🚀 go-pugleaf

lew...@cs.ualberta.ca (Michael Lewchuk) writes:
>A few comments before I start:
>Please please please if you want to spend time praising each others' comments
>or yelling at each other, take it to email.  I have no desire to read
>a Falcon vs. others squabble mixed in with useful info/speculation about CPUs.

	My humblest apologies.  I ahve written to this effect in my last such
post.


>ak...@crux4.cit.cornell.edu (Allen Kim) writes:
>>Nah, Intel already began work on the P6 before the marriage with HP.  The
>>fruits of that alliance will be the P7, which is due around the time I get
>>out of grad school (considering I go).

>You mean HP+Intel = a really nice CPU?  Btw, what improvements will the 686
>have over the 486 and 586 class processors?

	Yeah, I would also like to know what the heck the 686 is supposed to
be.  It seemed logical for the P7's "specs" to be those of the P6, but according
to what you say above...
	Let's just say you have managed to confuse me.


>Oh, btw, regarding Falcon's comments about the PowerPC coming out: although
>the PowerPC is competitive with the 586 line, it has its own instruction set
>and is not (to my knowlege) 586 compatible.  It may be 486 compatible with
>translation software.  I'm not sure, though.  Any PowerPC people out there?

	That may be hard to get an answer there.  As far as I know down here
the PowerPc is STILL not out. (The PowerMac is, but that's a diff. kettle of
fish).


>In other words, I don't think the PowerPC will make Origin program for
>Pentiums, and I don't think Origin will go Pentium-only any time soon
>because of the vast number of powerful 486 class processors which can
>sometimes outperform the low-grade pentiums, at a significantly lower price.

	If games like WC3 are going to become the norm for Origin (and
assuming they do survive the gamble) then I cannot see them programming for the
486 much.  Already there are people with dx2's who are saying that the SVGA
full fight mode is not fast enough.


>>Some people have called the Pentium "two 486's working together."  That's not
>>true; it's a lot more complicated than that, and it's pretty tough to explain.

	I know its not that simple, but it's fairly close.  I said it was 2
chips rather than 2 486's. (If the latter was the case, I think one would need
to borrow a few heatsinks from some Mechanoids!) ;-)


>How about "A 586 can perform at a rate of 1 to 2 486s depending on the
>nature and optimization of code"?

	Makes sense to me.


>>Oh, but you CAN use these new DX4 Overdrives in most existing boards.  If
>>you have a DX/33 or DX2/66 with a ZIF socket, chances are good that you can
>>remove the chip and plop in the DX4, since both chips operate on a 33 MHz
>>bus.  If you have an SX/25 or a DX/50 (in my case), usually there will be
>>a jumper that can change the bus speed to a 33 MHz setting.

	But you will need the board to run at a clock-triple speed (or something
like that - and yes I know that a DX4 is only clock-3x INSIDE the chip).  If
this was the case then they would NOT had brought out motherboards which take
advantage of the dx2 chips - and I mean specific motherboards which are
"downward compatible" - I know they exist for I have seen them).


>>And if you're wondering about voltage, yes the DX4 is 3.3V while existing
>>motherboards are 5V.  That's why the DX4 has an on-chip voltage regulator
>>so that it can work on 5V motherboards.
>Ain't it wunnerful to be an engineer? :)
>					Michael Lewchuk

	*smile* Not if you have to listen to all those engineer jokes. ;-)

							- Falcon