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 : IDTI - an IC Play on Growth Markets -- Ignore unavailable to you. Want to Upgrade?

To: srvhap who wrote (8479)	5/23/1998 1:12:00 PM
From: Hippieslayer	Read Replies (1) \| Respond to of 11555

There is a Cnet article--sorry I don't have the link--about mobile CPU's which didn't mention the C6 at all. What I got from the article is that INtel is going to make a major push with their wilted celery chip in the mobile marker. IDT better get their shit together and attack this market cuz I think it will the only area that they might be able to get a better foothold in than on the desktop platform

To: srvhap who wrote (8479)	5/23/1998 1:37:00 PM
From: Rob S.	Read Replies (1) \| Respond to of 11555

X86 refers to all legacy 3X86, 4X86, 5X86, and Pentium class processors - basically anything that runs windows. MMX and a few Pentium specific instructions have been added to the legacy X86 instructions but most programs that use these newer additions can run on processors that don't have them. For instance, many games that take advantage of MMX instructions will still run on non-MMX machines only they run slower. The reason that the Pentium generations of processors are so expensive and so complex is because they must run the legacy X86 instructions. uP technology has evolved tremendously over the past several years since X86 was developed and although the uPs now use many of the design conventions of more advanced architectures, they are still limited by the requirement to use the legacy instructions. Bottom line is that newer, more efficient uP architectures are possible today than the processors that must handle the legacy software. The Merced is designed "from the floor up" to be used in multi-processor systems and with multiple execution units either on the same piece of silicon or segmented onto separate chips. It is supposed to have much truer linearity of the scaling of performance. What that means can be explained by what happens when you add multiple processors in a Pentium system. The Pentium Pro can be used with two or more processors on the same motherboard. The idea is that the operating system can split up the program to run separate parts of it on the individual processors and then combine the results to achieve increased overall performance. The reality is that the resulting increase is typically only 25% better than a single uP. Two times the uPs gets you only 1.25 X the performance - marginal, complex, and costly. Linear scalability should result in much closer to 2X-performance improvement. Up to several Merced parts will be used in a system to deliver very high-end performance in workstations, servers, and mainframe capable systems. You will probably see a lot of built in redundancy so that system failures are more easily monitored and shifted over to other paths in near "real time". The Merced won't be competition for mainstream PC computing for at least three or four years. Although the first units will run X86 (Windows) programs about as well as Pentiums, they will sell for a much higher price. Mainstream software to take advantage of the Merced will likely take several years to evolve. I'm not clear at all about how this will occur or if the Merced will migrate down to the desktop or into portables and Internet devices. Merced will open up large, high margin markets for Intel and should become the thrust of their growth strategy for the next several years.