http://hardware.itmanagersjournal.com/hardware/05/03/03/0226235.shtml?tid=78
Will Big Blue open source the coming Cell chip?
Thursday March 03, 2005 (06:00 PM GMT)
Topic: Components
By: Jay Lyman
There has been far more speculation than actual details of the Cell chip -- a next-generation computer processor being developed jointly by IBM, Sony, and Toshiba. However, beyond the general information on the chip's architecture, capabilities and potential applications revealed last month, the publication of more about Cell on the Power.org site leads one close observer of the technology, Jim Trounson, to predict IBM will open source it.
Trounson, who is among a group from industry, academics, and finance forming Cell Industries, contends that the increased amount of information on Cell appearing on Power.org forecasts IBM's course for Cell.
"IBM cares about nothing other than IT systems for the Fortune 500," he wrote in an email to ITMJ. "IBM's 'official' hobby is open source software. IBM's 'secret' hobby is the Cell processor. IBM will mix open source software with the Cell processor to disrupt Microsoft and Intel and prevent them from playing in the Fortune 500."
Cell vision
Amid speculation of the Cell strategy, Trounson said he had been concerned about some players being locked out of using Cell hardware. After joining Power.org and looking at what IBM has done disclosing information on Cell and on building a complete PowerPC 750 system, Trounson said he is convinced that the Cell hardware will come ready for Linux and not for Redmond.
"The best outcome for IBM is: Intel is reduced to a Cell chip producer under license and Microsoft fades away as the Cell takes the PC market running Linux, OpenOffice, and Firefox," Trounson wrote. "All the engineering and back office applications port to the Cell and run 50 times faster. Dell, HP, and Tiger Direct switch to Cell hardware."
Trounson -- who plans to build hardware, software or both for Cell -- conceded that he had not received any response to his queries on the theory from IBM, despite sending email. He said others in his group attended Cell events at ISSCC but did not receive any information on timing, pricing, or open sourcing.
IBM responded to an ITMJ query on the matter by indicating it was "too premature to get into those specifics."
Big Blue's big move
Nevertheless, Trounson said he thinks the different players behind Cell will each get their rewards, with Big Blue bringing home the profits of PC domination while Linux vendors split the role of Microsoft.
The Cell observer said Sony, however, would prefer that all software developers other than its own be locked out of cell programming as long as possible.
"Because Sony wants to develop and sell content," Trounson said. "Sony views any programmable Cell system as a threat to their game console strategy."
But Trounson indicated Sony will only get a running start so it can establish its Cell-powered PlayStation3 as a viable product. If the PS3 sells well, it will consume all available Cell chip production for a year or more, according to Trounson, who added unauthorized programming will be thwarted by DRM built into the hardware.
As for Toshiba, the Cell project adds technology and credibility to the Japanese company, which will sell the chips and Cell memory, but will not have system-level efforts, Trounson predicted.
"Toshiba views Sony as their largest Cell customer and will not cross up Sony," he said.
IBM, on the other hand, will "recruit an army of developers" during the first year of Cell production by supplying software development systems -- as many as 100,000 -- to major application developers and large companies, as Trounson told ITMJ.
"Then IBM will publish an open source hardware and BIOS reference design which was used to develop all the application software," he wrote. "IBM will then allow the market to just fight it out. All the existing motherboard manufacturers will produce Cell motherboards. Dell and HP, etc. will assemble systems and sell them through their existing network. If any of the big players hesitate or stumble, someone like me will be right there to take their place.
"Some development machines will go to open source programmers who will start with open source code and create alternatives to all the expensive engineering of back office applications," he continued. "New classes of software will appear which are made possible by economics of the Cell technology. Cell systems will be designed to avoid the plague of viruses and spyware caused by overly complex hardware and an OS designed only with one objective -- get paid for every copy, who cares if it works."
While some might doubt his speculation, Trounson did not stop there, indicating that Cell may help IBM supplant Intel as the dominant chip supplier.
"IBM will not need to provide much help," he said. "If they publish a reference design, the world will take it from there just like was done with the PC. IBM does a really good job with documentation."
Cell Industries predicts that Intel will be building Cell with cooperation from IBM within a year.
Cell, software, and Microsoft's demise
For the anticipated finale, and the end of Microsoft dominance as we know it, Trounson forecast that IBM will not give Microsoft hardware to work with, and will cash in on its support for open source and Linux.
He said the initial software development tools and support for Cell will be C++ compiler and assembler with extensions to handle software product evaluations (SPEs). System clock based timing and tracing will allow programmers to follow the action, Trounson predicted. He added he hopes to see a compiler that could add code to a running application and link to that code.
"Applications should be able to produce code, compile that code, add that code to itself, and test the new code in a box," he wrote.
In terms of other software vendors, Trounson said Red Hat and SUSE/Novell will also figure into the Cell strategy.
"They both have very close association with IBM and will realize they are being groomed to replace Microsoft," he wrote. "Red Hat and SUSE/Novell will not be a monopoly like Microsoft and will not chalk up huge cash reserves. They will make a reasonable living packaging and supporting Linux as it evolves."
Cell Industries forecasts that as Intel begins producing Cell chips, Microsoft will try to port its operating system to the new processor. However, Linux will have a significant head start and Microsoft will in turn "fall apart."
"When hardware is commercially available, Windows will take two to three years to get the first version going," Trounson said. "IBM already has Linux running on the Cell [at that point]."
Adding that Cell chips will be in short supply for years, Trounson acknowledged that the prediction represents the unprecedented.
"The world has never seen a step change in technology like what is about to occur," Trounson said.