Jump to content

fermi/gtx 480/gtx 470

cmf

By cmf,
in nVidia

 Share
1,729 posts in this topic

Recommended Posts

rabbit74
rabbit74
Posted
Posted

Crashy crashy

 

Install went fine. Everything works great but...

 

When I run QuicktimeX on the second monitor after a few minutes the system locks up. It's not a KP just a lock up. My system has never crashed in over a year, otherwise.

 

Anybody else?

 

System is:

OSX-10.6.5 - GTX-470

Link to comment
Share on other sites
Maryusosx
Maryusosx
Posted
Posted
Ok so the official drivers are out any news about OpenCL and CUDA? I use my hackintosh as a workstation, i'm working with maya a lot, octane render which is CUDA, Premiere CS5 which is also CUDA, now vray RT for maya 2011 mac edition is released and is OpenCL and it works only with nvidia cards (ATi has problem with textures and a lot o bugs). Does any body know that GTX 460 with 2 GB of ram is 100% supported? Right now i have GTX 275 with 1792 MB and nvenabler doesn't fully recognize the card and the OpenCL\CUDA and OpenGL performance is about the same as a 9600GT . Can please anyone update the kext for GTX 460 2 GB plese, is possible?

 

Any body? About CUDA and OpenCL. You can test with Galaxies (OpenCL) and nBody (CUDA). I beg you.

Link to comment
Share on other sites
jijiji
jijiji
Posted
Posted

cuda works on gtx470.

nbody - 86 fps | 18.1BIPS | 361 GFLOP/s | single precision

 

no support for opencl right now as far as I know

 

*edit*

ran the benchmark (nbody)

 

> Windowed mode

> Simulation data stored in video memory

> Single precision floating point simulation

> Compute 2.0 CUDA device: [GeForce GTX 470]

14336 bodies, total time for 10 iterations: 951.536 ms

= 2.160 billion interactions per second

= 43.198 single-precision GFLOP/s at 20 flops per interaction

Link to comment
Share on other sites
Peter R.
Peter R.
Posted
Posted

Well, I installed 6.3, patched to 6.4, did [url="http://www.insanelymac.com/forum/topic/279450-why-insanelymac-does-not-support-tonymacx86/"]#####[/url], rebooted, patched to 6.5, installed Chammeleon 619, rebooted, installed the drivers and cuda, rebooted, and lo and behold, it works.

 

I still need to speed it up by doing something about sleep or speedcontrol or something, but I forget what it is, and I have to get the blasted alc 889 to work.

Link to comment
Share on other sites
georgegeorge
georgegeorge
Posted
Posted
So far confirmed working with the latest drivers are GTX 460, 465 (w/HDMI), 470, 480 (w/HDMI sound even). That accounts for all GF100 and GF104 chip cards.

Wrong. GTX 480M is not yet confirmed.

 

- System is Clevo D900F laptop

- I'm only able to boot up Snow Leopard if installing iATKOS V3 S2 disc with PCJKernel and AsereBLN v1.1.9 bootloader selected

- AquaMac installer doesn't work. Messes up the system requiring reinstallation of Snow

- Video card is GTX 480M (GTX 480 mobile!)

- Device ID: 0x06ca

 

With all that in mind, what can be done? Do CMF's or AquaMac's driver files need to be installed with the official nVidia driver or is the nVidia alone sufficient? Do I still need to swap cards? I've got a laptop and only 1 card. I don't have the option of swapping.

Link to comment
Share on other sites
Maryusosx
Maryusosx
Posted
Posted
cuda works on gtx470.

nbody - 86 fps | 18.1BIPS | 361 GFLOP/s | single precision

 

no support for opencl right now as far as I know

 

*edit*

ran the benchmark (nbody)

 

> Windowed mode

> Simulation data stored in video memory

> Single precision floating point simulation

> Compute 2.0 CUDA device: [GeForce GTX 470]

14336 bodies, total time for 10 iterations: 951.536 ms

= 2.160 billion interactions per second

= 43.198 single-precision GFLOP/s at 20 flops per interaction

 

 

10x a lot man :)

Link to comment
Share on other sites
cakesson33
cakesson33
Posted
Posted

I am uber stoked about the prospect of not having to pull my card everytime I want to boot into OSX.

 

I stumbled into this thread and I am a complete noob when it comes to kext and general modification to the underpinnings of OSX.

