gezzas525
Full Member level 3
ITANIUM a mistake
Intel's gamble on high-speed computer chip not paying off
By Therese Poletti and Dean Takahashi
Mercury News
In its biggest strategic mistake in a decade, Intel has spent an estimated $2 billion creating a high-speed computer chip, the Itanium, that most customers don't want and don't need.
Once intended to help the Santa Clara chip giant extend its dominance from personal computers to corporate computing, the Itanium has been such a flop that many industry insiders call it the ``Itanic.''
Moreover, Intel's insistence on supporting the Itanium over a cheaper, more popular technology that accomplishes many of the same aims has opened the door for its archrival, Advanced Micro Devices, to make inroads into the market for high-power machines companies use to manage financial transactions and run corporate networks.
Even Hewlett-Packard, which co-developed the Itanium with Intel, has bowed to customer demand and said it would sell AMD-powered servers.
About-face
Two weeks ago, Intel did an about-face, announcing that it would incorporate technology that mimics AMD's fast-selling Opteron chip into cheaper Intel chips being introduced later this year.
The Itanium itself isn't dead but appears destined to be a niche product for extremely high-end scientific tasks like simulating global weather patterns and designing drugs.
That leaves Intel with a conundrum: As AMD improves its Opteron chip, Intel will face pressure to enhance its own Opteron-like chip. At some point, that Intel chip could be competing with Itanium for the same customers.
``The only thing Itanium does any better is scientific applications,'' said Linley Gwennap, principal analyst with the Linley Group in Mountain View. ``I don't see how Intel can continue to sink money into it.''
It wasn't supposed to be that way.
Intel and HP began working on the Itanium in 1994 to address the needs of customers running ``big iron,'' the huge servers that manage large databases and networks. It was a lucrative market dominated by players like Sun Microsystems and IBM.
The idea was to build a microprocessor -- the brains of a computer -- that would be far more efficient at handling data than the technology used in the 386, 486 and Pentium chips that are the heart of PCs even today.
32 vs. 64 bits
Since the 1980s, the design of those PC chips has only allowed them to handle 32 bits, or pieces, of data at a time. The Itanium design handles 64 bits at a time. By processing data in 64-bit chunks, a computer can access much more memory. That allows it to tackle problems a gulp at a time, rather than nibble by nibble.
But Intel made a crucial decision: It used a new technology that wasn't compatible with its other chip designs.
That meant if a customer wanted to upgrade to the faster chip, it would have to invest in new software, too.
Still, Intel, the world's leading chip maker, bet that its high-volume manufacturing could produce cheaper chips than those from IBM and Sun Microsystems.
``It seemed natural they would dominate,'' said Gordon Haff, senior analyst at Illuminata, a market analysis firm in Nashua, N.H.
Instead, by the time the Itanium came to market in 2001, the economy was in the doldrums, and computer administrators who once spent heavily on new machines were now more concerned with extending the life of older software.
The result: Intel sold just 100,000 Itaniums last year at an average price of $2,000 -- small change for a company that churns out hundreds of thousands of chips a day for PCs and small servers.
Sunnyvale-based AMD, Intel's longtime rival, also made a move toward 64-bit technology, but it chose a design compatible with the older 32-bit chips customers were already using.
Last April, it launched the Opteron, a low-cost chip that could do some of what Itanium promised. The performance was good enough for most customers, and the Opteron chips cost an average of just $600. According to market research firm IDC, 35,000 servers were sold in 2003 with Opteron chips, compared with 19,000 servers with Itanium chips. But IDC notes that the Itanium servers typically had four processors, and were higher priced.
Not declaring victory
AMD, a perennial second to Intel, isn't declaring victory yet, but it's pleased by its success. ``It is clear the momentum is on our side,'' said Dirk Meyer, senior vice president of AMD's computation-products group.
Intel executives say there is no reason to rush to the lifeboats.
``In the second half of 2003, we saw usage triple,'' said Lisa Graff, director of Intel's Itanium group, referring to corporate use of Itanium. She said Intel expects sales to double in 2004.
SGI is using the Itanium in supercomputers it has sold to NASA, which has used them to create a model of how the world's oceans behave. And while HP has adopted Opteron, executives said they also remain committed to making Itanium-based machines.
Scott McNealy, chief executive of Santa Clara-based Sun, which offers servers using its own 64-bit chips as well as AMD's Opterons, said Intel went out on a limb with Itanium. And now, ``I think I hear some wood cracking,'' he said.
Intel hopes to catch up with AMD on the technology front by the summer. It can recover so quickly only because a few renegade engineers at Intel's Oregon facilities thought the Itanium was a mistake and started developing a 64-bit technology that was compatible with Intel's older chip designs.
But it will be too late for many customers. The San Diego Supercomputer Center, which committed to Itanium chips in its supercomputers, plans to buy Opteron-based servers for less complex tasks.
``The Itanium is being priced into the grave, and that's too bad,'' said Mason Katz, a manager at the federally funded research center. ``Just 18 months ago, it seemed like the Itanium was the future of 64-bit computing.''
