Why would a top secret feature in Leopard be a top secret feature in hardware - Leopard is software. I was thinking more along the lines of a fundamentally different Core Animation dependent Finder, a completely new UI and something totally unexpected with Time Machine type of shock-factor.
As mentioned already, to take full advantage of such a chip you need OS support. Every PC maker out there could put a hardware decoding chip in the PC (motherboard or videocard). To use it besides a limited number of applications, and to guarantee future compatibility with system upgrades, patches etc. it needs OS support. OS support also means that developers will get new hardware support for free. This is significant advantage. In Wintel world it is a catch 22 situation: why add additional chip without support, or why support nonexistent chip. In the past, Microsoft was not very keen to support such kind of improvements on a timely basis (SSE, videocard acceleration etc.). This time around they will feel a pressure to do this faster. For these reason Apple would want to keep the secret closer to the release date.
In science and statistics, Error Propagation can be defined in a number of ways. One is how relatively small errors can spread through a network and cause significantly larger errors.
If you read (and research) all the comments posted to the Cringley article on his web site, you will see that there are a number of significant factual errors in his article which are progating to this and other sites. I won't enumerate all of them here. Just go and read those comments, check them out and draw your own conclusions.
In science and statistics, Error Propagation can be defined in a number of ways. One is how relatively small errors can spread through a network and cause significantly larger errors.
If you read (and research) all the comments posted to the Cringley article on his web site, you will see that there are a number of significant factual errors in his article which are progating to this and other sites. I won't enumerate all of them here. Just go and read those comments, check them out and draw your own conclusions.
I saw a number of raw simplifications, but not any standout errors of fact so incontrovertible that they cause any long-term damage to the premise of the article. Don't squint quite so hard.
That doesn't mean this is true, but having true hardware h.264 encode/decode is a huge plus if you want ANY piece of Apple hardware to talk to AppleTV without stuttering. Remember AppleTV does not do uploadable codecs, it's just a dumb box very good a playing content in some "undisclosed" format. And anyone betting that "undisclosed" format isn't h.264 is barking up the wrong tree. So how do else you get EVERYTHING stored in iTunes to the AppleTV if you need to transcode it?
I thought most current video cards these days does already have the hardware decoder for MPEG2/4. Even the integrated chipset intel G965 would include this feature. Apple probably wasn't using this feature because it was not supported with software drivers. Anyway, added hardware and software support for MPEG4/AVC is a good thing. I just hope that new apple dvd player in Leopard would allow HD-DVD/BD play back.
I looked for a few more multimedia IC beyond the 3Dlabs DMS-02 and couldn't find much. It looks like for media co-processors it's the current darling of the industry.
It encodes 720p/24 in realtime and utilizes as little as a watt. It is intended for mobile applications like phones and GPS and whatever you need good portable power.
Imagine if Apple did choose a co-processor like this. Imagine if the laptops utilized the chip for GPS display as well.
$40 in lots of 1000. I suppose buying millions would lead ot a nice price reduction. Frankly I've always liked the idea of co-processors. A dedicated chip does better than a general purpose CPU within it's respective field. Video is important now and people want to be able to playback video without slowing their machine down to a crawl.
I think the battle is how portable the frameworks like Quicktime are. If Apple wants to change to a new co-processor how much work would be involved with supporing the new chip?
Lastly it's a bit odd that we've not gotten iLife 07 announced. Apple could have easily done a special event and announced iLife and iWork but yet it's almost the middle of March and they're still not here. Perhaps these programs will indeed be Leopard only. Perhaps the 2007 line of hardware will have encode/decode dedicated chips. I'm just trying to understand why EVERY Mac has been delayed. Something is up and Apple has yet to discuss the Top Secret features.
I'm just trying to understand why EVERY Mac has been delayed.
Aside from the quad core chip for the Mac Pro, I don't think there has been an update in any of the Intel chips that go into a Mac. Since Apple is depending on Intel's notebook chips for the rest of the line-up, I would guess that maybe they are waiting for the volume roll-out of the Santa Rosa platform.
I think that any Mac Pro update might be waiting for the Stoakley chipset (or platform?) or to make a big splash to show off some hopefully major new features in Final Cut Studio next month.
Aside from the quad core chip for the Mac Pro, I don't think there has been an update in any of the Intel chips that go into a Mac. Since Apple is depending on Intel's notebook chips for the rest of the line-up, I would guess that maybe they are waiting for the volume roll-out of the Santa Rosa platform.
I think that any Mac Pro update might be waiting for the Stoakley chipset (or platform?) or to make a big splash to show off some hopefully major new features in Final Cut Studio next month.
