Main Restorations Software Audio/Jukebox/MP3 Everything Else Buy/Sell/Trade
Project Announcements Monitor/Video GroovyMAME Merit/JVL Touchscreen Meet Up Retail Vendors
Driving & Racing Woodworking Software Support Forums Consoles Project Arcade Reviews
Automated Projects Artwork Frontend Support Forums Pinball Forum Discussion Old Boards
Raspberry Pi & Dev Board controls.dat Linux Miscellaneous Arcade Wiki Discussion Old Archives
Lightguns Arcade1Up Try the site in https mode Site News

Unread posts | New Replies | Recent posts | Rules | Chatroom | Wiki | File Repository | RSS | Submit news

  

Author Topic: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)  (Read 184669 times)

0 Members and 4 Guests are viewing this topic.

Shoegazer

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 135
  • Last login:May 10, 2015, 10:37:58 am
    • NonMAME - The ultimate site for the OTHER arcade emulators!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #80 on: November 08, 2007, 10:23:26 pm »
@taz-nz: Thanks for the helpful and informative post.  What you wrote sounds sensible, and I'll definitely be testing more in the near future with this new information in mind.

Regards,
Shoegazer
NonMAME: Your best source for the OTHER arcade emulators!
http://nonmame.retrogames.com

taz-nz

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 147
  • Last login:June 12, 2019, 10:12:13 am
  • If all else fails burn the manual!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #81 on: November 08, 2007, 11:58:51 pm »
Can anyone tell me what available switches are there while compiling with mingw and how do I specify them?  I also have a PIII that I would like to use on my future cabinet and would like to test different builds on it.
I could not figure this out from the compiling instructions.

You just need to edit the "makefile" in the MAME source folder before you compile it. Just open the "makefile" with notepad snd scroll down until you find this section of the test.

#-------------------------------------------------
# specify build options; see each option below
# for details
#-------------------------------------------------

# uncomment one of the next lines to build a target-optimized build
# NATIVE = 1
# ATHLON = 1
# I686 = 1
# P4 = 1
# PM = 1
# AMD64 = 1
# G4 = 1
# G5 = 1
# CELL = 1


Just remove the "# " from infront on the "PM = 1" and save and close file then compile MAME as normal, you will note that the MAME exe produced is now called mamepm.exe


TheManuel

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 805
  • Last login:August 07, 2020, 10:15:48 am
  • On and off hobbyist
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #82 on: November 09, 2007, 09:13:51 am »
Thanks.
I am actually compiling a MAME Plus version which looks a little different than that and did change i686=1 lines to pm=1 lines.  In the end I got a mame file called mamepmp.exe which I imagine is the PM optimized build but it did not make a difference on game performance.  Oh well.  I'll keep looking.
"The Manuel"

xmenxmen

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 290
  • Last login:September 16, 2012, 03:05:35 pm
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #83 on: November 09, 2007, 02:32:55 pm »
For those interested, most of the optimized versions are available online:

http://redump.emubase.de/mame.php

Wonder if there's much differene between the P4 and PM version?

TheManuel

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 805
  • Last login:August 07, 2020, 10:15:48 am
  • On and off hobbyist
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #84 on: November 09, 2007, 02:58:37 pm »
Thank you.
I had seen that website but there is no PM version.
"The Manuel"

taz-nz

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 147
  • Last login:June 12, 2019, 10:12:13 am
  • If all else fails burn the manual!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #85 on: November 09, 2007, 03:13:19 pm »
For those interested, most of the optimized versions are available online:

http://redump.emubase.de/mame.php

Wonder if there's much differene between the P4 and PM version?

More than you might expect, the Pentium M/Core Duo family of processors have more in common with the Pentium 3 than they do with Pentium 4 in many ways, the Netburst architecture used in the P4 was ment to scale all the way to 5Ghz, but due to pressure from the verious versions of the AMD Athlons Intel had to accelerate the scaling of the P4s clock speed, thus they hit a clock speed wall at 3.8-4ghz where they couldn't go any higher because the manufacture processes where now 3 years behind the clock speed curve of the P4. So instead they gave us two years of bandaid fixes with the prescott cores while they when back and developed the Core 2 Duo. (There is a lot more to the history of this but I don't want to bore everyone, and AMD still has the better system architecture, but Intel has the better Core Architecture at the moment)

I knew of the website, but I figured it was better if everyone learned make theier own builds. The pre optimised builds tend to get out of date, like mame is now at 0.120u3 and the site still at 0.120u2 but it will catch up in a few days.


u_rebelscum

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 3633
  • Last login:April 21, 2010, 03:06:26 pm
  • You rebel scum
    • Mame:Analog+
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #86 on: November 09, 2007, 06:34:14 pm »
A recommendation for the futue is to run a fixed number of frames rather than a fixed number of seconds.  The reason is that if someone wants to compare his benchmark to yours (if you care, anyway), and his/her computer is slower, that computer will run a shorter sequence of the game in 240sec so it will not be a 100% apples to apples comparison since some parts of the attract mode of a game are more taxing than others.

Alread done.  Mame's "seconds" in the benchmark is emulated seconds and Mame switched the benchmark/test option to "-seconds_to_run" from "-frames_to_run" a little while ago.

IOW, a game with 60 Hz refresh runs 600 frames if -str 10, no matter if the game runs at 500% or 50%.

The only "differences" between -ftr and -str changes between games with different refresh rates.  On 53 Hz games, -ftr 600 equals about -str 11, and -str 10 equals about -ftr 530; on 61 Hz games, -ftr 600 falls between -str 9 & -str 10, and str 10 equals about -ftr 610.
Robin
Knowledge is Power

TheManuel

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 805
  • Last login:August 07, 2020, 10:15:48 am
  • On and off hobbyist
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #87 on: November 09, 2007, 09:07:30 pm »
Thank you for correcting my ignorance.  Things have changed in last three an a half years I have been away from the hobby.

