*DAMN R6 Forum

We were having a talk about how much RAM SK should get for his new G5 earlier on GR today.

A couple people were talking about how 512 and 768 were enough, that 2GB was too much.  

Well, we all know that the more RAM you have, the faster OSX likes it.  And sure, I know that there's a point on that curve where it's not worth buying more.  But where is that point?

I just did a little looking and found this.  This may explain why some of you have trouble with RvS as well.

I'm looking at a dual G4 867 with 1.5GB of RAM.  It's been running for a couple days, and has had a little over 14,000 page outs.  That means it had to read from the disc 14,000 times instead of RAM because RAM was full (or so the story goes).

Hmm, 1.5GB and still using virtual memory.  BTW, it's using 4.2 GB of disc space for that virtual memory (so you guys with less RAM may be using more disc space).

Ok, let's look even more close.  How much is each app running actually using (or blocking off, as the case may be).

kernel is using 99MB
WindowServer is using 59MB
Finder is using 19MB
coreservicesd is using 9MB
SystemUI is using 8MB
Dock is using 8MB
Loginwindow is using 6MB
A bunch of other core services add up to another 46MB

This is not counting stuff like ARD agent, or Palm agent, that are extras.  That means just my bare system is using up ~254MB.  This is actual RAM, btw, not counting the VM they are using.  The kernel that is using 99MB of RAM is also using 790MB of VM.  

So, 256 just to run the core let's say.  Now, I open Safari.  125MB or RAM for that (468MB VM).  hmm.  So if I leave Safari running while I'm gaming, that's another 125MB RvS can't use.

Ok, I like to leave iTunes playing often.  There's another 22MB (154MB VM).  iPhoto started down around 20MB, till I actually used it, then it jumped to 66MB (172MB VM).

I started up JA in server mode.  There was another 100MB without anyone actually playing (I'll need to see if it goes up, I'm guessing yes).

So what does this all mean?

First, the obvious.  It means turn off other apps when you are gaming.  If you are getting page outs, you are going to the disc, and that slows you down.  Give the game as much possible RAM as you can.  I don't know if it will grab more physical RAM, but give it the opportunity.

Second, check your page outs, if you have a lot of them, you need more RAM.  Page Ins aren't as important.

Third, how slow is your HD?  Since you can't turn VM off in OS X like you could in earlier OS's, it's going to access the HD.  The faster your HD, the better.  7200RPM with an 8MB cashe is good.  Don't have a 133 bus in your old computer, you can add a bus card with the HD (since almost all of us have open PCI ports).  Most people ignore their HD speed when they talk about gaming, when it's also related to performance.  For those of you trying to scrape a little more life out of an iMac, yank out those 5400RPM drives (or slower) and put in a nice fast HD.  A 100GB 7200RPM with 8MB cashe will run about $100 on sale at most computer stores.  For those with towers, put in two drives and set up a RAID when you load them.  You'll be amazed how much faster that transfers information.  

The moral of this story is, you really can't have too much RAM, and 2GB is not going overboard.  Add to that, not to neglect your HD when looking at performance.  We can't do a lot about video cards, but we can get enough RAM and a fast HD or two in our machines.  Oh, and remember to always quit Safari.  It's not just a RAM hog, but it's also a CPU hog, even when it's minimized.  

P.S.  Next time I'm running MP Server 2 and it has 16 people on it, I'll try to remember to ssh in and see how much CPU and RAM it is using.  Only way I know how to check RvS is from another machine SSH'ing to it.

Hmmm. Just my humble opinion.

I have 1.5GB and have had over 105000 pageouts in 5 days. One problem with your theory is that (to the best of my knowledge) OS X caches everything. Even if you quit an app it may still be in ram. This can be demonstrated by opening apps a second time. First time is slow. Second time is fast. Even if you have ignored an app for days it can still be cached in the memory.

OS X is very different in how it handles memory than OS 9. It will assign an appropriate ammount to the frontmost application if at all possible. Therefore even with 512MB you should still have very reasonable performance. With the exception of the G5's which seem to spike up in some apps (photoshop) once you get above 1.GB or so.

The main problems you'll witness first when running low on memory are serious paging when switching apps or using fast user switching.

cheers

Martin

how do u findout pageout info/??

Type "top" in terminal (in your utilities folder)

you can also type "uptime" to see how long your puter's been running without crash/restart.

Martin

Its very true martin, the 'on second boot' makes a massive difference, you can also get hold of utilities that 'force' all ram and CPU to your front application - it claims to make a massive difference (i tried it and wasn't very inpressed though)

I have 2.5GB RAM in my G5 and it runs everything just great.  Photoshop screams!

my apple II pwns ur g5

Well, my 128k Mac pwns your Apple II.....it is in like new condition without even the slightest blemish.....I believe at last count it is worth roughly 70 times what your Apple II is worth.

[figuring the 128k Mac at $3500 and the Apple II at $50 for you math wizards out there]

To the best of my knowledge, you are still correct.  However, that accounts for page ins.  It's reading back from the disc that is the real slowdown, the page outs, unless I've read something horribly wrong.

Revolt, instead of using TOP in the shell, you are better off using Activity Monitor in the utilities folder.  You can sort the lists with it's nice GUI interface.

Unless your doing something like professional dv editing you don't need 2.5gigs of ram, or even more than 1gig. You wasted quite a bit of money GS.

I'd say that only professionals (people who's work depends on creative programs) need more than a gig of ram

 Up until I crashed about an hour ago, I had well over 200,000 pageouts; this from about 6 days uptime, with 1 gig or RAM.

Now, all this time, I'd been compressing video ( MPEG4 ), and of course doing other shit while it compressed.

If you want to be running any background tasks ( video compression, servers, etc ), you'll want to have at least 2 gigs of RAM.

Even in my Dual 1.25 with 2 gigs of RAM ( PC2700 ), I pick up a fair amount of pageouts.

If you plan to use your machine for one thing at a time, 1GB is fine. If you want to have lots of stuff going at once - shoot for more RAM, cheapest way to give yourself a great performance boost.  

I wasted quite a bit of money???  I paid $129 for each Gigabite of RAM....that included shipping.  So, I spent a total of $258 for 2.5GB of RAM (it came with 512MB RAM).  Plus, I do a LOT of image editing in Photoshop for work (as we don't have a Macintosh at work, and I REFUSE to work on Photoshop on a PC), so I really do use the RAM....this G5 screams with 2.5GB.

No offense to you GhostSniper, but you'll find G5s scream with anything over 1GB...

I just find it so hard to believe that 1 GB has really become the minimum for what can be seen as "enough" memory.

No more than two years ago, half a gig was more than enough.

Also, just curious, how much of a difference can be seen when it's reading from the disk vs. reading from memory? Is it very noticeable?

Lastly, screw apple for not letting iMac's be upgraded substantially. There I said it.... again....

Yes.

Yeah, but too bad your 8 MHz m68k can only address 16 MB of that RAM.

Oh come on!!!  My Apple Commodore 64 Amiga still runs and it easily owns anything you guys have.  Plus, it has that cool program where it speaks what you type.  Although, you have to type everything phonetically and with some numbers thrown in there.   8)

.::|N| Noto

You would see a major difference in photoshop if you were working on a psd thats 200megs with dozens of layers, paths and some crazy masks...  If your just cleaning up digital pictures you won't see any noticeable difference.  So yes i'd say that you did waste money on 2 "extra" gigs of ram, but don't worry gs thats still better than wasting money on an imac.

