Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Highest PowerPC OpenMark score IN THE WORLD: what is it? I think I have it :)

Jethryn Freyman

Jethryn Freyman

macrumors 68020
Original poster
Aug 9, 2007
2,329
3
Australia
Curious at to what it is.

Reason is that I believe I hold it :p

Power Mac G5, [June 2004], dual 1.8GHz, 256MB ATI Fire GL X3.

j8uNd.png


Clocked and cooled by me :)

>>also: >> I did have a 22K+ score, but I can't find the screenshot :( will look for it now

----------

Oh wait, found it :)
9Xy2G.png

https://forums.macrumors.com/posts/14799408/

----------

Also, this machine is a BEAST.

Graphics on a Macbook 4 year newer with 6GB memory and an SSD, well, they suck compared to this. You can see the frames skipping. On this G5, everything is smooth as melted butter.

Even when running two displays at 2048x1152 resolution from both DVI ports.

Seriously, if you have a G5 Mac and are looking at a GPU upgrade, you can't beat a flashed Fire GL X3. These things *brutalize* 3D graphics.
 
Jethryn Freyman said:
Curious at to what it is.

Reason is that I believe I hold it :p

Power Mac G5, [June 2004], dual 1.8GHz, 256MB ATI Fire GL X3.

Image

Clocked and cooled by me :)

>>also: >> I did have a 22K+ score, but I can't find the screenshot :( will look for it now

----------

Oh wait, found it :)
Image
https://forums.macrumors.com/posts/14799408/

----------

Also, this machine is a BEAST.

Graphics on a Macbook 4 year newer with 6GB memory and an SSD, well, they suck compared to this. You can see the frames skipping. On this G5, everything is smooth as melted butter.

Even when running two displays at 2048x1152 resolution from both DVI ports.

Seriously, if you have a G5 Mac and are looking at a GPU upgrade, you can't beat a flashed Fire GL X3. These things *brutalize* 3D graphics.
Click to expand...


Mhm, Sure it is. I would like to see your score pinned against a G5 Quad w/ X1900XT. I do find it strange that, in light of your "Worlds Highest OpenMark Score" that you fail to run a slightly modified Quake 1 Engine at anywhere above 1FPS. Bravo. I would like to see a GeekBench score.
 
CAINuKe said:
Mhm, Sure it is. I would like to see your score pinned against a G5 Quad w/ X1900XT. I do find it strange that, in light of your "Worlds Highest OpenMark Score" that you fail to run a slightly modified Quake 1 Engine at anywhere above 1FPS. Bravo. I would like to see a GeekBench score.
Click to expand...
GB=~1560. But Geekbench is a *CPU* test.

Yeah, it's pretty clear that TenebraeQuake and a bunch of rushed third party texture mods are not optimized well, OpenMark is a raw, brutal, total test of GPU rendering power.

And yes, if you'd followed the Openmark score thread, you'd see that my score handily beats Radeon X1900 XTs and Quadro FX 4500s ;P

----------

And also, when it wasn't rendering multiple vispatched water textures plus many explosions with >4x intensity effects, it was rendering at >120frames/sec in an upgraded 2012 version of Quake.
:)
 
Jethryn Freyman said:
GB=~1560. But Geekbench is a *CPU* test.

Yeah, it's pretty clear that TenebraeQuake and a bunch of rushed third party texture mods are not optimized well, OpenMark is a raw, brutal, total test of GPU rendering power.

And yes, if you'd followed the Openmark score thread, you'd see that my score handily beats Radeon X1900 XTs and Quadro FX 4500s ;P

----------

And also, when it wasn't rendering multiple vispatched water textures plus many explosions with >4x intensity effects, it was rendering at >120frames/sec in an upgraded 2012 version of Quake.
:)
Click to expand...

My Geekbench:
http://browser.primatelabs.com/geekbench2/794828

I couldn't find the thread you were referring to, but it doesn't matter.
I still highly doubt your claim.

Powerbook managed a valiant 5879 in OpenMark.
Given this GPU is voltage modded and is 42%/45% (GPU/VRam)
 
Jethryn Freyman: All of your posts are really, really annoying.
 
Last edited:
I doubt that is the world record holder. The FireGL X3 is not the fastest PowerPC card.
 
Really?

I call BS, there is no way that a Dual 1.8G5, can compete with a Quad 2.5 with a XT1900 in it.

Hell, my Dual 2.7 with a 7800 rapes the hell out of any Dual 1.8. Including my own dual 1.8.
 
Openmark is BS on ppc.. I wouldent piss on that app if it was on fire.
Not possible a x800 card (agp) has legs over 7800gtx. (pcie) and I say that in a friendly manner, I have currently a x800 in a dp 2.3 and have owned a quad with fx4500 and the fx4500 blows the x800 outa the water.
 
G51989 said:
Really?