Yes indeed the wait for Santa Rosa and Stoakley crossed my mind. I do wonder about the delays for iLife though. Apple's gotten pretty damned good at plugging leaks. It'll be interesting to see what updates they cook up with or without dedicated video processors.
Does this mean that it can decode 3 720p streams at once?
I cannot find anything that tells me how many HD streams can be decoded so I'm going to assume 1 for now.
It's nice to see OpenGL ES supported as well. Unlike a DSP it seem like it preserves alot of current programming principals. I guess Creative will be using the chip in some unnamed portable device as they of course own 3Dlabs.
The Apple TV (ATV) really does have a dearth of supported codecs which I find surprising even by Apple's standards. Since the computer does all the heavy lifting anyways the side benefit of such a media processor would be on the fly transcoding of non iTunes content into AVC for display. The ATV has 40GB hard drive which would work out nicely. I figure that if transcoding other QT CODECS into AVC can at near realtime then a quick buffering for non AVC content is a small price to pay for transparently being able to play most Quicktime video CODECS through ATV.
What secret features? The new OS won't have secret features. Do you understand why they do pre-releases of the OS? If so, then you understand why there aren't secrets.
Imagine a release of the new OS X in April which breaks Microsoft Word and Adobe Photoshop due to 'secret' features that hadn't had proper public or developer testing.
What secret features? The new OS won't have secret features. Do you understand why they do pre-releases of the OS? If so, then you understand why there aren't secrets.
Imagine a release of the new OS X in April which breaks Microsoft Word and Adobe Photoshop due to 'secret' features that hadn't had proper public or developer testing.
Apple has already done this before. Remember when they showed Aqua for the first time. No one outside of Apple had seen it yet their applications looked and functioned fine in Aqua.
Now in context imagine if Apple were to announce that Media co-processors were in fact included in every new Mac coming in 2007 and the way to take advantage is simply to use QTkit what's the issue? Apple consistently tells Developers to use the new API for "future proofing" yet many don't because they don't see the changes coming down the road. The last few years have been better with Developers getting on the ball a lot faster.
Clearly any Top Secret features that come most likely wont be in the form of a brand new API. Apple has to piggyback onto an existing API and somehow "enable" new functionality that won't be too hard for Developers to implement. Apple said "Top Secret" features are coming. They're not lying. I'm just speculating that this media co-processor could be considered a Top Secret feature that does dovetail with enhanced Leopard features.
Apple has already done this before. Remember when they showed Aqua for the first time. No one outside of Apple had seen it yet their applications looked and functioned fine in Aqua.
Now in context imagine if Apple were to announce that Media co-processors were in fact included in every new Mac coming in 2007 and the way to take advantage is simply to use QTkit what's the issue? Apple consistently tells Developers to use the new API for "future proofing" yet many don't because they don't see the changes coming down the road. The last few years have been better with Developers getting on the ball a lot faster.
Clearly any Top Secret features that come most likely wont be in the form of a brand new API. Apple has to piggyback onto an existing API and somehow "enable" new functionality that won't be too hard for Developers to implement. Apple said "Top Secret" features are coming. They're not lying. I'm just speculating that this media co-processor could be considered a Top Secret feature that does dovetail with enhanced Leopard features.
Or ZFS. Aqua was not included in the first Developer previews but was shown 3 months before the release. I am afraid it is too late for any changes which imply new interface , including new Finder interface, user controllable resolution independence or the like. It could be "under the hood" change, like co-processor support or bootable ZFS. Or just some new Application May be those Top Secret features were a stretch. Jobs likes to hyperbolize!
The fruits finally born from the Raycer Graphics acquisition perhaps?
Hmm... this means that apple has to re-design or modify Intel reference system. This would be too much work and would slow down the product release time. I some how think that many here are reading too much into the it and expecting more than what apple is willing to do. In any case, if Apple manages to add a hardware AVC/VC-1 decoder to improve HTPC experience with the factory setup, I would more than welcome the effort. Actually, it would make me want to get another mac.
I'm not clear on something; does the Apple TV have a dedicated H.264 chip or not?
IF all this hogwash were true. That may explain the TV delay. If TV does have it people would speculate that the rest of the lineup had it, and if it didn't and the rest of the lineup did have it every one will ask why Apple left it out of TV.
I think it's BS.
I do like the mention of Raycer again though. I think we all forgot about that. 8)
By the same token, if the rumor IS true and such a co-processor ships in upcoming hardware, then we're in for some extraordinary machines.
No doubt about that, but that's why I have some hard time to believe it is coming so suddenly with no hints from elsewhere. Sure, one time all we know today about multimedia processing will be museum pieces, but saying that this is going to happen so soon and abruptly is rather stretching it.