Well well nothing new here, I said all of this before.
Intel's gamble on high-speed computer chip not paying off
By Therese Poletti and Dean Takahashi
Mercury News
In its biggest strategic mistake in a decade, Intel has spent an estimated $2 billion creating a high-speed computer chip, the Itanium, that most customers don't want and don't need.
Once intended to help the Santa Clara chip giant extend its dominance from personal computers to corporate computing, the Itanium has been such a flop that many industry insiders call it the ``Itanic.''
Moreover, Intel's insistence on supporting the Itanium over a cheaper, more popular technology that accomplishes many of the same aims has opened the door for its archrival, Advanced Micro Devices, to make inroads into the market for high-power machines companies use to manage financial transactions and run corporate networks.
Even Hewlett-Packard, which co-developed the Itanium with Intel, has bowed to customer demand and said it would sell AMD-powered servers.
About-face
Two weeks ago, Intel did an about-face, announcing that it would incorporate technology that mimics AMD's fast-selling Opteron chip into cheaper Intel chips being introduced later this year.
The Itanium itself isn't dead but appears destined to be a niche product for extremely high-end scientific tasks like simulating global weather patterns and designing drugs.
That leaves Intel with a conundrum: As AMD improves its Opteron chip, Intel will face pressure to enhance its own Opteron-like chip. At some point, that Intel chip could be competing with Itanium for the same customers.
``The only thing Itanium does any better is scientific applications,'' said Linley Gwennap, principal analyst with the Linley Group in Mountain View. ``I don't see how Intel can continue to sink money into it.''
It wasn't supposed to be that way.
Intel and HP began working on the Itanium in 1994 to address the needs of customers running ``big iron,'' the huge servers that manage large databases and networks. It was a lucrative market dominated by players like Sun Microsystems and IBM.
The idea was to build a microprocessor -- the brains of a computer -- that would be far more efficient at handling data than the technology used in the 386, 486 and Pentium chips that are the heart of PCs even today.
32 vs. 64 bits
Since the 1980s, the design of those PC chips has only allowed them to handle 32 bits, or pieces, of data at a time. The Itanium design handles 64 bits at a time. By processing data in 64-bit chunks, a computer can access much more memory. That allows it to tackle problems a gulp at a time, rather than nibble by nibble.
But Intel made a crucial decision: It used a new technology that wasn't compatible with its other chip designs.
That meant if a customer wanted to upgrade to the faster chip, it would have to invest in new software, too.
Still, Intel, the world's leading chip maker, bet that its high-volume manufacturing could produce cheaper chips than those from IBM and Sun Microsystems.
``It seemed natural they would dominate,'' said Gordon Haff, senior analyst at Illuminata, a market analysis firm in Nashua, N.H.
Instead, by the time the Itanium came to market in 2001, the economy was in the doldrums, and computer administrators who once spent heavily on new machines were now more concerned with extending the life of older software.
The result: Intel sold just 100,000 Itaniums last year at an average price of $2,000 -- small change for a company that churns out hundreds of thousands of chips a day for PCs and small servers.
Sunnyvale-based AMD, Intel's longtime rival, also made a move toward 64-bit technology, but it chose a design compatible with the older 32-bit chips customers were already using.
Last April, it launched the Opteron, a low-cost chip that could do some of what Itanium promised. The performance was good enough for most customers, and the Opteron chips cost an average of just $600. According to market research firm IDC, 35,000 servers were sold in 2003 with Opteron chips, compared with 19,000 servers with Itanium chips. But IDC notes that the Itanium servers typically had four processors, and were higher priced.
Not declaring victory
AMD, a perennial second to Intel, isn't declaring victory yet, but it's pleased by its success. ``It is clear the momentum is on our side,'' said Dirk Meyer, senior vice president of AMD's computation-products group.
Intel executives say there is no reason to rush to the lifeboats.
``In the second half of 2003, we saw usage triple,'' said Lisa Graff, director of Intel's Itanium group, referring to corporate use of Itanium. She said Intel expects sales to double in 2004.
SGI is using the Itanium in supercomputers it has sold to NASA, which has used them to create a model of how the world's oceans behave. And while HP has adopted Opteron, executives said they also remain committed to making Itanium-based machines.
Scott McNealy, chief executive of Santa Clara-based Sun, which offers servers using its own 64-bit chips as well as AMD's Opterons, said Intel went out on a limb with Itanium. And now, ``I think I hear some wood cracking,'' he said.
Intel hopes to catch up with AMD on the technology front by the summer. It can recover so quickly only because a few renegade engineers at Intel's Oregon facilities thought the Itanium was a mistake and started developing a 64-bit technology that was compatible with Intel's older chip designs.
But it will be too late for many customers. The San Diego Supercomputer Center, which committed to Itanium chips in its supercomputers, plans to buy Opteron-based servers for less complex tasks.
``The Itanium is being priced into the grave, and that's too bad,'' said Mason Katz, a manager at the federally funded research center. ``Just 18 months ago, it seemed like the Itanium was the future of 64-bit computing.''
Well well nothing new here, I said all of this before.