This demo was of Ontario - AMD's low power solution for netbooks and low-end notebooks. This will be using the low power Bobcat cores and probably something similar to an HD5450 graphics-wise.
Intel's on chip GPU is just that - a GPU, and a primitive one at that. It can't even do OpenCL. It's certainly not a competitor to anything that AMD will release. Never mind Intel's appalling graphics drivers (and consistent history of poor driver releases), and benchmark cheating (so that they look competitive in reviews).
Apple audit their factories every year, since 2006 when there were problems.
Apple and HP are the only companies to have set a working conditions standard with their suppliers.
That means that all the other companies that use Foxconn (pretty much most electronics companies do) don't even have that minimum standards requirement with Foxconn.
Now Apple is taking further steps to improve the workers' lot by paying them some more. Let's hope it helps with the problem.
What you need is the ability to define the content semantically to identify the content, to style the content, and to allow resizing, reflow, etc, based upon the viewing device. In addition a reality is that publishers will want DRM.
A format that is basically HTML doesn't have the semantics. You can extend/bastardise the HTML to include <chapter>, <footnote> , and so on - normal HTML renderers would ignore them, dedicated readers could use them.
PDF is just horrible for eBooks, unless you really really don't care about semantics, restyling, reflow, usability, etc, and just want the pages presented as pages as published. Obviously there are some cases where this is useful, however on a phone I might just want to click on an icon (or even better, an embedded reference in the text) to get an image, rather than be forced to scroll around a large document.
I don't think it would take a group of eBook people long to define a comprehensive XML format for books that supports most book formats. Allow hooks in the XML to extend it, and all you need them is support in eBook applications. Oh.
I would imagine an archive format, like ePub, with SVG support for diagrams and.png and.jpeg support for images, is good enough. Add a DRM wrapper and who wouldn't be happy?
It's simple - apps that are preinstalled by the carrier are already pre-approved to access the telephony APIs (for non-international, non-roaming calls).
The security levels to access telephony APIs are (at least);
1) No Access (default) 2) Access to call national numbers (whilst not roaming) with user confirmation 3) Access to call national numbers (whilst not roaming) without confirmation 4) Access to call all numbers with confirmation 5) Access to call all numbers.
I don't see why there should be a feature to auto-dial, without confirmation, an international number, unless you've rewritten the dialer application. I guess that a dialer application could be written with a trojan caller inside it - but such an application probably shouldn't be running as a service style app. So (5) could be restricted to foreground (visible) applications only.
I have. The OS was terrible. It was like using Windows 3.1 on a tiny screen. Configuration settings everywhere, nothing centralised, nothing simple to configure, complexity abounding, stupid interface.
The Psion 5 got it right for the 'mini-netbook' style devices. PalmOS got it right for the PDA style devices.
And I'm hardly surprised that new phones beat year old phones in terms of specifications. Sheesh.
There are a lot of British ex-pats (5.5 million http://news.bbc.co.uk/1/hi/uk/6210358.stm). If the Times can get one in ten to pony up for a website subscription (with kindle/iPad/iPhone access) for a quid or two a week then they will get direct income of £25m-£50m/year. I assume they can probably also get the same amount of subscribers within the country. They will probably lose any income from paper-based versions of the newspaper to these people, but sales were probably low with the ex-pats already, and most of the cost of a paper version is the paper and printing itself.
The things that derails this grand plan is the other, ad-funded, websites that will provide the same news to these people. Will The Guardian (making money from their ad-funded website) risk losing it all (£40m a year according to a post above) by becoming a paysite? Not at all.
I look forward to Murdoch's other papers going behind a paywall and effectively removing their vile rhetoric from the public internet.
The Computing A Level isn't designed to teach you specific languages, it is designed to give a broad spread of (pre-degree level) computer science knowledge, part of which is programming. It is the programming concepts that matter most, and they can be learned with Java, Pascal or Python quite well. I did Pascal when I did my A Levels um, 16 years ago.
Getting rid of PHP is good, in my opinion.
Anyway, most UK Computer Science degrees say they would rather no A Level Computing because they have to spend time un-teaching bad habits. Looking at the VB6 option, I can see why.
The company hired you to improve an already existing piece of work.
It is irrelevant whether or not you created that work in the first place, as you say that there is no agreement to transfer copyright, etc, to the university of the system you had written before. Therefore if the previous system was GPL, the university should make its changes GPL, or negotiate with the author of the original system (you) for its own license.
Using GPL libraries shouldn't be a problem however. Indeed as an internal project they can make their own changes and not release them anyway.