I do plenty of freelance webdesign work in my spare time on my g5, and extra ram is essential.
Try using photoshop, dreamweaver, freehand, transmit, cpp, firebird & IE while leaving background programs running, such as peak 4, or even mencoder for kicks. Thats when you notice more than 768megs of ram.
I also use an edit station at work which is a dual 1.25ghz g4 with only 768megs of ram. And work with final cut on projects  that are 10gigs+ of uncompressed dv, and can still maintain great speeds and response.

I'd say for general computer use 1gig is a good ceiling, however if your going to put your computer to the limit on a daily basis buying more won't hurt.

Snipe, an imac is the ford mustang of macs, its for secretaries.

Two years ago, the fastest mac wasnt half the speed of the current Macs.  Two years ago, standard HD's were smaller too.

Things change.

I'd say 1GB would be the minimum I'd have on a Mac right now.  If I take my mac down to 512, it runs much slower, not even pushing it.  1GB would be my low end with 2GB being my high end.  1.5 being what I have, that sort of makes sense =D.

I remember my first PC.  50MHz processor, 200MB HD, and 8MB Ram.  I couldn't run anything past Windows 95 due to the lack of a math co-processor.  Also, it would take about 10 minutes to boot up since most of the HD was taken up by the OS itself.  It only contained 3 other programs.   :(

.::|N| Noto

My first Mac was a whole 8Mhz machine that maxed out at 4MB of RAM and had a HUGE 20MB hard drive that we thought we'd never fill (no shit, we really didn't think we could fill it).

I agree with Bucc.  RAM for computers is like water for athletes, you can never get enough.  I have 2 GB RAM in my new 15" powerbook and it runs soooo smooth while I have so many apps running in the background.  And as Martin said, OS X works differently with RAM than as OS 9 did.  It auto allocates the RAM to all the apps running, with the most amount of RAM allocated to the foremost active app your working with, to run smoother.  Whereas in OS 9, you can set it manually.  Its really a matter of opinion, OS X does the work for you, but you can't control it and OS 9 you can control but you have to put in the work.

Hi All

OK I reckon we should all:

Ignore pagein, pageouts when running OS X. It seems that even with heaps of ram you can't stop them as the OS is designed to use all the ram anyway.

Buy as much ram as you can afford. This has been good advice for years.

BTW that is not an excuse to put 16GBs in your Duallie G5's guys.

:-)

Martin

Tis true, i even remember Apple saying how differently osx used ram... that the system was designed to all the ram as much as possible! The More RAM the merrier, if you can afford it...

Speaking of Ram. I found out last night that my G4 Quicksilver which should have PC133 in it infact came preloaded with PC100...... Cheers apple i really appreciate it i don't think :(

As a matter of fact, I do work with some really huge files in Photoshop.  The work I'm doing for the dealership includes creating advertisments such as posters and brochures and I'm normally playing around with pictures that are 1200 to 2400 dpi.  I do this stuff at home because I just don't have enough time (or a Mac) to do it at work with all the Fleet Sales that I do.

And nice jab at the Mustang there.....but just remember, it's the only game left in town now that General Motors stopped production on the Firebird and Camaro.  Find me another car that gives you rear-wheel drive, a V-8, is made in America, and can be sold for UNDER $20k.  And the All-New 2005 Mustang is gonna knock your socks off.

talking of using RAM FCP loves the stuff. i just finnished authoring a DVD. During editing the raw DV folders we were cutting from came in at just over 120GB... There is quite a difference cutting on a duel G5 with 3Gigs of RAM and cutting on a duel G4 with 512Mb RAM... Final cut just loves memory.. i think it eats it for breakfast, lunch and tea..

.. although im sure it liked the g5 processors as well ;)

If your doing anything for print using 1200 or 2400dpi your crazy, and you shouldn't be using photoshop for any type of layout. Use a vector based program for that.
As for the difference between photoshop on mac or pc, there is hardly any (in contrast to something like flash mx).

and GS, you should know that it not a jab, its the truth. It was an economy sports car to begin with. A 60s mustang looks like the purple minivan that spaz drives in comparison to something like a GTO.

Now you've just crossed the line Jeb.  A '65 Mustang convertable is a beautiful machine.  The '68 and '69 Mustangs made a GTO look like a Buick.

(http://www.classic-cars-worldwide.com/gallery/ford-68-mustang-007.jpg)

(http://www.classic-cars-worldwide.com/gallery/ford-68-mustang-014.jpg)

(http://members.shaw.ca/altec/carads/shelby/sh1969gt350feb18.jpg)

(http://community.webshots.com/s/image3/3/14/10/27931410skqbfqbIWV_ph.jpg)

You've been smoking too much of Fassst's stash Jeb.  The Mustang didn't become an "economy" sports car until the "Mustang II", which most Mustang owners refuse to recognize as ever existing.

How you can call cars that ran with 302 BOSS, 351 Clevelands, and up to 429's economy is beyond me.  They may not has cost as much as a 'vette, but they weren't competing with 'vette's either.  The T-Bird and later the AC Cobra (which crushes a 'vette) were made to compete with it.

Are there any good programs to test the speed of your computer? Like benchmarking?

Haxial has a benchmark program for free, and there was one on the web as well (Jeb or Ace may remember the URL for it, I don't right now).

Only problem with most benchmarking apps for Mac is they don't take advantage of dual processors.  So a dual 1GHz will test at the same speed as a single 1GHz, all other things being equal.

The web site was cool, because you could track your mac against others of the same type, or against other people you know (which is how I know my iBook scored better than Ace's powerbook, though just, and it is 300MHz faster) =D.

XBench (http://www.xbench.com/)

By far the best Mac benchmarking program I have seen.

What does everyone score on Xbench?

I use xbench to get this data, can anyone decipher this for me:


Results   99.97   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.1 (7C107)
      Physical RAM      512 MB
      Model      PowerBook3,5
      Processor      PowerPC G4 @ 1.00 GHz
         Version      7455 (Apollo) v3.3
         L1 Cache      32K (instruction), 32K (data)
         L2 Cache      256K @ 1.00 GHz
         L3 Cache      1024K @ 201 MHz
         Bus Frequency      134 MHz
      Video Card      ATY,RV250M9
      Drive Type      FUJITSU MHS2060AT
   CPU Test   120.63   
      GCD Loop   116.38   4.54 Mops/sec
      Floating Point Basic   122.53   443.12 Mflop/sec
      AltiVec Basic   122.42   3.56 Gflop/sec
      vecLib FFT   123.87   1.92 Gflop/sec
      Floating Point Library   118.26   4.73 Mops/sec
   Thread Test   87.35   
      Computation   63.65   859.31 Kops/sec, 4 threads
      Lock Contention   139.18   1.75 Mlocks/sec, 4 threads
   Memory Test   99.21   
      System   101.77   
         Allocate   647.49   422.36 Kalloc/sec
         Fill   76.52   609.11 MB/sec
         Copy   67.27   336.35 MB/sec
      Stream   96.77   
         Copy   93.36   682.45 MB/sec [altivec]
         Scale   94.14   694.76 MB/sec [altivec]
         Add   98.59   630.95 MB/sec [altivec]
         Triad   101.46   619.93 MB/sec [altivec]
   Quartz Graphics Test   143.93   
      Line   116.32   2.96 Klines/sec [50% alpha]
      Rectangle   139.30   9.80 Krects/sec [50% alpha]
      Circle   140.58   3.24 Kcircles/sec [50% alpha]
      Bezier   129.29   1.40 Kbeziers/sec [50% alpha]
      Text   242.94   3.96 Kchars/sec
   OpenGL Graphics Test   89.53   
      Spinning Squares   89.53   62.65 frames/sec
   User Interface Test   196.92   
      Elements   196.92   63.34 refresh/sec
   Disk Test   58.80   
      Sequential   61.94   
         Uncached Write   62.78   26.17 MB/sec [4K blocks]
         Uncached Write   48.94   20.04 MB/sec [256K blocks]
         Uncached Read   96.65   15.30 MB/sec [4K blocks]
         Uncached Read   55.95   22.61 MB/sec [256K blocks]
      Random   55.97   
         Uncached Write   54.90   0.82 MB/sec [4K blocks]
         Uncached Write   51.86   11.70 MB/sec [256K blocks]
         Uncached Read   58.43   0.39 MB/sec [4K blocks]
         Uncached Read   59.33   12.21 MB/sec [256K blocks]


Its all german to me. btw: This is a powerbook g4 with 512ram.

