Click Here
home features news forums classifieds faqs links search
6071 members 
Amiga Q&A /  Free for All /  Emulation /  Gaming / (Latest Posts)
Login

Nickname

Password

Lost Password?

Don't have an account yet?
Register now!

Support Amigaworld.net
Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
Donate

Menu
Main sections
» Home
» Features
» News
» Forums
» Classifieds
» Links
» Downloads
Extras
» OS4 Zone
» IRC Network
» AmigaWorld Radio
» Newsfeed
» Top Members
» Amiga Dealers
Information
» About Us
» FAQs
» Advertise
» Polls
» Terms of Service
» Search

IRC Channel
Server: irc.amigaworld.net
Ports: 1024,5555, 6665-6669
SSL port: 6697
Channel: #Amigaworld
Channel Policy and Guidelines

Who's Online
24 crawler(s) on-line.
 159 guest(s) on-line.
 0 member(s) on-line.



You are an anonymous user.
Register Now!
 miggymac:  24 mins ago
 Gunnar:  53 mins ago
 pixie:  2 hrs 13 mins ago
 DiscreetFX:  2 hrs 53 mins ago
 DWolfman:  3 hrs 2 mins ago
 cncparts:  4 hrs 35 mins ago
 saipaman4366:  5 hrs 21 mins ago
 Beajar:  5 hrs 40 mins ago
 Rob:  5 hrs 43 mins ago
 agami:  6 hrs 46 mins ago

/  Forum Index
   /  Amiga OS4 Hardware
      /  PowerPC notebook - Status update
Register To Post

Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 Next Page )
PosterThread
tlosm 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 11:19:23
#81 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2746
From: Amiga land

@WolfToTheMoon

I continue use My Quad G5 and had a Dual P4 3.0 ghz .
Belive Me the quad perform everything better than the Dual P4 and on the dual P4 i had a Radeon 7750HD 2gb ddr3 and 4 gb of ram and ssd

If you dont belive me i hope you will belive in this


http://www.primatelabs.com/blog/2008/06/mac-performance-june-2008/

and let me know if a P4 can do the same please watch all this video

https://www.youtube.com/watch?v=so35SrThHUs

Last edited by tlosm on 12-Jun-2015 at 11:20 AM.
Last edited by tlosm on 12-Jun-2015 at 11:20 AM.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32;
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
#nomorea-eoninmyhome

 Status: Offline
Profile     Report this post  
WolfToTheMoon 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 11:31:45
#82 ]
Super Member
Joined: 2-Sep-2010
Posts: 1351
From: CRO

@tlosm

You believe what you want, but benchmarks show that G5 is not "much faster" than an P4.
At primatelabs, a water cooled dual core G5 2.7 GHz is slower than a Core Duo mobile in a laptop. 'nough said.
And FYI, both G5 and P4 were terrible CPUs compared to Athlon64.

_________________

 Status: Offline
Profile     Report this post  
olegil 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 12:06:05
#83 ]
Elite Member
Joined: 22-Aug-2003
Posts: 5895
From: Work

@WolfToTheMoon

Not so strange considering it's a whole different beast. A CD is MUUUUUCH faster than a P4.
http://www.cpubenchmark.net/compare.php?cmp[]=734&cmp[]=1077

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

 Status: Offline
Profile     Report this post  
olegil 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 12:56:32
#84 ]
Elite Member
Joined: 22-Aug-2003
Posts: 5895
From: Work

@olegil

Since I wanted to check this coremark thing out for myself, I downloaded it and tested my laptop. T9900 (Core 2 Duo 3068MHz). Score 18100 (single thread 9600), QorIQ P2041@1500MHz gets a score of 18600. That's actually pretty good. Since a P3041 is the same core at the same frequency but with more PCIe, this tells us a WHOLE lot about what to expect from the LOWEST performer of the upcoming AmigaOne range.

Unfortunately this really really means we need some form of Multi Processing fairly soon.

