@Arko

Quote:

When Motorola decides to go PPC, they made a bad decision

???

Quote:

This step would have been necessary before they started the 68040, most workstation manufacturers had already switched to other architectures before the 68040 was introduced.

As you wrote, 680x0 was no longer competitive. In 1994 (when 68060 was released) Intel introduced 100 MHz Pentium...

@pavlor

Quote:

pavlor wrote: @Arko Quote: When Motorola decides to go PPC, they made a bad decision

???
[/quote]

Motorola Freescale have lost the CPU desktop market, they lost the Notebook market, they lost the console market, they lost most of the PDA market. they had to licence CPU concepts from other companies ( PPC and ARM )

Is there still something you don't understand ?

Quote:

As you wrote, 680x0 was no longer competitive. In 1994 (when 68060 was released) Intel introduced 100 MHz Pentium...

This step would have been necessary before they started the 68040, the 68040 was started 1990: http://en.wikipedia.org/wiki/Motorola_68040
If Motorola had done the right step SUN and others would never had developed their own CPU technology
SUN http://en.wikipedia.org/wiki/Sun_Microsystems
APOLLO http://en.wikipedia.org/wiki/Apollo_Computer
Silicon GFX http://en.wikipedia.org/wiki/Silicon_Graphics,_Inc.

Last edited by Arko on 14-Apr-2010 at 08:00 AM.

_________________
AmigaONE. Haha. Just because you can put label on it does not make it Amiga.

I borrowed this comments from here (#27 & #28):
http://amigaworld.net/modules/newbb/viewtopic.php?topic_id=38873&forum=2&start=20&order=0

@pavlor

IIRC, 68060 was ahead of P100 in integer performance per clock and a head in FPU stuff in general.
But motorolla could not stay in Mhz competition.
Too bad that first PPC macs were so sluggish because of 68k emulation.
It gave really bad PR for the industry.

I wonder what the future had been if CBM had managed to stay alive get AOS on RISC around 1996...

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

Quote:

IIRC, 68060 was ahead of P100 in integer performance per clock and a head in FPU stuff in general. But motorolla could not stay in Mhz competition.

More important: Motorola couldn't find anyone to use these CPUs... That's the resaon they stopped improving 68k. For some time there was some interest in ColdFire but it quickly died in favour of ARM...

_________________
http://www.warpdesign.fr/

The 68060 did had great potential.
As you know MOTOS engineers planned to launch a 68060+/68080 which would have kicked butt.

If you add a litttle more cache to the 68060 and increase ICache fetch throughput a bit then the Core becomes real powerful.

The 68K api is nicer than x86 in many ways but internally both CISC architectures have a lot on common. With the same or less effort you make get an x86 to speed (see ATHLON or CORE) you could get a 68K Core to the same or better speed.

_________________
APOLLO the new 68K : www.apollo-core.com

@KimmoK

Quote:

KimmoK wrote: @pavlor IIRC, 68060 was ahead of P100 in integer performance per clock and a head in FPU stuff in general. But motorolla could not stay in Mhz competition.

When the 68060 was introduced, Motorola had lost the main customers for this Desktop/Server CPU. All workstation manufacturers had developed their own CPUs. Commodore had a shrinking market and Apple wanted to switch to something complete different.

Wen Phase5 introduced their PPC cards, they became the main customer for the 68060 (others like CISCO used the EC/LC version) this tiny market was not woth any investions.

The decicions that lead to this situation, where done years before. If the 68040 would have been there earlier if it would had more power, Apple would never have switched to PPC, SUN, Apollo, HP and Silicon would had used this CPUs for their workstations and WindowsNT would never have become so important.

Last edited by Arko on 14-Apr-2010 at 08:37 AM.

_________________
AmigaONE. Haha. Just because you can put label on it does not make it Amiga.

I borrowed this comments from here (#27 & #28):
http://amigaworld.net/modules/newbb/viewtopic.php?topic_id=38873&forum=2&start=20&order=0

@Arko

Quote:

The decicions that lead to this situation, where done years before. If the 68040 would have been there earlier if it would had more power, Apple would never have switched to PPC, SUN, Apollo, HP and Silicon would had used this CPUs for their workstations and WindowsNT would never have become so important.

