Click Here
home features news forums classifieds faqs links search
5637 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
Channel: #Amigaworld
Channel Policy and Guidelines

(Uses JAVA Applet and Port 1024)
Visit the Chatroom Website

Who's Online
 55 guest(s) on-line.
 0 member(s) on-line.



You are an anonymous user.
Register Now!
 corto:  5 mins ago
 wolfe:  8 mins ago
 AdvancedFollower:  25 mins ago
 Plexus:  46 mins ago
 clint:  1 hr 7 mins ago
 Blade2019NL:  1 hr 12 mins ago
 wawa:  1 hr 24 mins ago
 Tuxedo:  1 hr 34 mins ago
 BigD:  1 hr 42 mins ago
 Yssing:  1 hr 57 mins ago

/  Forum Index
   /  Amiga OS4 Hardware
      /  first experiment of x5000 on qemu
Register To Post

Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 Next Page )
PosterThread
Spectre660 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 16:01:23
#101 ]
Elite Member
Joined: 4-Jun-2005
Posts: 3569
From: Unknown

@Srtest

Observations based on use with a SI card.
Will test with a NI card and report .


Quote:

Srtest wrote:
@Spectre660

LXDE vs MATE on the X5k makes a different use of the gpu? you are reffering to both (non-working) SI and (working) NI cards?

_________________

 Status: Offline
Profile     Report this post  
tlosm 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 16:08:21
#102 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2339
From: Amiga land

the different windows manager use different compositors
, if the user are using a gpu with radeon (NI) drivers you dont see and have the issue of slow mooving window this because the graphic is made by the gpu in gpu vram.

If the user use the software resterizer driver of mesa the vram of this driver is the main ram and the graphic are made by the cpu.
more faster is machine ram transfer rate better will be all the graphic system



mate use Marco and Compiz
Lxde and xfce use xfvm.

edit: if this thing come more with a si card it can be a bus problem.
note: im using a NI card on x5000

Last edited by tlosm on 14-Sep-2017 at 04:37 PM.
Last edited by tlosm on 14-Sep-2017 at 04:33 PM.
Last edited by tlosm on 14-Sep-2017 at 04:33 PM.
Last edited by tlosm on 14-Sep-2017 at 04:11 PM.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32; A4k.
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
AmigaOne X5000/40 16GB

 Status: Offline
Profile     Report this post  
Spectre660 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 16:45:42
#103 ]
Elite Member
Joined: 4-Jun-2005
Posts: 3569
From: Unknown

@Srtest


SI card works well with X500/020 + Ubuntu LXDE 16.04.
Window movement is good. No lagging .
Radeon HD 6570 .

Quote:

Srtest wrote:
@Spectre660

LXDE vs MATE on the X5k makes a different use of the gpu? you are reffering to both (non-working) SI and (working) NI cards?

Last edited by Spectre660 on 14-Sep-2017 at 04:46 PM.

_________________

 Status: Offline
Profile     Report this post  
tlosm 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 16:50:19
#104 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2339
From: Amiga land

@Spectre660

that strange why i have the same issue with 6570? im using 16.10

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32; A4k.
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
AmigaOne X5000/40 16GB

 Status: Offline
Profile     Report this post  
Spectre660 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 17:00:22
#105 ]
Elite Member
Joined: 4-Jun-2005
Posts: 3569
From: Unknown

@tlosm

Try
(1) 16.04

(2) Simple xorg.conf

Section "Monitor"
Identifier "Monitor0"
EndSection

Section "Device"
Identifier "Device0"
#Driver "fbdev"
Driver "radeon"
EndSection

Section "Screen"
Identifier "Screen0"
Device "Device0"
Monitor "Monitor0"
DefaultDepth 24
SubSection "Display"
Viewport 0 0
Depth 24
EndSubSection
EndSection

(3) Same kernel as me (4.13.2)
https://drive.google.com/open?id=0BxoZdOdjH0QYdHBFZXdfSEwzb3c

Quote:

tlosm wrote:
@Spectre660

that strange why i have the same issue with 6570? im using 16.10

Last edited by Spectre660 on 14-Sep-2017 at 05:05 PM.
Last edited by Spectre660 on 14-Sep-2017 at 05:05 PM.
Last edited by Spectre660 on 14-Sep-2017 at 05:03 PM.

