So if we can confirm this, then we maybe able to use the external TMDS to drive non-coherent displays beyond the pixel clock limits of the R9200.
What's the proper way to confirm it? I own both kinds of Minis..
Last edited:
So if we can confirm this, then we maybe able to use the external TMDS to drive non-coherent displays beyond the pixel clock limits of the R9200.
You'll have to pull them apart and look for a TMDS.What's the proper way to confirm it? I have both kinds of Minis..
Oh why did it click it!Behold the sight of the Matrox G200 Quad MMS with four discrete GPUs... and four TMDS transmitters: don't click me!
I'm not knocking its passthrough abilities, they're actually really impressive. But when I want to run Windows XP on a new system, I run VMWare because it will emulate a graphics card, and use the one I already have in any system I'm running to do so. Until QEMU can do the same, I wouldn't really count on it.Well, I mean Qemu-ppc can do PCI passthrough in a Linux host. I've fired up and run Quake3 and JK2 just fine in Mac OS X Tiger on a PCI Rage128 16MB.
It wasn't the fastest thing in the world, but it was playable. The host was a 3.75Ghz AMD Ryzen 7 1700. Qemu's integer performance was about on par with a 1.25Ghz G4 but FPU and AltiVec suffer greatly.
A Radeon 9200 PCI, or even a PCI-E graphics cards should be doable. I did pass a GeForce 6600 PCI-E to it, but we need to do some work with OpenBios to get those cards working. With SLOF I was able to boot PPCLE linux in Qemu-PPC and I did get display form the GF6600 in text console it worked perfect, but got all garbled when I tried to run X. I think I just had an improperly configured X.
Anyway, nobody other than me has ever run OpenGL games in Qemu-PPC on the Mac OS, so I just didn't continue with the work of updating OpenBios to load FCODE ROMs for later cards than the Rage128.
You can run PCI Passthough on any Mac Tower from the MacPro 3,1 to I assume the 7,1 if you are running Linux as the host OS. You can ran as many GFX or other PCI devices as you have slots for.
I also ran a PCI FireWire card and passed it to OS 9 and Tiger. Tested my FW iSight in Tiger and a FW HD in OS 9. They worked just fine.
You can't do PCIe pass thru with a PCIe slot connected to Thunderbolt?Besides, the question was in reference to if you could do it on a modern Mac, unless you have a current Mac Pro, you wouldn't even have PCIE slots to do passthrough. It's honestly amazing what you can do with QEMU, but modern Macs are mostly not built for that.
You know what? That probably would work. It still fails the clunkiness test for me, but can you tell I haven't really worked with Thunderbolt? I never even really considered it. Good Thunderbolt enclosures were pricy the last time I checked, but maybe that changed. It would be nice if that was the case.You can't do pass PCIe pass thru with a PCIe slot connected to Thunderbolt?
You'll have to pull them apart and look for a TMDS.
If you ever want to do that, snap so high res photos of both sides of both logic boards and we will help you.
I don't see a TMDS, other side please.OK, this is Mac Mini G4 1.42 GHz, bottom side.
I don't see a TMDS, other side please.
https://www.digchip.com/datasheets/parts/datasheet/435/SII1162.phpI see a Silicon Image chip there that isn't present on 1.42 GHz board
https://www.digchip.com/datasheets/parts/datasheet/435/SII1162.php
That is a TMDS transmitter. Bingo.
Nevermind that. The important bit is that it does 165 MHz pixel clock — enough for 1600×1200 using non-reduced blanking (e.g. CVT or GTF) and 1920×1200 using reduced blanking (e.g. CVT-RB or CVT-RBv2) assuming a 60 Hz refresh rate.1600 x 1200? That's strange.
That does not apply to the Silent Upgrade.So, what about "Mac Mini has a 135 MHz pixel clock limit" then. Still applies![]()
That does not apply to the Silent Upgrade.
165Mhz TMDS means 165MHz.
I'm not sure what I can do, we can try some of the later 'NDRV's for the Silent Upgrade and see if they don't have the issue I ran into that forced me to use the one from 10.3.7?( I think ).@DearthnVader , do you mind revisiting OS9 ATI drivers for Mac Mini G4 issue?![]()
For each dual-link DVI output: Two TMDS transmitters (or one internal and one external one I guess, given the GeForce 6800 GT DDL seems to have two external ones but two dual-link DVIs) but crucially, a fully wired up DVI output (pins 4, 5, 12, 13, 20, 21 in the picture below enable dual-link). Does the Mac mini G4 have that? If both transmitters can be combined to exceed 165 MHz via single-link DVI, higher-res displays with a HDMI input should™ work.What's to full bandwidth of Dual Link DVI on the 30" Apple display?
We need some high res shots of both Mini's DVI connections on the logic board.....For each dual-link DVI output: Two TMDS transmitters (or one internal and one external one I guess, given the GeForce 6800 GT DDL seems to have two external ones but two dual-link DVIs) but crucially, a fully wired up DVI output (pins 4, 5, 12, 13, 20, 21 in the picture below enable dual-link). Does the Mac mini G4 have that? If both transmitters can be combined to exceed 165 MHz via single-link DVI, higher-res displays with a HDMI input should™ work.
![]()
We need some high res shots of both Mini's DVI connections on the logic board.....
Interestingly the OM( Old Mini ) has this nice circuit pad for something, but not a TMDS as it's only 8 pins and that is not enough for any TMDS I know of.
I'm thinking it maybe for an LVDS?