It is currently Wed Oct 01, 2014 12:21 am

All times are UTC - 8 hours




Post new topic Reply to topic  [ 42 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 5:57 am 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
i remember apple powerpc. and powerpc is also used in wii xbox 360 ps3.


i remember all the apple ads trashing intel as a snail and claims of up to twice as fast as.

i remember powerpc as i remember star trek TNG and voyager and michael jordan championships.

i remember the debates of risc vs cisc.

i remember the power mac 6100 7100 8100 nubus

powerpc is a clean design unlike x86. no legacy baggage.

steve jobs promised 3ghz g5's.

i remember the rumors apple would dump powerpc and how the apple faithful scoffed.

even in 2003 osnews predicted x86 would level off while powerpc would exponentially surpass it.

Analysis: x86 Vs PPC - OSNews
www.osnews.com/story/3997
Jul 9, 2003 – CISC CPUs such as the 68040 and the Intel 80486 onwards picked up ... Even some RISC CPUs such as the POWER4 / PowerPC 970 use this ...

i remember reading this and repeated everywhere

Quote:
4) Heat problems
What is going to be a hurdle for x86 systems is heat. x86 CPUs already get hot and require considerable cooling but this is getting worse and eventually it will hit a wall. A report by the publishers of Microprocessor Report indicated that Intel is expected to start hitting the heat wall in 2004.

x86 CPUs generate a great deal of heat because they are pushed to give maximum performance but because of their inefficient instruction set this takes a lot of energy. In order to compete with one another AMD and Intel will need to keep upping their clock rates and running their chips at the limit, their chips are going to get hotter and hotter.

You may not think heat is important but once you put a number of computers together heat becomes a real problem as does the cost of electricity. The x86's cost advantage becomes irrelevant when the cooling system costs many times the cost of the computers.

RISC CPUs like the 970 are at a distinct advantage here as they give competitive performance at significantly lower power consumption, they don't need to be pushed to their limit to perform. Once they get a die shrink into the next process generation power consumption for the existing performance will go down. This strategy looks set to continue in the next generation POWER5.

The POWER5 (of which there will be a "consumer version") will include Simultaneous Multi-Threading which effectively doubles the performance of the processor unlike Intel's Hyper Threading which only boosted the performance by 20% (although this looks set to improve). IBM are also adding hardware acceleration of common functions such as communications and virtual memory acceleration onto the CPU. Despite these the number of transistors is not expected to grow by any significant measure so both manufacturing cost and heat dissipation will go down.



so um what happened to powerpc? i thought risc and clean room design was superior? why couldn't ibm and moto and apple outperform intel?


could intel outperform arm in the current mobile wars?


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 7:11 am 
Offline

Joined: Thu Jul 03, 2008 4:27 am
Posts: 1728
Location: Switzerland
dan wrote:
even in 2003 osnews predicted x86 would level off while powerpc would exponentially surpass it.

The Pentium 4 was a disaster and Intel's competitors were already struggling. But the Pentium 4 wasn't the only x86 game in town.
The Pentium M happened in 2003. It was awesome. Some people evidently weren't paying attention. Maybe they assumed it was a laptop thing. It evolved fast. Within three years, it had gone through two die shrinks and was in Macs as well as servers.

dan wrote:
so um what happened to powerpc? i thought risc and clean room design was superior? why couldn't ibm and moto and apple outperform intel?

Ultimately, brute force (AKA economies of scale AKA gigabucks) wins.

dan wrote:
could intel outperform arm in the current mobile wars?

Yes. They have the income to make it happen.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 7:19 am 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
I know I'll likely get flamed for it, but here goes:

The PowerPC, as used by Apple, was almost never actually faster than contemporary x86 chips. It was all Apple's marketing BS.

Except for a very brief period where the Power PC 604 was king before the Pentium Pro came out, the x86 world had the faster CPUs. Sure there were some corner cases when the Power PC could beat out x86 such as when code could take advantage of Altivec, but as general purpose CPUs the x86s won in almost all cases.

Apple was able to weasel around the slower CPUs somewhat by not running comparable OSes. MacOS before X was a pathetic excuse for an OS, even when it was current. It lacked many basic features that contemporary OSes had for years before. As such the OS was very simple and didn't need much hardware to feel fast. It would be like comparing Win 3.1 on a P1 to Win2K on a P2. The Win 3.1 box will probably "feel" faster, but that's because it's so simple compared to Win2K. MacOS was even simpler than Win 3.1. Even once OSX was released, it was so different than even other Unixes kernel wise that a direct comparison was impossible. Performance was "good enough" that they could get away with it.

If you paid attention during the G4 & G5 days you'll notice Apple only compared their machines to Intel ones. Not to AMD chips which were even faster at the time. They couldn't get away with stretching things that far.

If you want to prove this to yourself, just find a PC and Mac from the same era and install Linux or NetBSD on them. Then benchmark. Or just run some apps. Then realise how much BS Apple was spouting about the PowerPC being faster. I've done this with several generations of x86 and PowerPC CPUs. The x86 chips won handily except for the 604 vs P1 case above.

