davide445

I did understand USB Type C need to implement an "alt mode" to carry on PCIe signals. This will need a laptop vendor to use extra hw components and not just sw? If a laptop does have a type c port that didn't implement the needed hw, a connected expansion card might be able to enable this feature?

Main problem for TB diffusion is cost, due to Intel and Apple royalties AFAIK. If the situation didn't change will be hard to see TB3 widspread on sub $1500 laptop. In that case I was thinking about the wireless Skylake connection trough WiGig WBE, does anyone have more information about his availability?

More interesting the vGRID solution for enterprise, where you can remotely access to 3D applications trough visualization solutions such as Citrix or VMware, using remote GPU power. Cloud gaming it's a different product, specific for Nvidia hardware. The same concept is also available under AMD hardware. Nobody tested it? Searching on the net there are example of usage with Maya, Autocad and others. With Nvidia building available data center in USA Northwest, USA Southwest, USA East Coast, Northern Europe, Central Europe, and Asia Pacific the latency might be acceptable. DIY eGPU problems are well described in this forum and in my understandings are: lack of general availability of laptop external interface suitable for eGPU (limited to TB right now), no official support of eGPU from OEM manufacturers and OS producers, that make it something enthusiasts or indie professional can use, but not general pro users. Limitations of cloud GPU solutions are in my understanding: need of a low latency and fast internet connection, cost structure shifting to a perpetual cost where you remain without service if you didn't want to pay, lock-in effect once you get used to this solution, dependency from third party services.

Reading about this news NVIDIA I was wondering if in the future can become a solution to overcome the lack of external interfaces for eGPU. Right now Nvidia cloud GPU appear to be limited for gaming (only Nvidia hardware as client) but curios if anyone tested a GPU cloud service provider using not just a game but a 3D application.

Searching for replacement of my aging X220 was reading about this 12.5" Dell model Latitude 12 5000 Series Thin and Light Business Laptop | Dell featuring an Smart Card reader. Can be used with ExpressCard eGPU interface?

Shift my post here due the more general topic. Reading about upcoming Intel architecture happen to notice the emphasis about wireless connectivity with peripheral. Is planned to use WiGig (IEEE 802.11ad) standard, that contains Display (WiGig WDE), Bus (WBE) and Serial (WSE) Extensions. WBE is focused on providing up to 4.7 Gbps low latency wireless access to PCIe bus, so in theory good (even if not perfect due to bandwith restriction, similar to EC) to avoid current (and probably future also) lack of TB port in standard laptop. My concerns are: - reading various presentations appear Intel focus is more on WDE, so maybe there will be not a widespread of WBE enabled laptop - in my understanding extensions will be implemented on hardware to lower power draw, so will be up to OEM to push it inside laptop - still maybe using mPCIe expansion cards (I didn't find a way to know how common are the slot in current laptop) will be possible to add WBE to standard laptop, avoiding the hassle of current eGPU cabling using this interface - there was an early WBE implementation with Dell Latitude 6430u laptop and his wireless docking station, that apparently was plagued with many problems, need to see if these will be resolved in upcoming implementations - world leader in WiGig silicon production is Wilocity, acquired by Qualcomm in July 2014, can't find in Qualcomm products info about the availability of tri-band chips to enable third party laptop - all of that is coming late 2015 and probably with availability 2016, so we need to wait What so you think about? Maybe someone does have better informations to share.

To be deleted

Interesting thanks. Appear Win notebook with TB are bulky gaming/workstation version. Can't find any mPCIe spec, no way to find this kind of select box based filter?

Planning my next laptop upgrade was looking also for the eGPU market. My interest is to find small and midrange cost Win laptop, not a mobile workstation or gaming rig, to be powered with eGPU. Can't find a real solution. Thunderbolt ports are only on costly workstation or high end laptop, also can't find any statistics about mPCIe diffusion on the market. So here my questions: - there are any reliable source to find laptops with mPCIe port, without digging into laptop spec sheet? - TB port limit is cost, that push his adoption only to high cost laptop. Did you see any reason for upcoming change in this situation? - any upcoming free standard that can bypass previous problems? Did read about USB 3.1 to PCIe adapter but appear to me a DIY solution that can't be standardized

Mmmm the only suggestion I can give you is be sure you installed the new drivers always uninstalling first the old one, and restarting the installation sequence as new one. In my case the problem (similar to yours) was only the sequence and the presence of old driver version. To be sure uninstall divers, CCC, all and restart from step 1. If didn't work hope for more expert to give help.

