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 : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Windsock who wrote (169572)	8/18/2002 11:07:04 AM
From: greg s	Read Replies (1) \| Respond to of 186894

Windsock, You have hut the nail on the head, re: The indicator of "IPC" is almost exclusively used to explain away the low and embarrassing clock speed of the Athwiper. AMD had to come up with a performance rating that would help put their, shall we say, lacking performance in a more positive light. Intel has the opposite problem. They have to come up with a way to convey exceptional performance at a relative clock speed. They may try to use a performance rating of some kind, but that's not the only way to achieve the goal. IMHO, Intel's position is much more tenable than AMD's. Greg.

To: Windsock who wrote (169572)	8/18/2002 2:27:42 PM
From: Ali Chen	Read Replies (1) \| Respond to of 186894

"Unless you are looking at a table in a software reference manual describing individual instructions" I must be a while when you looked at those manuals. Since incorporation of Out-Of-Order architectures, Pentium-Pro/Pentium-II/K6, both Intel and AMD stopped publishing any clock numbers for individual instructions, e.g. developer.intel.com because they make no sense, are not individually measurable, and sometimes appear as absurdly negative. "The indicator of "IPC" is almost exclusively used to explain away the low and embarrassing clock speed of the Athwiper." There is not much of embarrassment if AMD processors still post equal or better scores on user applications than the majority of mass-produced P4/SDRAM systems. What is really embarrassing is that the bloated frequency of P4 does not automatically translate into higher user application performance, unless you recompile all software. Which is a pretty expensive undertaking. As result, an average user may expect higher speedup on his software investments when upgrading to AMD, or at least it is more cost effective. - Ali

To: Windsock who wrote (169572)	8/18/2002 4:58:21 PM
From: wanna_bmw	Respond to of 186894

Windsock, Re: "Unless you are looking at a table in a software reference manual describing individual instructions, the concept of IPC is not even a fixed number -- it a range." Excellent point, Windsock. The AMDroids have not yet realized this fact, which is why they continue to support the AMD notion that IPC can be summed up in a single QuantiSpeed number. Some of the brighter AMDroids may try to argue in favor of an "average" IPC, but only when the data points that are used for the average are of their choosing. wbmw