Edit: Since the best i7 in the coremark statistics gets a single core (two threads) rating of 25k and a P2041 would get a single core (one thread) rating around 5k, the fact that the i7 is clocked more than twice as fast is also fairly interesting. Two times the clock and twice the threads gives 5 times the performance on something said to be VASTLY better. Right. Because 5 is SO EXTREMELY much better than 4.

I honestly think the e6500 is gonna be good enough, especially if we can exploit the extra cores and threads.

Edit2: I just remembered, I'm running two sticks of 800MHz DDR2 in my laptop, as I found this to be _slightly_ faster in memtest than one stick of 1033MHz DDR2. A P2041 would likely be running one stick of 1333MHz DDR3. So the laptop should be ahead also here, but it isn't.

Last edited by olegil on 12-Jun-2015 at 02:25 PM.
Last edited by olegil on 12-Jun-2015 at 01:01 PM.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

 Status: Offline
Profile     Report this post  
pavlor 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 14:44:10
#85 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9578
From: Unknown

@olegil

I don´t think Coremark numbers are directly comparable between different CPU architectures (we had discussion about this topic...).

We now know SpecInt2006 result (base) of T4240:

6.86 at 1666 MHz


Nearly exactly as I expected.

 Status: Offline
Profile     Report this post  
tlosm 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 14:52:40
#86 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2746
From: Amiga land

@WolfToTheMoon

CD is faster then G5 Quad 2005 ?


iMac (Mid 2007)
Intel Core 2 Duo T7700 2.4 GHz (2 cores) 3142

Power Mac G5 (Late 2005)
PowerPC G5 2.5 GHz (4 cores) 3290


Compare with my (late 2005) 2006 PowerMac G5 Quad

32 bit
http://i1249.photobucket.com/albums/hh511/tlosm/BenchmarkG5/geek1_zps15ed5bcd.jpg

64 bit
http://s1249.photobucket.com/user/tlosm/media/BenchmarkG5/geek2_zpsc20d2812.jpg.html?sort=3&o=7


Usually i belive in what i have not in what the other say :P :P :P :P

Here all my Bench and tests
http://s1249.photobucket.com/user/tlosm/library/BenchmarkG5

Last edited by tlosm on 12-Jun-2015 at 02:57 PM.
Last edited by tlosm on 12-Jun-2015 at 02:55 PM.
Last edited by tlosm on 12-Jun-2015 at 02:55 PM.
Last edited by tlosm on 12-Jun-2015 at 02:54 PM.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32;
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
#nomorea-eoninmyhome

 Status: Offline
Profile     Report this post  
pavlor 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 15:02:34
#87 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9578
From: Unknown

@tlosm

So, in your benchmark, one Core2 core is 2x faster than one G5 core at same clock speed.

 Status: Offline
Profile     Report this post  
tlosm 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 15:04:38
#88 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2746
From: Amiga land

@pavlor

You speak about the Machine bench of Geek bench or my Benchmark on cinebench and Blender?

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32;
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
#nomorea-eoninmyhome

 Status: Offline
Profile     Report this post  
pavlor 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 15:06:44
#89 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9578
From: Unknown

@tlosm

Numbers you posted here, I assume it is Geekbench.

 Status: Offline
Profile     Report this post  
tlosm 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 15:09:20
#90 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2746
From: Amiga land

@pavlor

That is Machine Benchmark for single core performance check here

http://s1249.photobucket.com/user/tlosm/library/BenchmarkG5
there is cinebench and blender

In single core the g5 perform near like an I7 860 but i dont have sense make this comparisons

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32;
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
#nomorea-eoninmyhome

 Status: Offline
Profile     Report this post  
olegil 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 15:10:06
#91 ]
Elite Member
Joined: 22-Aug-2003
Posts: 5895
From: Work

@pavlor

Why would coremark not be comparable? The whole POINT of coremark is to be comparable across different arches. Maybe you're thinking of dhrystone?

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

 Status: Offline
Profile     Report this post  
pavlor 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 15:25:18
#92 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9578
From: Unknown

@olegil

Quote:
Why would coremark not be comparable?


What is measured by Coremark? Eg. in Spec CPU 2006 you have some application benchmarks (gcc, bzip2, h264 etc.).