If Mustangs actually looked like minivans, I'm sure you would be all over them Jewb. Short of a nice dualie, how else are you going to haul around the rolling jelly donuts you call lady friends? Even if you managed to squeeze on of your woman's fat rolls in the Mustang, I don't see how you could drive with the shift knob permanently pushed over in the neutral spot between first and second.

I scored 194.17 with a Dual 2 G5 runnig 512 RAM. I would like to see what others with about the same computer as me are running.

Results   194.17   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      512 MB
      Model      PowerMac7,2
      Processor      PowerPC 970x2 @ 2.00 GHz
         L1 Cache      64K (instruction), 32K (data)
         L2 Cache      512K @ 2.00 GHz
         Bus Frequency      1 GHz
      Video Card      ATY,RV350
      Drive Type      Maxtor 6Y160M0
   CPU Test   137.31   
      GCD Loop   99.05   3.87 Mops/sec
      Floating Point Basic   204.99   741.33 Mflop/sec
      AltiVec Basic   88.14   2.56 Gflop/sec
      vecLib FFT   141.97   2.20 Gflop/sec
      Floating Point Library   327.69   13.12 Mops/sec
   Thread Test   188.84   
      Computation   133.56   1.80 Mops/sec, 4 threads
      Lock Contention   322.15   4.04 Mlocks/sec, 4 threads
   Memory Test   322.74   
      System   347.82   
         Allocate   754.06   491.87 Kalloc/sec
         Fill   306.62   2440.71 MB/sec
         Copy   247.67   1238.37 MB/sec
      Stream   301.03   
         Copy   253.86   1855.73 MB/sec [G5]
         Scale   263.83   1947.08 MB/sec [G5]
         Add   351.72   2250.99 MB/sec [G5]
         Triad   368.30   2250.34 MB/sec [G5]
   Quartz Graphics Test   246.27   
      Line   256.57   6.53 Klines/sec [50% alpha]
      Rectangle   221.48   15.58 Krects/sec [50% alpha]
      Circle   218.38   5.03 Kcircles/sec [50% alpha]
      Bezier   228.08   2.48 Kbeziers/sec [50% alpha]
      Text   341.70   5.57 Kchars/sec
   OpenGL Graphics Test   168.83   
      Spinning Squares   168.83   118.15 frames/sec
   User Interface Test   322.42   
      Elements   322.42   103.71 refresh/sec
   Disk Test   137.19   
      Sequential   141.41   
         Uncached Write   163.72   68.24 MB/sec [4K blocks]
         Uncached Write   133.58   54.70 MB/sec [256K blocks]
         Uncached Read   135.14   21.39 MB/sec [4K blocks]
         Uncached Read   137.13   55.41 MB/sec [256K blocks]
      Random   133.21   
         Uncached Write   160.93   2.41 MB/sec [4K blocks]
         Uncached Write   182.70   41.21 MB/sec [256K blocks]
         Uncached Read   99.67   0.66 MB/sec [4K blocks]
         Uncached Read   120.38   24.77 MB/sec [256K blocks]

Some1 tell me what it means if i have 0 page outs? Im gettin lots of pageins but 0 pageouts. Does this mean im not sending/uploading anything?

I scored 153.86 with a Single 1.8GHz G5 with 2.5GB RAM.

Few notes here.

Did some checking on |MP|Server 2, which has been running RvS as a server.

It's been up for 10 days now, running RvS often, and has only had a little over 4000 page outs.  This is a very good thing.

RvS with 15 players on was also using between 70% and 87% of the CPU.  Never peaking out like Ghost Recon had.  It was also using 217MB or RAM and just under 400MB of VM.

MP Server 2 has been the second best host on GR that I know of (Ace's Dual G5 on his 10MBs connection kicking it's ass of course).

Other things.  

Voo, I'll try to explain what the results mean as I go through these, and you can apply them to yours.

First, the numbers themselves mean little.  They tell you how fast your computer is, but you need to put it in context.  You need to compare your numbers to those of other machines.

I just ran xBench on two machines here.  They are close to the same spec, though one is a 933 with 2MB L3, and the other is a dual 867 with 1MB L3.  

Over all scores were:
867 - 121
933 - 114
Voo - 100
*Makes sense that the scores are close, and that while having a faster processor, Voo's is in 3rd.

CPU Test:
867 - 107
933 - 115
Voo - 120.63
* This relates right to MHz, though cashe make a little difference too.  Notice it doesn't really take the dual CPU into account here.
*The numbers it gives here are how many millions (or billions) of operations per second your processor runs.  Along with break downs of different operations.

Thread Test
867 - 120.31
933 - 83.41
Voo - 87.35
*Now one of the two tests takes into account multiple processors, which is why the 867 moves up.

Memory Test
867 - 136
933 - 132
Voo - 99
*Since the BUS speeds and RAM speeds are very close on mine, this result is also very close.  Voo's laptop using slower RAM gets worse marks on this.

Quartz Graphics Test
867 - 135
933 - 125
Voo - 144
*This test measures how well your video card does at drawing vector graphics using Quartz.  ATI's laptop video card beat the GeForce4MX's.
*The second generation GeForce4MX with only 32MB beat out the older one with 64MB.  More on this later.

Open GL Graphics Test
867 - 95
933 - 101
Voo - 89.53
*This test measured how well each video card handled OpenGL instructions.  Here, the GeForce4's beat the ATI, and the extra MB's (933 has the older 64MB card, not the newer 32MB DDR card) let the 933 beat out the 867

User Interface Test
867 - 193
933 - 217
Voo - 196
*this is the most true video test.  Running a bunch of UI functions and seeing what the refresh rate is.  Again, the older GeForce4 MX 64MB beats out the other two video cards.  
*I will try this again later with a much lower screen resolution and see if it makes a difference.

Disk Test
867 - 99
933 - 92
Voo - 59
*here we see the difference between good HD's, and the ones they put in laptops.
*Compare the read and write rates here with the read and write rates of RAM, Then you'll understand why VM sucks and slows down your machine so much.  

So, out of the three machines we have here, the one with the least amount of MHz got the best score.  Three reasons for that.  1) it's RAM and HD are slightly faster than the 933, and much much faster than the laptop.  2) one test did take into account both processors, which made for a jump there.  3) Three?  I meant two =D.  The 933 catches up because of it's better video card and it's close numbers everywhere else.  The 1GHz laptop fails due to is much slower RAM (even though it is the same bus speed) and it's horrible HD.

Now, look at Voo's numbers for RAM and HD.  He's getting in the 600MB/sec for reading and writing to RAM.  On his HD this falls to under 1MB/sec for some (and those are some of the important ones in regards to how VM uses it).  So, in Voo's case, His HD is between 35 and 1700 times slower than his RAM (and you can bet it's closer to the latter).

On my machine I get like numbers (just higher).  My RAM fills at over 1.4 GB/sec and runs in the 800MB/sec for the other tests.  While my HD runs from 40MB/sec down to 1MB/sec.

So, as you can see, even with faster HD's, they are a poor substitute for RAM.  

