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 : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: Charles Gryba who wrote (78592)	4/27/2002 10:03:08 PM
From: wanna_bmw	Read Replies (1) \| Respond to of 275872

Constantine, ok - no more sarcasm. Here is how your argument appears to me. It sounds like you think Intel is being "wasteful" by putting excessive amounts of cache on their processors, rather than coming up with more "innovative" ways of increasing performance. In my opinion, there are plenty of legitimate ways to increase performance in a microprocessor. Integrating the memory controller is one of them. It's a sure way to decrease latency in reading data from memory, thus relieving the memory bottleneck. AMD took that risk, and if it pays off, then they deserve whatever returns they get. Intel, on the other hand, decided to increase the cache size of their microprocessors, thus reducing the amount of traffic to memory, which also relieves the memory bottleneck. They do it this way because it's easier, it requires less validation, less design work, less complexity, and it offers more flexibility. They can do this because they have the manufacturing capacity to produce larger chips. It's an advantage, but not an "unfair" advantage. Yet some AMDroids (no names) like to look at this as "cheating", as if AMD "innovates", while Intel sneaks by with the "quick and dirty" method. I've heard similar arguments before. These same people are the ones that complain about the Pentium 4 memory bandwidth, since quad-pumping the bus isn't a "fair" way to increase performance. Intel isn't allowed to showcase optimize applications either, because it doesn't reflect performance in "real world" applications. And others - Intel doesn't really make money because of "accounting tricks", rather than good management, they only retain market share through "deceptive marketing" and "strongarming", rather than salesmanship and good customer relations, yatta, yatta, etc. The same people used to complain that Intel's cache size was too small, and that AMD had the advantage with a larger L1 cache. And now, the cache size is too big, and Intel's unfair advantage adds too much cost to the end user. What a crock. Nothing is "illegitimate" or "unfair" in engineering. If you can get more performance with a different design and you are able to manufacture it, then you do it. I don't know what you think McKinley will get in terms of benchmark scores, but it's been my projection that they will be quite good, even with respect to Intel's x86 line. If I am wrong, and Intel continues to strike out with IA-64, then maybe you are right, and the architecture simply can't compete with the progress made with x86. However, if I am right, then Intel is justified in pursuing this architecture, proprietary or not. And with that, I am done. It's no use explaining the obvious to someone whose preconceived notions have already convinced them one way or another. wbmw