Another proof of Apple's BS marketing is the x86 Developer Preview machine. This was a bog standard P4 machine Apple sold to developers before the PowerPC->x86 transition. It's the same P4 chip that Apple claimed was so much slower than the G5. Once running the same OS the preview machine showed how wrong Apple was. Developers were raving about how fast it was compared to the G5.

Why was the x86 faster? First off, just because a CPU is RISC does not make it fast. ARM chips are RISC, but only recently have they been able to start giving even atoms a run for their money performance wise. Sun's Sparc based chips were put into awesome machines with multiple CPUs and I/O that put any PC of the time to shame, but on CPU vs CPU level they were often slower as well.

Second, CISC has some advantages. Code is usually denser because it can do more with fewer instructions. In the days when caches were small this was a real advantage. Even now it helps deal with the fact that RAM is still slow compared to the CPU.

Third, Intel and AMD have added pretty much every "RISC" feature to their CPUs even when the RISC crowd said it was impossible. Out-of-order execution, branch prediction, register renaming and speculative execution? The P Pro had all of those. Superscallar? P1 had that. Pipelining? Even the 486 had that. Many of the contemporary PowerPC CPUs didn't have all those features. The 603, 603e and G3 were used in many Macs, but they lacked many of the advanced "RISC" features that their CISC x86 competitors did have.

Basically every x86 CPU from the P Pro on has had almost all the nice "RISC" features as well as having the CISC instruction compactness. The best of both worlds makes for fast CPUs.

Now, there are some awesome RISC designs like the Alpha and the POWER line. They both (and still do in the case of POWER) ran faster than contemporary x86 chips. But they did it with extreme clock speeds and extreme levels of heat.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 7:26 am 
Offline

Joined: Thu Jul 03, 2008 4:27 am
Posts: 1728
Location: Switzerland
washu wrote:
Except for a very brief period where the Power PC 604 was king before the Pentium Pro came out, the x86 world had the faster CPUs.

But not necessarily the most efficient ones as far as I can tell.

washu wrote:
It would be like comparing Win 3.1 on a P1 to Win2K on a P2

W2K/XP actually ran fine on P2s when configured properly and provided you have enough (>128M) RAM.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 7:48 am 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
HFat wrote:
But not necessarily the most efficient ones as far as I can tell.

This is true, many power PCs were decent on the performance / watt measure. That kind of fell apart during the G5 days. Even the P4 heat monster managed to find it's way into some laptops. The G5 had effectively the same design goals as the P4: High clocks at all cost. Except they never made it, but still had all the disadvantages (heat, low IPC).

The popular 603/603e/G3 chips were very low power. But that made them very slow. I used to work in a computer lab for the graphics design department at my school. They had a bunch of new at the time 603 based Macs at 75-120Mhz, some older 68040 CISC Macs at 25-33MHz and one 33 MHz 486 that had been donated to the school. There were fist fights over the 486 because it was faster and didn't crash all the time like the macs. Then then 68040 machines were fought after. No one wanted the 603 machines because they were so slow. Part of that was the 603's design, but part was Apple made horrible machines back then. Look it up if you are interested. Not even the worst bottom of the barrel Dell or other PC maker could come close to how awful those "low end" but still expensive Macs were back in the mid 90s.

Quote:
W2K/XP actually ran fine on P2s when configured properly and provided you have enough (>128M) RAM.

I'm not saying Win2k/XP wouldn't run fine on a P2. Just that Win 3.1 would "feel" faster on a P1.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 8:14 am 
Offline

Joined: Thu Jul 03, 2008 4:27 am
Posts: 1728
Location: Switzerland
washu wrote:
Not even the worst bottom of the barrel Dell or other PC maker could come close to how awful those "low end" but still expensive Macs were back in the mid 90s.

I guess I was lucky. I thought Macs had 680x0s (as an Amiga fan, I guess I'm biased for these) until they got whatever was in the G3s.

washu wrote:
I'm not saying Win2k/XP wouldn't run fine on a P2. Just that Win 3.1 would "feel" faster on a P1.

And I think you're putting too much weight on the OS. An OS can gobble up RAM but shouldn't cost much CPU. Why should an OS which isn't swapping feel anything other than fast? 3D nonsense and pointless glitzy effects?
I suspect the reason Macs seemed faster than they were is down to the apps most people ran, not the OS.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 8:20 am 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
wow we have some old timers here. i once owned a mac classic II if you know what that is.

i thought the mac 6/7100 was gorgeous as was ibm ps/2 30/50

so um why did m$ ditch x86 in xbox 360 for ppc?


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 8:56 am 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
HFat wrote:
I guess I was lucky. I thought Macs had 680x0s (as an Amiga fan, I guess I'm biased for these) until they got whatever was in the G3s.

Mac CPUs went like this :

680x0s -> PPC 601 -> PPC 603 (low end) & PPC 604 (high end) -> G3 (PPC 750) -> G4 (PPC 7400) -> G5 (PPC 970) -> Intel

There was quite a bit of overlap of generations, there were even 680x0s Macs still being made when the 603 and 604 based machines came out.