 Status: Offline
Profile     Report this post  
cdimauro 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 20:26:13
#93 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3621
From: Germany

@olegil

Quote:

olegil wrote:
@cdimauro

Quote:

cdimauro wrote:
@olegil: they still report "three instructions", whereas the G4 can process up to 4 (one should be a branch).

So, they don't consider a branch as an instruction.


Well, it isn't. If the branch predictor is implemented sanely

That's a different thing which is only an optimization to save some pipeline stages.

But apart this, a branch is an instruction and should be:
- decoded;
- sent to the proper unit;
- tracked;
- processed (here you can use some optimization);
- committed.

You cannot avoid all this, but just optimize something.

Regarding the G4, it can decode up to 4 instructions, but only 3 can be sent to the proper queues, and another one be processes as a branch. It means that, in any case, you need 4 decoders (fetch 16 byte -> 4 longwords -> 4 instructions) to do that job, because even the branch instruction has to be decoded and properly recognized. Plus all other actions above.
Quote:
Basically you either take the most likely path and have to clear out the pipeline if you're wrong (PPC 603e does this, always guesses loop), OR you take BOTH paths (x86 does this). One of these approaches cost a LOT of transistors, shouldn't be hard to guess which. As for G4, I don't remember what it does.

That's not correct. x86/x64 doesn't execute both paths, but uses the branch predictor, trying to "guess" the real/future target and taking actions in advance (load if needed, decode if needed, then execute). Of course, it's depends on the specific microarchitecture, but that's what a modern x86 does.

Maybe you confused x86 with Itanium, which instead does what you stated, using predication (not prediction). The same which ARM (non-Thumb) does from very long time, but using conditional instructions execution; however ARM dropped it with ARMv8/ARM64.
Quote:
Edit: It seems to have "dynamic branch prediction", so I'm thinking it's NOT the 603e approach (as that is not very dynamic).

Just to add something from what I reported above, x86/x64 has special prefixes to be applied to (conditional) branch instructions, in order to give a "static" hint (branch strongly taken, or not-taken).
Quote:
Edit2: AND reading further it's all explained in the next chapter:
[...]
So it keeps statistics of the branch (well, two bits, but anyway), it keeps a cache of targets, the branch doesn't take any cycles if it hits the cache.

Like what x86/x64 processors do.

BTW, Intel introduced a more advanced and much more efficient branch predictor some years ago, but there are little information about (albeit what's public is very interesting).
Quote:
This gives you close to "three instructions and a branch" each cycle.

See above: that's not the case. Branch instructions have to be decoded, keep track, etc. So you need resources for that: they aren't free.
Quote:
The e6500 manual is phrased a little bit differently (actually not that much), but it seems to do more or less exactly the same thing. Basically all the good stuff from G4 has now been put in the e5500, this is why they named it e6500, as far as I can tell. Now I need to work.

I don't think so. The e5500 is reported to decoded up to 2 instructions, and 3 for the e6500 (but it has to be seen how decoders are shared/used by the two hardware threads). If this already counts possible branches, the G4/e600 has still advantage over them.

 Status: Offline
Profile     Report this post  
cdimauro 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 20:37:44
#94 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3621
From: Germany

@KimmoK

Quote:

KimmoK wrote:
@cdimauro

>Better than old Altivec it's very likely, but why do you think that it's better than SSE2

I did not save any link, just read a study (non apple .pdf) where P4 systems were tested against G5 systems.
In that document 2Ghz G5 was slower than much higher clocked P4 with compiler optimized code.
(I think it took long time before compilers started to fully optimize for G5)
And with hand optimized assembler (on that document/study) G5 became a lot faster in multimedia related operations . IIRC, 150% faster than the higher clocked P4.

Please, take a look at this benchmark with read (and heavy also) applications.
Quote:
But as we know, Altivec has not had big updates since G4, while Intel SIMD has got at least 4...5 updates to it's SIMD technology since 2001.