I think you'll have to suck it up for now, the university owns the work you've done. Of course if the university is publicly funded, then really it is the taxpayer that owns it, but know how that goes in general. Just take the money they're paying.
Once you've finished doing the changes, you can go back to the previous version that is yours, and add in the new features again, from scratch, building upon the knowledge and experience you got from doing it before!
Flash video will not be the primary means of video distribution in a year's time. It will be a means, and it'll probably be available for quite a few years, but the browsers bring it in, and websites will inevitably provide a HTML5 alternative for compatible browsers.
Do you think that these other sites will chose to deliberately not use Flash for video distribution? Maybe Adobe will sell them on DRM.
YouTube already has a native app on the iPhone, so is already irrelevant. Maybe blip.tv will have a native app within a year. Not that I've ever heard of it.
Irrelevant. Within a year 90%+ of online videos will have a HTML5 alternative. Other services that currently use Flash for video will have apps in the App Store as well.
2. Put the core Flash technologies into the standards bodies
I thought the Flash format was open, hence Gnash, etc. Standardise it, fine. It still won't change anything about Apple's stance.
3. Create an iPhone-certified SWF exporter for Creative Suite
Basically you're asking for a new Cocoa Touch component - a "Flash Canvas" but with native code instead of ActionScript. I'm sure that the existing (but now disallowed) Flash->iPhone OS App compiler did this. I don't think Apple will care to implement that in any form though, they want to push their OS, not a cross-platform subset of functionality. They already have a HTML5 Canvas which can be controlled via Javascript. http://www.canvasdemos.com/ has some examples.
4. Explore a Flash app certification process
If the app is worthwhile, it will get ported to Apple's platform anyway. Apple would rather the apps all shared the same UI conventions and used the same underlying UI kit.
In my opinion Adobe should put their effort into leveraging their creative suite tools to generate HTML5 apps, via HTML5 Canvas, Video and so on. If they don't, Apple surely will eventually have such offerings!
The ARM Cortex A8 core is roughly on par with an Intel Atom, clock for clock (maybe a little slower). Of course it tops out at 1GHz, and the Atom is available up to 1.86GHz.
The ARM Cortex A9 is faster than the A8, and will be available at up to 2GHz in single, dual and quad-core versions. A quad-core 2GHz A9 can do 10,000 DMIPS in 1.9W. I would hope that such a CPU would be coupled with decent caches and memory interface, otherwise the cores will get starved.
ARM also needs a 64-bit ISA, but I'm sure that they're working on it, or have even completed it.
And it's already licensed, and TI will also have an ARM Cortex A9 license, and probably already have a license for whatever is coming after that.
These licenses will not be reversible.
And as the A9, etc, will be viable for a good two or three years in SoC designs, there will be plenty of time for the SoC manufacturers to invest in MIPS based systems for future designs. In addition MIPS already has a 64-bit variant, Android has been ported, etc.
In addition Qualcomm and Marvell have architecture licenses, so they can continue to create their own versions of their ARM cores, improving clock speeds and features.
Makes sense to me to do both frame renderings in parallel, so you're generating both eyes' views for each vertex in a scene at the same time, applying the same texture at the same time, and so on. This could reduce the bandwidth required because you would be utilising at the same time in rendering the same vertexes, textures, etc, when rendering both views. Especially when you have something like render-to-texture which is then applied elsewhere (e.g., a TV screen) - it makes no sense to re-render that for each viewpoint, when the eyes are seeing the same scene.
But it would double the time to render each view, require twice the video memory for rendering, etc.
I think this will really take off in the next generation. When the PS4 comes with a HD6870 equivalent graphics card, and 4GB RAM.
Damn right. And European Law sides very closely with customers.
All it would require is a concerted combined effort by EU PS3 owners getting refunds for the retailers (who have to legally make the refund) to sue Sony, and force the Linux option back into the EU fat PS3 firmware.
If there was, it would need a lot of memory, as binary translation does.
Sadly the iPad only has 256MB of RAM, the same as the 3GS. With an emulator you'd have far less available memory than on a 3GS!
Also the system files are ARM (nobody has said they're PowerPC). Or maybe the emulator is built into the chip! That could account for the extra transistors!!! lol.
The Chipworks analysis is solid. It is a 45nm die. It's 53.3mm^2, compared to the 65nm 73mm^2 3GS die.
The author assumed perfect scaling of transistor density, thus he came up with an additional ~50m transistors on the new die.
1MB of L2 cache (6T) is 6 * 8 * 2^32 transistors, + L2 cache tags. That's 50m transistors at least. If the 3GS CPU had a 256KB L2, this one could have a 1MB L2. It would account for a lot of the performance increase as well. If there are more spare transistors, then maybe the power management is more aggressive (requiring transistors), or the graphics are better, or a myriad of other options.