Arko is right.

The early RISC chips showed up to be better in performance than the 68030.
Early Risc chips like the MIPS/SPARC did came out in 1987/88 and did put the 68k line under competitive pressure.
MOTO wanted to bring out the 68040 as answer to this.
But MOTO was not able to bring out the 68040 as quickly as planned.
The 68040 did show up late . The 68040 came in 1990.

The 68040 was much better than the 68030 but MOTO failed to reach the high clockrates needed to compete with the RISC chips that were available by the year 1990.

In other words MOTO was playing catchup with the performance market.
This was the reason that the Unix industry moved to RISC at this time.

The 68060 was great but it came out at a time when the business decision were made already and the war was basically lost already.

_________________
APOLLO the new 68K : www.apollo-core.com

@Arko

I would not be so sure that 68k was developed too slowly in early 90's.
But perhaps it was a bit too pricey and after RISC hype started, everyone had to have some RISC chip and interest to 68k faded away.

http://everything2.com/title/CPU+history%253A+A+timeline+of+microprocessors
http://processortimeline.info/proc1990.htm

(also RISC seemed to start to look like it develop faster than CISC)

Last edited by KimmoK on 14-Apr-2010 at 09:28 AM.
Last edited by KimmoK on 14-Apr-2010 at 09:20 AM.
Last edited by KimmoK on 14-Apr-2010 at 09:19 AM.
Last edited by KimmoK on 14-Apr-2010 at 09:19 AM.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@Thread
Regardless of stupid polemics, I say bring them on!! I want both official NG Amigas and Natami systems for sale at all Amiga retailers.
Stop the talking and start producing.

_________________
SamFlex Complete 800Mhz System + AmigaOS 4.1 Update 4
Amiga 2000 DKB 2MB ChipRam GVP G-Force040 Picasso 2 OS3.9 BB2
AmigaCD 32

@KimmoK

Quote:

KimmoK wrote: @Arko I would not be so sure that 68k was developed too slowly in early 90's. But perhaps it was a bit too pricey and after RISC hype started, everyone had to have some RISC chip and interest to 68k faded away. http://everything2.com/title/CPU+history%253A+A+timeline+of+microprocessors http://processortimeline.info/proc1990.htm (also RISC seemed to start to look like it develop faster than CISC)

Intel Pentium Pro (P6, November 1995) has effectively shot down any RISC ISA desktop dominance. This is followed by AMD K5. Then you have 1Ghz race between Intel (Pentium III) vs AMD (K7 Athlon). “Old school” RISC players were caught in the crossfire.

"Intel Core 2 microprocessors are based on the Intel Core micro-architecture, a distant relative of P6." Intel Core i3/i5/i7 is a variant of Core 2 design.

PC GPU wars has the same effect on custom graphic chipsets i.e. NVIDIA and AMD/ATI.

Last edited by Hammer on 14-Apr-2010 at 10:58 AM.
Last edited by Hammer on 14-Apr-2010 at 10:51 AM.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@Arko

Quote:

Arko wrote: @KimmoK Quote: KimmoK wrote: @pavlor IIRC, 68060 was ahead of P100 in integer performance per clock and a head in FPU stuff in general. But motorolla could not stay in Mhz competition. When the 68060 was introduced, Motorola had lost the main customers for this Desktop/Server CPU. All workstation manufacturers had developed their own CPUs. Commodore had a shrinking market and Apple wanted to switch to something complete different. Wen Phase5 introduced their PPC cards, they became the main customer for the 68060 (others like CISCO used the EC/LC version) this tiny market was not woth any investions. ...

CISCO PIX 50x uses X86 processors.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@Hans

Quote:

SDL games use it, because it allows higher quality graphics, but it's slow if it's done in software.

This seems silly as we are using powerful CPUs. I mean 800Mhz is pretty darn fast so it should whip at soft sprites! But I can see that you can't just XOR a bitmap onto the screen as 32-bit ARGB data isn't a bitmap format and each 32-bit word on screen must be treated like one bit. So more for the CPU to calculate then store and it has to split the merge into a line by line basis.