Sorry for PowerPCs/Altivec, but the progress cannot wait them.
Quote:
(and intel has also been superior in developing compilers for their SIMD, in PowerPC there never was such power in compiler development, so compiler seem to be far behind what latest cores could do)

It took years for Intel to develop his auto-vectorization compiler. Certainly at the time of introduction it wasn't read (and for long time).
Quote:
((unless I'm mistaken, current compilers for AOS4 do not yet fully support PA6T, just as an example))

GCC supports auto-vectorization, but I don't know what's the situation for PowerPCs.
Quote:
>(which is an old technology,

Sure, SSE2 is from y2001.
Apple people had a lot of fun in optimizing for G4 and G5 so that it started to beat P4 with SSE2 (per Mhz and more.

Not even that: see above.

 Status: Offline
Profile     Report this post  
pavlor 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 20:39:43
#95 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9578
From: Unknown

@cdimauro

Quote:
The e5500 is reported to decoded up to 2 instructions, and 3 for the e6500


In e500mc and e6500 cores it is similar (two simple instruction units (SFX0, SFX1), a multiple-cycle instruction unit (MU)). I don´t have documentation for e5500, but I doubt it is different.

Edit: e500mc

Last edited by pavlor on 12-Jun-2015 at 08:41 PM.

 Status: Offline
Profile     Report this post  
cdimauro 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 20:40:24
#96 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3621
From: Germany

@pavlor

Quote:

pavlor wrote:
@olegil

Quote:
Why would coremark not be comparable?


What is measured by Coremark? Eg. in Spec CPU 2006 you have some application benchmarks (gcc, bzip2, h264 etc.).

I agree. SPEC is a more general benchmark. In fact, it's the most used benchmark, even by companies.

I don't understand why people doesn't want to use it, and continue to stick to a more limited benchmark like Coremark.

 Status: Offline
Profile     Report this post  
cdimauro 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 20:42:53
#97 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3621
From: Germany

@pavlor

Quote:

pavlor wrote:
@cdimauro

Quote:
The e5500 is reported to decoded up to 2 instructions, and 3 for the e6500


In e500mc and e6500 cores it is similar (two simple instruction units (SFX0, SFX1), a multiple-cycle instruction unit (MU)). I don´t have documentation for e5500, but I doubt it is different.

Edit: e500mc

Pay attention that I'm talking about the number of instructions that can be decoded, and not about the type of execution units.

 Status: Offline
Profile     Report this post  
pavlor 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 20:43:05
#98 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9578
From: Unknown

@cdimauro

Quote:
I don't understand why people doesn't want to use it,


Not free...

Quote:
and continue to stick to a more limited benchmark like Coremark.


Because it shows nice numbers for their products.

 Status: Offline
Profile     Report this post  
pavlor 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 20:47:04
#99 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9578
From: Unknown

@cdimauro

Quote:
Pay attention that I'm talking about the number of instructions that can be decoded, and not about the type of execution units.


Executions units are the same. However, I´m no expert in this field.

 Status: Offline
Profile     Report this post  
cdimauro 
Re: PowerPC notebook - Status update
Posted on 12-Jun-2015 20:55:39
#100 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3621
From: Germany

@pavlor

Quote:

pavlor wrote:
@cdimauro

Quote:
I don't understand why people doesn't want to use it,


Not free...

But numbers are already available.
Quote:
Quote:
and continue to stick to a more limited benchmark like Coremark.


Because it shows nice numbers for their products.

Can we call it cheating?

Quote:

@cdimauro

Quote:
Pay attention that I'm talking about the number of instructions that can be decoded, and not about the type of execution units.


Executions units are the same.

Yes. And integer/load-store/branch are duplicated on the e6500 (for each thread).
Quote:
However, I´m no expert in this field.


Let's say this: the number of instructions that can be decoded each cycle is a severe limit to the number of instructions that can be executed (each cycle).

The G4 has 13 execution units, but (greatly simplifying) can decoded up to 4 (3 generic + one branch ). So he cannot execute 13 instructions per clock, but 4 at maximum (again, greatly simplifying).

 Status: Offline
Profile     Report this post  
Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 Next Page )

[ home ][ about us ][ privacy ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright (C) 2000 - 2019 Amigaworld.net.
Amigaworld.net was originally founded by David Doyle