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.
Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?


To: grok who wrote (51748)3/6/1999 1:10:00 AM
From: Cirruslvr  Read Replies (1) | Respond to of 1572749
 
"Perhaps Intel's legal department should be called Stupid Inside."

I thought these were good reads not yet posted here.

"Intel puts legal screws on Ugeek"
theregister.co.uk

"Movement to halt Pentium III grows"
news.com

"Hacking Intel's CPU ID"
hardwarecentral.com

On a side note, that Next Generation Playstation is going to have a very large processor.

"Gate width 0.18 micron

VDD Voltage 1.8 V

Power Consumption 15 Watts

Metal Layers 4

Total Transistors 10.5 Million

Die Size 240 mm2

Package 540pin PBGA"

240mm2 on .18 micron!? Could one of the reasons be because it has 4 layers?

ars-technica.com



To: grok who wrote (51748)3/6/1999 3:28:00 AM
From: Ali Chen  Read Replies (1) | Respond to of 1572749
 
<Alpha is said by its designers to be short tick because it has an extremely high clock rate but doesn't do a whole lot in a cycle. It also has a long pipeline.>
Nice introduction, but everything is backwards.
Alpha has a long pipeline because they WANT every stage
to be "short tick". To make the tick "short", they had to
reduce amount of sequential gates in stage logic,
and therefore they can't do "a whole lot" in a cycle,
as you correctly noticed. As a result, the pipeline
must have more stages to execute same instructions.
But because the number of sequential gates in stages
is strictly limited by design, the stage propagation
time is short, and that is why the frequency is high.
This is called a "superpipelined" design. It was invented
by DEC and apparently stolen by Intel.

As it follows from AMD optimization guide, K6 executes
most instructions in just 4 stages. Now compare this
with 12-14 stages in Pentium-II/Pro, and you may arrive
to a conclusion that the stage complexity in K6 must be
much higher, not to say 3 times, but around that.
However, the overall K6 frequency is just 20% below
Pentiums. Now draw your own conclusion about who makes
better FETs. Only idiots from Intel Supremacy Klan with
"make is so Phd" cannot see this "tiny architecture
thing" and blatantly try to compare incomparable things.

Take care,
- Ali
<end of the lecture>