By the way, it is good to see you (your web persona I guess) around.  You were very helpful when I was building my control panel years ago.
"The Manuel"

TheManuel

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 805
  • Last login:August 07, 2020, 10:15:48 am
  • On and off hobbyist
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #88 on: November 09, 2007, 09:12:11 pm »
Hit "post" button too early...

I am a little frustrated.
I have tested i686, PM and P3 builds of MAME and they all yield virtually the same result, at least on the game I am benchmarking against which is Mace (it sucks, by the way).
Could I be missing something?  Even the -mt parameter makes no difference and I did set osdprocessors=2.

I am using the same command line parameters as tar-nz on my Core 2 Duo e4300 overclocked to 2.4GHz.  I would like to see the same kind of improvements (proportional to my processor speed) tar-nz saw with the PM build but it is not working for me.

Any pointers will be appreciated.
"The Manuel"

divemaster127

  • Trade Count: (+60)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 1977
  • Last login:December 02, 2018, 08:05:08 pm
  • My webstore is arcadeemulator.net
    • arcadeemulator.net
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #89 on: November 10, 2007, 12:23:03 pm »
Hey guys could I get a couple of opinions I have stopped setting up my new quad until i get some good feedback used to video did not matter but now which video card would you guys use
1. pci-e arcadevga or I have a extra 8800 gts laying around.
My monitor is a betson I did some performance test & the 8800 just runs all over the arcadevga.
thanks for your help
dm
I carry both ultimarc & happ items, all brand new & I ship from the united states. My online store is ARCADEEMULATOR.NET, pm if I can help in anyway.

taz-nz

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 147
  • Last login:June 12, 2019, 10:12:13 am
  • If all else fails burn the manual!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #90 on: November 11, 2007, 02:56:20 am »
CPU
ROM
Driver
      0.119   
     0.120   
     0120u1 
     0120u2 
   0.120(64)
   0.120u2(64)
   0.120u3(64)
E6400 @ 3.4ghz    crusnusa   midvunit.c
212.24%
212.12%
178.27%
182.84%
  n/a
220.81%   
  n/a
E6850 @ 4.0ghzcrusnusa   midvuint.c
257.69%
257.32%
218.20%
214.79%
271.81%
269.31%   
277.86%   
E6400 @ 3.4ghz    starblad   namcos21.c
234.33%
120.83%
116.97%
119.37%
  n/a
127.74%   
  n/a
E6850 @ 4.0ghzstarblad   namcos21.c
299.30%
144.10%
144.00%
142.10%
153.57%
150.71%   
150.75%   
E6400 @ 3.4ghz    gauntleg   vegas.c
47.92%
114.97%
132.38%
139.10%
  n/a
163.50%   
  n/a
E6850 @ 4.0ghzgauntleg   vegas.c
68.91%
166.17%
173.07%
166.03%
183.42%
191.32%   
200.33%   
E6400 @ 3.4ghz    blitz   seattle.c
122.25%
87.86%
111.59%
113.12%
  n/a
145.63%   
  n/a
E6850 @ 4.0ghzblitz   seattle.c
153.43%
138.59%
143.28%
136.99%
164.15%
185.35%   
172.56%   
E6400 @ 3.4ghz    daytona   model2.c
   n/a
86.18%
87.22%
89.09%
  n/a
120.89%   
  n/a
E6850 @ 4.0ghzdaytona   model2.c
   n/a
106.17%
106.43%
104.94%
142.72%
140.42%   
141.41%   
E6400 @ 3.4ghz    gradius4   hornet.c
60.36%
79.34%
   n/a
97.97%
  n/a
79.57%   
  n/a
E6850 @ 4.0ghzgradius4   hornet.c
116.72%
164.94%
   n/a
180.32%
122.16%
125.69%   
128.15%   
E6400 @ 3.4ghz    radikalb   gaelco3d.c
77.46%
76.11%
95.72%
109.88%
  n/a
163.93%   
  n/a
E6850 @ 4.0ghzradikalb   gaelco3d.c
94.25%
93.97%
116.77%
129.44%
150.07%
200.22%   
196.07%   
E6400 @ 3.4ghz    ridgerac   namcos22.c
65.56%
64.73%
90.77%
92.37%
  n/a
147.79%   
  n/a
E6850 @ 4.0ghzridgerac   namcos22.c
80.00%
79.54%
114.89%
112.44%
106.77%
194.55%   
177.79%   
E6400 @ 3.4ghz    scud   model3.c
34.41%
32.91%
33.00%
33.88%
  n/a
28.48%   
  n/a
E6850 @ 4.0ghzscud   model3.c
41.21%
39.60%
39.56%
38.93%
33.81%
33.89%   
30.36%   

No big preformance changes with 0.120u3, the biggest change in my mind is that SCUD no longer looks like a strobe light, but still has a lot of graphics issues.


I am a little frustrated.
I have tested i686, PM and P3 builds of MAME and they all yield virtually the same result, at least on the game I am benchmarking against which is Mace (it sucks, by the way).
Could I be missing something?  Even the -mt parameter makes no difference and I did set osdprocessors=2.

I am using the same command line parameters as tar-nz on my Core 2 Duo e4300 overclocked to 2.4GHz.  I would like to see the same kind of improvements (proportional to my processor speed) tar-nz saw with the PM build but it is not working for me.

Any pointers will be appreciated.