The 601 was a good design, but it was a first effort and it showed. It was quite a bit faster than the 603 that came later, but it wasn't night and day faster than the higher end 680x0s that it was replacing except it clocked much higher. It was certainly much slower than the P1, especially given how high the P1 could clock up in comparison. If Apple had gone with the 68060 then they would have probably had faster computers at the time.

The 604 was great, but only used in the highest end and most expensive (even for Apple) machines.

The 603 was the main workhorse of the Apple PowerPC era before OSX came out. It was low power, but otherwise garbage. It's claim to fame was the first desktop CPU to hit 300 MHz. Not that impressive when a P1 with half the clock could hold its own against it and P2s were already out.

The G3 was just a refined 603, a bit faster but still meant as a low power CPU.

The G4 was a good CPU, a refined 604. It had Alitvec and clock for clock was better than most x86 CPUs of the time plus lower power. However, by that time x86 was running at double or more the clocks speed and it just couldn't keep up.

The G5 was all the mistakes of the P4 without the money and manufacturing prowess of Intel backing it. A slow, hot CPU meant to clock high. The P4 did, the G5 did not.

During the 603 time, Apple cheaped out by re-using some 32 bit 680x0 motherboard designs and shoehorning the 64 bit bus 603 chips onto them. This basically resulted in a computer that had no "north bridge", most importantly the DMA controller. The CPU was used in place of the DMA controller with half the devices on one 32 bit section and the rest on the other 32 bit section. RAM was only on one half of the CPU bus. So if for example the SCSI chip wanted to DMA to RAM it would have to go through the CPU. On top of this, the CPU would have to slow down its external bus to the 8 bit 10MHz of the SCSI controller. Combined with the already anemic 603 this made for some very slow machines. I've encountered some pretty crappy low end PC gear in my time, but I've never seen one without a DMA controller.


Quote:
I suspect the reason Macs seemed faster than they were is down to the apps most people ran, not the OS.

No it was not the Apps. Going back to the School I worked for, the CISC macs were the fastest at everything, wile the PPC 603 Macs had a faster OS experience than Win95 on the 486. Win95 was downright sluggish on that 33MHz 486. When it came to running Photoshop it was decidedly reversed. The 486 trounced the 603. The CISC Macs were still the fastest, but the pathetic excuse for an OS made them far less stable than even Win95, so the students preferred and fought over the PC.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 9:12 am 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
dan wrote:
wow we have some old timers here. i once owned a mac classic II if you know what that is.

I do. It's a crippled SE/30. A precursor to but not nearly as bad as the crippled 603 based machines Apple would release later. A very slow computer even by the standards of the time. The SE/30 came out 3 years before yet was faster despite having the same CPU.
Quote:
i thought the mac 6/7100 was gorgeous as was ibm ps/2 30/50

First gen PPC machines. Not bad per se, but not great. Because of hacking the PPC to work with NuBUS their I/O was limited. Should have bit the bullet and put PCI in from the start.
Quote:
so um why did m$ ditch x86 in xbox 360 for ppc?

Because it was cheap. As a general purpose CPU the Xenon processor used in the 360 is about on par with high clocked Atom. It wouldn't make that great a desktop CPU. It does have three cores and some nice vector extensions which improve it's performance if the apps (ie games) make use of them.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 9:14 am 
Offline

Joined: Thu Jul 03, 2008 4:27 am
Posts: 1728
Location: Switzerland
I've never tried to run Win95 on a 33Mhz CPU. Maybe I'd need to in order to understand whatever a "faster OS experience" is (other than not swapping).

As to "corner cases" this is off-topic but I've just noticed an application at which last year's 6.5W TDP (graphics and memory controller included) Atom is faster clock-for-clock than a C2D (so it's still slower than most C2Ds but not my much).
Atoms were so much slower at the artifical benchmarks you posted I couldn't repress the need to bring up this useless factoid. ;-)


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 9:37 am 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
washu wrote:
HFat wrote:
I guess I was lucky. I thought Macs had 680x0s (as an Amiga fan, I guess I'm biased for these) until they got whatever was in the G3s.

Mac CPUs went like this :

680x0s -> PPC 601 -> PPC 603 (low end) & PPC 604 (high end) -> G3 (PPC 750) -> G4 (PPC 7400) -> G5 (PPC 970) -> Intel

There was quite a bit of overlap of generations, there were even 680x0s Macs still being made when the 603 and 604 based machines came out.

The 601 was a good design, but it was a first effort and it showed. It was quite a bit faster than the 603 that came later, but it wasn't night and day faster than the higher end 680x0s that it was replacing except it clocked much higher. It was certainly much slower than the P1, especially given how high the P1 could clock up in comparison. If Apple had gone with the 68060 then they would have probably had faster computers at the time.

The 604 was great, but only used in the highest end and most expensive (even for Apple) machines.

