``Such capabilities are technically possible thanks to the high speeds of the Thunderbolt port,''
Such capabilities are technically possible, and most likely probable thanks to the high speeds of DisplayPort 1.4.
I see Apple dropping Thunderbolt in the long run.
TB ends up just being another protocol that the drivers send over the wires. DisplayPort 1.4 will end up running over the same wires. This is what I see Apple moving to, one port that automagically configures to the proper protocol based on what is hooked up at the time. Apple will stay with TB due to its ability to packetize multiple protocols so that things like a monitor can have several functions like being a hub or being a lan connection beyond its monitor function.
I say they dump Thunderbolt to no longer be tied to Intel.
And a little video card called the GTX 1080 was just released. Dare to dream . . .
In all seriousness, for a single card solution at an affordable price (meaning less than $700 for a single card), the new 1070 and 1080 are about the only game in town. If Apple were serious (and I really doubt this rumor as they've never had decent GPU's in their computers) this would coincide nicely with their release. I can see them bull**)thing about how "there has never been the power to do what we wanted to do in the graphics and gaming space . . . until now."
AMD's offerings in the 14nm space are better, significantly better in fact.
Yes they most certainly are. Most importantly is the quality of the Radeon RX 480 and its Compute power, and lower power consumption, with only 51% utilization right now, to the 98% from Nvidia's 1070/1080. In short, Apple suddenly becomes VR Extreme ready at a very low price point.
Dump those across Apple's lines and the products value jumps up heavily.
SAN JOSE, Calif. – March 1, 2016 – The Video Electronics Standards Association (VESA®) today announced it has published version 1.4 of the DisplayPort (DP) audio/video standard. The first major update to DisplayPort since version 1.3 was released in September 2014, DP 1.4 is also the first DP standard to take advantage of VESA’s Display Stream Compression (DSC) technology. DSC version 1.2 transport enables up to 3:1 compression ratio and has been deemed, through VESA membership testing, to be visually lossless. Together with other new capabilities, this makes the latest version of DP ideally suited for implementation in high-end electronic products demanding premier sound and image quality.
DisplayPort is a packet-based, extensible protocol for transporting video and audio data. Initially introduced as a new external interface, its flexibility has enabled its adaptation to embedded displays and incorporation into other connectors like the new reversible USB Type-C™ interface and Thunderbolt™. Its Multi-Stream Transport (MST) capability enables high-resolution support of multiple monitors on a single display interface.
In September 2014, VESA published DP 1.3, which has been the baseline for new system development. DP 1.3 increased the maximum link bandwidth to 32.4 Gbps, with each of four lanes running at a link rate of 8.1 Gbps/lane, a 50-percent increase over the previous DP 1.2a specification. DP 1.3 added extra protocol flexibility to enable more seamless operation over the USB Type-C connector in the form of the DisplayPort Alt Mode. The increased link rate increased the uncompressed display resolution support up to 5K x 3K (5120×2880), and also upped the MST resolution, enabling simultaneous use of two 4K UHD monitors, each with a pixel resolution of 3840×2160, when using VESA Coordinated Video Timing.
DP 1.4 further builds on the capabilities of the standard’s prior incarnations. The use of video transport compression enhances the ability to take advantage of the USB Type-C connector, enabling both high-definition video and SuperSpeed USB, while also facilitating High Dynamic Range (HDR) and 8K video across the DisplayPort or USB-C connector. Examples of increased display resolution with the new standard include 8Kp60Hz HDR deep color and 4Kp120Hz HDR deep color. Other key new features include:
Forward Error Correction – FEC, which overlays the DSC 1.2 transport, addresses the transport error resiliency needed for compressed video transport to external displays.
HDR meta transport – HDR meta transport uses the “secondary data packet” transport inherent in the DisplayPort standard to provide support for the current CTA 861.3 standard, which is useful for DP to HDMI 2.0a protocol conversion, among other examples. It also offers a flexible metadata packet transport to support future dynamic HDR standards.
Expanded audio transport – This spec extension covers capabilities such as 32 audio channels, 1536kHz sample rate, and inclusion of all known audio formats.
Makes me wonder if I could eventually just buy the Razer Blade instead and use any monitor I want with one of these new Mackbook Pro's.
In theory you could use the Razer Core with any Thunderbolt 3 equipped device. Intel eGPU (AMD XConnect) is a standard for external GPUs, but it needs proper driver and firmware support to work. Meaning that Apple would need to decide to support it.
Dave2D has a good overview of the Razer Core on YouTube.
How about using what every they have planned for the A10x?
The A9x in the iPadPro is driving a screen have the size of a of the 5K and can drive a second video stream as. They also added PCIe support in the A9 series for usb-3/lightening interface and NVMe storage. So must be getting fairly close to having enough drive to 5k of of pixels.
Add another generation on that and you have a monitor that could run lightweight standalone apps Email, web broswers, AppleTV functions when not connected. Yet plug in a Mac bang workhorse.