I call BS, there is no way that a Dual 1.8G5, can compete with a Quad 2.5 with a XT1900 in it.

Hell, my Dual 2.7 with a 7800 rapes the hell out of any Dual 1.8. Including my own dual 1.8.
Click to expand...

OpenMark isn't a CPU test. It's purely GPU. An example, if one could put Jethryn Freyman's X800 into a Sawtooth PowerMac G4 with a 400Mhz CPU, it'd probably still get similar speeds or speeds that would completely saturate the Sawtooth's slow APG bus. Of course a quad would beat a dual at CPU tests. But if the quad has a poor preforming GPU and the dual has a high end one, the dual would win.

goMac said:
I doubt that is the world record holder. The FireGL X3 is not the fastest PowerPC card.
Click to expand...

The FireGL X3 and the X800/X850 may not be the fastest, but they can overclock well beyond other cards' stock speeds. If other cards were overclocked to their limits, they may reach scores of 25,000 or greater and make an overclocked X800/X850/FireGL X3 look pokey.

VanneDC said:
Not possible a x800 card (agp) has legs over 7800gtx. (pcie) and I say that in a friendly manner, I have currently a x800 in a dp 2.3 and have owned a quad with fx4500 and the fx4500 blows the x800 outa the water.
Click to expand...

While the FX4500 is more powerful then a X800, Jethryn Freyman doesn't have an X800, despite what OpenMark says. A X800 or the faster X850 may come close to FX4500 powers, but the two cards are designed for different markets.
 
Intell said:
OpenMark isn't a CPU test. It's purely GPU. An example, if one could put Jethryn Freyman's X800 into a Sawtooth PowerMac G4 with a 400Mhz CPU, it'd probably still get similar speeds or speeds that would completely saturate the Sawtooth's slow APG bus. Of course a quad would beat a dual at CPU tests. But if the quad has a poor preforming GPU and the dual has a high end one, the dual would win.



The FireGL X3 and the X800/X850 may not be the fastest, but they can overclock well beyond other cards' stock speeds. If other cards were overclocked to their limits, they may reach scores of 25,000 or greater and make an overclocked X800/X850/FireGL X3 look pokey.



While the FX4500 is more powerful then a X800, Jethryn Freyman doesn't have an X800, despite what OpenMark says. A X800 or the faster X850 may come close to FX4500 powers, but the two cards are designed for different markets.
Click to expand...

My thoughts are that it is a PC Flashed X800 to be a FireGL X3. I know that you can convert GeForce cards to Quattro cards vis VBIOS flashing. I think OpenMark recognizes the GPU core as a X800 (which it was originally) and not the workstation equivalent card FireGL X3. That, or it really is a genuine FireGL X3, The core is the same as a X800 and OpenMark just treats it as such.
 
goMac said:
ATI made other faster cards for PowerPC.
Click to expand...
The X1900 for the PCI final G5 tower?
OK, yeah, it should be faster, but show me a faster score :)

VanneDC said:
Openmark is BS on ppc.. I wouldent piss on that app if it was on fire.
Click to expand...
Why?

Not possible a x800 card (agp) has legs over 7800gtx
Click to expand...
It's not, it's an overclocked FireGL X3. That's the pure hardware.

. (pcie) and I say that in a friendly manner, I have currently a x800 in a dp 2.3 and have owned a quad with fx4500 and the fx4500 blows the x800 outa the water.
Click to expand...
Thank for for the civility :),
though again, hardware-wise, what I have is a Fire GL X3, not a Radeon, and then it has a significant overclock and a [really dodgy] 'cooling system' (i use the term loosely) blasting on it while running the overclock in a 3D app.

Intell said:
OpenMark isn't a CPU test. It's purely GPU. An example, if one could put Jethryn Freyman's X800 into a Sawtooth PowerMac G4 with a 400Mhz CPU, it'd probably still get similar speeds or speeds that would completely saturate the Sawtooth's slow APG bus. Of course a quad would beat a dual at CPU tests. But if the quad has a poor preforming GPU and the dual has a high end one, the dual would win.
Click to expand...
Openmark scales well when you overclock the ATI GPU; I've got a 466MHz Digital Audio G4, when I replace this G5 and no longer need it for work, I'll stick the FireGL into the old G4 and see what happens

The FireGL X3 and the X800/X850 may not be the fastest, but they can overclock well beyond other cards' stock speeds. If other cards were overclocked to their limits, they may reach scores of 25,000 or greater and make an overclocked X800/X850/FireGL X3 look pokey.
Click to expand...
Ohhh man, that sounds like a challenge to me :)

But yes, at the moment I'm looking into some proper ways of cooling the card, I'm hoping to land an 8-core Mac Pro soonish, in that case I'll be using this G5 to see how far I can push the limits.