Hope that explained it enough for you Voo.  If not, let me know.  Like I said, the important thing is to compare it.  You want to know why some machines run things like shit, while others run it fine, see what some of the differences here are, and maybe that will point you in the right direction.  Bitch as much as we do about video cards, our game performance is based on these other factors as well.

Oh, i forgot two of my "clanmates" are devout ford sluts.
My original analogy comparing imacs to mustangs was just fine, cheap & not the best.

Besides, how could i not post a slam at the expense of fords when ghost sniper is around, his  mission on the forums is to tell people that he sells fords, or which wars his relatives were in.

and to provide a purpose to this post so it won't be deleted as a flame,
My xbench result was 158 points i believe, 1.6ghz g5. My G4 466mhz scored a 60, and that hosts rvs just fine (dedicated).

Bucc and Jeb,

By far the best/most beautiful stang created was the '67 Hardtop.  Better looking than the Fastback or the Mach's or any other stang.  Agree?

I agree completely....and the new 2005 Mustang was built to resemble the 1967 model.....it's gonna be great!

Disagree.  Convertible is much sexier than the hard top.

Nah, I dont like convertibles, hardtop with mb a sunroof is just fine for me.  Except that the 67' stangs didnt come with sun roof...I don't even know if any cars at that time had sunroofs.

Only reason you don't like convertibles is because you could only have the top down 1 month out of the year up there in the great white north =P

2005 Mustang is hot.  They are even using some of the T-Bird interior in it =D

Plus and rather conveniently most of the last G4's could take a max of 1.5 gigs anyway... :)

OK

I just ran benchmakrs using the latest version of Xbench. I tested with no other apps open, then with a lot of other apps open but hidden, then with other apps open but visible. Here are the results.

---------------------------------------------------------------------------------------
No other apps open:

Results   202.79   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      1536 MB
      Model      PowerMac7,2
      Processor      PowerPC 970x2 @ 2.00 GHz
         L1 Cache      64K (instruction), 32K (data)
         L2 Cache      512K @ 2.00 GHz
         Bus Frequency      1 GHz
      Video Card      ATY,RV350
      Drive Type      ST3160023AS
   CPU Test   194.29   
      GCD Loop   118.08   4.61 Mops/sec
      Floating Point Basic   318.31   1.15 Gflop/sec
      AltiVec Basic   137.35   3.99 Gflop/sec
      vecLib FFT   225.48   3.50 Gflop/sec
      Floating Point Library   415.19   16.62 Mops/sec
   Thread Test   211.62   
      Computation   154.15   2.08 Mops/sec, 4 threads
      Lock Contention   337.41   4.24 Mlocks/sec, 4 threads
   Memory Test   341.40   
      System   379.21   
         Allocate   725.33   473.13 Kalloc/sec
         Fill   303.08   2412.50 MB/sec
         Copy   309.32   1546.59 MB/sec
      Stream   310.45   
         Copy   270.33   1976.14 MB/sec [G5]
         Scale   268.99   1985.12 MB/sec [G5]
         Add   354.08   2266.09 MB/sec [G5]
         Triad   378.29   2311.36 MB/sec [G5]
   Quartz Graphics Test   267.06   
      Line   259.16   6.60 Klines/sec [50% alpha]
      Rectangle   246.83   17.36 Krects/sec [50% alpha]
      Circle   254.11   5.86 Kcircles/sec [50% alpha]
      Bezier   229.41   2.49 Kbeziers/sec [50% alpha]
      Text   397.10   6.47 Kchars/sec
   OpenGL Graphics Test   179.14   
      Spinning Squares   179.14   125.36 frames/sec
   User Interface Test   338.43   
      Elements   338.43   108.86 refresh/sec
   Disk Test   105.98   
      Sequential   111.10   
         Uncached Write   133.09   55.48 MB/sec [4K blocks]
         Uncached Write   114.44   46.86 MB/sec [256K blocks]
         Uncached Read   88.04   13.94 MB/sec [4K blocks]
         Uncached Read   119.14   48.14 MB/sec [256K blocks]
      Random   101.31   
         Uncached Write   95.40   1.43 MB/sec [4K blocks]
         Uncached Write   99.79   22.51 MB/sec [256K blocks]
         Uncached Read   98.88   0.65 MB/sec [4K blocks]
         Uncached Read   112.80   23.21 MB/sec [256K blocks]

---------------------------------------------------------------------------------------------
Lots of apps open but hidden:

Results   199.95   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      1536 MB
      Model      PowerMac7,2
      Processor      PowerPC 970x2 @ 2.00 GHz
         L1 Cache      64K (instruction), 32K (data)
         L2 Cache      512K @ 2.00 GHz
         Bus Frequency      1 GHz
      Video Card      ATY,RV350
      Drive Type      ST3160023AS
   CPU Test   193.90   
      GCD Loop   118.07   4.61 Mops/sec
      Floating Point Basic   319.03   1.15 Gflop/sec
      AltiVec Basic   137.69   4.00 Gflop/sec
      vecLib FFT   221.74   3.44 Gflop/sec
      Floating Point Library   415.09   16.61 Mops/sec
   Thread Test   207.71   
      Computation   152.93   2.06 Mops/sec, 4 threads
      Lock Contention   323.62   4.06 Mlocks/sec, 4 threads
   Memory Test   342.01   
      System   379.17   
         Allocate   659.43   430.15 Kalloc/sec
         Fill   317.39   2526.43 MB/sec
         Copy   308.18   1540.92 MB/sec
      Stream   311.48   
         Copy   271.24   1982.77 MB/sec [G5]
         Scale   269.66   1990.06 MB/sec [G5]
         Add   355.16   2273.05 MB/sec [G5]
         Triad   380.09   2322.32 MB/sec [G5]
   Quartz Graphics Test   263.10   
      Line   254.17   6.47 Klines/sec [50% alpha]
      Rectangle   241.75   17.01 Krects/sec [50% alpha]
      Circle   250.04   5.76 Kcircles/sec [50% alpha]
      Bezier   227.62   2.47 Kbeziers/sec [50% alpha]
      Text   393.59   6.42 Kchars/sec
   OpenGL Graphics Test   178.48   
      Spinning Squares   178.48   124.90 frames/sec
   User Interface Test   328.08   
      Elements   328.08   105.52 refresh/sec
   Disk Test   103.50   
      Sequential   105.58   
         Uncached Write   132.10   55.06 MB/sec [4K blocks]
         Uncached Write   114.94   47.07 MB/sec [256K blocks]
         Uncached Read   75.67   11.98 MB/sec [4K blocks]
         Uncached Read   119.02   48.09 MB/sec [256K blocks]
      Random   101.51   
         Uncached Write   96.42   1.45 MB/sec [4K blocks]
         Uncached Write   99.17   22.37 MB/sec [256K blocks]
         Uncached Read   99.15   0.65 MB/sec [4K blocks]
         Uncached Read   112.80   23.22 MB/sec [256K blocks]

-----------------------------------------------------------------------------------------------------
Other apps open and visible:

