Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : HWP -- Hewlett Packard -- Ignore unavailable to you. Want to Upgrade?

To: Jody Ritchie who wrote (2813)	1/10/1999 11:07:00 AM
From: William Partmann	Read Replies (1) \| Respond to of 4722

Jody, as you know, Merced is not dead. Though Ramsey may wish it so. The following has a few nuggets on its timelines. Singer on the Merced David Lammers Gadi Singer, who with Steve Smith runs the Merced project at Intel, will tell you that cross-training works. A few years after graduating from Technion University in Israel in 1983, Singer went to work for Intel on a 80386 commercial compiler project, and then moved on to CAD tool development. His EDA development has been interspersed with work as a chip designer, notably on the Pentium team from 1990 to 1993. From late 1993 until the autumn of 1998, Singer was Intel's corporate CAD manager, working on technologies to develop Merced. "It is extremely effective to move from CAD to processor design. When you work in CAD, you learn what can be done, and if you are just in design sometimes it is hard to know what the possibilities are. Then when you go back from design to CAD you know what needs to be done. It is very complementary." Singer also has a good view of what can be left in the hands of commercial EDA vendors, and what tools must be developed at Intel. "We have very clear criteria about what to make or buy." As chairman of the EDA Industry Council for the past two years, Singer observed "a very significant shift in the openness of the EDA vendors, in the development of standards and quality standards. Whoever is elected as the next chairman will need to be very strong to make sure that the practice of this trend toward openness becomes established, in a repeatable fashion." Greg Spirakis-after working "two in a box" with Singer for some time-has taken on sole management of the design-technology division at the microprocessor technology group. Singer has been working "two in a box" with Smith for the past four or five months to co-manage the Merced effort. Merced samples will go to OEMs at midyear, and production starts in mid-2000, he said, with everything on track to meet those dates. A decade ago, Singer helped develop a technique that allows a high level of formal validation between the RTL description and the schematics. And he was one of four engineers to develop the Intel hardware-description language, Intel's version of Verilog. That work helps explain why Singer feels confident: at the Intel Developer's Forum in November, Intel disclosed that the Merced design-as described in the Intel RTL language-had booted up successfully. "I think because of my earlier work, I understand how significant that is. The RTL is very closely coordinated and validated with the physical implementation." When the first Merced samples ship, that claim will be tested.