_________________

 Status: Offline
Profile     Report this post  
tlosm 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 17:14:40
#106 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2339
From: Amiga land

@Spectre660

i will check but will need to do the same on G5 quad for true comparison.

have your kenrel the marvel sata driver and more core than the two of x5000/20?

Last edited by tlosm on 14-Sep-2017 at 05:18 PM.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32; A4k.
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
AmigaOne X5000/40 16GB

 Status: Offline
Profile     Report this post  
Spectre660 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 17:22:32
#107 ]
Elite Member
Joined: 4-Jun-2005
Posts: 3569
From: Unknown

@tlosm

I can also try with Ubuntu 16.10 .
Need to copy my 16.04 and then do an upgrade.
But after work .

Config has these .

CONFIG_SMP=y
CONFIG_NR_CPUS=4

CONFIG_SATA_MV=y

Quote:

tlosm wrote:
@Spectre660

i will check but will need to do the same on G5 quad for true comparison.

have your kenrel the marvel sata driver and more core than the two of x5000/20?

_________________

 Status: Offline
Profile     Report this post  
tlosm 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 17:29:26
#108 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2339
From: Amiga land

@Spectre660

ok will i will try

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32; A4k.
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
AmigaOne X5000/40 16GB

 Status: Offline
Profile     Report this post  
Srtest 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 17:53:40
#109 ]
Regular Member
Joined: 15-Nov-2016
Posts: 158
From: Israel, Haderah

@tlosm

Another question: if both the gpu and cpu are working and you do transfers from ram to vram and back (like the pc dxva copy-back function in video playback), does a certain limitation come into play? Does the modern gui make those transfers regularlly?

On the pc the transfer from ram to vram and back while playing videos allows you to use awesome post processing while that is happening. That might be even more important than simple memory speed.

Last edited by Srtest on 14-Sep-2017 at 05:56 PM.

 Status: Offline
Profile     Report this post  
tlosm 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 18:13:48
#110 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2339
From: Amiga land

@Srtest

x5000 start become more slower when for example i open more "graphic operations" and this make the system choppy ... this slow down made audio start made strange noise too.
For example if i open two mame at same time i start have big issue in the radeon audio . this issue is not present on G5 with radeon too.
Right now im building on G5 qemu with -j4 and i was open two mame with two ghost and goblins n o issue plus writing here

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32; A4k.
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
AmigaOne X5000/40 16GB

 Status: Offline
Profile     Report this post  
Srtest 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 18:29:44
#111 ]
Regular Member
Joined: 15-Nov-2016
Posts: 158
From: Israel, Haderah

@tlosm

Maybe you can try ScummVM with opengl vs software renderer. I would also try to use different video decoders on vlc or smplayer (with mplayer or mpv). Then there is this open source space strategy game which performance fluctuated for me that used hardware acceleration. Maybe there is also a benchmark for web hardware acceleration for FF.

 Status: Offline
Profile     Report this post  
tlosm 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 20:03:25
#112 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2339
From: Amiga land

@Srtest

FF on g5 quad result much more better than Ps3 and Xbox one , on X5000 like ps3

using this peacekeeper.futuremark.com

this is the mate result


2017-09-14 G5 quad Firefox 45.0 Mate 17.04 (dirt)
Click for details

364

2012-03-19 Samsung Galaxy Tab 10.1
Click for details

356

2012-06-01 HTC One X
Click for details

351

2012-11-02 Microsoft Surface
Click for details

348

2012-03-16 Eee Pad Transformer Prime
Click for details

344

2012-11-12 Nokia Lumia 920
Click for details

334

2012-11-17 Microsoft Xbox 360
Click for details

193

2012-11-17 Sony PlayStation 3
Click for details

Last edited by tlosm on 14-Sep-2017 at 08:26 PM.
Last edited by tlosm on 14-Sep-2017 at 08:04 PM.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32; A4k.
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
AmigaOne X5000/40 16GB

 Status: Offline
Profile     Report this post  
amimaniac 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 20:36:09
#113 ]
Member
Joined: 9-Nov-2008
Posts: 56
From: Poland

@tlosm

Hi tlosm, can you run this on X5000 http://aminet.net/package/util/shell/stream_os4
and post single thread stream score here?