Results   202.22   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      1536 MB
      Model      PowerMac7,2
      Processor      PowerPC 970x2 @ 2.00 GHz
         L1 Cache      64K (instruction), 32K (data)
         L2 Cache      512K @ 2.00 GHz
         Bus Frequency      1 GHz
      Video Card      ATY,RV350
      Drive Type      ST3160023AS
   CPU Test   194.04   
      GCD Loop   118.09   4.61 Mops/sec
      Floating Point Basic   319.28   1.15 Gflop/sec
      AltiVec Basic   137.70   4.00 Gflop/sec
      vecLib FFT   222.42   3.45 Gflop/sec
      Floating Point Library   415.06   16.61 Mops/sec
   Thread Test   206.65   
      Computation   151.11   2.04 Mops/sec, 4 threads
      Lock Contention   326.72   4.10 Mlocks/sec, 4 threads
   Memory Test   338.43   
      System   372.16   
         Allocate   641.02   418.14 Kalloc/sec
         Fill   304.24   2421.79 MB/sec
         Copy   311.11   1555.55 MB/sec
      Stream   310.31   
         Copy   270.66   1978.50 MB/sec [G5]
         Scale   268.37   1980.58 MB/sec [G5]
         Add   353.57   2262.86 MB/sec [G5]
         Triad   378.64   2313.46 MB/sec [G5]
   Quartz Graphics Test   272.37   
      Line   248.14   6.32 Klines/sec [50% alpha]
      Rectangle   236.30   16.62 Krects/sec [50% alpha]
      Circle   244.81   5.64 Kcircles/sec [50% alpha]
      Bezier   224.14   2.44 Kbeziers/sec [50% alpha]
      Text   645.69   10.52 Kchars/sec
   OpenGL Graphics Test   180.82   
      Spinning Squares   180.82   126.54 frames/sec
   User Interface Test   330.61   
      Elements   330.61   106.34 refresh/sec
   Disk Test   105.91   
      Sequential   110.97   
         Uncached Write   132.44   55.21 MB/sec [4K blocks]
         Uncached Write   115.31   47.22 MB/sec [256K blocks]
         Uncached Read   87.42   13.84 MB/sec [4K blocks]
         Uncached Read   119.28   48.19 MB/sec [256K blocks]
      Random   101.29   
         Uncached Write   95.46   1.43 MB/sec [4K blocks]
         Uncached Write   100.11   22.58 MB/sec [256K blocks]
         Uncached Read   98.93   0.65 MB/sec [4K blocks]
         Uncached Read   112.15   23.08 MB/sec [256K blocks]

------------------------------------------------------------------------------------------------------

As you can see the difference in results is so close it's not funny. It is easily in the error range of the benchmark program. Therefore I don't think we have to worry about closing other apps before doing things like games, etc. However obviously if you have apps doing stuff in the background (playing itunes or rendering) you may still see some slowdown.



--------------------------------------
BTW The open apps were
Safari
Appleworks
Mellel
Filemaker Pro
Xcode
Photoshop
iTunes
Quicktime player

BFG:
Small corection.  The last G4's could run 2GB, but only since the MDD (Mirror Drive Doors).  The Quicksilvers were limited to 1.5GB.

Martin:

All nice except for your conclusion Martin.  Since the average gamer here, and the people I'm talking to are the ones that DON'T HAVE A DUAL 2GHz G5 WITH 1.5GB OF RAM!

The whole point of looking at it, and closing apps, etc, was to people having TROUBLE RUNNING THE GAME.

BTW, you also have the same amount of RAM that I use and recommend.  So I have no idea what you are basing your conclusions off of, other than maybe you didn't read the whole thing.  You have to take it in context.

Your in Detroit, your in the same boat, no?

Hey, true convertible lovers will scream down the highway at 80mph with the top down in sub-zero weather.

With the heat on full!

But only if I'm not on the motorcycle already.

(http://www.sandersonsales.com/EleanorColors/blackwhite.gif)

this my friends...is the ultimate mustang. the 1967 shelby gt 500 E "eleanor".  they start at $80,000+  

http://www.sandersonsales.com/

Thanks for that comparison Buc, you explained it to a "T" for me.

***Has his credit card ready for those speed bumped G5's***

My xbench results on a iMac 800 MHz 17" Flat Panel with 768 RAM...

Results   79.40   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      768 MB
      Model      PowerMac4,5
      Processor      PowerPC G4 @ 800 MHz
         Version      7450 (V'ger) v2.1
         L1 Cache      32K (instruction), 32K (data)
         L2 Cache      256K @ 800 MHz
         Bus Frequency      100 MHz
      Video Card      GeForce4 MX
      Drive Type      Maxtor 4D080H4
   CPU Test   91.91   
      GCD Loop   82.59   3.23 Mops/sec
      Floating Point Basic   96.97   350.69 Mflop/sec
      AltiVec Basic   98.28   2.86 Gflop/sec
      vecLib FFT   98.69   1.53 Gflop/sec
      Floating Point Library   85.67   3.43 Mops/sec
   Thread Test   65.95   
      Computation   47.97   647.60 Kops/sec, 4 threads
      Lock Contention   105.50   1.32 Mlocks/sec, 4 threads
   Memory Test   65.69   
      System   56.89   
         Allocate   466.80   304.49 Kalloc/sec
         Fill   37.18   295.94 MB/sec
         Copy   42.21   211.03 MB/sec
      Stream   77.72   
         Copy   73.66   538.48 MB/sec [altivec]
         Scale   73.01   538.83 MB/sec [altivec]
         Add   78.68   503.57 MB/sec [altivec]
         Triad   87.08   532.06 MB/sec [altivec]
   Quartz Graphics Test   89.18   
      Line   84.12   2.14 Klines/sec [50% alpha]
      Rectangle   88.56   6.23 Krects/sec [50% alpha]
      Circle   95.79   2.21 Kcircles/sec [50% alpha]
      Bezier   86.10   935.51 beziers/sec [50% alpha]
      Text   92.31   1.50 Kchars/sec
   OpenGL Graphics Test   84.21   
      Spinning Squares   84.21   58.93 frames/sec
   User Interface Test   93.85   
      Elements   93.85   30.19 refresh/sec
   Disk Test   76.03   
      Sequential   77.49   
         Uncached Write   84.51   35.23 MB/sec [4K blocks]
         Uncached Write   62.63   25.65 MB/sec [256K blocks]
         Uncached Read   104.42   16.53 MB/sec [4K blocks]
         Uncached Read   70.22   28.37 MB/sec [256K blocks]
      Random   74.62   
         Uncached Write   71.33   1.07 MB/sec [4K blocks]
         Uncached Write   77.60   17.50 MB/sec [256K blocks]
         Uncached Read   75.39   0.50 MB/sec [4K blocks]
         Uncached Read   74.45   15.32 MB/sec [256K blocks]


Anyone else with an iMac similar to mine please do the test and submit your results, I know there are a lot of people with the same computer as me.

Cutter... If i was gonna buy one... id most definatly be up for buying that one! That is one very nice car. yes please!

Oh elenor... i bet she makes one hell of a noise when u start her up! ;)

Yeah I know Bucc.

I agree about my conclusion being not very useful :-). I seem to have lost track of the original topic starting points. (Though not as much as some others I believe)

If it was a linear extrapolation it could useful be but I doubt it is.

I'll repeat with my iBook however and see what that gets. I'll post them later

Martin

Hey Buc, the following is an xbench of my G4 sawtooth 1ghz. You will see that it beat out my powerbook 1ghz G4 in all test but the disk test. If you look the brand new PB kills the sawtooth in the disk test. Do you think if I go out and buy the fastest HD i can put in it that my fps in RvS will improve?