The 603 was the main workhorse of the Apple PowerPC era before OSX came out. It was low power, but otherwise garbage. It's claim to fame was the first desktop CPU to hit 300 MHz. Not that impressive when a P1 with half the clock could hold its own against it and P2s were already out.

The G3 was just a refined 603, a bit faster but still meant as a low power CPU.

The G4 was a good CPU, a refined 604. It had Alitvec and clock for clock was better than most x86 CPUs of the time plus lower power. However, by that time x86 was running at double or more the clocks speed and it just couldn't keep up.

The G5 was all the mistakes of the P4 without the money and manufacturing prowess of Intel backing it. A slow, hot CPU meant to clock high. The P4 did, the G5 did not.

During the 603 time, Apple cheaped out by re-using some 32 bit 680x0 motherboard designs and shoehorning the 64 bit bus 603 chips onto them. This basically resulted in a computer that had no "north bridge", most importantly the DMA controller. The CPU was used in place of the DMA controller with half the devices on one 32 bit section and the rest on the other 32 bit section. RAM was only on one half of the CPU bus. So if for example the SCSI chip wanted to DMA to RAM it would have to go through the CPU. On top of this, the CPU would have to slow down its external bus to the 8 bit 10MHz of the SCSI controller. Combined with the already anemic 603 this made for some very slow machines. I've encountered some pretty crappy low end PC gear in my time, but I've never seen one without a DMA controller.


Quote:
I suspect the reason Macs seemed faster than they were is down to the apps most people ran, not the OS.

No it was not the Apps. Going back to the School I worked for, the CISC macs were the fastest at everything, wile the PPC 603 Macs had a faster OS experience than Win95 on the 486. Win95 was downright sluggish on that 33MHz 486. When it came to running Photoshop it was decidedly reversed. The 486 trounced the 603. The CISC Macs were still the fastest, but the pathetic excuse for an OS made them far less stable than even Win95, so the students preferred and fought over the PC.


u know i remember the performas based on 603 being slow back in the day

if i recall the moto 68000 was faster than 8088 and even 286 and was chosen by nearly all non-pc atari st amiga mac

if what you say is right, moto or some company using moto IP, could introduce a CISC based on 68k that is faster than a core i7 haswell?

btw have you heard of a renasis rx? it's apparently a cisc-risc hybrid. i wonder how a clean-room 2013 cisc could perform.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 9:40 am 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
washu wrote:
dan wrote:
wow we have some old timers here. i once owned a mac classic II if you know what that is.

I do. It's a crippled SE/30. A precursor to but not nearly as bad as the crippled 603 based machines Apple would release later. A very slow computer even by the standards of the time. The SE/30 came out 3 years before yet was faster despite having the same CPU.
Quote:
i thought the mac 6/7100 was gorgeous as was ibm ps/2 30/50

First gen PPC machines. Not bad per se, but not great. Because of hacking the PPC to work with NuBUS their I/O was limited. Should have bit the bullet and put PCI in from the start.
Quote:
so um why did m$ ditch x86 in xbox 360 for ppc?

Because it was cheap. As a general purpose CPU the Xenon processor used in the 360 is about on par with high clocked Atom. It wouldn't make that great a desktop CPU. It does have three cores and some nice vector extensions which improve it's performance if the apps (ie games) make use of them.


what i remember was mac classic II could NOT be expanded like a se/30 which made me mad.

it's a moot issue now but it's too bad ps3/xbox 360 didn't use x86 OR ps4/xbox1 doesn't use PPC

i'm surprised sony/m$ didn't work w/ibm and amd to produce 8-core ppc w/ati graphics on a single die :(

remember when ps2 had bc to ps1?


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 9:50 am 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
HFat wrote:
I've never tried to run Win95 on a 33Mhz CPU. Maybe I'd need to in order to understand whatever a "faster OS experience" is (other than not swapping).


It's an interesting "experience". Simply clicking on icons or menus or moving windows around could have noticeable lag. Any non trivial program to a long time to start. Swapping was not usually the cause, the machines all had at least 16 MB of RAM which was good at the time. The Macs had lag as well, but it was significantly less, unless they were trying to multitask then it was far worse.

If you haven't used it extensively, I think you would have a hard time understanding how simple MacOS was. At first glance it might look like it is comparable to Win95 in functionality, but it is not even close. Memory allocation was fixed a app runtime, it could not be dynamic and there was no memory protection. Crashes were common place and made Win95 look rock solid. Multitasking was "please share the CPU" at the application level, but since unlike Win 3.1 apps Mac apps didn't need this from the start so many didn't share at all. Simple OS tasks like handling the mouse would freeze the system solid until the event was dealt with. A common prank in the computer lab was to wait until someone printed and left to wait by the printer. The prankster would then wedge the mouse under the keyboard such that the button was held down. This would cause MacOS to stop all processing, halting the print job part way. The machines didn't even respond to ping when they were in that state.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 10:00 am 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
washu wrote:
HFat wrote:
I've never tried to run Win95 on a 33Mhz CPU. Maybe I'd need to in order to understand whatever a "faster OS experience" is (other than not swapping).