Thanks in advance.

 Status: Offline
Profile     Report this post  
tlosm 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 20:41:07
#114 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2339
From: Amiga land

@amimaniac

i cant on os4 because x5000 os4 isnt ready for 5040
i will ask one friend of mine to make it on x5000/20

Last edited by tlosm on 14-Sep-2017 at 08:41 PM.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32; A4k.
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
AmigaOne X5000/40 16GB

 Status: Offline
Profile     Report this post  
Spectre660 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 21:04:39
#115 ]
Elite Member
Joined: 4-Jun-2005
Posts: 3569
From: Unknown

@amimaniac

X5000/020

10.AmigaOS_4.1:> Data:Downloads/stream.os4
-------------------------------------------------------------
This system uses 8 bytes per DOUBLE PRECISION word.
-------------------------------------------------------------
Array size = 2000000, Offset = 0
Total memory required = 45.8 MB.
Each test is run 10 times, but only
the *best* time for each is used.
-------------------------------------------------------------
Your clock granularity/precision appears to be 1 microseconds.
Each test below will take on the order of 27413 microseconds.
(= 27413 clock ticks)
Increase the size of the arrays if this shows that
you are not getting at least 20 clock ticks per test.
-------------------------------------------------------------
WARNING -- The above is only a rough guideline.
For best results, please be sure you know the
precision of your system timer.
-------------------------------------------------------------
Function Rate (MB/s) Avg time Min time Max time
Copy: 1083.3883 0.0297 0.0295 0.0301
Scale: 1043.1970 0.0314 0.0307 0.0356
Add: 1144.2197 0.0424 0.0419 0.0428
Triad: 1152.1560 0.0420 0.0417 0.0425
-------------------------------------------------------------
Solution Validates
-------------------------------------------------------------

Last edited by Spectre660 on 14-Sep-2017 at 09:04 PM.

_________________

 Status: Offline
Profile     Report this post  
tlosm 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 21:13:27
#116 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2339
From: Amiga land

@Spectre660

Quote:

Spectre660 wrote:
@amimaniac

X5000/020

10.AmigaOS_4.1:> Data:Downloads/stream.os4
-------------------------------------------------------------
This system uses 8 bytes per DOUBLE PRECISION word.
-------------------------------------------------------------
Array size = 2000000, Offset = 0
Total memory required = 45.8 MB.
Each test is run 10 times, but only
the *best* time for each is used.
-------------------------------------------------------------
Your clock granularity/precision appears to be 1 microseconds.
Each test below will take on the order of 27413 microseconds.
(= 27413 clock ticks)
Increase the size of the arrays if this shows that
you are not getting at least 20 clock ticks per test.
-------------------------------------------------------------
WARNING -- The above is only a rough guideline.
For best results, please be sure you know the
precision of your system timer.
-------------------------------------------------------------
Function Rate (MB/s) Avg time Min time Max time
Copy: 1083.3883 0.0297 0.0295 0.0301
Scale: 1043.1970 0.0314 0.0307 0.0356
Add: 1144.2197 0.0424 0.0419 0.0428
Triad: 1152.1560 0.0420 0.0417 0.0425
-------------------------------------------------------------
Solution Validates
-------------------------------------------------------------


now compare it with single cores machine
https://www.cs.virginia.edu/stream/peecee/Bandwidth.html

2000.12.23 Dell_Pentium4-1400 1 1352.5 1364.3 1572.1 1555.3

Last edited by tlosm on 14-Sep-2017 at 09:17 PM.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32; A4k.
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
AmigaOne X5000/40 16GB