Results   95.43   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      1024 MB
      Model      PowerMac3,1
      Processor      PowerPC G4 @ 1.00 GHz
         Version      7455 (Apollo) v2.1
         L1 Cache      32K (instruction), 32K (data)
         L2 Cache      256K @ 500 MHz
         L3 Cache      2048K @ 286 MHz
         Bus Frequency      100 MHz
      Video Card      ATY,R200
      Drive Type      QUANTUM FIREBALLP KX20.5
   CPU Test   121.47   
      GCD Loop   117.52   4.59 Mops/sec
      Floating Point Basic   121.21   438.35 Mflop/sec
      AltiVec Basic   122.68   3.56 Gflop/sec
      vecLib FFT   125.76   1.95 Gflop/sec
      Floating Point Library   120.49   4.82 Mops/sec
   Thread Test   88.35   
      Computation   63.89   862.55 Kops/sec, 4 threads
      Lock Contention   143.15   1.80 Mlocks/sec, 4 threads
   Memory Test   99.42   
      System   153.40   
         Allocate   653.72   426.42 Kalloc/sec
         Fill   194.55   1548.61 MB/sec
         Copy   77.60   388.00 MB/sec
      Stream   73.54   
         Copy   72.17   527.55 MB/sec [altivec]
         Scale   75.01   553.54 MB/sec [altivec]
         Add   73.65   471.37 MB/sec [altivec]
         Triad   73.40   448.48 MB/sec [altivec]
   Quartz Graphics Test   138.55   
      Line   111.77   2.85 Klines/sec [50% alpha]
      Rectangle   122.04   8.59 Krects/sec [50% alpha]
      Circle   129.99   3.00 Kcircles/sec [50% alpha]
      Bezier   138.28   1.50 Kbeziers/sec [50% alpha]
      Text   248.63   4.05 Kchars/sec
   OpenGL Graphics Test   100.60   
      Spinning Squares   100.60   70.40 frames/sec
   User Interface Test   209.11   
      Elements   209.11   67.26 refresh/sec
   Disk Test   45.86   
      Sequential   40.47   
         Uncached Write   34.42   14.35 MB/sec [4K blocks]
         Uncached Write   33.94   13.90 MB/sec [256K blocks]
         Uncached Read   87.30   13.82 MB/sec [4K blocks]
         Uncached Read   34.64   14.00 MB/sec [256K blocks]
      Random   52.91   
         Uncached Write   44.14   0.66 MB/sec [4K blocks]
         Uncached Write   45.68   10.30 MB/sec [256K blocks]
         Uncached Read   87.69   0.58 MB/sec [4K blocks]
         Uncached Read   50.89   10.47 MB/sec [256K blocks]

Let me know because ill go buy a new HD tomorrow. The only other that that hampers me is the 100mhz bus speed. Wish there was a way to upgrade that to 133mhz.

Voo,

All I can find on your current HD is that it's an ATA/66 drive (and probably one of the fastest of those @ 7200 RPM).  I could be wrong, but that's what I've seen.

If this is the case, unless you really want to spend money to extend the life of this Mac, I'd just save for your next Mac.

If you want to boost your HD performance, you'd have to drop in a PCI card to interface with faster HD's (ATA/100's or ATA/133's if you can use them - have to look it up).  Then I'd drop in two drives and set them up as a stripped RAID (using disk utilities), which will give you about as fast of performance as you can get on that machine without putting in a RAID card and SCSI drives (which would cost way too much).

You are talking between $100 and $200 for the PCI card and then add $200-$300 for some fast HD's, so $300 - $500 total if you want to bother.  I don't know how much of a difference it will make, especially with your 100MHz bus, but it's your money if you want to try it.

A sawtooth @ 1ghz.... what processor uprgrade card did you use voodoo? Ive put an 800mhz card into an early sawtooth 350mhz comp and i was seiously inpressed. Of course the limitations of the bus etc are still their but it genually feels like its kicked it higher that 800hz.

And who beats this? ;D

Results   15.07   
   System Info      
      Xbench Version      1.1.3
      System Version      10.2.8 (6R73)
      Physical RAM      224 MB
      Model      Power Macintosh
      Processor      PowerPC 60? @ 384 MHz
         L1 Cache      16K (instruction), 16K (data)
         L2 Cache      512K
         Bus Frequency      48 MHz
      Drive Type      IBM DGHS09Z
   CPU Test   14.27   
      GCD Loop   25.08   979.41 Kops/sec
      Floating Point Basic   61.20   221.33 Mflop/sec
      vecLib FFT   4.88   75.81 Mflop/sec
      Floating Point Library   51.91   2.08 Mops/sec
   Thread Test   25.04   
      Computation   20.46   276.22 Kops/sec, 4 threads
      Lock Contention   32.26   404.93 Klocks/sec, 4 threads
   Memory Test   7.15   
      System   6.47   
         Allocate   34.12   22.26 Kalloc/sec
         Fill   5.26   41.89 MB/sec
         Copy   4.10   20.48 MB/sec
      Stream   7.98   
         Copy   6.95   50.80 MB/sec
         Scale   6.88   50.80 MB/sec
         Add   9.22   59.02 MB/sec
         Triad   9.63   58.84 MB/sec
   Quartz Graphics Test   19.38   
      Line   12.55   319.41 lines/sec [50% alpha]
      Rectangle   19.53   1.37 Krects/sec [50% alpha]
      Circle   41.90   965.74 circles/sec [50% alpha]
      Bezier   22.90   248.81 beziers/sec [50% alpha]
      Text   16.79   273.62 chars/sec
   User Interface Test   19.47   
      Elements   19.47   6.26 refresh/sec
   Disk Test   22.14   
      Sequential   16.22   
         Uncached Write   13.78   5.75 MB/sec [4K blocks]
         Uncached Write   20.23   8.28 MB/sec [256K blocks]
         Uncached Read   15.85   2.51 MB/sec [4K blocks]
         Uncached Read   16.26   6.57 MB/sec [256K blocks]
      Random   34.84   
         Uncached Write   33.19   0.50 MB/sec [4K blocks]
         Uncached Write   32.96   7.43 MB/sec [256K blocks]
         Uncached Read   65.53   0.43 MB/sec [4K blocks]
         Uncached Read   25.59   5.27 MB/sec [256K blocks]

No OpenGL test because it crashed upon testing - well, you could see the rendered image updating while dragging the mouse over the window...

Lol. thats inpressive Elan... problem is at the  moment my comps in such a bad way i doubt it would do much better!

It was originally a G4 400. I upgraded with a Sonnet 1ghz processor chip and ATI 8500 64ram graphic  card. I have a feeling that the 100mhz bus is whats really slowing everything down. If there were only a way to over clock my bus speed to 133mhz. :(  Ah well, I hear that they will be releasing a dual G5 2.7 with even faster FSB speeds. Ill be getting one. Supposedly it will be one inch longer also to fit the new logic boards.

Yay

I did the xbench test on my iBook. 3 results as before.

----------------------------------------------------------------------
No Other Apps Open

Results   42.95   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      640 MB
      Model      PowerBook4,3
      Processor      PowerPC G3 @ 800 MHz
         Version      750FX v2.2
         L1 Cache      32K (instruction), 32K (data)
         L2 Cache      512K @ 800 MHz
         Bus Frequency      100 MHz
      Video Card      ATY,RageM7
      Drive Type      IBM-IC25N030ATCS04-0
   CPU Test   24.71   
      GCD Loop   59.10   2.31 Mops/sec
      Floating Point Basic   56.19   203.22 Mflop/sec
      vecLib FFT   8.53   132.35 Mflop/sec
      Floating Point Library   101.54   4.06 Mops/sec
   Thread Test   53.36   
      Computation   42.18   569.37 Kops/sec, 4 threads
      Lock Contention   72.62   911.58 Klocks/sec, 4 threads
   Memory Test   25.86   
      System   28.63   
         Allocate   626.46   408.64 Kalloc/sec
         Fill   28.46   226.54 MB/sec
         Copy   14.69   73.46 MB/sec
      Stream   23.58   
         Copy   21.77   159.14 MB/sec
         Scale   21.59   159.33 MB/sec
         Add   25.26   161.64 MB/sec
         Triad   26.44   161.58 MB/sec
   Quartz Graphics Test   74.14   
      Line   73.27   1.87 Klines/sec [50% alpha]
      Rectangle   66.33   4.67 Krects/sec [50% alpha]
      Circle   65.97   1.52 Kcircles/sec [50% alpha]
      Bezier   80.47   874.37 beziers/sec [50% alpha]
      Text   89.84   1.46 Kchars/sec
   OpenGL Graphics Test   57.70   
      Spinning Squares   57.70   40.38 frames/sec
   User Interface Test   104.64   
      Elements   104.64   33.66 refresh/sec
   Disk Test   40.42   
      Sequential   44.65   
         Uncached Write   39.39   16.42 MB/sec [4K blocks]
         Uncached Write   40.34   16.52 MB/sec [256K blocks]
         Uncached Read   67.25   10.65 MB/sec [4K blocks]
         Uncached Read   40.75   16.46 MB/sec [256K blocks]
      Random   36.93   
         Uncached Write   23.19   0.35 MB/sec [4K blocks]
         Uncached Write   41.19   9.29 MB/sec [256K blocks]
         Uncached Read   51.72   0.34 MB/sec [4K blocks]
         Uncached Read   46.34   9.54 MB/sec [256K blocks]