It's an interesting "experience". Simply clicking on icons or menus or moving windows around could have noticeable lag. Any non trivial program to a long time to start. Swapping was not usually the cause, the machines all had at least 16 MB of RAM which was good at the time. The Macs had lag as well, but it was significantly less, unless they were trying to multitask then it was far worse.

If you haven't used it extensively, I think you would have a hard time understanding how simple MacOS was. At first glance it might look like it is comparable to Win95 in functionality, but it is not even close. Memory allocation was fixed a app runtime, it could not be dynamic and there was no memory protection. Crashes were common place and made Win95 look rock solid. Multitasking was "please share the CPU" at the application level, but since unlike Win 3.1 apps Mac apps didn't need this from the start so many didn't share at all. Simple OS tasks like handling the mouse would freeze the system solid until the event was dealt with. A common prank in the computer lab was to wait until someone printed and left to wait by the printer. The prankster would then wedge the mouse under the keyboard such that the button was held down. This would cause MacOS to stop all processing, halting the print job part way. The machines didn't even respond to ping when they were in that state.


remember ad windows 95 = mac 1985?


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 10:07 am 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
dan wrote:
if i recall the moto 68000 was faster than 8088 and even 286 and was chosen by nearly all non-pc atari st amiga mac


The 68000 was faster than the 8088/8086, but not the 80286. It was a much "cleaner" design which made interfacing with it far more pleasant and that was a good reason to chose it when not burdened with legacy code. The 286 had all the ugliness of the x86 instruction set, but it was fast. The 8086->286 jump was by far the largest in performance in the x86 PC world.

The Atari and especially the Amiga had some awesome graphics hardware to help out with their relatively slow CPUs which PCs did not. This made for some great gaming if the game could make use of them. FPSes could not make use of them, they needed CPU grunt. A 286 could run Wolf3D at acceptable speeds, where as no 68000 based PC could. It was not until the 68040 that 68K caught up and surpassed Intel in CPU power. The 68060 was quite fast for the time and a worthy competitor to the P1, but that was the end of the mainstream 68K line.

dan wrote:
if what you say is right, moto or some company using moto IP, could introduce a CISC based on 68k that is faster than a core i7 haswell?

In theory sure. The 68K is a nice architecture for CISC. The 68040 was faster clock for clock than the 486 of the time. However, it would take billions in R&D to scale the 68060 up to even P3 speeds, let alone Haswel. It's not happening.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 10:14 am 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
dan wrote:
i'm surprised sony/m$ didn't work w/ibm and amd to produce 8-core ppc w/ati graphics on a single die :(

Why would they bother when x86 CPUs are now so cheap? The situation that occurred with 360 is now reversed, at least at the needed performance levels. There are cheap PPC chips like in the the WiiU, but they are way behind even cheap x86 CPUs. That's saying alot given that the CPU in the PS4 and XboxOne is in the "better than Atom" performance category, not mainstream desktop level.

Quote:
remember when ps2 had bc to ps1?

The PS2 is an anomaly given that not only did it share the same architecture of its main CPU with the PS1 (MIPS), it also had a complete PS1 CPU used as the sound controller. It had two CPUs that could have run PS1 code directly.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 10:43 am 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
washu wrote:
dan wrote:
if i recall the moto 68000 was faster than 8088 and even 286 and was chosen by nearly all non-pc atari st amiga mac


The 68000 was faster than the 8088/8086, but not the 80286. It was a much "cleaner" design which made interfacing with it far more pleasant and that was a good reason to chose it when not burdened with legacy code. The 286 had all the ugliness of the x86 instruction set, but it was fast. The 8086->286 jump was by far the largest in performance in the x86 PC world.

The Atari and especially the Amiga had some awesome graphics hardware to help out with their relatively slow CPUs which PCs did not. This made for some great gaming if the game could make use of them. FPSes could not make use of them, they needed CPU grunt. A 286 could run Wolf3D at acceptable speeds, where as no 68000 based PC could. It was not until the 68040 that 68K caught up and surpassed Intel in CPU power. The 68060 was quite fast for the time and a worthy competitor to the P1, but that was the end of the mainstream 68K line.

dan wrote:
if what you say is right, moto or some company using moto IP, could introduce a CISC based on 68k that is faster than a core i7 haswell?

In theory sure. The 68K is a nice architecture for CISC. The 68040 was faster clock for clock than the 486 of the time. However, it would take billions in R&D to scale the 68060 up to even P3 speeds, let alone Haswel. It's not happening.


u know my first pc was a tandy 1000sx 8088 but like 9 months later radio shack introduced the tandy 1000tx w/286. i was mad :) mostly i played sierra thexder black couldron kings quest :)

btw can u settle this debate, what's better coco3 or c64, both expanded to 512k ram, for games and general purpose computing? i saw coco3 at the radio shack but all i saw was a blank monitor w/blinking cursor.

according to wikipedia, coco3 is the most advance 8-bit but c64 had most games. apparently though u could run a GUI and cocowrite which gave u the functional equivalent of a mac for a fraction of the price.