I'm not sure why your seeing no difference with the -mt, there may be something else in your system that is limiting your prefomance, CPU FSB/Cache, memory speed or amount, graphics card. I did a little testing on the PM optimisation and it seams it lost some of it's edge with the new builds, probably due to the fact the drivers are now so much better optimised than the were before.

I compile two 32bit builds of mame 0.120u3 with and with out the PM swtich inable in the make file it then bench marked a selection of ROMs with both version with and without the -MT swtch.

ROM            STANDARD          STD -MT                  PM                   PM -MT

carneveil    220.38%            229.76%            222.15%            229.87%
a51site4     178.56%            186.05%            181.06%            187.67%
starblad      130.53%            136.82%            135.86%            142.34%
mace           193.00%            195.97%            189.39%            189.75% (32bit graphics bug)
gradius4     115.47%            120.49%            111.49%            115.96% (32bit graphics bug)
propcycl        68.68%              79.93%             71.29%             73.39%
crusnusa     223.62%            242.82%            226.68%            235.00%
gauntdl       110.27%            113.19%            108.82%            111.18% (32bit graphics bug)

As you can see some roms gain a few % with the PM optimisation, all the drivers with the u3 Build 32bit graphics bug drop a little.
It looks like the PM optimised build may have become obsolete with the last few release of MAME, the early build gain alot more for the optimised builds.

Hey guys could I get a couple of opinions I have stopped setting up my new quad until i get some good feedback used to video did not matter but now which video card would you guys use
1. pci-e arcadevga or I have a extra 8800 gts laying around.
My monitor is a betson I did some performance test & the 8800 just runs all over the arcadevga.
thanks for your help
dm

If your Betson will run ok off the 8800 gts and you don't have another use for it already I probably use that, It total over-kill for MAME I only run a 8800 gts as I also PC game alot on my system. The 8800gt may have a preformance advantage with Vista, I wouldn't think you would see much difference in MAME benchmarks between the two cans in Windows XP, but as Vista likes to render the Aero desktop in the backgrond even when it's not visible,  the 8800gts may have an advantage there.


manny2003

  • Trade Count: (0)
  • Newbie
  • *
  • Offline Offline
  • Posts: 1
  • Last login:November 14, 2007, 12:24:33 am
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #91 on: November 11, 2007, 05:28:08 pm »
i remember reading that the guy who wrote the software SLI driver for rendering (he called in SLI because it splits the rendering between 2 or more cpus) said on his site that .120 builds through to u2 if you have a quad using OSDPROCESSOR=2 would be faster on a quad than using OSDPROCESSORS=4 because of the way the render code worked.

The work in progress for the 120u3 build of mame states as a feature that the renderer supports quad processors. (maybe that means there isn't a penalty for using OSDPROCESSORS=4,  or it could mean that it works properly.) He has also only added support for this SLI rendering for some of the drivers at this point (hence the resaon why some of them do no really get an improvement from the switch.)

-Manny

TheManuel

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 805
  • Last login:August 07, 2020, 10:15:48 am
  • On and off hobbyist
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #92 on: November 12, 2007, 11:22:07 am »
Thank you for all the information and advice, taz-nz.  It sounds like recent versions of mame are already pretty well optimized so perhaps I should not concern myself to much with that.

One final question to all, though:
Is there a way to set the environment variable osdprocessor permanently in the registry or do we always have to set it before running mame to get the benefit?

Thanks.
"The Manuel"

tommy

  • Guest
  • Trade Count: (0)
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #93 on: November 12, 2007, 11:30:15 am »
I didn't read this whole thread. Did anyone find Gauntlet Legends to be anymore playable with a good dual processor PC?

Tiger-Heli

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 5447
  • Last login:January 03, 2018, 02:19:23 pm
  • Ron Howard? . . . er, I mean . . . Run, Coward!!!
    • Tiger-Heli
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #94 on: November 12, 2007, 11:45:23 am »
Is there a way to set the environment variable osdprocessor permanently in the registry or do we always have to set it before running mame to get the benefit?
I would expect there to be a mame.ini file line for it, but haven't checked.
It's not what you take when you leave this world behind you, it's what you leave behind you when you go. - R. Travis.
When all is said and done, generally much more is SAID than DONE.

TheManuel

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 805
  • Last login:August 07, 2020, 10:15:48 am
  • On and off hobbyist
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #95 on: November 12, 2007, 01:43:57 pm »
Well, this is not a MAME setting per se but a parameter that is set from a command prompt windows so I don't know if it could be taken care of from the MAME ini file.
"The Manuel"

taz-nz

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 147
  • Last login:June 12, 2019, 10:12:13 am
  • If all else fails burn the manual!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #96 on: November 12, 2007, 02:35:59 pm »
I didn't read this whole thread. Did anyone find Gauntlet Legends to be anymore playable with a good dual processor PC?

-gauntleg          0.120u2 64bit   152.79%         87.09fps        (Core 2 Duo E6850 @ 4ghz)

Totally playable, but still tended to lock up at random in 0.120u2, not sure if this bug is fixed in 0.120u3.





Carnevil

  • Trade Count: (0)
  • Newbie
  • *
  • Offline Offline
  • Posts: 1
  • Last login:November 13, 2007, 03:36:01 pm
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #97 on: November 13, 2007, 02:57:23 pm »
I am curious. Has anyone tried 120u3 with a quad core processor to see how well the new quad rendering in polynew.c works?
I plan on buying a Core2Duo real soon and I would like to know if I should go with a dual core or quad core.
Thanks

u_rebelscum

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 3633
  • Last login:April 21, 2010, 03:06:26 pm
  • You rebel scum
    • Mame:Analog+
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #98 on: November 14, 2007, 04:13:52 pm »
Is there a way to set the environment variable osdprocessor permanently in the registry or do we always have to set it before running mame to get the benefit?

