Power 7 CPU - 8 cores

[vox]

Yet again Vox shows the true depths of his ignorance.

What ignorance? That X1000 has sold its small offerings in despite of having the price in range of fake Amiga mini?

[gazgod]

What ignorance? That X1000 has sold its small offerings in despite of having the price in range of fake Amiga mini?

And if you believe that X1000 are running a nuclear power station its time to take your meds.

Power 7 is generations more advanced than the CPU in the X1000, hell my power 5 box would run rings round it.

[zylesea]

The 68k is old and it was heavily influenced by the PDP-11/VAX-11 which would be ancient and quite an innovative design back then. I think it would be difficult to make fast on modern hardware but I think the 68k could be modernized and run well enough. A modern 68k would be the easiest to use and have the best code density of any "modernized" CPU (x86, ARM and PPC are old designs too). I think it could compete with ARM for small electrical devices. What do you think of this modernized 68k ISA:
 
OpenOffice Writer
http://www.heywheel.com/matthey/Amiga/68kF_PRM.odt

PDF
http://www.heywheel.com/matthey/Amiga/68kF_PRM.pdf
   
html
http://www.heywheel.com/matthey/Amiga/68kF_PRM.html

I know there is no multiprocessing or caching instructions at this point but they are more dependent on the implementation. What do you like and dislike? Any love for the 68k besides me?

http://www.freescale.com/webapp/sps/site/homepage.jsp?code=PC68KCF

[Iggy]

The 68k is old and it was heavily influenced by the PDP-11/VAX-11 which would be ancient and quite an innovative design back then. I think it would be difficult to make fast on modern hardware but I think the 68k could be modernized and run well enough. A modern 68k would be the easiest to use and have the best code density of any "modernized" CPU (x86, ARM and PPC are old designs too). I think it could compete with ARM for small electrical devices. What do you think of this modernized 68k ISA:
 
OpenOffice Writer
http://www.heywheel.com/matthey/Amiga/68kF_PRM.odt
 
PDF
http://www.heywheel.com/matthey/Amiga/68kF_PRM.pdf
 
html
http://www.heywheel.com/matthey/Amiga/68kF_PRM.html
 
I know there is no multiprocessing or caching instructions at this point but they are more dependent on the implementation. What do you like and dislike? Any love for the 68k besides me?

Well Matt,
Unless there was a lot of revision (and  I love the 68K and used to sell 68K based hardware), no contest.
An in order, cacheless processor vs a modern CPU?
We'd lose.

[Iggy]

http://www.freescale.com/webapp/sps/site/homepage.jsp?code=PC68KCF

Coldfire is cool Zylesea, but they won't sell you a V5 and they never produced the V6.

[matthey]

@zylesea
I'm aware of the ColdFire. Freescale weakened and stripped the 68k so much that they ruined it and lost most of the 68k supporters and programmers. They did finally add some useful instructions back in but what a feeble effort. I guess they were successful in their marketing attempt to drive 68k people to the PPC (or ARM and x86 more likely). I was talking about a robust powerful 68k ISA. The 68kF ISA I showed would increase the code density considerably and remove many short branches while allowing for nearly 100% backward compatibility of the 68k family. The CF accomplished none of these.

Well Matt,
Unless there was a lot of revision (and  I love the 68K and used to sell 68K based hardware), no contest.
An in order, cacheless processor vs a modern CPU?
We'd lose.

I think full OoOE is wasteful, especially for an electricity conserving device. Doing a few instructions like division out of order can make sense though. The 68060 is a proven excellent processor and could be scaled up with today's tech, modernized and enhanced. The 68kF has a lot of what I'm talking about. I need to finish the addressing modes and add floating point. The 68k has some great advantages that Motorola/Freescale never explored and instead through away.

[commodorejohn]

I'd love to see an updated 68k, myself - it's a terrifically friendly architecture. That's one of the biggest reasons I hope Natami does come to fruition - the massive chipset improvements are neat, but I really would like to see the 68k core in action.

[Iggy]

I will agree with you about code density Matt.
 
But then I come from a 6809 background.
 
That processor was designed with position independant, reentrant code in mind.
 
I can still write code for that that takes up a small fraction of what a "modern" CPU would use.

[matthey]

@Iggy
Almost every choice made with the 68kF improves code density. I would expect 5%-15% better code density than 68020 or ColdFire code. The ColdFire has some code density improvements (MVS/MVZ, MOV3Q, BYTEREV, etc.) but they are offset by what they took away (Byte and Word instruction sizes, addressing modes, bitfield instructions, etc.)

[Iggy]

The only problem with the arguement about the importance of code density is memory is cheap.

[jj]

Just use I7s , much better bang for buck.