Hopefully Chipworks will be able to grind the A4 die down to a lower metal layer, and get to see the actual functional blocks on the die.
The article is dreamy bullshit, but not for what you write about multitasking (especially since OS4 will provide for it, and designing hardware to cope with future demands is sensible).
The performance analysis shows the product's CPU power matches a 1GHz Cortex A8, compared with scaling up from the 600MHz A8 in the 3GS.
The article links to the Chipworks A4 die dissection, and the product code which is just a higher version of the 3GS product code. That certainly doesn't fit in with putting in a PowerPC core instead of the ARM core in the previous product, never mind fitting the PowerPC core to the ARM-specific internal bus and peripherals. The code name would be completely different. If there's anything that can be guaranteed, it is that the A4 utilised an ARM core.
The Apple A4 is a 45nm version of the 3GS Samsung CPU, rebranded by Apple (because they bought Intrinsity, who developed/enhanced/tweaked the Samsung product originally). The extra transistors are accounted for by having a wider memory bus, probably more L2 cache, and maybe higher performance graphics.
Also the guy assumed perfect transistor scaling, which doesn't happen.
Generally some days I feel like I do around one hour of actual work, and others up to four hours. Of course you can tack on an hour or two of meetings on top of that. And, of course, browsing the web is actually just a way to sharpen your mind as you think about your current task.
Most of them just do buzzword matching on CVs rather than actual filtering by skill, so you'll get some really rubbish dregs turn up with inflated CVs.
Also, try to get one going through a relationship break-up (especially an expensive divorce), or one with criminal/drug addict children / wife. These will increase their productivity as they will want to stay in work.
Slashdot Mirror
Hence "a two generation old integrated graphics core".
With a 320x240 display, you're better off using pretty much anything else for remote desktop (most smartphones are at least 480x320).
I guess you can have an 80x30 terminal though, with a 4x8 pixel font.
Otherwise, C64, Spectrum, CPC, Atari 8-bit, etc, emulators would be good because of the full keyboard.
This demo was of Ontario - AMD's low power solution for netbooks and low-end notebooks. This will be using the low power Bobcat cores and probably something similar to an HD5450 graphics-wise.
I seriously doubt heat is going to be an issue.
"Trading blows" with a two generation old integrated graphics core that is going to get replaced early next year with one around 5x - 10x faster.
This is like a wannabe thug trying to beat up an elderly gentleman.
And failing.
Intel's on chip GPU is just that - a GPU, and a primitive one at that. It can't even do OpenCL. It's certainly not a competitor to anything that AMD will release. Never mind Intel's appalling graphics drivers (and consistent history of poor driver releases), and benchmark cheating (so that they look competitive in reviews).
Apple audit their factories every year, since 2006 when there were problems.
Apple and HP are the only companies to have set a working conditions standard with their suppliers.
That means that all the other companies that use Foxconn (pretty much most electronics companies do) don't even have that minimum standards requirement with Foxconn.
Now Apple is taking further steps to improve the workers' lot by paying them some more. Let's hope it helps with the problem.
What you need is the ability to define the content semantically to identify the content, to style the content, and to allow resizing, reflow, etc, based upon the viewing device. In addition a reality is that publishers will want DRM.
A format that is basically HTML doesn't have the semantics. You can extend/bastardise the HTML to include <chapter>, <footnote> , and so on - normal HTML renderers would ignore them, dedicated readers could use them.
PDF is just horrible for eBooks, unless you really really don't care about semantics, restyling, reflow, usability, etc, and just want the pages presented as pages as published. Obviously there are some cases where this is useful, however on a phone I might just want to click on an icon (or even better, an embedded reference in the text) to get an image, rather than be forced to scroll around a large document.
I don't think it would take a group of eBook people long to define a comprehensive XML format for books that supports most book formats. Allow hooks in the XML to extend it, and all you need them is support in eBook applications. Oh.
I would imagine an archive format, like ePub, with SVG support for diagrams and .png and .jpeg support for images, is good enough. Add a DRM wrapper and who wouldn't be happy?
It's simple - apps that are preinstalled by the carrier are already pre-approved to access the telephony APIs (for non-international, non-roaming calls).
The security levels to access telephony APIs are (at least);
1) No Access (default)
2) Access to call national numbers (whilst not roaming) with user confirmation
3) Access to call national numbers (whilst not roaming) without confirmation
4) Access to call all numbers with confirmation
5) Access to call all numbers.