Meaning when you extend the screen the external monitor didn't power on and remain in standby due to no signal? From what you wrote appear your card was not correctly recognized after all. Did you install drivers downloading it directly from AMD website or using amddriverdownloader?

My suggested troubleshooting process: - try to reboot again - the monitoring process are launched inside the external monitor attached to eGPU? Since I did noticed the laptop monitor its still driven by the iGPU. - try to load the eGPU with some benchmark can let you free to monitor the activity, such as Cinebench OpenGL test

Indigo it's an unbiased renderer same as Octane. Hybrid CPU + GPU so far, the pure OpenCL version it's just passed to full time development, I did expect first public beta beginning next year. Will work on any OpenCL available device such as CPU core, multiple GPU (one reason I did go for it), first on the market if Thea didnt catch up. Did have plugins for 3dsm, Blender, C4D, Maya, Sketchup, Revit, I just tested it on C4D.

Standard GTX 650 it's stated with a 65W TDP, if you ordered the EXP GDC with the 220w Dell DC-2 power brick you didn't need a different PSU. In any case you need an external power supply since the ExpressCard interface didn't provide any power.

Just for you to know with a TDP of 145W for that specific card you don't need an ATX PSU and you can use the Dell DA-2 power brick sold with the adapter.

Hwtools table is too conservative/outdated, i.e. state HD 7950 is not suggested and I'm using it without any problem. Look at http://forum.techinferno.com/diy-e-gpu-projects/6578-implementations-hub-tb-ec-mpcie.html or Guides section if you can find others using your card or laptop and follow their experiences.

From my experience on these card: - did you first uninstall other card drivers - you need to connect the card the first time only after the laptop is up and running - after the card is recognized you need to reboot with the card already connected - didn't try to extend desktop before boot is fully complete and catalyst cc is loaded

Others more expert than me can answer if I'm wrong, but what you ask is not possible. The card in your desktop will be connected to desktop motherboard using his PCIe interface, that's exactly what you need to make it work with your laptop. So to use it you will need to disconnect from the mobo and connect to laptop using the adapters you look at, meaning to bring it outside the desktop. If that was your idea you can't also connect it using the DisplayPort interface, that's just a specialized interface to transmit graphics data and not control graphics card logic.

TDP of this card appear to be 244w and will be not at 100% all the time, so even a 350w PSU need to be enough.

My idea was not to avoid Tb but to avoid Intel to block a TB eGPU solution by limiting from beginning his potential market. Intel didnt want TB eGPU to cut his CPU market, so refuse licensing for TB. Asking to Intel TB license for a eGPU solution with limited market (ie as before: restricted by GPU type, or application used) can reassure Intel about his indirect costs of this license. Not ideal for eGPU but better than nothing. Question is if there are a technical way to really block a eGPU adapter to work with a specific GPU brand or just with some applications.

To prevent Intel ostracism you think will be possible to create some limited version of TB adapter? Such as something limited only for AMD or Nvidia GPU or a specific market (geography) or applications?

@garytyler another interested to eGPU for rendering here. My experience started with the mindset of upgrading my business subnotebook for 3D modelling, animation, VFX, rendering. For cost/benefit reason I choose an AMD card (HD 7950 Boost 3GB), that is working good with C4D, HitFilm and Indigo Renderer. Problem is I underestimated the time for the availability of OpenCL pure GPU renderer able to work with C4D, so will look next year if maintain my AMD card or go for Nvidia. How did you work with Octane? Able to offload totally rendering to GPU?

Thanks, will decide for Nvidia path beginning next year, when will be clearer the availability time of some OpenCL pure GPU renderer. In fact I didn't like the idea to pass to Nvidia since will receive less at same price: smaller bus, less shaders and much worse double precision floating performance. But currently almost all GPU renderer are CUDA only, with few exceptions that didnt work with my workflow.

My current AMD eGPU setup is working flawlessly, problem is for some projects will possibly need a CUDA GPU. Question is if anyone did find or foresee problems working with GTX 670 4GB GPU.

Not so nice to see another proprietary technology from Nvidia, but better for eGPU anyway.