-----------------------------------------------------------------------------------
Lots of apps open but hidden

Results   40.23   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      640 MB
      Model      PowerBook4,3
      Processor      PowerPC G3 @ 800 MHz
         Version      750FX v2.2
         L1 Cache      32K (instruction), 32K (data)
         L2 Cache      512K @ 800 MHz
         Bus Frequency      100 MHz
      Video Card      ATY,RageM7
      Drive Type      IBM-IC25N030ATCS04-0
   CPU Test   20.99   
      GCD Loop   45.51   1.78 Mops/sec
      Floating Point Basic   36.96   133.67 Mflop/sec
      vecLib FFT   7.64   118.54 Mflop/sec
      Floating Point Library   94.55   3.78 Mops/sec
   Thread Test   53.20   
      Computation   41.90   565.59 Kops/sec, 4 threads
      Lock Contention   72.86   914.70 Klocks/sec, 4 threads
   Memory Test   24.07   
      System   26.20   
         Allocate   532.57   347.40 Kalloc/sec
         Fill   27.21   216.63 MB/sec
         Copy   13.18   65.89 MB/sec
      Stream   22.27   
         Copy   20.55   150.18 MB/sec
         Scale   20.51   151.38 MB/sec
         Add   23.77   152.10 MB/sec
         Triad   24.91   152.20 MB/sec
   Quartz Graphics Test   76.33   
      Line   75.69   1.93 Klines/sec [50% alpha]
      Rectangle   67.77   4.77 Krects/sec [50% alpha]
      Circle   67.87   1.56 Kcircles/sec [50% alpha]
      Bezier   87.44   950.07 beziers/sec [50% alpha]
      Text   87.95   1.43 Kchars/sec
   OpenGL Graphics Test   55.66   
      Spinning Squares   55.66   38.95 frames/sec
   User Interface Test   95.79   
      Elements   95.79   30.81 refresh/sec
   Disk Test   40.77   
      Sequential   44.27   
         Uncached Write   39.28   16.38 MB/sec [4K blocks]
         Uncached Write   40.39   16.54 MB/sec [256K blocks]
         Uncached Read   64.14   10.15 MB/sec [4K blocks]
         Uncached Read   40.74   16.46 MB/sec [256K blocks]
      Random   37.79   
         Uncached Write   23.78   0.36 MB/sec [4K blocks]
         Uncached Write   42.53   9.59 MB/sec [256K blocks]
         Uncached Read   52.14   0.34 MB/sec [4K blocks]
         Uncached Read   47.37   9.75 MB/sec [256K blocks]


-------------------------------------------------------------------------------

Lots of apps open and visible

Results   40.77   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      640 MB
      Model      PowerBook4,3
      Processor      PowerPC G3 @ 800 MHz
         Version      750FX v2.2
         L1 Cache      32K (instruction), 32K (data)
         L2 Cache      512K @ 800 MHz
         Bus Frequency      100 MHz
      Video Card      ATY,RageM7
      Drive Type      IBM-IC25N030ATCS04-0
   CPU Test   21.99   
      GCD Loop   59.89   2.34 Mops/sec
      Floating Point Basic   35.03   126.70 Mflop/sec
      vecLib FFT   7.89   122.47 Mflop/sec
      Floating Point Library   101.05   4.04 Mops/sec
   Thread Test   52.33   
      Computation   40.18   542.48 Kops/sec, 4 threads
      Lock Contention   74.98   941.27 Klocks/sec, 4 threads
   Memory Test   24.43   
      System   26.25   
         Allocate   615.52   401.50 Kalloc/sec
         Fill   26.04   207.26 MB/sec
         Copy   13.46   67.32 MB/sec
      Stream   22.84   
         Copy   21.09   154.20 MB/sec
         Scale   21.04   155.31 MB/sec
         Add   24.41   156.20 MB/sec
         Triad   25.51   155.85 MB/sec
   Quartz Graphics Test   75.95   
      Line   79.02   2.01 Klines/sec [50% alpha]
      Rectangle   68.08   4.79 Krects/sec [50% alpha]
      Circle   66.80   1.54 Kcircles/sec [50% alpha]
      Bezier   84.44   917.47 beziers/sec [50% alpha]
      Text   85.66   1.40 Kchars/sec
   OpenGL Graphics Test   55.55   
      Spinning Squares   55.55   38.87 frames/sec
   User Interface Test   97.88   
      Elements   97.88   31.48 refresh/sec
   Disk Test   40.38   
      Sequential   44.44   
         Uncached Write   40.70   16.97 MB/sec [4K blocks]
         Uncached Write   39.20   16.05 MB/sec [256K blocks]
         Uncached Read   65.11   10.31 MB/sec [4K blocks]
         Uncached Read   40.70   16.45 MB/sec [256K blocks]
      Random   36.99   
         Uncached Write   24.04   0.36 MB/sec [4K blocks]
         Uncached Write   41.03   9.25 MB/sec [256K blocks]
         Uncached Read   52.16   0.34 MB/sec [4K blocks]
         Uncached Read   43.49   8.95 MB/sec [256K blocks]

Hopefully this might be more useful. Make of it what you will.

yeah thats the bugger with these upgrade cards.. They are only really any good up to a limit where u get a Bus bottleneck..
I still hear rumours of Duel G5 3.0Ghz By late July....

Xbench of my G5 I'll do a test of my iMac G3 and Blue&White G3 asap.

