SI
SI
discoversearch

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: Paul Engel who wrote (64839)9/16/1998 12:22:00 AM
From: dumbmoney  Read Replies (3) | Respond to of 186894
 
KNI (nee MMX-2) details (from comp.sys.intel):

----------
From: Erik Corry <erik@arbat.com>
Newsgroups: comp.sys.intel
Subject: Katmai/MMX2 details released

According to c't, Intel has released details of MMX2 in Katmai.
heise.de
(English link anyone?).

The Katmai has 8 128-bit MMX2 registers usable as 4x32bit
fp each. They are not aliased to the normal fp or MMX
registers. This is a correct decistion IMHO, would have
been better for MMX, too. It means the OS needs to know
about the extra registers, but upgrading your OS is no big
deal (if you are upgrading your processor), and if they do
it right, the application cannot tell that MMX2 is there
at all unless the OS activates the support, in which case
it is reasonable to expect the OS to also handle saving
and restoring the registers on context switches. (Not sure
whether W95 will support it. Seems beyond doubt that W98
will, as will Linux for both 2.0 and 2.1/2.2).

The advantage is that you can do MMX2 in parallel with
MMX or fp work, and it probably simplifies the internal
data paths.

Did AMD get this one wrong for 3D-Now?

There are conditional move commands, conversion functions
for MMX and FP, memory-streaming (cache control), distance
calculation and min-max commands.
----------

My comment: The new context-switching instructions were introduced in Deschutes (i.e. current PII) and are already used by Win98.



To: Paul Engel who wrote (64839)9/16/1998 1:09:00 AM
From: Dale J.  Read Replies (1) | Respond to of 186894
 
Paul, There is no end to the games these analysts play.

Amen. It is really something. Kurlak continues to talk down Intel just before options expiration. It is amazing he can still kill an Intel rally.

Dale



To: Paul Engel who wrote (64839)9/16/1998 6:51:00 AM
From: GVTucker  Read Replies (2) | Respond to of 186894
 
<< Kurlak has issued several sets of numbers. That is typical for some
sleazy operators.

Now, whatever Intel posts, he can say that his numbers were REAL CLOSE.

He just won't specify which sets of numbers he used.>>

Paul, this isn't true. When was the last time you read one of these reports that you so criticize?