 

At this point there is an overwhelming amount of different instructions spread throughout this thread :) . Could some kind soul post a noob tutorial on how to get a gtx 460 recognized in OSX? I would be forever greatful! :blink:

 

I run a MacPro 3.1 (early 2008). I want to have my 8800GT as well as the added gtx 460 2gb Palit used when i Octane Render on win7 64.

 

Thanks so much!

/Christian

 

MacPro 2008 8 core | 14 GB Ram | nVidia 8800GT & nVidia GTX 460 (2GB) | Windows 7 (64) & OsX 10.6.5

Link to comment
Share on other sites
Gringo Vermelho
Gringo Vermelho
Posted
Posted

Lighten up dude. And for your own sake, grow a thicker skin.

 

I don't know how to help you with that, I can only point you to the directions already given in this thread.

 

Cakesson - use Chameleon 2.0 RC5 w/ GraphicsEnabler=y and install the latest nvidia drivers, that's all. Hopefully.

 

Having your other video card working at the same time requires (as far as I know) injection via DSDT. Use google and the forum search to find examples.

Link to comment
Share on other sites
Flyday
Flyday
Posted
Posted

Installation guide. If you do not have another video card.

1. Install 10.6.5 update http://support.apple.com/kb/DL1324, do not reboot.

2. Install Chameleon 2 RC5 http://www.insanelymac.com/forum/index.php?showtopic=231075

3. Download driver:

http://www.mediafire.com/?5sq55pw31055ra1

4. Copy all kexts to /System/Library/Extensions (replace old ones)

5. Add this to your boot plist if you haven't done so already:

<key>GraphicsEnabler</key>
	<string>Yes</string>

6. Clear caches, repair permissions and reboot.

Link to comment
Share on other sites
georgegeorge
georgegeorge
Posted
Posted
Lighten up dude. And for your own sake, grow a thicker skin.

 

I don't know how to help you with that, I can only point you to the directions already given in this thread.

I wonder how many more times you're going to keep editing that stupid post.

 

Flyday, what if I already have a bootloader installed? The one I have currently was installed automatically with iATKOS. I'm not sure how to install/uninstall a bootloader manually even though I've read much about it.

 

Is that driver you uploaded the official Nvidia one or the driver from the first post in the thread?

Link to comment
Share on other sites
Gringo Vermelho
Gringo Vermelho
Posted
Posted
I wonder how many more times you're going to keep editing that stupid post.

As many times as I feel I have to in order to get my point across. :)

I'm not sure how to install/uninstall a bootloader manually even though I've read much about it.

You don't have to do it manually if you don't feel comfortable with it. It's not difficult though, just type some commands in Terminal. But some people are scared of Terminal for some reason. You can use iFabio's automated Chameleon 2.0 RC5 installer. Use the forum search to find it.

Is that driver you uploaded the official Nvidia one or the driver from the first post in the thread?

I'm guessing neither. The mediafire upload most likely contains a modified version of the latest nvidia drivers that removes the model identifier check in the installer. Without modification the installer will only run on MacPro3,1 4,1 and 5,1 (IIRC) so if you use one of those model identifiers on your Hackintosh (You shouldn't be, since it's a laptop) you can run the installer as it is delivered by nvidia.

Link to comment
Share on other sites
georgegeorge
georgegeorge
Posted
Posted
But some people are scared of Terminal for some reason. You can use iFabio's automated Chameleon 2.0 RC5 installer. Use the forum search to find it.

Do I need to uninstall the current bootloader before running the automated installer or will it be overwritten?

 

I'm guessing neither. The mediafire upload most likely contains a modified version of the latest nvidia drivers that removes the model identifier check in the installer. Without modification the installer will only run on MacPro3,1 4,1 and 5,1 (IIRC) so if you use one of those model identifiers on your Hackintosh (You shouldn't be, since it's a laptop) you can run the installer as it is delivered by nvidia.

So what you're saying is every time Nvidia releases a new Fermi driver update, I'll have to ask someone to upload a modified version?

Link to comment
Share on other sites
Sasan
Sasan
Posted
Posted
I was not successful on my GT 430, I did following steps:

 

1-Clean installation of Mac OS 10.6 and then combo update to 10.6.5

2- Installing chameleon RC5 Rev.619 with GraphicsEnabler=Yes

3-Installed the Nvidia Official Driver.

4- In NVDAGF100Hal.kext in Info.plist, I added this line of Code according to my card

 

0x0de110de&0xffc0ffff

 

to IOPCIPrimaryMatch section.

Saved and restarted and no effect at all.