CAINuKe said:
My thoughts are that it is a PC Flashed X800 to be a FireGL X3.
Click to expand...
It's a PC FireGL X3 256MB, flashed with Mac ROM for either an X800 XT or X800 Pro. Either way it's got the hardware, it's all there, and I can run it at a 60% overclock.

I know that you can convert GeForce cards to Quattro cards vis VBIOS flashing. I think OpenMark recognizes the GPU core as a X800 (which it was originally) and not the workstation equivalent card FireGL X3. That, or it really is a genuine FireGL X3, The core is the same as a X800 and OpenMark just treats it as such.
Click to expand...
OK, don't know if you're right or wrong, I just don't know that much and how much ATI screwed around with their cards, but:

-I do know that this card got those benchmarks without problem at a 60% overclock, and that nobody has posted a faster PowerPC OpenMark score here.

Therefore my original claim as having the fastest raw GPU power in any PowerPC Mac in the world still stands.

;), :p
 
Jethryn Freyman said:
The X1900 for the PCI final G5 tower?
OK, yeah, it should be faster, but show me a faster score :)


Why?


It's not, it's an overclocked FireGL X3. That's the pure hardware.


Thank for for the civility :),
though again, hardware-wise, what I have is a Fire GL X3, not a Radeon, and then it has a significant overclock and a [really dodgy] 'cooling system' (i use the term loosely) blasting on it while running the overclock in a 3D app.


Openmark scales well when you overclock the ATI GPU; I've got a 466MHz Digital Audio G4, when I replace this G5 and no longer need it for work, I'll stick the FireGL into the old G4 and see what happens


Ohhh man, that sounds like a challenge to me :)

But yes, at the moment I'm looking into some proper ways of cooling the card, I'm hoping to land an 8-core Mac Pro soonish, in that case I'll be using this G5 to see how far I can push the limits.


It's a PC FireGL X3 256MB, flashed with Mac ROM for either an X800 XT or X800 Pro. Either way it's got the hardware, it's all there, and I can run it at a 60% overclock.


OK, don't know if you're right or wrong, I just don't know that much and how much ATI screwed around with their cards, but:

-I do know that this card got those benchmarks without problem at a 60% overclock, and that nobody has posted a faster PowerPC OpenMark score here.

Therefore my original claim as having the fastest raw GPU power in any PowerPC Mac in the world still stands.

;), :p
Click to expand...



Here is the big kicker, without a timestamp/TXTEdit with your name and a full desktop screenshot how can you prove that it is actually "Yours" ;D

I know that, since far, far back. ATi & Nvidia use the same GPU cores in their workstation class and gaming class graphics cards. They are just optimized to different standards. Fire being OpenGL optimised for rendering and X/HD series for various game engines.


As for cooling, I would replace the thermal paste and look on newegg for an aftermarket cooler. I am pretty sure its a standardized mounting system.
 
CAINuKe said:
Here is the big kicker, without a timestamp/TXTEdit with your name and a full desktop screenshot how can you prove that it is actually "Yours" ;D

I know that, since far, far back. ATi & Nvidia use the same GPU cores in their workstation class and gaming class graphics cards. They are just optimized to different standards. Fire being OpenGL optimised for rendering and X/HD series for various game engines.


As for cooling, I would replace the thermal paste and look on newegg for an aftermarket cooler. I am pretty sure its a standardized mounting system.
Click to expand...
:) Thanks for the tips there.

rossmosis said:
You realize that there are people who own PowerMacs that *don't* belong to this forum, right?
Click to expand...
Shhhhhh!
 
GermanyChris said:
Here is the thread..

https://forums.macrumors.com/threads/1217251/

He trounces all over all of us including the x1900's in the quads and not by a few points either 6000 to 7000..
Click to expand...

I still call shenanigans, mainly because.
Openmark is ONLY a OpenGL benchmark.
A workstation card is optimized specifically for OpenGL based applications.

This benchmark only shows one level of performance, thats it.
"Hurr Durrr, I have the highest score, check it noobs"
Geekbenchs GPU section has more validity than this post.
 
CAINuKe said:
My thoughts are that it is a PC Flashed X800 to be a FireGL X3.
Click to expand...
It's got the standard "FireGL" sticker on the fan - it was advertised as a Radeon X800 XT though, that's what System Profiler and ATI Displays call it, OpenMark called it an X800 Pro, and yeah I do know that in the past people have flashed PC FireGL X3s' with Mac X800 XT ROMs, since they're the same card more or less. Same RV420 core.[/QUOTE]

CAINuKe said:
I still call shenanigans, mainly because.
Openmark is ONLY a OpenGL benchmark.
A workstation card is optimized specifically for OpenGL based applications
Click to expand...
Yep, OK, I can settle for the title of "fastest PowerPC for OpenGL performance" :)
or something like that
 
Since cooling and an SSD will clearly affect the performance of the graphics card... :rolleyes:
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back