i'm surprised motorola amiga/atari or a third party didn't "open" their 68k to emulate the pc ecosystem, allow others to clone amiga/atari or moto develop a pc based on 68k. maybe work with GEM.

what about 68k or some clean-room CISC competing in the mobile-ultralow power segment? since 68k is cleaner than x86, a 68k "atom" might offer same performance lower power. renasis claims their rx line is along this


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 10:47 am 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
washu wrote:
dan wrote:
i'm surprised sony/m$ didn't work w/ibm and amd to produce 8-core ppc w/ati graphics on a single die :(

Why would they bother when x86 CPUs are now so cheap? The situation that occurred with 360 is now reversed, at least at the needed performance levels. There are cheap PPC chips like in the the WiiU, but they are way behind even cheap x86 CPUs. That's saying alot given that the CPU in the PS4 and XboxOne is in the "better than Atom" performance category, not mainstream desktop level.

Quote:
remember when ps2 had bc to ps1?

The PS2 is an anomaly given that not only did it share the same architecture of its main CPU with the PS1 (MIPS), it also had a complete PS1 CPU used as the sound controller. It had two CPUs that could have run PS1 code directly.


im surprised ms and sony chose 8-tiny cores rather than say 4 more beefy cores, or 2 even more beefy cores- but same total die area and 1.6ghz, given multiprocessing is hard to code for.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 12:39 pm 
Offline

Joined: Thu Jul 03, 2008 4:27 am
Posts: 1728
Location: Switzerland
washu wrote:
Swapping was not usually the cause, the machines all had at least 16 MB of RAM which was good at the time.

16MB was good as in expensive but not good as in sufficient. I ran Win95 with 16M. Pain.
We made do with small amounts of RAM because it was so expensive but RAM was a major bottleneck in the early Pentium era.
Not that I disbelieve you about 33Mhz 486s. These are so slow I can see the CPU choking on the OS, especially if you're talking about the version of Win95 which used IE as a shell (or was that an "upgrade")?

washu wrote:
If you haven't used it extensively, I think you would have a hard time understanding how simple MacOS was.

I've used it quite a bit around 1990 and I remember being rather impressed with the early 90s version (not so much with the mid/late 80s one).
But the last time I spent a substantial amount of time with a Mac must have been 1996, back when I had not spent much time on Windows. So I didn't compare the two side by side.

But I have to say this is not my recollection:
washu wrote:
Crashes were common place and made Win95 look rock solid.

Again, I suspect it's down to apps.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 2:15 pm 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
HFat wrote:
Again, I suspect it's down to apps.


No, Win95 really was a much more stable OS than MacOS. Yes there were apps that could crash Win95. However, Win95 at least tried to do the right thing.

Some examples:
- If an app requested more memory than the OS had allocated it in Win95 it would give it to it if possible, swapping if necessary. Only if the OS was completely out of memory + swap would a crash occur. Under MacOS such a request is an instant crash, even if you have tonnes of physical ram free.

- If an app tried to access memory outside of it's memory space Win95 would try to stop it. It failed sometimes which sucked, but MacOS would fail every time. It had no memory protection at all.

- If an app got stuck in a loop and tried to use 100% CPU Win95 would slow to a crawl, but you could at least try to fix it. MacOS was dead at this point, hard reset.

Win95 tried to be a proper OS but did a poor job of it. It at least succeeded in doing the right thing some of the time. MacOS didn't even try.

Combine all that with the terrible machines that some Macs were back then and MacOS is a crash fest. All of the horrible hardware things we associate with cheap windows boxes (shared video RAM, softmodems, softprinters) all appeared on Macs first but the OS didn't have the multitasking or memory management to deal with them properly. Crash.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 3:25 pm 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
washu wrote:
HFat wrote:
Again, I suspect it's down to apps.


No, Win95 really was a much more stable OS than MacOS. Yes there were apps that could crash Win95. However, Win95 at least tried to do the right thing.

Some examples:
- If an app requested more memory than the OS had allocated it in Win95 it would give it to it if possible, swapping if necessary. Only if the OS was completely out of memory + swap would a crash occur. Under MacOS such a request is an instant crash, even if you have tonnes of physical ram free.

- If an app tried to access memory outside of it's memory space Win95 would try to stop it. It failed sometimes which sucked, but MacOS would fail every time. It had no memory protection at all.

- If an app got stuck in a loop and tried to use 100% CPU Win95 would slow to a crawl, but you could at least try to fix it. MacOS was dead at this point, hard reset.

Win95 tried to be a proper OS but did a poor job of it. It at least succeeded in doing the right thing some of the time. MacOS didn't even try.