Yes.
Not sure about Vista, but XP: open Control Panel, Open System, click the Advanced tab, click the Environment Variables button, click New in the System Variables (lower half).
Robin
Knowledge is Power

AaronGiles

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 26
  • Last login:May 17, 2008, 09:59:11 pm
  • I want to build my own arcade controls!
    • Aaron's Home Page
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #99 on: November 14, 2007, 08:01:40 pm »
Let's clear up some misinformation:

1. You don't need to set OSDPROCESSORS, unless you are debugging a problem. By default, MAME detects the number of processors in your system and uses that. The default implementation should be good. Leave it alone.

2. The -mt switch does not affect the multicore 3D acceleration. The multicore 3D code is enabled always. The -mt switch merely controls whether or not MAME uses a second thread to manage windows and blitting. Generally if you have multiple cores, this is a good thing; however, it is not enabled by default, so you will need to enable it in your mame.ini.

3. "Quad rendering" has nothing to do with quad core. The quad rendering functions mentioned in the latest whatsnew refer to rendering a polygon with 4 vertices (called a "quad").

4. The 3D acceleration code scales well to 2 processors. Above 2 you get diminishing returns, and at some point it will start to get slower rather than faster.

divemaster127

  • Trade Count: (+60)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 1977
  • Last login:December 02, 2018, 08:05:08 pm
  • My webstore is arcadeemulator.net
    • arcadeemulator.net
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #100 on: November 14, 2007, 08:14:53 pm »
Well that makes my decision I will throw my dual core 6750 in my mame & use the quad core in my racer plus I cannot overclock my quad no matter what I do, 2 exact quad cores, 1 overclocks perfect the 2nd one is terrible, same motherboards, cpu coolers etc
thanks
dm
I carry both ultimarc & happ items, all brand new & I ship from the united states. My online store is ARCADEEMULATOR.NET, pm if I can help in anyway.

TheManuel

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 805
  • Last login:August 07, 2020, 10:15:48 am
  • On and off hobbyist
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #101 on: November 14, 2007, 09:19:02 pm »
Is there a way to set the environment variable osdprocessor permanently in the registry or do we always have to set it before running mame to get the benefit?

Yes.
Not sure about Vista, but XP: open Control Panel, Open System, click the Advanced tab, click the Environment Variables button, click New in the System Variables (lower half).

Thanks, man.
You have just made my life easier...

... and thanks Aaron for clearing all of that up.
Now it makes more sense why I don't see much difference with some of the settings I tried.
« Last Edit: November 14, 2007, 09:21:55 pm by TheManuel »
"The Manuel"

taz-nz

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 147
  • Last login:June 12, 2019, 10:12:13 am
  • If all else fails burn the manual!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #102 on: November 15, 2007, 12:39:30 am »
Let's clear up some misinformation:

1. You don't need to set OSDPROCESSORS, unless you are debugging a problem. By default, MAME detects the number of processors in your system and uses that. The default implementation should be good. Leave it alone.

2. The -mt switch does not affect the multicore 3D acceleration. The multicore 3D code is enabled always. The -mt switch merely controls whether or not MAME uses a second thread to manage windows and blitting. Generally if you have multiple cores, this is a good thing; however, it is not enabled by default, so you will need to enable it in your mame.ini.

3. "Quad rendering" has nothing to do with quad core. The quad rendering functions mentioned in the latest whatsnew refer to rendering a polygon with 4 vertices (called a "quad").

4. The 3D acceleration code scales well to 2 processors. Above 2 you get diminishing returns, and at some point it will start to get slower rather than faster.

Thanks for that Aaron, I had figured as much from what I seen testing different builds etc, but wasn't sure enough to start making statements about any of it, I would have asked but I didn't want to waste your time.

Red

  • Trade Count: (+14)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 464
  • Last login:November 07, 2017, 08:37:51 am
  • I want to build my own arcade controls!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #103 on: November 15, 2007, 01:22:46 am »
2. The -mt switch does not affect the multicore 3D acceleration. The multicore 3D code is enabled always. The -mt switch merely controls whether or not MAME uses a second thread to manage windows and blitting. Generally if you have multiple cores, this is a good thing; however, it is not enabled by default, so you will need to enable it in your mame.ini.

"Generally if you have multiple cores, this is a good thing; however, it is not enabled by default, so you will need to enable it in your mame.ini."

Any plans to make this enabled by default, so one doesn't have to enable it in your mame.ini file?  Thanks.

Tiger-Heli

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 5447
  • Last login:January 03, 2018, 02:19:23 pm
  • Ron Howard? . . . er, I mean . . . Run, Coward!!!
    • Tiger-Heli
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #104 on: November 15, 2007, 08:31:46 am »
Any plans to make this enabled by default, so one doesn't have to enable it in your mame.ini file?  Thanks.
At risk of stating the obvious, if he planned to, I think they would have, it's an easy enough change to make.
It's not what you take when you leave this world behind you, it's what you leave behind you when you go. - R. Travis.
When all is said and done, generally much more is SAID than DONE.

Red

  • Trade Count: (+14)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 464
  • Last login:November 07, 2017, 08:37:51 am
  • I want to build my own arcade controls!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #105 on: November 15, 2007, 10:38:44 am »
Hopefully then it will be set to Default on later versions.

ozzi22

  • Trade Count: (0)
  • Jr. Member
  • **
  • Offline Offline
  • Posts: 9
  • Last login:May 12, 2008, 07:09:57 pm
  • I'm a llama!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #106 on: November 17, 2007, 07:24:04 pm »
how does using one of these builds in a frontend affect the bech results?  examples mala or game ex

taz-nz

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 147
  • Last login:June 12, 2019, 10:12:13 am
  • If all else fails burn the manual!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #107 on: November 18, 2007, 04:14:08 am »
Hopefully then it will be set to Default on later versions.

