Amiga.org

Amiga News and Community Announcements => Amiga News and Community Announcements => Amiga Hardware News => Topic started by: Kees on April 22, 2002, 12:11:34 AM

Title: Motorola's MHz = Mega Hurts ?
Post by: Kees on April 22, 2002, 12:11:34 AM
There is an interesting article on OSopinion.com (http://www.osopinion.com) concerning the PPC processors.

"What Motorola doesn't want you to know is that obsolescence was planned from the beginning of its involvement in the PowerPC fiasco."

Click here (http://www.osopinion.com/perl/story/17368.html) for the article ...

Title: Re: Motorola's MHz = Mega Hurts ?
Post by: The_Editor on April 22, 2002, 12:36:22 AM
Makes you wonder how fast Motorola could have made the "060" had they continued development like Intel did with the x86.    :-o
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: yssing on April 22, 2002, 01:01:28 AM
Yep... That would probalby have been the better choice, or alternative...
Geesss. I did not knew that about the G3/G4...
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: Chathurawind on April 22, 2002, 01:21:25 AM
This article  is crap. Even I can see that and I'm no HW expert...
Title: Re: Motorola
Post by: SlimJim on April 22, 2002, 01:39:31 AM
I like this paragraph:
 
"Apple and IBM were pleased, but Motorola's M68060
was running circles around both the Pentium and the
PowerPC 601 in a beast called the Amiga. To put it
in perspective, the '040 and '060 Amiga lines actually
used PowerPC chips as text co-processors!"
 
That is not really the way I recall it...:-)
.
SlimJim
Title: Re: Motorola
Post by: System on April 22, 2002, 02:15:22 AM
Neither do I!

This fellow who write the article isn´t an expert on his area, I agree totaly with that.
But it seems very impressive if some outsider reads it so why bother ?

  :-D
All publicity is good publicity isn´t it?
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: Kent on April 22, 2002, 02:34:45 AM
The article reminds me of the diagramed plot line in the Scarlet letter...  it looks something like this:

_/\_

It starts out flat, increases to a point of OMFG and falls rapidly back down to where it started from.

Another example is that of a dying patient.  The pulse starts out as non existant and after a couple plateens of electro shock treatment to get the heart going again the flat line doesn't falter from its course.

Yawn.  I think I'm going to go and do something useful like watching the TeleTubbies with my kids... what did that nu-nu do this time!?  :-D

:pint:
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: Elektro on April 22, 2002, 03:06:19 AM
Please don't do that to your kids...  :-D
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: Chathurawind on April 22, 2002, 04:18:56 AM
That article is appallingly inaccurate and pretty baseless.

IIRC, the 620 was not even released.

The 601 was the first in the PPC family, and started at 60MHz in the Apple lines. It was fast, the OS was still mainly 68k though, hence emulation slowed the system down. I don't know if Apple ever used the 602, but it was the chip between the 601 and the 603, obviously!

Yes, one wonders where the 68060 would have got to if Apple had used that processor. It would have made faster speeds for sure, maybe up to 200MHz in 1996/1997 time period. But the PPC was much faster than the x86 in 1998.

It is only the recent history of the PPC that it has lost any speed advantage - 1999 onwards in essence when the G4 failed to scale beyond 500MHz and the x86 processors (basically an x86 decoder with RISC core) scales upwards and onwards.

And why? Because Motorola don't care about desktop processors any longer. Embedded is their market. Without the money that Apple is putting into future PPCs, there would be very few upcoming fast processors. As it is, the ones that come out will be solely for Apples use for a long time, we won't see them in the Amiga market.
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: System on April 22, 2002, 07:12:52 AM
The article states this:
"The 602 was used in stadium scoreboards, remote-controlled Transformers and the popular Nintendo 64."

The bit about N64 is simply not true, and I imagine most people on this board know this.  Doesn't do much for the reliability of the author.
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: Chathurawind on April 22, 2002, 10:04:30 AM
>620

Waiting for this CPU is similar to "Waiting for Godot"! :( This CPU is another failed Bull (France) project that bring down the company.

There is another mythical PowerPC 615.

Also there is an old PowerPC compatible vendor: Exponential with its failed PowerPC 704 project. The company went under several years ago.

http://bwrc.eecs.berkeley.edu/CIC/announce/1996/x704-533.html
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: yssing on April 22, 2002, 11:28:01 AM
[IRONIC]
No really.. after reading this article..
who wanst RISC ???
why not just go for x86 ???
[/IRONIC]
 :-D  :-D  :-D
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: whabang on April 22, 2002, 12:13:49 PM
Quote
who wanst RISC ???
why not just go for x86 ???