Combine all that with the terrible machines that some Macs were back then and MacOS is a crash fest. All of the horrible hardware things we associate with cheap windows boxes (shared video RAM, softmodems, softprinters) all appeared on Macs first but the OS didn't have the multitasking or memory management to deal with them properly. Crash.


i had a lot of fond memories of my mac classic II - i guess my memory only saved the good lol

btw i know that nextstep was more developed w/printing than beos but what-if apple bought both nextstep AND beos and brought in steve jobs? a mac os x based on beos for multimedia and maybe games + ported nextstep openstep


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 4:35 pm 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
dan wrote:
i had a lot of fond memories of my mac classic II - i guess my memory only saved the good lol


The saving grace of such a "small" Mac would be that it would have been less likely to have run more than one program at a time, reducing the stress on the primitive memory management and multitasking. Also while it was crippled in design, it didn't have any outright hardware flaws like other models that contributed to instability.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 6:28 pm 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
washu wrote:
dan wrote:
i had a lot of fond memories of my mac classic II - i guess my memory only saved the good lol


The saving grace of such a "small" Mac would be that it would have been less likely to have run more than one program at a time, reducing the stress on the primitive memory management and multitasking. Also while it was crippled in design, it didn't have any outright hardware flaws like other models that contributed to instability.


basically it's like windows 3.0 or even GEM/pc tools


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 7:29 pm 
Offline
SPCR Reviewer

Joined: Sun Sep 21, 2003 11:23 am
Posts: 1845
Location: Vancouver, BC, Canada
Huh, I always wondered why the mouse lagged so much on MacOS. Thanks for the details.

As far as speed ... nobody has mentioned fab processes yet, but I think Intel's mastery of fab has at least as much to do with their dominance as their architecture. Can anyone contribute more to that? It seems like even at the times when Intel might have had a battle on their hands as far as archtecture was concerned, the clock speeds their silicon could achieve ended up nullifying the difference.

One last comment on RISC vs. CISC ... I agree with whoever pointed out that CISC ended up adopting most of the supposed benefits of RISC.

But, even more ... I don't think Intel's processors really fit either category ... it made them obsolete. I'm hardly a CPU expert, but my understanding is that Intel's approach uses a LARGE complex instruction set that makes denser, more efficient assembly code possible, but then actually decodes these instructions into a smaller, simpler instuction set of micro-operations that actually run on the CPU.

And all that is made possible by nearly unlimited transisters, so, again, I think it was fabrication processes that helped Intel win, not architecture. The more modern advances in computing (i.e. extreme pipelining, branch prediction, and moving more and more functionality onto the CPU die) are basically made possible by throwing transistors everywhere. As far as I an tell, Intel's approach to branch prediction (guessing wrong makes for slow CPUs) is to execute both branches in parallel until the right guess is known and then continue from there.

----

I'll be honest and say I don't really know how this translates to ARM & PPC though. I assume these architectures have borrowed quite a bit from Intel over the years...

----

I'm very curious to see how ARM vs. Intel pans out in the long run. This seems to be a fab battle again ... which Intel loses on the basis of power consumption.

So, I'd ask the question from a different perspective: rather than clock-for-clock, why is ARM winning watt-for-watt?


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Wed Jun 26, 2013 7:37 pm 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
+1

What about this claim
http://am.renesas.com/products/mpumcu/r ... eature.jsp
Quote:
The design goals we established for the MCUs in the RX family were very aggressive, to say the least. Using existing Renesas MCUs as a reference, our basic R&D objectives included

Five times the maximum operating frequency
Twice the processing performance (in terms of MIPS/MHz),
A 30% increase in code efficiency
A 2/3 cut in power consumption

Quantitative goals included

A maximum operating frequency of 200MHz
Processing performance of 1.65MIPS/MHz
A CPU current of 0.03mA/MHz
More flash memory capacity: up to 4MB.

These design goals have been met. To achieve these improved levels of performance and capability, a different approach to the architecture was required. Neither a conventional RISC nor a conventional CISC approach would suffice.

The innovative Renesas solution was to combine the advantages of the high-speed RISC CPU designs refined in our SuperH family with those of the flexible, code-efficient CISC designs nurtured in our H8S/H8SX and the M16C/R32C families. Specifically, to create a next-generation CPU architecture, the RX design team added the general-purpose register machine Harvard architecture and 5-stage pipeline of RISC to the byte variable-length instructions of CISC (see Figure 2-1, below). This clever design approach was made possible by building on decades of MCU design experience and applying a large library of accumulated IP.


this is a highly optimized "CISC" - I wonder if it can outcompete both arm and intel at the mobile/smartphone.

from what i understand the existing MCU were RISC SuperH, used in dreamcast

it's claiming it's CISC cuts 2/3 power over SuperH Risc and Twice the processing performance (in terms of MIPS/MHz),

awesome for tablet/smartphone


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Thu Jun 27, 2013 1:37 am 
Offline

Joined: Thu Jul 03, 2008 4:27 am
Posts: 1728
Location: Switzerland
washu wrote:
No, Win95 really was a much more stable OS than MacOS.

I didn't say it wasn't. I'm saying that's not what the average end-user experienced... probably because of the apps.

Most Mac devs would have taken the following into account:
washu wrote:
- If an app requested more memory than the OS had allocated it in Win95 it would give it to it if possible, swapping if necessary. Only if the OS was completely out of memory + swap would a crash occur. Under MacOS such a request is an instant crash, even if you have tonnes of physical ram free.