I think you'll find in hurts performance on single core processors and thus is defaultly off, that doesn't mean someone couldn't have mame turn it on automatically if a multi-core processors was detected, it probably wouldn't be that hard to do since Mame already does this for Arron's software SLI code.

how does using one of these builds in a frontend affect the bech results?  examples mala or game ex

Frontends are just ROM launchers, they take up very little system resources, and are basically inactive while a ROM is being emulated, so I wouldn't expect to see any noticeably difference. There are dozens of others tasks running the background on average windows PC, and most of them have little to no effect on MAMEs performance, my Antivirus on the other hand has a bad habit of kicking in while I'm trying to benchmark ROMs, that is very noticeable.

I would like to see how a 64bit version of SDLMAME benchmarked, running on a 64bit version of Linux, it would be interest to see if Windows resource usage has any negative effect MAMEs performance.  I'm  looking into this currently, but as soon as you say linux everything get 10x more complex than it needs to be, so there probably will not be results on this one until atleast the next stable build of MAME is released.

Looks like I might have to get a E8500 after all in the new year, below are overclocking results with on pre-production sample, air cooler with a less than fantastic Thermal Take V1 heatsink (lot of room to improve there)

497X9.5=>4722Mhz on air cooler   


For full details jump over to http://forums.tweaktown.com/showthread.php?t=25544

Wish I had that CPU to play with.



ozzi22

  • Trade Count: (0)
  • Jr. Member
  • **
  • Offline Offline
  • Posts: 9
  • Last login:May 12, 2008, 07:09:57 pm
  • I'm a llama!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #108 on: November 22, 2007, 10:38:34 am »
Any new bench results with mame 121?

taz-nz

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 147
  • Last login:June 12, 2019, 10:12:13 am
  • If all else fails burn the manual!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #109 on: November 23, 2007, 02:55:11 am »
Any new bench results with mame 121?

Yeah will be up tomorrow, I'll post a full set once I re-run them with 0.120u1, vegas.c was broken in 0.120  :cry:.
But you can see my 0.120 less gauntleg, gauntdl, tenthdeg over at http://benchmark.mameworld.net/ where I currently hold the title of "speed king" >:D just click on "Benchmark Database", and then select "0.121" under "MAME Build", then hit "Show me the list" this will give you my lastest results.

Update:

Had some PC issues this weekend lots of random lockups, BSODs and restarts, figured it was a heat or stablility issue with the overclocking, wasted a day trying to fix a problem that wasn't there  :banghead:, turned out it was my DVD-Writer on it's way out  :hissy:  unplugged it and all was OK again but then I had to undo everything from the previous day, so no benchmarks until tommorow at the earliest, sorry.

 
« Last Edit: November 25, 2007, 05:08:25 am by taz-nz »

Tiger-Heli

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 5447
  • Last login:January 03, 2018, 02:19:23 pm
  • Ron Howard? . . . er, I mean . . . Run, Coward!!!
    • Tiger-Heli
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #110 on: November 27, 2007, 10:58:44 am »
So for all those of you wanting to build a preformance M.A.M.E. PC on a budget, I did a quick shopping list and then found some prices over at www.clubit.com (There may be better and cheaper places out there, but they had the right gear at what looked like the right price)

Taz-nz - thanks for all the helpful info in this thread.  I'm really not looking to replace my MAME machine, but was curious why you picked some of the components you did (I understand the theory, but not all the requirements)., and my budget is much less than $360.

    GIGABYTE GA-P35-DS3R Rev. 2.0 LGA 775 Intel P35 ATX Intel Motherboard Retail     
$128.50
Is it important to use this particular board, or just a P35 board.  For instance, Newegg has a Open box MSI P35-Neo2 FR for $65, or the Gigabyte GA-P35-DS3L for $90.  Club-It also has similar (slightly better) prices on P35 boards.
    Cooler Master RR-CCH-LB12-GP Hyper 212 CPU Cooler   
$33.99
I guess the cooler is pretty critical to stable overclocking, but $35 seems a bit high.  I don't know of good alternates, though.  I've like Arctic Cooling in the past (but that was on Athlon XP systems), and the new fanless towers (like Tuniq) look cool, but again not sure how they'd be for overclocking.  Any idea how the stock Intel cooler would do?
    Intel Core 2 Duo E4500 M0 Stepping Conroe 2.2GHz 2MB L2 LGA 775 Processor Retail BX80557E4500 SLA95   
$123.50
How would the E2140, E2160, or E2180 compare with this processor - they are all 800 FSB parts and priced at $70-$90.  (Also, some of these mention a particular "Stepping", but I don't know how this factors in.)
   CORSAIR XMS2 2GB (2 x 1GB) DDR2-800 CL 5-5-5-12 Dual Channel Kit Retail TWIN2X2048-6400
$74.50
I'm not sure 2GB is required for a MAME machine - Newegg and Club-It have 1GB DDR2-800 Cas 5 for $25-$35 and 2GB (matched pair) for $54.  Would this be acceptable?  Would DDR2-1066 offer any advantages?

If the pieces above would work, you could knock $200 off the price above and put together a system for around $160 (65+70+25)

Thanks again for all your help!!!
It's not what you take when you leave this world behind you, it's what you leave behind you when you go. - R. Travis.
When all is said and done, generally much more is SAID than DONE.

