MaximumBurrito

Thanks goalque, that is useful - I didn't know it was necessary to add IOPCITunnelCompatible true to IONDRVSupport.kext as well. I have edited AMDSupport per your instructions, yes. Which block of the IONDRVSupport.kext Info.plist file to I need to edit? The main text or the "1", "2", or "3" block? Power brick is the same one nando recommended. Perhaps it doesn't really deliver 120W... I can try another.

Hi all, still trying to understand why my Powercolor HD 7750 with Akitio is not showing up in OS X 10.10.2 on my 2013 Air. Here's the status: Akitio starts up when I boot the Air (Thunderbolt is connected prior to start-up, not plugged in during start-up). I'm using a 120W external power brick to provide 25W for the Akitio board + 75W for the HD 7750. The HD 7750 doesn't need any power connectors besides the PCI slot itself. HD 7750 fan starts on boot, so the card is powered. I did edit the .kexts per goalque's instructions here, and I made sure to edit the AMDRadeonX4000 .kext block for AMDVerdeGrapicsAccelerator, to match the 0x683f PCI ID of the HD 7750. I disabled .kext signing and rebuilt the cache. Checking the console boot log, there are no messages with "IOPCITunnelCompatible needed" or such. But the AMD7000Controller and AMDRadeonX4000 kexts are not loaded when I check System Report > Software > Extensions, and System Report > Graphics only shows the internal Intel gpu. I also booted using Command-S to see if I could see any boot messages about the card, but I didn't see any relevant messages. The Akito itself is showing up as a Thunderbolt device in System Report, so it's not a problem with the Akitio. It simply appears the card itself isn't showing up. Let me ask a question to those of you who might have used an eGPU for compute or rendering; if you don't edit the .kexts with <key>IOPCITunnelCompatible</key> <true>, does the card still show up in the system? Is this modification only needed if you want the eGPU to be able to drive the display? I did try to boot with the original un-edited .kexts, and the card did not show up in this case either. I'm getting close to returning the Akitio and cutting my losses. Maybe the HD 7750 just doesn't work. P.S. I got a delayed message from OWC tech support about the power extension cable mentioned in the Helios manual. I had hoped that this cable would connect the Helios' mini-4 pin PCB-mounted connector to a 6-pin graphics card connector. It turns out there IS a cable included, but it has a SATA connector at the other end, not a 6-pin. Guess the Helios doesn't offer much advantage for those looking for a plug-and-play closed-box eGPU option, unless you want to wire up your own internal connector cable.</true>

I've been trying to get a plug-and-play eGPU set-up working on a 2013 Macbook Air under OS X 10.10.2. The idea is to make a nice portable OpenCL 2.0 development and computing tool. I'd like to keep everything inside the Akitio box and do a bare minimum of modding of the .kexts and (hopefully) no modding of the actual Akitio. I was inspired by goalque's success with AMD and a Macbook Pro and AMD R9 280X and JimJ740's success with a Mac Mini and AMD R9 290X. They made it look relatively easy, but so far I've been a bit stuck myself. I recently returned an AMD R7 250X I had tried out previously (wouldn't run off Akitio's power without modification), and replaced it with an AMD HD 7750 which is able to run on the 75W provided by the Akitio. But though the card is powered (card fan is spinning) and the AMD7000Controller, AMDRadeonX4000 and AMDSupport .kexts have been edited and sudo kextcache used, the card is not showing up in system report > extensions. I have disabled .kext signing and rebooted before using sudo kextcache, and have double-checked that the .kexts are edited in the right places. I made sure to pick a card that already has a built-in Apple driver by looking up the PCI number in the .kexts (0x683f1002 for the HD 7750). Is there a specific time I need to plug in the Akitio to the Air? I've read some windows installations that required quite a bit of fussing to find the right time during boot-up to plug the Akitio's Thunderbolt cable into the Mac, but nothing about this in OS X. I don't know that this would be the problem, but I can't think of what else I might be doing wrong. Thanks in advance.

Hi Nando, I checked with OWC customer service and it seems no such cable is provided with the Helios, despite the fact that it's mentioned in that spec table in the manual. They said it only comes with the power brick and a thunderbolt cable, like the Akitio. I'm trying to get through to an OWC engineer to see if they have any appropriate cables available, but right now it looks like I'd have to wire it together myself even if I bought the Helios. This wouldn't be too big a deal, but I don't exactly know which wire to connect to which. Are the mini 4-pin leads the same as a larger 4-pin molex connector? If so, it might be possible to replace a molex-to-6-pin cable's molex connector with the smaller mini 4-pin. Hi goalque, going to a lower-power card (one that can draw enough power only through the PCIe connector and doesn't need an additional cable) seems like a good option. The problem is that I do want to get a card that supports OpenCL version 2.0, since it will make my code development easier. Per Wikipedia, the only lower-wattage AMD cards that support OpenCL 2.0 are the R7 250 you mentioned, the R7 240, and the FirePro W5100. But when I looked through the Air's .kext files to see if either of these cards' PCI numbers were there, I didn't see them. I'll doublecheck the .kexts to make sure I didn't miss a card, but those seem to be the only options based what I saw earlier. Unless you guys have some ideas on how to wire up my own cable or another lower-power OpenCL 2.0 card that could be used by the Mac, neither option looks good right now. If neither work, then it looks like I'm stuck pulling the Akitio apart and using an external power supply :/