Newer x86 compatibles is RISC! Well, sort of anyway...
It uses microcode to emulate missing instructions.
Rather effective, since a program optimized for Pentium, K5, 6x86 and later runs as kinda RISCy, but they are still compatible with older CPU's wich still runs the as CISC. They simply removed some instructions and made others more effective.
Correct me if I'm wrong, but I thinks it is that way...
 :-D
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: System on April 22, 2002, 12:41:32 PM
Yeah so there are still stupid persons who think Mhz does matter ... well look at this rc5-64 speeds:

http://n0cgi.distributed.net/speed/query.cgi?cputype=79&cpumhz=2556&recordid=1&contest=rc5&multi=0 (http://n0cgi.distributed.net/speed/query.cgi?cputype=79&cpumhz=2556&recordid=1&contest=rc5&multi=0)

Pentium :

Record id | CPU Name | MHz | OS | Client | Speed
8390 | Intel Pentium 4 | 2556 | Windows 2000 | 2.8015 RC5 | 3,644,544

And now the PPC :

http://n0cgi.distributed.net/speed/query.cgi?cputype=99&cpumhz=1600&recordid=1&contest=rc5&multi=0 (http://n0cgi.distributed.net/speed/query.cgi?cputype=99&cpumhz=1600&recordid=1&contest=rc5&multi=0)

Record id | CPU Name | MHz | OS | Client | Speed
8437 | Power PC 7450/7455 G4 | 1600 | MacOS X 10.1 | 2.8016 RC5 | 16,991,648

So a 2556 Mhz Pentium proccessor ONLY does 3,644,544 Kkeys/sec , while THE LOWERED MHZ PPC
AT 1600 Mhz , DOES 16,991,648 Kkeys/sec !!
What i can say is that is is like 6 times faser then this stupid Puntium 4.

So F*ck Mhz intel-wintel-lovers...

 :-P
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: Chathurawind on April 22, 2002, 12:44:34 PM
Hmm, that makes one wonder it it wasn't all planned to make Amiga go down.
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: yssing on April 22, 2002, 01:22:26 PM
ohh nnoo...
Conspiracy... so Motorola did way back then, start to plan the amiga downfall...
Lets hope not...
But ooh my, they did inhibited the MAC in reacing its full potential... what will this do to amiga...

The voices say NOOO....  :-D  :-D  :-D
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: Bobsonsirjonny on April 22, 2002, 01:42:50 PM
I've just been thinking - so please correct me if I am wrong.... If Apple have PPC pretty much to themselves - and we wont get much of a looking....


If X86 chips are cheaper than PPC - and if X86 are just RISC chips that emulate some CISC instructions - is it possible to write Amiga OS so that it only accesses the X86 RISC instructions - and not the CISC ones? Would that provide cheaper, faster more efficient hardware? Of course come future OS's Hardware will become agnostic!!!!


Just a thought - I dont know much about that stuff.
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: anarchic_teapot on April 22, 2002, 03:02:54 PM
Quote
I've just been thinking - so please correct me if I am wrong.... If Apple have PPC pretty much to themselves - and we wont get much of a looking....


Which was one of the reasons givens for not going for a G4 processor in the AmigaOne: Apple hogs all the top-of-the-range chips.

OTOH, the A1 uses an IBM chip, not Motorola.

Title: Re: Motorola's MHz = Mega Hurts ?
Post by: redrumloa on April 22, 2002, 03:56:44 PM
Quote
OTOH, the A1 uses an IBM chip, not Motorola.


Yes! Now we need IBM to lay their nuts on the table and start cranking up Mhz speed. :-o
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: ido on April 22, 2002, 04:02:50 PM
Quote
So a 2556 Mhz Pentium proccessor ONLY does 3,644,544 Kkeys/sec , while THE LOWERED MHZ PPC AT 1600 Mhz , DOES 16,991,648 Kkeys/sec !! What i can say is that is is like 6 times faser then this stupid Puntium 4.


Nico, don't believe everything you read. you can probebly find some other article that says exactly the opposite... i don't think we can actually know for sure how fast this cpu is, but i don't think the G4 is 7.44 times faster then the P4 Mhz to Mhz.
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: whabang on April 22, 2002, 05:06:55 PM
@bobsonsirjonny

Even if current x86 compatibles are using a reduced instruction set, it is still not a real RISC.

Sure, there has been a significant performance increase when switching to this technique ( just try to compare a 486 120 to a Pentium 120 ), but it is still not optimized for RISC operation. The old instructions are still there. Some of them are just emulated, and some of the are optimized, but as long as it is to be backward compatible it will be slower than PPC and other RISC CPU's. Try running a DivX movie on a Pentium 200 MMX - Impossible! On the other hand, a 206 MHz StrongARM does the job without problems.
Naahh... the 486 to Pentium step was a good idea, and considering the MHz race between AMD and Intel one can get rather decent performance out of it, but it won't las for ever...
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: Dr_Righteous on April 22, 2002, 09:13:16 PM
Quote

Even if current x86 compatibles are using a reduced instruction set, it is still not a real RISC.


Um, isn't that a paradox? RISC = Reduced Instruction Set Computing.

And incidently, AMD processors are FULL RISC processors with extra transitors to emulate x86 instructions. AMD did just release a chip without the x86 emulation.

Something tells me this is all part of the grand plan of Amiga Inc. I'm not exactly sure how, but it's a gut feeling.
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: Chathurawind on April 23, 2002, 11:18:56 AM
Quote
Um, isn't that a paradox? RISC = Reduced Instruction Set Computing.


Yah, Yah, Yah!!! I talked before I thought... Stupid me!!!
*bangs my head into my desk* :-D  :-D  :-D

Didn't know that bout the athlons tho'.
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: whabang on April 23, 2002, 11:19:59 AM
Yikes! Forgot to log in!
Title: Re: Motorola's MHz = Mega Hurts ?
Post by: ido on April 23, 2002, 12:13:14 PM
Quote
And incidently, AMD processors are FULL RISC processors with extra transitors to emulate x86 instructions. AMD did just release a chip without the x86 emulation.


you mean the alchemy? that's not a PC class CPU, it's just for PDAs and stuff. i really hope they'll release their hammer CPU in a non-x86 version!