TheManuel

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 805
  • Last login:August 07, 2020, 10:15:48 am
  • On and off hobbyist
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #111 on: November 27, 2007, 11:43:21 am »
I have played with different memory timings and speed on my main computer and none of it does squat.  I went from 533 to 800 FSB and 4-4-4-12 to 5-5-5-15 and it only looks pretty in the Sandra graphs but MAME does not care.  I only cares for one thing: processors speed. 

The tests above where on a Core 2 Duo E4300 running at 2.4GHz.  I ran it at 3.0GHz for a while and the FPS did not quite scale linearly to 25% faster but to about 22% faster.  The games I was looking at were mace and umk3.

On my intended MAME cab machine (which I have yet to fully configure) I tried 384 vs 512 and PC100 vs PC133 and so no difference there either but I only benchmarked umk3, mvsc and garou.  I don't have enough evidence but I tend to think that you probably don't get added performance above 512MB of memory even for the larger neogeo games.  I don't know if this is the case with hard drive games because that machine will not run it at any decent speed so I did not even bother to benchmark it.

I suppose I will have to use an older version of MAME for umk3 and other of the more demanding games.  Luckily, I don't care for any of the hard drive games but, like I said a few posts ago, I'm out of luck for SFIII.
"The Manuel"

metahacker

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 166
  • Last login:January 17, 2024, 03:09:39 am
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #112 on: November 27, 2007, 08:53:18 pm »
can you try a comparison to your regular benchmarks with the following in your Makefile ARCH= line ?

-march=nocona -mtune=nocona -minline-all-stringops -maccumulate-outgoing-args -mmmx -msse -msse2 -msse3 -mfpmath=sse -funroll-loops -fomit-frame-pointer

and perhaps even compare that to

-march=pentium-m -mtune=pentium-m -minline-all-stringops -maccumulate-outgoing-args -mmmx -msse -msse2 -msse3 -mfpmath=sse -funroll-loops -fomit-frame-pointer

divemaster127

  • Trade Count: (+60)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 1977
  • Last login:December 02, 2018, 08:05:08 pm
  • My webstore is arcadeemulator.net
    • arcadeemulator.net
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #113 on: November 27, 2007, 11:14:31 pm »
Ok, i finally have my mame e6750 w/vista 64, 4 gig ram running on mame, 121.1, carnevil & mace run perfect, Gauntlet legends & blitz run good but they still have a slight sound jumping.  Taz how does blitz run on your 6850, & 121 have you noticed a slowdown if so i may go back to v120.  thanks for the help also i overclocked to 3.4
dm
I carry both ultimarc & happ items, all brand new & I ship from the united states. My online store is ARCADEEMULATOR.NET, pm if I can help in anyway.

taz-nz

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 147
  • Last login:June 12, 2019, 10:12:13 am
  • If all else fails burn the manual!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #114 on: November 28, 2007, 01:13:08 am »
Taz-nz - thanks for all the helpful info in this thread.  I'm really not looking to replace my MAME machine, but was curious why you picked some of the components you did (I understand the theory, but not all the requirements)., and my budget is much less than $360.

When I compiled this list of parts, I wanted give people a good starting point I knew would overclock well, so I went with parts I deal with on a daily basis.

    GIGABYTE GA-P35-DS3R Rev. 2.0 LGA 775 Intel P35 ATX Intel Motherboard Retail     
$128.50
Is it important to use this particular board, or just a P35 board.  For instance, Newegg has a Open box MSI P35-Neo2 FR for $65, or the Gigabyte GA-P35-DS3L for $90.  Club-It also has similar (slightly better) prices on P35 boards.

Any P35 chipset board is a good start, but they are not all created equal, the Gigabyte P35-D** series of board are proven overclockers, and they use better quality parts, see Ultra Durabilty 2. The GA-P35-DS3L should be fine stay away from the GA-P31-DS3L the chipset is rubbish. I'd stay away from the MSI unless you can find plenty of proof it's a good overclocker, I not a fan of the brand they fail to often for my liking. (The main reason for going with a P35 chipset board is that they are very stable at high FSBs which is very important when overclocking, extra cooling of the northbridge doesn't hurt any)

    Cooler Master RR-CCH-LB12-GP Hyper 212 CPU Cooler   
$33.99
I guess the cooler is pretty critical to stable overclocking, but $35 seems a bit high.  I don't know of good alternates, though.  I've like Arctic Cooling in the past (but that was on Athlon XP systems), and the new fanless towers (like Tuniq) look cool, but again not sure how they'd be for overclocking.  Any idea how the stock Intel cooler would do?

You really need the heatsink if you want to get any really good overclocking results, the factory heatsink is overkill for these CPUs out of the box, but soon fails to keep up as you start to overclock them. There are plenty of good coolers out there, but currently the Hyper 212 is the best value for money as far as I'm concerned, it's performance is only slightly below that of theThermalright 120 Ultra I used in my system, at about half the price. There are loads of coolers out there that don't live up to the hype or their cost, the Hyper 212 just gets the job done and is affordable.

    Intel Core 2 Duo E4500 M0 Stepping Conroe 2.2GHz 2MB L2 LGA 775 Processor Retail BX80557E4500 SLA95   
$123.50
How would the E2140, E2160, or E2180 compare with this processor - they are all 800 FSB parts and priced at $70-$90.  (Also, some of these mention a particular "Stepping", but I don't know how this factors in.)

The E21** series overclock well like all the Core 2 Duo, but the 1mb cache has a noticable effect of their preformance when compared to a E4*00 or E6**0 series Core 2 Duo of the same clock speed, they are also limited in most cases to around the 3ghz overclocked from what i've seen, with a few managing 3.2-3.4ghz. The E4*00 series processors, are more likely to hit the 3.2-3.6ghz mark and have the large cache which helps their preformance.

