andIBM have thrown their support behind Itanium, but customers will have to beconvinced that Itanium systems can demonstrate the sort of reliability andsecurity now available on high-end servers. That will take time, so few analystsexpect an Itanium rush.

Meanwhile, rival Advanced Micro Devices Inc. is designing its own 64-bitx86 chip, although that project has also been delayed and isn't expected until atleast next year.

Now that the agonies of the Merced development are behind them, Intelofficials say it's been worth the effort, and they've learned valuable lessons. Mr.Yu, for instance, says Intel has learned to limit the initial euphoria of its chipdesigners and to avoid packing a new development project with too manyuntested engineers.

Mr. Crawford, the architect, has felt the stings of Itanium's many setbacks morepersonally than most. "We're headed toward the thrill of victory, but wesuffered a number of the agonies of defeat," he says.