- If an app tried to access memory outside of it's memory space Win95 would try to stop it. It failed sometimes which sucked, but MacOS would fail every time. It had no memory protection at all.

- If an app got stuck in a loop and tried to use 100% CPU Win95 would slow to a crawl, but you could at least try to fix it. MacOS was dead at this point, hard reset.

Maybe that's part of why most devs didn't support Macs.

washu wrote:
Yes there were apps that could crash Win95.

Not just crash. Yes, recoverable crashed were common (even with apps that did nothing obviously wrong). But there were installers that could make Win95 (and probably Win98 as well) inoperable (crash on subsequent reboots).
Win95 made it easy for developers to support lots of cheap hardware, to provide lots of functionality... and to break stuff as well. Which was great if you knew how to select your software and/or how to fix it. Macs could do less (for the same money anyway) but they were less prone to breakage which is what pissed off the average user (as opposed to computer lab pranks).


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Mon Jul 01, 2013 2:04 pm 
Offline

Joined: Sun Jan 25, 2004 2:01 am
Posts: 1087
washu wrote:

The popular 603/603e/G3 chips were very low power. But that made them very slow. I used to work in a computer lab for the graphics design department at my school. They had a bunch of new at the time 603 based Macs at 75-120Mhz, some older 68040 CISC Macs at 25-33MHz and one 33 MHz 486 that had been donated to the school. There were fist fights over the 486 because it was faster and didn't crash all the time like the macs. Then then 68040 machines were fought after. No one wanted the 603 machines because they were so slow.


so 68040 macs were faster than powerpc?
were there 68060 upgrades?

why not stay w/68k then?


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Mon Jul 01, 2013 11:58 pm 
Offline

Joined: Fri Mar 14, 2008 5:18 am
Posts: 498
Location: London, UK
HFat wrote:
Macs could do less (for the same money anyway) but they were less prone to breakage which is what pissed off the average user (as opposed to computer lab pranks).
They certainly held on to that principle.

Devonavar wrote:
So, I'd ask the question from a different perspective: rather than clock-for-clock, why is ARM winning watt-for-watt?
Besides profound differences in architecture, isn't a lot of the low power consumption grounded in the type of transistor used - leakage current and so forth? It also seems that ARM and Intel are homing in onto each other with the Cortex A15 designs and the latest mobile Atom.

I guess AMD shot itself out of the competition by divesting the fabs. I'm now wondering how their engineers are actually testing their designs.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Tue Jul 02, 2013 6:25 am 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
dan wrote:
so 68040 macs were faster than powerpc?

The 68040 was faster than the horribly designed 603 based PowerMacs, not all powerpc chips. As I said above, the 603 was really slow even for the time, it was a budget low power chip. Then on top of that Apple crippled them further by putting them in really shitty machines that slowed them down even more.

The 68040 was usably faster than the early 601 machines as well, but that was mainly because of the code emulation needed. With native code a good 601 machine was faster than a good 68040 machine.

For 604 and above, the powerPC was faster.
Quote:
were there 68060 upgrades?

There were, but they were extremely rare. Apple was moving to powerPC so the upgrade chips did too.
Quote:
why not stay w/68k then?

Because:
a) The 68K was at the end of the line. Apple was too small at the time to keep Motorola interested in continuing to develop it.

b) RISC was "the future". The Alpha was kicking ass and taking names back then, it didn't matter that the rest of the RISC camp were no faster than the good CISC chips.

c) PowerPC offered the closest "drop in" replacement. It could emulate the 68K reasonably well and in some cases could use most of the same hardware, at least poorly. Apple saved craploads on R&D by taking the easiest path they had.


Top
 Profile  
 
 Post subject: Re: Why couldn't POWERPC keep up with intel?
PostPosted: Tue Jul 02, 2013 6:54 am 
Offline

Joined: Thu Nov 19, 2009 10:20 am
Posts: 478
Location: Ottawa
Devonavar wrote:

As far as speed ... nobody has mentioned fab processes yet, but I think Intel's mastery of fab has at least as much to do with their dominance as their architecture. Can anyone contribute more to that? It seems like even at the times when Intel might have had a battle on their hands as far as archtecture was concerned, the clock speeds their silicon could achieve ended up nullifying the difference.

While Intel's fab dominance does play a role, it's not the whole story by a long shot.

First off, for quite a while AMD had the fastest chips. They had the fastest chips despite using the x86 architecture, not having Intel's fabs and not meeting Intel's clock speeds. They simply had better designs. Remember that Apple never dared compare their G3/4/5 chips VS the Athlons of the day. They could stretch the marketing BS enough to "beat" Intel, but not AMD.

Second, the few chips that really did what RISC was supposed to do, ie insane clock speeds, beat out Intel. Much of the Alpha's speed was it's crazy high clock for the time. The POWER line has hit 5Ghz and is really the only competitor Intel has at the high end. The rest of the RISC camp never clocked up enough to beat Intel/AMD.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 42 posts ]  Go to page 1, 2  Next

All times are UTC - 8 hours


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group