No doubt about that, but that's why I have some hard time to believe it is coming so suddenly with no hints from elsewhere. Sure, one time all we know today about multimedia processing will be museum pieces, but saying that this is going to happen so soon and abruptly is rather stretching it.
It would definitely be a surprise. To mix frozen metaphors, I won't hold my breath but I will cross my fingers.
Comments
Why would a top secret feature in Leopard be a top secret feature in hardware - Leopard is software. I was thinking more along the lines of a fundamentally different Core Animation dependent Finder, a completely new UI and something totally unexpected with Time Machine type of shock-factor.
As mentioned already, to take full advantage of such a chip you need OS support. Every PC maker out there could put a hardware decoding chip in the PC (motherboard or videocard). To use it besides a limited number of applications, and to guarantee future compatibility with system upgrades, patches etc. it needs OS support. OS support also means that developers will get new hardware support for free. This is significant advantage. In Wintel world it is a catch 22 situation: why add additional chip without support, or why support nonexistent chip. In the past, Microsoft was not very keen to support such kind of improvements on a timely basis (SSE, videocard acceleration etc.). This time around they will feel a pressure to do this faster. For these reason Apple would want to keep the secret closer to the release date.
If you read (and research) all the comments posted to the Cringley article on his web site, you will see that there are a number of significant factual errors in his article which are progating to this and other sites. I won't enumerate all of them here. Just go and read those comments, check them out and draw your own conclusions.
In science and statistics, Error Propagation can be defined in a number of ways. One is how relatively small errors can spread through a network and cause significantly larger errors.
If you read (and research) all the comments posted to the Cringley article on his web site, you will see that there are a number of significant factual errors in his article which are progating to this and other sites. I won't enumerate all of them here. Just go and read those comments, check them out and draw your own conclusions.
I saw a number of raw simplifications, but not any standout errors of fact so incontrovertible that they cause any long-term damage to the premise of the article. Don't squint quite so hard.
That doesn't mean this is true, but having true hardware h.264 encode/decode is a huge plus if you want ANY piece of Apple hardware to talk to AppleTV without stuttering. Remember AppleTV does not do uploadable codecs, it's just a dumb box very good a playing content in some "undisclosed" format. And anyone betting that "undisclosed" format isn't h.264 is barking up the wrong tree. So how do else you get EVERYTHING stored in iTunes to the AppleTV if you need to transcode it?
It encodes 720p/24 in realtime and utilizes as little as a watt. It is intended for mobile applications like phones and GPS and whatever you need good portable power.
Imagine if Apple did choose a co-processor like this. Imagine if the laptops utilized the chip for GPS display as well.
$40 in lots of 1000. I suppose buying millions would lead ot a nice price reduction. Frankly I've always liked the idea of co-processors. A dedicated chip does better than a general purpose CPU within it's respective field. Video is important now and people want to be able to playback video without slowing their machine down to a crawl.
I think the battle is how portable the frameworks like Quicktime are. If Apple wants to change to a new co-processor how much work would be involved with supporing the new chip?
Lastly it's a bit odd that we've not gotten iLife 07 announced. Apple could have easily done a special event and announced iLife and iWork but yet it's almost the middle of March and they're still not here. Perhaps these programs will indeed be Leopard only. Perhaps the 2007 line of hardware will have encode/decode dedicated chips. I'm just trying to understand why EVERY Mac has been delayed. Something is up and Apple has yet to discuss the Top Secret features.
I'm perplexed.
I'm just trying to understand why EVERY Mac has been delayed.
Aside from the quad core chip for the Mac Pro, I don't think there has been an update in any of the Intel chips that go into a Mac. Since Apple is depending on Intel's notebook chips for the rest of the line-up, I would guess that maybe they are waiting for the volume roll-out of the Santa Rosa platform.
I think that any Mac Pro update might be waiting for the Stoakley chipset (or platform?) or to make a big splash to show off some hopefully major new features in Final Cut Studio next month.
Aside from the quad core chip for the Mac Pro, I don't think there has been an update in any of the Intel chips that go into a Mac. Since Apple is depending on Intel's notebook chips for the rest of the line-up, I would guess that maybe they are waiting for the volume roll-out of the Santa Rosa platform.
I think that any Mac Pro update might be waiting for the Stoakley chipset (or platform?) or to make a big splash to show off some hopefully major new features in Final Cut Studio next month.
Yes indeed the wait for Santa Rosa and Stoakley crossed my mind. I do wonder about the delays for iLife though. Apple's gotten pretty damned good at plugging leaks. It'll be interesting to see what updates they cook up with or without dedicated video processors.
VideoStream Ports
3x high bandwidth, bi-directional digital video I/O ports
Multi-function (LCD, Camera, TV encode/decode, etc.)
Resolutions up to 1280x1024 at 24bpp
Does this mean that it can decode 3 720p streams at once?