Plug 4 of them into a MacPro so 4 users can share the same machine and you have a brilliant set up for CAD/BIM where a small team is sharing a highly co-ordinated high volume dataset (ie Building Model). All that info stays on the central machine. All the users personal workspace stays on the iDisplay.
And a little video card called the GTX 1080 was just released. Dare to dream . . .
In all seriousness, for a single card solution at an affordable price (meaning less than $700 for a single card), the new 1070 and 1080 are about the only game in town. If Apple were serious (and I really doubt this rumor as they've never had decent GPU's in their computers) this would coincide nicely with their release. I can see them bull**)thing about how "there has never been the power to do what we wanted to do in the graphics and gaming space . . . until now."
"Never had decent GPU's"
lol ok
For the serious readers of AI, Apple uses AMD graphics because they support OpenCL, which is leveraged by Final Cut, etc. nVidia doesn't support OpenCL very well, so Apple doesn't waste their time with them.
And Apple doesn't sell these as gaming machines. When someone says "serious graphics", chances are they're someone who can't see a use for a GPU beyond a game.
How about using what every they have planned for the A10x?
The A9x in the iPadPro is driving a screen have the size of a of the 5K and can drive a second video stream as. They also added PCIe support in the A9 series for usb-3/lightening interface and NVMe storage. So must be getting fairly close to having enough drive to 5k of of pixels.
Add another generation on that and you have a monitor that could run lightweight standalone apps Email, web broswers, AppleTV functions when not connected. Yet plug in a Mac bang workhorse.
Plug 4 of them into a MacPro so 4 users can share the same machine and you have a brilliant set up for CAD/BIM where a small team is sharing a highly co-ordinated high volume dataset (ie Building Model). All that info stays on the central machine. All the users personal workspace stays on the iDisplay.
Not quite the same. The PowerVR GPU's in iOS devices render things differently, and I don't know if they even support OpenCL.
How about using what every they have planned for the A10x?
The A9x in the iPadPro is driving a screen have the size of a of the 5K and can drive a second video stream as. They also added PCIe support in the A9 series for usb-3/lightening interface and NVMe storage. So must be getting fairly close to having enough drive to 5k of of pixels.
Add another generation on that and you have a monitor that could run lightweight standalone apps Email, web broswers, AppleTV functions when not connected. Yet plug in a Mac bang workhorse.
Plug 4 of them into a MacPro so 4 users can share the same machine and you have a brilliant set up for CAD/BIM where a small team is sharing a highly co-ordinated high volume dataset (ie Building Model). All that info stays on the central machine. All the users personal workspace stays on the iDisplay.
Not quite the same. The PowerVR GPU's in iOS devices render things differently, and I don't know if they even support OpenCL.
Would it need handle OpenCL? I would think the main GPU in the host machine would handle openCL. Being able to offload the systems GPU to the external GPU should give more capacity on the main GPU for application work on the main GPU.
How about using what every they have planned for the A10x?
The A9x in the iPadPro is driving a screen have the size of a of the 5K and can drive a second video stream as. They also added PCIe support in the A9 series for usb-3/lightening interface and NVMe storage. So must be getting fairly close to having enough drive to 5k of of pixels.
Add another generation on that and you have a monitor that could run lightweight standalone apps Email, web broswers, AppleTV functions when not connected. Yet plug in a Mac bang workhorse.
Plug 4 of them into a MacPro so 4 users can share the same machine and you have a brilliant set up for CAD/BIM where a small team is sharing a highly co-ordinated high volume dataset (ie Building Model). All that info stays on the central machine. All the users personal workspace stays on the iDisplay.
5K monitor has close to 3x the pixels. 5.5 million vs 14.7 million.
This would be awesome. Even more so if you could replace the card with whatever you wanted, though that is unlikely. The $999 price would be somewhat more justified then too.
It would require design changes, and you would need drivers. I don't think it would be a simple drop in solution.
Yes, this is what I was wondering if it could happen. Solves all the problems with the rMB.
many many problems actually. Like Apple moving to integrated video on everyone except the 15" MBP.. If this is the case, we might see some people opt for lighter laptops for travel and invest more in a display.
My only concern is what video card, and I'll bet it's not upgradable. That can be a sticking point when .. since most monitors you can put on any system when you upgrade to a new laptop. This LOCKS that monitor to that generation of hardware (Thunderbolt 3 laptops, desktops, etc) only.
Really what they are doing is taking the idea of Razer Blade GPU Thunderbolt Dock and sticking it into a iMac type housing. Makes me wonder if I could eventually just buy the Razer Blade instead and use any monitor I want with one of these new Mackbook Pro's.
It may not be used as a gpu in the traditional sense. It could just be a way of getting MST to work with whatever topology Apple is using without exposing a full set of utilities. Note that Apple favored 5K over 4K, because it allows for a cleaner solution alongside existing hardware. Pixel count doubles with respect each dimension. Of course this only made it into displayport standards as of displayport 1.3, which hasn't really rolled out at this point. It's possible that Apple influenced its inclusion in displayport 1.3, but for now it's not implemented as part of a standard.