I don't see why there should be a feature to auto-dial, without confirmation, an international number, unless you've rewritten the dialer application. I guess that a dialer application could be written with a trojan caller inside it - but such an application probably shouldn't be running as a service style app. So (5) could be restricted to foreground (visible) applications only.
I have. The OS was terrible. It was like using Windows 3.1 on a tiny screen. Configuration settings everywhere, nothing centralised, nothing simple to configure, complexity abounding, stupid interface.
The Psion 5 got it right for the 'mini-netbook' style devices. PalmOS got it right for the PDA style devices.
And I'm hardly surprised that new phones beat year old phones in terms of specifications. Sheesh.
There are a lot of British ex-pats (5.5 million http://news.bbc.co.uk/1/hi/uk/6210358.stm). If the Times can get one in ten to pony up for a website subscription (with kindle/iPad/iPhone access) for a quid or two a week then they will get direct income of £25m-£50m/year. I assume they can probably also get the same amount of subscribers within the country. They will probably lose any income from paper-based versions of the newspaper to these people, but sales were probably low with the ex-pats already, and most of the cost of a paper version is the paper and printing itself.
The things that derails this grand plan is the other, ad-funded, websites that will provide the same news to these people. Will The Guardian (making money from their ad-funded website) risk losing it all (£40m a year according to a post above) by becoming a paysite? Not at all.
I look forward to Murdoch's other papers going behind a paywall and effectively removing their vile rhetoric from the public internet.
Only from the types of service provider that you go to! You know, the ones' whose tubes are clogged with viruses, worms, and the like. ;-)
Seriously, I'm sure that such foreign DNA doesn't hang around in the mouth for long.
The Computing A Level isn't designed to teach you specific languages, it is designed to give a broad spread of (pre-degree level) computer science knowledge, part of which is programming. It is the programming concepts that matter most, and they can be learned with Java, Pascal or Python quite well. I did Pascal when I did my A Levels um, 16 years ago.
Getting rid of PHP is good, in my opinion.
Anyway, most UK Computer Science degrees say they would rather no A Level Computing because they have to spend time un-teaching bad habits. Looking at the VB6 option, I can see why.
The company hired you to improve an already existing piece of work.
It is irrelevant whether or not you created that work in the first place, as you say that there is no agreement to transfer copyright, etc, to the university of the system you had written before. Therefore if the previous system was GPL, the university should make its changes GPL, or negotiate with the author of the original system (you) for its own license.
Using GPL libraries shouldn't be a problem however. Indeed as an internal project they can make their own changes and not release them anyway.
I think you'll have to suck it up for now, the university owns the work you've done. Of course if the university is publicly funded, then really it is the taxpayer that owns it, but know how that goes in general. Just take the money they're paying.
Once you've finished doing the changes, you can go back to the previous version that is yours, and add in the new features again, from scratch, building upon the knowledge and experience you got from doing it before!
Flash video will not be the primary means of video distribution in a year's time. It will be a means, and it'll probably be available for quite a few years, but the browsers bring it in, and websites will inevitably provide a HTML5 alternative for compatible browsers.
Do you think that these other sites will chose to deliberately not use Flash for video distribution? Maybe Adobe will sell them on DRM.
YouTube already has a native app on the iPhone, so is already irrelevant. Maybe blip.tv will have a native app within a year. Not that I've ever heard of it.
1. Create a Flash video player plug-in
Irrelevant. Within a year 90%+ of online videos will have a HTML5 alternative. Other services that currently use Flash for video will have apps in the App Store as well.
2. Put the core Flash technologies into the standards bodies
I thought the Flash format was open, hence Gnash, etc. Standardise it, fine. It still won't change anything about Apple's stance.
3. Create an iPhone-certified SWF exporter for Creative Suite
Basically you're asking for a new Cocoa Touch component - a "Flash Canvas" but with native code instead of ActionScript. I'm sure that the existing (but now disallowed) Flash->iPhone OS App compiler did this. I don't think Apple will care to implement that in any form though, they want to push their OS, not a cross-platform subset of functionality. They already have a HTML5 Canvas which can be controlled via Javascript. http://www.canvasdemos.com/ has some examples.
4. Explore a Flash app certification process
If the app is worthwhile, it will get ported to Apple's platform anyway. Apple would rather the apps all shared the same UI conventions and used the same underlying UI kit.
In my opinion Adobe should put their effort into leveraging their creative suite tools to generate HTML5 apps, via HTML5 Canvas, Video and so on. If they don't, Apple surely will eventually have such offerings!
The ARM Cortex A8 core is roughly on par with an Intel Atom, clock for clock (maybe a little slower). Of course it tops out at 1GHz, and the Atom is available up to 1.86GHz.