 Status: Offline
Profile     Report this post  
Spectre660 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 21:14:50
#117 ]
Elite Member
Joined: 4-Jun-2005
Posts: 3569
From: Unknown

@Spectre660

Sam460ex

AmigaOS_4.1:> Data:Downloads/stream.os4
-------------------------------------------------------------
This system uses 8 bytes per DOUBLE PRECISION word.
-------------------------------------------------------------
Array size = 2000000, Offset = 0
Total memory required = 45.8 MB.
Each test is run 10 times, but only
the *best* time for each is used.
-------------------------------------------------------------
Your clock granularity/precision appears to be 2 microseconds.
Each test below will take on the order of 68514 microseconds.
(= 34257 clock ticks)
Increase the size of the arrays if this shows that
you are not getting at least 20 clock ticks per test.
-------------------------------------------------------------
WARNING -- The above is only a rough guideline.
For best results, please be sure you know the
precision of your system timer.
-------------------------------------------------------------
Function Rate (MB/s) Avg time Min time Max time
Copy: 380.2062 0.1002 0.0842 0.2007
Scale: 416.6932 0.0843 0.0768 0.1214
Add: 378.0625 0.1815 0.1270 0.5355
Triad: 465.2196 0.1186 0.1032 0.1588
-------------------------------------------------------------
Solution Validates
-------------------------------------------------------------

_________________

 Status: Offline
Profile     Report this post  
Srtest 
Re: first experiment of x5000 on qemu
Posted on 14-Sep-2017 23:10:53
#118 ]
Regular Member
Joined: 15-Nov-2016
Posts: 158
From: Israel, Haderah

By the way guys, I didn't know the X5k/020 was supposed to be the X1k successor. It didn't appear to me as faster just like a better and more reiable candidate for aos4.1 based on what was available after pa6t became an endangered species.

I found the core2quad in those results and guess what? the X1k comes close when pitted against the 9-10 years old quad pc. That's interesting considering all the fury I encountered when I first came here to decide about it, how it was slow and a bad choice. Turns out it was a darn good one for the current crop of A1s.

And yet they say the X5k/020 manages to be faster than the X1k on aos. Based on these results when the X1k beats the X5k in memory performance, how can it be? the single +200mhz thread? I wonder. Maybe that performance is not as important as vram to ram and vice-versa? What are the deciding factors?

 Status: Offline
Profile     Report this post  
tlosm 
Re: first experiment of x5000 on qemu
Posted on 15-Sep-2017 0:33:47
#119 ]
Elite Member
Joined: 28-Jul-2012
Posts: 2339
From: Amiga land


there isnt not cpu faster than P5040 for single core/tread available in powerpc scene by freescale. other are really much expensive Power8/9.
i think was this the only choice for have the last best Aos4 experience.
Others are much slower in mips for thread.
for example e6500 will be best choice for linux but is slower for thread compared P5040.

And i repeat again and i hope is the last time this machine will gave great aos4 and mos experience but is good only for this oses

Last edited by tlosm on 15-Sep-2017 at 12:48 AM.
Last edited by tlosm on 15-Sep-2017 at 12:35 AM.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32; A4k.
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
AmigaOne X5000/40 16GB

 Status: Offline
Profile     Report this post  
Seiya 
Re: first experiment of x5000 on qemu
Posted on 15-Sep-2017 9:49:27
#120 ]
Super Member
Joined: 19-Aug-2006
Posts: 1373
From: Italia

Quote:

And i repeat again and i hope is the last time this machine will gave great aos4 and mos experience but is good only for this oses



i think many users will buy it only for OS4/MOS.
For any other OS it is too exxxxxxxpensive
If you want to use Linux, you won't use PPC

Last edited by Seiya on 15-Sep-2017 at 09:51 AM.

_________________

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

[ home ][ about us ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright 2000 - 2017 Amigaworld.net.

Amigaworld.net was originally founded by David Doyle