Results   196.41   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      1536 MB
      Model      PowerMac7,2
      Processor      PowerPC 970x2 @ 2.00 GHz
         L1 Cache      64K (instruction), 32K (data)
         L2 Cache      512K @ 2.00 GHz
         Bus Frequency      1 GHz
      Video Card      ATY,RV350
      Drive Type      Maxtor 6Y160M0
   CPU Test   147.13   
      GCD Loop   101.62   3.97 Mops/sec
      Floating Point Basic   317.77   1.15 Gflop/sec
      AltiVec Basic   89.81   2.61 Gflop/sec
      vecLib FFT   145.87   2.26 Gflop/sec
      Floating Point Library   332.74   13.32 Mops/sec
   Thread Test   182.97   
      Computation   127.08   1.72 Mops/sec, 4 threads
      Lock Contention   326.56   4.10 Mlocks/sec, 4 threads
   Memory Test   325.59   
      System   345.11   
         Allocate   671.26   437.86 Kalloc/sec
         Fill   277.66   2210.15 MB/sec
         Copy   277.66   1388.29 MB/sec
      Stream   308.15   
         Copy   268.61   1963.57 MB/sec [G5]
         Scale   267.26   1972.40 MB/sec [G5]
         Add   350.45   2242.89 MB/sec [G5]
         Triad   375.56   2294.66 MB/sec [G5]
   Quartz Graphics Test   254.06   
      Line   252.56   6.43 Klines/sec [50% alpha]
      Rectangle   220.19   15.49 Krects/sec [50% alpha]
      Circle   230.14   5.30 Kcircles/sec [50% alpha]
      Bezier   222.96   2.42 Kbeziers/sec [50% alpha]
      Text   425.68   6.94 Kchars/sec
   OpenGL Graphics Test   172.28   
      Spinning Squares   172.28   120.56 frames/sec
   User Interface Test   306.27   
      Elements   306.27   98.51 refresh/sec
   Disk Test   136.96   
      Sequential   140.40   
         Uncached Write   168.09   70.07 MB/sec [4K blocks]
         Uncached Write   130.16   53.30 MB/sec [256K blocks]
         Uncached Read   132.77   21.02 MB/sec [4K blocks]
         Uncached Read   136.50   55.15 MB/sec [256K blocks]
      Random   133.69   
         Uncached Write   159.28   2.39 MB/sec [4K blocks]
         Uncached Write   187.60   42.31 MB/sec [256K blocks]
         Uncached Read   100.29   0.66 MB/sec [4K blocks]
         Uncached Read   119.89   24.67 MB/sec [256K blocks]

When I did a second test without the disk tests I got:
Results   209.84   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      1536 MB
      Model      PowerMac7,2
      Processor      PowerPC 970x2 @ 2.00 GHz
         L1 Cache      64K (instruction), 32K (data)
         L2 Cache      512K @ 2.00 GHz
         Bus Frequency      1 GHz
      Video Card      ATY,RV350
      Drive Type      Maxtor 6Y160M0
   CPU Test   138.03   
      GCD Loop   98.33   3.84 Mops/sec
      Floating Point Basic   206.24   745.82 Mflop/sec
      AltiVec Basic   89.68   2.61 Gflop/sec
      vecLib FFT   142.87   2.22 Gflop/sec
      Floating Point Library   327.41   13.11 Mops/sec
   Thread Test   186.31   
      Computation   129.26   1.75 Mops/sec, 4 threads
      Lock Contention   333.48   4.19 Mlocks/sec, 4 threads
   Memory Test   326.52   
      System   348.55   
         Allocate   740.68   483.15 Kalloc/sec
         Fill   280.22   2230.56 MB/sec
         Copy   271.11   1355.57 MB/sec
      Stream   307.11   
         Copy   267.69   1956.84 MB/sec [G5]
         Scale   266.07   1963.57 MB/sec [G5]
         Add   349.60   2237.43 MB/sec [G5]
         Triad   374.52   2288.32 MB/sec [G5]
   Quartz Graphics Test   251.80   
      Line   247.92   6.31 Klines/sec [50% alpha]
      Rectangle   212.52   14.95 Krects/sec [50% alpha]
      Circle   226.30   5.22 Kcircles/sec [50% alpha]
      Bezier   228.63   2.48 Kbeziers/sec [50% alpha]
      Text   430.07   7.01 Kchars/sec
   OpenGL Graphics Test   172.28   
      Spinning Squares   172.28   120.56 frames/sec
   User Interface Test   318.26   
      Elements   318.26   102.37 refresh/sec

So mb I'll swap the HD as soon as its worn out, or when I need some more storage.
Oh btw, last week I opened a psd file of 615Mb in photoshop and it worked so smooth...

What with the size of these posts this must be getting on for the largest thread ever!
My results... on my poor old aging G4 DP  oh the score when up 10 points when i didn't run the disk test... new HD is on its way though :D i gotta have more space!

Results   130.38   
   System Info      
      Xbench Version      1.1.3
      System Version      10.3.2 (7D24)
      Physical RAM      512 MB
      Model      PowerMac3,5
      Processor      PowerPC G4x2 @ 1.00 GHz
         Version      7455 (Apollo) v2.1
         L1 Cache      32K (instruction), 32K (data)
         L2 Cache      256K @ 1.00 GHz
         L3 Cache      2048K @ 250 MHz
         Bus Frequency      134 MHz
      Video Card      GeForce4 MX
      Drive Type      ST380021A
   CPU Test   123.60   
      GCD Loop   123.69   4.83 Mops/sec
      Floating Point Basic   124.34   449.67 Mflop/sec
      AltiVec Basic   123.28   3.58 Gflop/sec
      vecLib FFT   122.53   1.90 Gflop/sec
      Floating Point Library   124.17   4.97 Mops/sec
   Thread Test   138.24   
      Computation   127.94   1.73 Mops/sec, 4 threads
      Lock Contention   150.33   1.89 Mlocks/sec, 4 threads
   Memory Test   135.81   
      System   198.81   
         Allocate   586.95   382.87 Kalloc/sec
         Fill   235.81   1877.07 MB/sec
         Copy   109.34   546.72 MB/sec
      Stream   103.13   
         Copy   99.69   728.71 MB/sec [altivec]
         Scale   98.17   724.51 MB/sec [altivec]
         Add   104.96   671.74 MB/sec [altivec]
         Triad   110.61   675.85 MB/sec [altivec]
   Quartz Graphics Test   152.23   
      Line   127.14   3.24 Klines/sec [50% alpha]
      Rectangle   135.04   9.50 Krects/sec [50% alpha]
      Circle   140.15   3.23 Kcircles/sec [50% alpha]
      Bezier   142.22   1.55 Kbeziers/sec [50% alpha]
      Text   293.47   4.78 Kchars/sec
   OpenGL Graphics Test   108.57   
      Spinning Squares   108.57   75.98 frames/sec
   User Interface Test   222.25   
      Elements   222.25   71.49 refresh/sec
   Disk Test   93.27   
      Sequential   98.44   
         Uncached Write   98.62   41.11 MB/sec [4K blocks]
         Uncached Write   96.85   39.66 MB/sec [256K blocks]
         Uncached Read   99.55   15.76 MB/sec [4K blocks]
         Uncached Read   98.78   39.91 MB/sec [256K blocks]
      Random   88.61   
         Uncached Write   85.53   1.28 MB/sec [4K blocks]
         Uncached Write   82.47   18.60 MB/sec [256K blocks]
         Uncached Read   90.61   0.60 MB/sec [4K blocks]
         Uncached Read   97.22   20.01 MB/sec [256K blocks]

Question, well, actually two of them.

1) What is the difference between a page-in & page-out
2) if ive got 24000 pageins in a 2 hour period of playing Halo, and 2000 some odd page-outs, does that bode well? ill? or does it mean my computer is slowly ticking down to total self destruct?

Im on a G4 1ghz 1gig ram 15" laptop.  

My suspicion (and humble opinion) is that your puter has therefore cached heaps of stuff in memory which has been moved to hard drive (pageins) and not asked for much back (page outs). Probably been caching Halo stuff which is not used again. (Assuming you are progressing through the game :-)

Ah ok, makes sense.  Anyone else who knows for certainty feel free (not to slight you Martin, just curious)  

And as of now, im making zero progress through the game, on Legendary, things get a weeeee bit difficult, so its more loading the last checkpoint over and over and over and...etc.  (stuck on Truth & Rec, trying to battle my way into the Covenant ship, but all other levels are completed on Legendary)

Actually tho, at the time i was playing Multiplayer, on just one server the entire time.  (This is because, *gasp* there were no TKing assholes or flat-out assholes in the game, so life was good!)

ahh yeah im stuck in the same place - or rather ive got onto the ship but i can't get past the ambush when the come flooding into the transporter room... Its those big invisible dudes with the swords.. just to damn many of them!