[Iggy]

Just use I7s , much better bang for buck.

Nah, Phenom II X4s (while they're still available).
I picked up a 3.2 GHz 955 for <$80.
Runs at 3.4-3.5.
 
X86 is cheap.

[matthey]

The only problem with the argument about the importance of code density is memory is cheap.

Code density matters for small electrical devices especially with batteries. That is why I talked about competing with ARM and not x86 on the desktop. I'm thinking of laptops, pads, netbooks, smart phones, embedded devices, fanless desktops where ARM leaves something to be desired and x86_64 is like taking a MAC truck to the grocery store. Better code density also means more instructions in the instruction cache and a smaller instruction fetch is needed. Less memory usage is still a small advantage in general, more so on low end electrical devices.

Just use I7s , much better bang for buck.

Can you show me how to program utilizing all the cores? I have a Windows component bug with the file sharing (which is used) at work that crashes. Can you disassemble the component if I send it to you and fix it for me? While you are at it, can you fix our GHz computers from stopping and being unresponsive for several seconds? I suppose I can upgrade to the i7 and it will probably be fast until Windows 9 comes out (the fix for the Windows 8 every other generation mistake). We still use Windows XP so we shouldn't need an i7 CPU but Windows slows down the more and more we use it. Can you fix that too? Intel can always quadruple the number of cores, move to 128 bits and add a few terrabytes of memory so we can finally have CPU Nirvana since we couldn't get there by GHz alone. I would rather have an efficient flexible CPU that I can program than a high latency, resource hogging DSP monster of a CPU in the same way as I would rather drive a nimble little sports car with 300HP than a fire breathing MAC truck with 3000HP. Unfortunately, most people choose the MAC truck because it's faster in a straight line for a few seconds (in a straight line on an open road) and 3000HP is a much bigger number than 300HP. There aren't very many true CPU connoisseurs any more just like sports car connoisseurs are a dying breed too .

[KimmoK]

Power 7 is generations more advanced than the CPU in the X1000, hell my power 5 box would run rings round it.

If I'm not mistaken, PA6T is a mobile PPC970 ... and IIRC, PPC970 is a Power6 stripped and modified for desktop?

So it is not generations behind, even if it runs circles around (PA6T might be better in performance per watt).



Other than that...
Nice to see that Freescale is finally doing something with the PPC. But as long as our niche OSs use only one core and one thread, we would get some 4000MIPS from a 130 000MIPS T4 chip. (would be like using dos in i7 system)

[AJCopland]

Can you show me how to program utilizing all the cores? I have a Windows component bug with the file sharing (which is used) at work that crashes. Can you disassemble the component if I send it to you and fix it for me? While you are at it, can you fix our GHz computers from stopping and being unresponsive for several seconds? I suppose I can upgrade to the i7 and it will probably be fast until Windows 9 comes out (the fix for the Windows 8 every other generation mistake). We still use Windows XP so we shouldn't need an i7 CPU but Windows slows down the more and more we use it. Can you fix that too?

None of that is down to the CPU. You'd encounter the same issues running a super-68k CPU because it's software. When you don't have to deal with a modern OS and the thousands of processes it has to manage then you can get plenty of performance out of ANY of these architectures.

Bog it down then it will go slow.

Honestly everyone goes on about AmigaOS vs Windows 7/8 but frankly AOS does absolutely nothing in comparison to it. If, and it's obviously a hypothetical if, AmigaOS development had continued in parity with Windows and x86 over the years then we'd all be whining about the same things. It's not some wonder-CPU-architecture that made AOS usable, it was simply because it was extremely primitive compared to modern operating systems.

[KimmoK]

Any CPU engineer will tell you instantly we have hit the wall and all this multi-core desktop CPU stuff is just a scam really. The maximum number of cores without losing efficiency of code execution is effectively 3. Not 4 not 8 but 3. You can not utilise much more than this without actually starting to waste cycles of CPU time delaying/setting up use of threads to run on other cores. ....For desktop computers 3 or 4 is about it. So Moore's law is f**ked well and trully unless we start seeing 5 and 6ghz CPUs QUICKLY!

It depends on desktop's use how it can share load. Some applications can split their doings to hundreds of small items to be processed parallelly, while some other taks run fastest in one pipe. Some jobs can even be split to SIMD and shader units of the system.

Easy to share/split are: renderings, video encoding, compiling, etc... Even I have used ten CPUs clusters that work very well indeed.

Latest TCOM chips have 32 cores, perhaps more. And on workstations you see things like multiple i7 chips on one motherboard. They are there because they work.


And finally... latest overclocked CPUs run faster than 9Ghz. Even if you can buy only 5Ghz parts from (IBM) shop.