Also regarding thunderbolt, I haven't seen it on any PCI card solution.
That is just sad news. Did not know about the compatibility problems with CC. But wouldn't create more compatibility issues if there were multiple options for gpu's?
That's a distinct possibility. Open CL is Apples game Nvidia doesn't like open CL neither does Adobe for that matter. Most 3D apps and games are probably the same. I guess what Apple really needs is more open CL support.
Not quite the same. The PowerVR GPU's in iOS devices render things differently, and I don't know if they even support OpenCL.
Would it need handle OpenCL? I would think the main GPU in the host machine would handle openCL. Being able to offload the systems GPU to the external GPU should give more capacity on the main GPU for application work on the main GPU.
The implementation here turns off the system's GPU, so that wouldn't work.
This would be very cool. Just a thought: Displays have a very long life, compared to computers, don't they? I think a 5K 27" screen would be good for 10+ years. I think it's important to take this into account when creating such a groundbreaking solution. To make sure that computers 3 generations from now can connect just fine.
This would be very cool. Just a thought: Displays have a very long life, compared to computers, don't they? I think a 5K 27" screen would be good for 10+ years. I think it's important to take this into account when creating such a groundbreaking solution. To make sure that computers 3 generations from now can connect just fine.
It looks like USB-C will be around for awhile, so this could provide the "future-proofing." If a monitor with an E-GPU has a USB-C connector with DisplayPort and Thunderbolt "alternate modes" enabled, theoretically, that's all that would be needed to allow connection to many future computers. If the computer has Thunderbolt over USB-C, then delivering display info to the monitor's embedded E-GPU should be possible. If the computer has DisplayPort (1.3 or better) over USB-C, then there would be the possibility of bypassing the E-GPU. For nearer term, and to hedge bets, it would be wise to include an HDMI 2.0 connection too.
Would it need handle OpenCL? I would think the main GPU in the host machine would handle openCL. Being able to offload the systems GPU to the external GPU should give more capacity on the main GPU for application work on the main GPU.
The implementation here turns off the system's GPU, so that wouldn't work.
Only to save battery and that isn't an issue in this situation. Given the computer will be connected to a a powered monitor by a USB-c cable that could otherwise charge the computer.
Comments
Yes they most certainly are. Most importantly is the quality of the Radeon RX 480 and its Compute power, and lower power consumption, with only 51% utilization right now, to the 98% from Nvidia's 1070/1080. In short, Apple suddenly becomes VR Extreme ready at a very low price point.
Dump those across Apple's lines and the products value jumps up heavily.
http://www.vesa.org/featured-articles/vesa-publishes-displayport-standard-version-1-4/DisplayPort 1.4
Dave2D has a good overview of the Razer Core on YouTube.
The A9x in the iPadPro is driving a screen have the size of a of the 5K and can drive a second video stream as. They also added PCIe support in the A9 series for usb-3/lightening interface and NVMe storage. So must be getting fairly close to having enough drive to 5k of of pixels.
Add another generation on that and you have a monitor that could run lightweight standalone apps Email, web broswers, AppleTV functions when not connected. Yet plug in a Mac bang workhorse.
Plug 4 of them into a MacPro so 4 users can share the same machine and you have a brilliant set up for CAD/BIM where a small team is sharing a highly co-ordinated high volume dataset (ie Building Model). All that info stays on the central machine. All the users personal workspace stays on the iDisplay.
lol ok
For the serious readers of AI, Apple uses AMD graphics because they support OpenCL, which is leveraged by Final Cut, etc. nVidia doesn't support OpenCL very well, so Apple doesn't waste their time with them.
And Apple doesn't sell these as gaming machines. When someone says "serious graphics", chances are they're someone who can't see a use for a GPU beyond a game.
I would think the main GPU in the host machine would handle openCL. Being able to offload the systems GPU to the external GPU should give more capacity on the main GPU for application work on the main GPU.
It may not be used as a gpu in the traditional sense. It could just be a way of getting MST to work with whatever topology Apple is using without exposing a full set of utilities. Note that Apple favored 5K over 4K, because it allows for a cleaner solution alongside existing hardware. Pixel count doubles with respect each dimension. Of course this only made it into displayport standards as of displayport 1.3, which hasn't really rolled out at this point. It's possible that Apple influenced its inclusion in displayport 1.3, but for now it's not implemented as part of a standard.
Also regarding thunderbolt, I haven't seen it on any PCI card solution.
Just a thought: Displays have a very long life, compared to computers, don't they? I think a 5K 27" screen would be good for 10+ years. I think it's important to take this into account when creating such a groundbreaking solution. To make sure that computers 3 generations from now can connect just fine.
Given the computer will be connected to a a powered monitor by a USB-c cable that could otherwise charge the computer.