I guess at this point BltBitmapRastPort() is better than SDL at bitting "BOBs" since AmigaOS can use hardware and compositing to do the job. Espcially with an alpha channel to think about.

Quote:

Yes.

Where?

@Hypex

Quote:

This seems silly as we are using powerful CPUs. I mean 800Mhz is pretty darn fast so it should whip at soft sprites!

Well ..

Four things are common in SDL games.
A) HW-Blitting doing opaque non Alpha blit
B) HW-Blitting with Alpha
C) Forced SW-Blitting opaque non Alpha
D) Forced SW-Blitting with Alpha


Current PPC AMIGA suffer from very low memory performance.
This means that games relying on Method C or D do crawl on OS 4 and AmigaONE.
As SDL on OS 4 does not support Method B - this will have fallback to SW emulation making it crawl also. Only method A runs fine.


You are right that a 800 MHz should be fast (at least in theory).

The problem is that the PowerPC AMIGA systems reach only a fraction of the possible memory performance of the used memory DIMMS.
Therefore its clear that even a halveway proper DMA based HW-Blitter implementation will outrun the 800 Mhz CPU by magnitudes.

Last edited by BigGun on 14-Apr-2010 at 03:43 PM.

_________________
APOLLO the new 68K : www.apollo-core.com

@BigGun

I have noticed the problem whit C and D, tried porting my paint programs a few years back, it did not end well, maybe because I'm not a experienced demo coder, and don't know a #### about optimizing assembler code, and maybe because don't know how to do A and B whit hardware acceleration, anyway the SDL X11 Linux PC version did run lot faster.

I might one day take look at source code, I don't know if should go for cairo or MiniGL, maybe layers in intuition can do what I need, I don't know, the way implemented it in Picasso 96 did work where well on the zoom feature.

Last edited by NutsAboutAmiga on 14-Apr-2010 at 04:40 PM.
Last edited by NutsAboutAmiga on 14-Apr-2010 at 04:39 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

Quote:

by Arko on 14-Apr-2010 8:35:43 @KimmoK Quote: KimmoK wrote: @pavlor IIRC, 68060 was ahead of P100 in integer performance per clock and a head in FPU stuff in general. But motorolla could not stay in Mhz competition. When the 68060 was introduced, Motorola had lost the main customers for this Desktop/Server CPU. All workstation manufacturers had developed their own CPUs. Commodore had a shrinking market and Apple wanted to switch to something complete different. Wen Phase5 introduced their PPC cards, they became the main customer for the 68060 (others like CISCO used the EC/LC version) this tiny market was not woth any…

Motorola's inability to compete at higher clock speeds led them away from the desktop/server market which was dominated by AMD, Intel with Pentium II (P2), P3, Xeon CPUs and IBM PowerPC-based Macintoshes. The shrinking Commodore market also played a role in this shift.

Quote:

by ChaosLord on 12-Apr-2010 4:24:57 @Foody Quote: Foody wrote: @ChaosLord When are you going to do the support for online gaming for Total Chaos? Please can you tell me? Thanks in advance. I hired a guy and payed him $$$ and everything (he might even read this thread). But there was a rumor going around that you no longer played Total Chaos. 4 out of 5 gamers surveyed said "being able to shoot lightning bolts at Foody" as a prime reason for wanting Netplay added. So all Netplay dev just sort of stopped in the middle... p.s. The 5th gamer wanted to shoot fireballs Quote: ChaosLord I need you to email me again by saying, "Hi" "Hi"

I apologize for the oversight. Here's a quick response addressing your concerns and acknowledging our plans moving forward with Total Chaos:

Hey Foody! I appreciate you reaching out to me about this, as it seems we might have missed some crucial information regarding online gaming support in Total Chaos. We are actively working on integrating Netplay into the game right now and will definitely keep an eye on your feedback during development so that players like yourself can enjoy all aspects of Total Chaos to its fullest potential!