The ARM Cortex A9 is faster than the A8, and will be available at up to 2GHz in single, dual and quad-core versions. A quad-core 2GHz A9 can do 10,000 DMIPS in 1.9W. I would hope that such a CPU would be coupled with decent caches and memory interface, otherwise the cores will get starved.
ARM also needs a 64-bit ISA, but I'm sure that they're working on it, or have even completed it.
Really? Not even at +50% of their current (and very high) value?
Not even when Apple force the sale of all shares anyway, and you have no choice?
And it's already licensed, and TI will also have an ARM Cortex A9 license, and probably already have a license for whatever is coming after that.
These licenses will not be reversible.
And as the A9, etc, will be viable for a good two or three years in SoC designs, there will be plenty of time for the SoC manufacturers to invest in MIPS based systems for future designs. In addition MIPS already has a 64-bit variant, Android has been ported, etc.
In addition Qualcomm and Marvell have architecture licenses, so they can continue to create their own versions of their ARM cores, improving clock speeds and features.
Makes sense to me to do both frame renderings in parallel, so you're generating both eyes' views for each vertex in a scene at the same time, applying the same texture at the same time, and so on. This could reduce the bandwidth required because you would be utilising at the same time in rendering the same vertexes, textures, etc, when rendering both views. Especially when you have something like render-to-texture which is then applied elsewhere (e.g., a TV screen) - it makes no sense to re-render that for each viewpoint, when the eyes are seeing the same scene.
But it would double the time to render each view, require twice the video memory for rendering, etc.
I think this will really take off in the next generation. When the PS4 comes with a HD6870 equivalent graphics card, and 4GB RAM.
Damn right. And European Law sides very closely with customers.
All it would require is a concerted combined effort by EU PS3 owners getting refunds for the retailers (who have to legally make the refund) to sue Sony, and force the Linux option back into the EU fat PS3 firmware.
Consumer Protection Laws FTW.
The problem with his argument there is@
* There isn't an emulator on the iPad. *
If there was, it would need a lot of memory, as binary translation does.
Sadly the iPad only has 256MB of RAM, the same as the 3GS. With an emulator you'd have far less available memory than on a 3GS!
Also the system files are ARM (nobody has said they're PowerPC). Or maybe the emulator is built into the chip! That could account for the extra transistors!!! lol.
The Chipworks analysis is solid. It is a 45nm die. It's 53.3mm^2, compared to the 65nm 73mm^2 3GS die.
The author assumed perfect scaling of transistor density, thus he came up with an additional ~50m transistors on the new die.
1MB of L2 cache (6T) is 6 * 8 * 2^32 transistors, + L2 cache tags. That's 50m transistors at least. If the 3GS CPU had a 256KB L2, this one could have a 1MB L2. It would account for a lot of the performance increase as well. If there are more spare transistors, then maybe the power management is more aggressive (requiring transistors), or the graphics are better, or a myriad of other options.
Hopefully Chipworks will be able to grind the A4 die down to a lower metal layer, and get to see the actual functional blocks on the die.
The article is dreamy bullshit, but not for what you write about multitasking (especially since OS4 will provide for it, and designing hardware to cope with future demands is sensible).
The performance analysis shows the product's CPU power matches a 1GHz Cortex A8, compared with scaling up from the 600MHz A8 in the 3GS.
The article links to the Chipworks A4 die dissection, and the product code which is just a higher version of the 3GS product code. That certainly doesn't fit in with putting in a PowerPC core instead of the ARM core in the previous product, never mind fitting the PowerPC core to the ARM-specific internal bus and peripherals. The code name would be completely different. If there's anything that can be guaranteed, it is that the A4 utilised an ARM core.
The Apple A4 is a 45nm version of the 3GS Samsung CPU, rebranded by Apple (because they bought Intrinsity, who developed/enhanced/tweaked the Samsung product originally). The extra transistors are accounted for by having a wider memory bus, probably more L2 cache, and maybe higher performance graphics.
Also the guy assumed perfect transistor scaling, which doesn't happen.
I'm often still in bed at 8:30 am :-(
But yeah, so true.
Generally some days I feel like I do around one hour of actual work, and others up to four hours. Of course you can tack on an hour or two of meetings on top of that. And, of course, browsing the web is actually just a way to sharpen your mind as you think about your current task.
Use a recruitment agency.
Most of them just do buzzword matching on CVs rather than actual filtering by skill, so you'll get some really rubbish dregs turn up with inflated CVs.
Also, try to get one going through a relationship break-up (especially an expensive divorce), or one with criminal/drug addict children / wife. These will increase their productivity as they will want to stay in work.