Does this mean that it can decode 3 720p streams at once?
I cannot find anything that tells me how many HD streams can be decoded so I'm going to assume 1 for now.
It's nice to see OpenGL ES supported as well. Unlike a DSP it seem like it preserves alot of current programming principals. I guess Creative will be using the chip in some unnamed portable device as they of course own 3Dlabs.
The Apple TV (ATV) really does have a dearth of supported codecs which I find surprising even by Apple's standards. Since the computer does all the heavy lifting anyways the side benefit of such a media processor would be on the fly transcoding of non iTunes content into AVC for display. The ATV has 40GB hard drive which would work out nicely. I figure that if transcoding other QT CODECS into AVC can at near realtime then a quick buffering for non AVC content is a small price to pay for transparently being able to play most Quicktime video CODECS through ATV.
Imagine a release of the new OS X in April which breaks Microsoft Word and Adobe Photoshop due to 'secret' features that hadn't had proper public or developer testing.
What secret features? The new OS won't have secret features. Do you understand why they do pre-releases of the OS? If so, then you understand why there aren't secrets.
Imagine a release of the new OS X in April which breaks Microsoft Word and Adobe Photoshop due to 'secret' features that hadn't had proper public or developer testing.
Apple has already done this before. Remember when they showed Aqua for the first time. No one outside of Apple had seen it yet their applications looked and functioned fine in Aqua.
Now in context imagine if Apple were to announce that Media co-processors were in fact included in every new Mac coming in 2007 and the way to take advantage is simply to use QTkit what's the issue? Apple consistently tells Developers to use the new API for "future proofing" yet many don't because they don't see the changes coming down the road. The last few years have been better with Developers getting on the ball a lot faster.
Clearly any Top Secret features that come most likely wont be in the form of a brand new API. Apple has to piggyback onto an existing API and somehow "enable" new functionality that won't be too hard for Developers to implement. Apple said "Top Secret" features are coming. They're not lying. I'm just speculating that this media co-processor could be considered a Top Secret feature that does dovetail with enhanced Leopard features.
Apple has already done this before. Remember when they showed Aqua for the first time. No one outside of Apple had seen it yet their applications looked and functioned fine in Aqua.
Now in context imagine if Apple were to announce that Media co-processors were in fact included in every new Mac coming in 2007 and the way to take advantage is simply to use QTkit what's the issue? Apple consistently tells Developers to use the new API for "future proofing" yet many don't because they don't see the changes coming down the road. The last few years have been better with Developers getting on the ball a lot faster.
Clearly any Top Secret features that come most likely wont be in the form of a brand new API. Apple has to piggyback onto an existing API and somehow "enable" new functionality that won't be too hard for Developers to implement. Apple said "Top Secret" features are coming. They're not lying. I'm just speculating that this media co-processor could be considered a Top Secret feature that does dovetail with enhanced Leopard features.
Or ZFS. Aqua was not included in the first Developer previews but was shown 3 months before the release. I am afraid it is too late for any changes which imply new interface
The fruits finally born from the Raycer Graphics acquisition perhaps?
Hmm... this means that apple has to re-design or modify Intel reference system. This would be too much work and would slow down the product release time. I some how think that many here are reading too much into the it and expecting more than what apple is willing to do. In any case, if Apple manages to add a hardware AVC/VC-1 decoder to improve HTPC experience with the factory setup, I would more than welcome the effort. Actually, it would make me want to get another mac.
I'm not clear on something; does the Apple TV have a dedicated H.264 chip or not?
IF all this hogwash were true. That may explain the TV delay. If TV does have it people would speculate that the rest of the lineup had it, and if it didn't and the rest of the lineup did have it every one will ask why Apple left it out of TV.
I think it's BS.
I do like the mention of Raycer again though. I think we all forgot about that. 8)
I think it's BS.
For some reason I tend to agree with this assessment.
For some reason I tend to agree with this assessment.
By the same token, if the rumor IS true and such a co-processor ships in upcoming hardware, then we're in for some extraordinary machines.
By the same token, if the rumor IS true and such a co-processor ships in upcoming hardware, then we're in for some extraordinary machines.
No doubt about that, but that's why I have some hard time to believe it is coming so suddenly with no hints from elsewhere. Sure, one time all we know today about multimedia processing will be museum pieces, but saying that this is going to happen so soon and abruptly is rather stretching it.
No doubt about that, but that's why I have some hard time to believe it is coming so suddenly with no hints from elsewhere. Sure, one time all we know today about multimedia processing will be museum pieces, but saying that this is going to happen so soon and abruptly is rather stretching it.
It would definitely be a surprise. To mix frozen metaphors, I won't hold my breath but I will cross my fingers.