I choose the E4500 M0 stepping as the 9x multiplier means you only need overclock to 400mhz FSB to get to 3.6ghz, and thus you only need DDR2-800 memory. The M0 stepping is the same as the G0 in the E6*50 series processors, these stepping overclock better than the older steppings of Core 2 Duo, and thus are worth having if your looking to hit big numbers. (steppings are small changes made in the CPU design to fix bugs or increase production yeilds, they tend to optimise the design toward stability and some make for great improvements in overclocking.)
   CORSAIR XMS2 2GB (2 x 1GB) DDR2-800 CL 5-5-5-12 Dual Channel Kit Retail TWIN2X2048-6400
$74.50
I'm not sure 2GB is required for a MAME machine - Newegg and Club-It have 1GB DDR2-800 Cas 5 for $25-$35 and 2GB (matched pair) for $54.  Would this be acceptable?  Would DDR2-1066 offer any advantages?

2gb is not a must for MAME, but the catch is that the best OS option for mame Preformance is Vista 64bit which is a memory hog, and really needs 2gb to run well, to give you an idea, while doing some of the CPU only benchmark up to 1.7gb of physical ram were in use, and that with nothing running but Vista and MAME. Windows XP 64bit might be easier on the memory usage but is not as well supported as Vista 64bit, and you can get Vista Home Prem 64bit a lot cheaper & easier than you can XP Pro 64bit.

In regard to ram speed I really haven't seen much difference in MAMEs preformance when using different preformance memory, but make sure you run dual channel no matter how much memory you run. Any half respectable DDR-800 ram will do just make sure it's compatible with the P35 chipset, there are a few Dimms out there that work great on the older Intel chipsets and with AM2 CPUs but are totally unstable withe P35 chipset.
If the pieces above would work, you could knock $200 off the price above and put together a system for around $160 (65+70+25)

Thanks again for all your help!!!

Preformace is just a question of money you have to ask yourself how fast do you want to go, and then speed the money need to get their. To give you an Idea my system for CPU, RAM, Motherboard, & Heatsink cost around $1300 NZD and number of months planning and saving.

Hope this helps fills in some of the blanks for you.

taz-nz

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 147
  • Last login:June 12, 2019, 10:12:13 am
  • If all else fails burn the manual!
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #115 on: November 28, 2007, 02:30:48 am »
Ok, i finally have my mame e6750 w/vista 64, 4 gig ram running on mame, 121.1, carnevil & mace run perfect, Gauntlet legends & blitz run good but they still have a slight sound jumping.  Taz how does blitz run on your 6850, & 121 have you noticed a slowdown if so i may go back to v120.  thanks for the help also i overclocked to 3.4
dm

Haven't noticed much change with 0.121, most of the big changes were in 0.121u# releases
gauntleg is about the same, runs fine speed wise but still likes to lock up at random.
blitz is down a little maybe 5%, I suck at this game totally so haven't tried in game lately.

Good to see you've got your PC up and running well, and welcome to the fast lane. what heatsink are you using by the way?

Any new bench results with mame 121?

Full set of in game benchmarks, run in Vista 64bit, use MAMEUI 64bit, with the command line:
-noautoframeskip -frameskip 0 -seconds_to_run 240 -nothrottle -nosleep -video ddraw -skip_gameinfo -effect none -nowaitvsync -noreadconfig -mt [ROM NAME]

- 1942                      0.121u1      6572.08%

- a51site4                0.121u1      235.35%

- airco22b                0.121u1      143.19%

- alpinerd                 0.121u1      88.50%

- biofreak                 0.121u1      145.12%

- blitz2k                   0.121u1      160.68%

- blitz99                   0.121u1      153.09%

- blitz                       0.121u1      167.30%

- calspeed                0.121u1      204.23%

- carnevil                  0.121u1      285.30%

- crusnusa                0.121u1      276.53%

- crusnwld                0.121u1      249.91%

- cybrcomm              0.121u1      135.27%

- Daytona                 0.121u1      212.79%

- gauntdl                  0.121u1      134.20%

-gauntleg                 0.121u1      159.81%

- gradius4                 0.121u1      96.85%       (64bit MAME /64bit Windows)
                                 0.121u1      110.77%      (32bit MAME /32bit Windows)      
                                 0.121u1      113.95%      (32bit MAME /64bit Windows)

- hyprdriv                  0.121u1      175.17%

- mace                      0.121u1      251.07%

- offroadc                  0.121u1      438.94%

- propcycl                  0.121u1      137.27%

- radikalb                  0.121u1      177.72%

- raveracw                0.121u1      121.21%

- ridgerac                  0.121u1      151.72%

- scud                        0.121u1      28.61%    (score keeps getting worse, but demo speed keeps getting better, can't wait for this one to be a working ROM)

- sfrush                     0.121u1      211.70%

- sidebs2                  0.121u1      215.90%

- speedup                 0.121u1      179.16%

- starblad                  0.121u1      145.58%

- surfplnt                   0.121u1      163.14%

- tenthdeg                 0.121u1      85.30%

- timecris                   0.121u1      145.30%

- wargods                 0.121u1      411.74%

- wg3dh                    0.121u1      375.54%

- xevi3dg                  0.121u1      371.37%

can you try a comparison to your regular benchmarks with the following in your Makefile ARCH= line ?

-march=nocona -mtune=nocona -minline-all-stringops -maccumulate-outgoing-args -mmmx -msse -msse2 -msse3 -mfpmath=sse -funroll-loops -fomit-frame-pointer

and perhaps even compare that to

-march=pentium-m -mtune=pentium-m -minline-all-stringops -maccumulate-outgoing-args -mmmx -msse -msse2 -msse3 -mfpmath=sse -funroll-loops -fomit-frame-pointer


I've been really busy of late, but if I get a chance in the next few days I give it a go, mind you it will have to be a 32bit build, 64bit builds are still beyond me.