 

Here is my GT 430 info :

 

Type: GPU

Bus: PCIe

PCIe Lane Width: x16

VRAM (Total): 32 MB

Vendor: NVIDIA (0x10de)

Device ID: 0x0de1

Revision ID: 0x00a1

Displays:

Display:

Resolution: 1024 x 768

Pixel Depth: 32-Bit Color (ARGB8888)

Main Display: Yes

Mirror: Off

Online: Yes

 

 

 

Do I miss any thing? or doing any step wrong?

 

 

Any thoughts? I would really appreciate if someone help me through this...

Link to comment
Share on other sites
Gringo Vermelho
Gringo Vermelho
Posted
Posted
So what you're saying is every time Nvidia releases a new Fermi driver update, I'll have to ask someone to upload a modified version?

Yes, if you don't want to spend the time to find out how to remove the model identifier check from the installer yourself.

 

Do I need to uninstall the current bootloader before running the automated installer

No, but things could get messy if you install Chameleon to somewhere else than the same place as the older version.

Make sure you know exactly where your present version of Chameleon is before running the installer. There'll be different options depending on whether you're using GUID or MBR partition style and whether you have Windows on the same drive as OS X or not. You need to be sure about these things before installing. That said I have never used the installer, maybe it has some sort of auto detection of these things. Fire it up and see.

Link to comment
Share on other sites
georgegeorge
georgegeorge
Posted
Posted
Yes, if you don't want to spend the time to find out how to remove the model identifier check from the installer yourself.

If it's not too complicated for a newbie to learn, I'd like to know how it's done.

 

No, but things could get messy if you install Chameleon to somewhere else than the same place as the older version.

Make sure you know exactly where your present version of Chameleon is before running the installer. There'll be different options depending on whether you're using GUID or MBR partition style and whether you have Windows on the same drive as OS X or not. You need to be sure about these things before installing. That said I have never used the installer, maybe it has some sort of auto detection of these things. Fire it up and see.

Yes, that's where I'm having the problem. iATKOS installed the bootloader for me and I can't figure out where it's located. I'm not afraid of using the terminal, it's just that I always find the instructions seemed to have missed a step and can't get from one point to another.

Link to comment
Share on other sites
Gringo Vermelho
Gringo Vermelho
Posted
Posted

Unplug external drives, eject CDs and flash drives, then run Terminal, type diskutil list and post the output here.

If it's not too complicated for a newbie to learn, I'd like to know how it's done.

Basically you remove the model check from a configuration file inside the installer.mpkg.

I don't remember the details, use the forum search or google.

Link to comment
Share on other sites
majaprinz
majaprinz
Posted
Posted
Any body? About CUDA and OpenCL. You can test with Galaxies (OpenCL) and nBody (CUDA). I beg you.

Adobe Premiere CS5 recognizes my GTX 470 but it seems to be slower than my old GTX 260, it doesn´t show even one video of my canon 5dmkII without problems, with GTX 260 two videos have been possible. Perhaps i made a mistake, the numbers in the OpenGL test are around 4700-5000, like others.

Link to comment
Share on other sites
LockHead
LockHead
Posted
Posted

how did you guys manage to have 4k, 5k scores in openglviewer? mine has 500'ish only.

 

i already edited the AppleGraphicsPowerManagement kext reflecting devide id 06cd but still the same results :(

 

help please.

 

EDIT: also, how to view GPU temps :P

 

EDIT2: nevermind, i already figured for the opengl thing but i need help how to check GPU temps :D

Link to comment
Share on other sites
the clash
the clash
Posted
Posted

I've been getting random lockups with a EVGA GTX460 1gb, whereas, my old 9800GTX+ was ok under 10.6.4.

The "only" diff is the injection. I've done under dsdt on the 9800, but, used chameleon on the 460.

 

Can you guys share a working dsdt solution (with 2 dvis, 1 hdmi)?

Link to comment
Share on other sites
Jocelyn84
Jocelyn84
Posted
Posted
how did you guys manage to have 4k, 5k scores in openglviewer? mine has 500'ish only.

 

i already edited the AppleGraphicsPowerManagement kext reflecting devide id 06cd but still the same results :(

 

help please.

 

EDIT: also, how to view GPU temps :D

 

 

 

Did you run Kext Utility and reboot after adding your id?

 

Its not some simple app, but here you go if you want temps.

http://www.projectosx.com/forum/index.php?showtopic=1246

Link to comment
Share on other sites
 Share
Go to topic listing
×
×
  • Create New...