MVC, I get that I made a dumb mistake. I did of lot of research to make sure my toaster was compatible with my toast and not enough on whether I could plug it into the wall. So, guys, help me avoid another dumb mistake. Is there a cable available that connects a mini 4 pin PCB-mounted power connector to the 6-pin connector required by the card? Not much point in modifying the Akitio to split the input power per Nando's schematic here, or getting a OWC Helios instead that already has this connector, if I can't actually get the power to the card. The whole point of this build is to keep this portable, which to me means that I can close up the case and throw it in my backpack. So the Dell DA2 solution is out, since the BPlus swex wouldn't fit into the case. The R7 250X draws 95 W max per Wikipedia, and the Akitio board draws 10W as discussed in the Dell DA2 thread, so my understanding is that it should be no problem to get this to work with an external 120W power brick plus split input power and an internal cable from the 4 pin to the card. So it's just this internal cable that's the limiting factor. I'm fine with stripping and crimping cable wires, so if I need to custom-build such a cable I would just need to know which wire to attach to which. Thanks in advance, M. Burrito

Thanks Izzard, I'm glad to know the system report warning messages aren't the issue. Thanks goalque, your guides have been very helpful in convincing me I could give eGPU for parallel computing a try. Regarding your questions, goalque, I did use kextcache after turning off kext signing, and my kexts do match your images, so it looks like they're edited correctly. So the issue must be the power supply. It looks like I had misinterpreted the Akitio board power discussion here and here and assumed that the Akitio had the additional 4-pin connector I would need to power the HIS R7 250X 6-pin power input. I think I must have looked at the picture in the thread of the OWC Helios with its extra 4-pin connector and thought it was a picture of the Akitio. Pure wishful thinking! When I got the Akitio and opened it up I realized I was wrong, but I hoped it might be possible for the card to power itself (on idle) through the PCI connector, so that I could just test whether the Air could recognize the card without having to add an external power supply or modify the Akitio. Nope! I should mention that the Akitio itself is powering up (blue light, fan) when I turn the Air on, registering as a Thunderbolt device in system report, and powering down when I turn it off, so it's not the Akitio. So at this point, it looks like my options are to either mail my Akitio to my pro-EE brother-in-law and have him kindly split the Akitio's input power as described at the bottom of the page here, or return it and get a OWC Helios instead and save him the trouble. I don't want to use an external power supply BTW, since I'd prefer to keep this as portable as possible. Thanks for the pointers, and sorry if the issue was obvious. Here's hoping I can get the mod or replacement done before the 30 day return period on the card so I can bail if it still doesn't work out!

Hi all, I've been lurking for a month or two now, reading up on how to put together an eGPU set-up for OpenCL computation and code development. I don't have a lot of EE / modding experience, so I've been trying to find a route that involves the minimum of case modification, re-wiring, and system file editing. The code I'm developing is difficult enough without putting myself through a lot of hardware setup first. I was inspired by goalque's post on 2 x AMD 280X "out-of-the-box" here to try my hand at the chEEp version. I have 2013 refurb Macbook Air, OS X 10.10.2 Akitio Thunder2 extra 120W power brick and an AMD R7 250x. Not an impressive card, I know, but I picked it 1. because it's cheap enough to make my initial jump into eGPU less risky and 2. because I compared the list of AMD PCI ID numbers at https://pci-ids.ucw.cz/read/<wbr>PC/1002 with the AMD7000Controller.kext and AMDRadeonX4000.kext plist files and saw that the 683d PCI ID number of the R7 250x was there, so it should already be supported with a built-in Apple driver on the Air. No set-up beyond kext editing required, right? I edited the .kexts and refreshed the cache per goalque's how-to (modified to edit the "Verde" portion of the AMDRadeonX4000.kext instead of the "Tahiti"), powered down and re-started with the AMD and Akitio hooked up and powered, hopefully anticipating success, and... kexts are not loading. Specifically, in About this Mac > System Report > Software > Extensions, I get "Kext signature validation error code -67030" for all the .kexts I edited, along with "Loadable: No". The weird thing is, I DID turn off .kext signing before I refreshed the cache, per JimJ740's comment here, but this didn't help. If I type nvram boot-args I get boot-args kext-dev-mode=1 , as expected, so kext signing should be turned off. I thought disabling .kext signing was supposed to get rid of the above error, but for some reason that's not the case for me. Is this something simple that I'm doing wrong? Or is there something weird about OS X 10.10.2? I can't imagine that's it, probably this is a basic n00b mistake. Thanks in advance for your help!