« Last Edit: November 28, 2007, 02:34:50 am by taz-nz »

divemaster127

  • Trade Count: (+60)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 1977
  • Last login:December 02, 2018, 08:05:08 pm
  • My webstore is arcadeemulator.net
    • arcadeemulator.net
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #116 on: November 28, 2007, 08:02:11 am »
Taz do you think it would be worth the jump to a 6850
dm
I carry both ultimarc & happ items, all brand new & I ship from the united states. My online store is ARCADEEMULATOR.NET, pm if I can help in anyway.

Tiger-Heli

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 5447
  • Last login:January 03, 2018, 02:19:23 pm
  • Ron Howard? . . . er, I mean . . . Run, Coward!!!
    • Tiger-Heli
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #117 on: November 28, 2007, 08:38:31 am »
Taz - that helped a lot.  What I am mainly looking for is the ability to play some of the more demanding games like Cruisin' and SFrush.  Based on your specs (150-200% in these games), I am hoping a E2100 at 3 Ghz or so might give 75% performance, which would still be around 100% (and I could live with less than 100%, but 60% is not playable).  I don't plan to go to 64-bit, and understand that it's not a totally linear scale.

I just have a couple of additional questions, but your replies already helped a lot!

As far as coolers, any idea how the Arctic Cooling Freezer 7 Pro would compare with the others?

Quote
but make sure you run dual channel no matter how much memory you run.
Still true with XP?  (I've read reviews that said dual-channel wasn't that important and was thinking of getting a single 1GB stick to keep some ram slots free - not sure the reviews were of dual-core systems, though).

Quote
Any half respectable DDR-800 ram will do just make sure it's compatible with the P35 chipset, there are a few Dimms out there that work great on the older Intel chipsets and with AM2 CPUs but are totally unstable withe P35 chipset.
How can you tell? (Or do you just read tons of reviews?)
It's not what you take when you leave this world behind you, it's what you leave behind you when you go. - R. Travis.
When all is said and done, generally much more is SAID than DONE.

TheManuel

  • Trade Count: (0)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 805
  • Last login:August 07, 2020, 10:15:48 am
  • On and off hobbyist
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #118 on: November 28, 2007, 09:31:04 am »
Quote
Quote
but make sure you run dual channel no matter how much memory you run.
Still true with XP?  (I've read reviews that said dual-channel wasn't that important and was thinking of getting a single 1GB stick to keep some ram slots free - not sure the reviews were of dual-core systems, though).


TigerHeli:

I have run my computer with and without dual channel configuration and it makes no difference whatsoever even if the bandwidth is improved in synthetic benchmarks.  I tested it with mace and with the "Splinter Cell: Pandora Tomorrow" computer game which I have installed merely for benchmarks as I don't play PC games (or mace for that matter but it makes for nice benchmarks).

My PC is an e4300 which I am running at 300FSB for 2.7GHz although I ran it at 3.0GHz for a week or so before it became unstable.  All of this was with stock voltage and cooler.  The motherboard is Abit IB9 which is the Intel P965 chipset.  Memory is 2GB DDR2-667 running stock although I can run it at DDR2-800 but I don't like to push the memory long term as I don't get any benefit.  Video card: old Geforce PX-6200TD unlocked to 4 pixel pipelines.

If you are interested, I can run a couple of configurations that you might want to consider as long as they don't exceed 3.0GHz for the CPU or 800MHz for the memory since I don't like to mess around with voltages,  to help you make up your mind about what you need to purchase.

If you don't plan to increase voltages, I don't think you need any fancy coolers.  My CPU runs normally with a 67C reserve to the maximum Tjunction temperature according to CoreTemp.  Under Orthos load, the reserve can get as low as 32C which is a bit scary but still safe when you consider that most normal applications will not stress the CPU that much even at 100% CPU load as reported by the Task Manager.  In the summer, the reserve is about 5C shorter due to the temperature difference inside my house.  What you need to make sure it that you seat cooler properly which is not easy to do the Core 2 Duo's socket layout.  I use Arctic Silver because it is cheap but I don't know if it makes a difference vs the stock paste.

Finally, to extend the life of the processor, I set my power scheme to laptop/portable so that EIST kicks in and the CPU multiplier is knocked down to 6X form 9X while the computer is idle or doing light duty like browsing the web or playing pacman :-) but immediately goes to full throttle when more processing power is required.  As a result, my 2.7GHz is running at 1.8GHz most of the time which happens to be the stock speed.

Good luck and let me know if you need help.
"The Manuel"

divemaster127

  • Trade Count: (+60)
  • Full Member
  • ***
  • Offline Offline
  • Posts: 1977
  • Last login:December 02, 2018, 08:05:08 pm
  • My webstore is arcadeemulator.net
    • arcadeemulator.net
Re: 4Ghz Core 2 Duo vs M.A.M.E. 0.120 (benchmark results)
« Reply #119 on: November 28, 2007, 10:14:20 am »
I am running the Rosewill Z-3 coolers I have 3 of these & they work great, the artic cooler you are looking at is fine.  But with the tags on it, it makes it a pain to install these coolers are not good for being used over & over.  That is why I like the rosewill or similar they use a back plate w/screws & nuts.  I am kicking around dumping my 6750 & going to the 6850 just for the extra performance.  I also prefer Gigabyte motherboards I have 3 of the GA-P35-DS3L, these boards are tough & great overclockers.
dm
I carry both ultimarc & happ items, all brand new & I ship from the united states. My online store is ARCADEEMULATOR.NET, pm if I can help in anyway.