Before you know it, Microsoft and Intel will start to get (more) hostile towards each other, with the result being that Microsoft has to rely more on AMD.
It's official. There is no longer any difference between 'good' and 'evil'. Just like how successful Democrats and Republicans are mostly just moderates with different names, good and evil have met in the middle in the tech industry.
It's too painful. Apple (good) with Intel (bad) and so Microsoft (bad) with AMD (good). I can't take it anymore!
The switch from PPC to x86 for Apple is an issue only for Mac developers and maybe some of the users. No one else. A Mac is still a Mac, even when it's got an Intel processor in it. The only effect of going to x86 is that Apple won't fall behind on performance.
The assumption seems to be that a Mac is somehow more attractive to developers with an Intel processor in it than with a PowerPC. Why!? What difference does it make? If you're developing for Linux and decide to also support the Mac, you still have to go through the same porting process to support OSX that you had to go through before. Sure, on those rare occasions when someone has some x86 assembly they would need to port, it helps, but otherwise, it's a non-issue.
Porting to the Mac is porting to OSX. OSX is the same on both architectures. Apple switching to Intel doesn't make OSX any more like Linux.
The switch from PPC to x86 for Apple is an issue only for Mac developers and maybe some of the users. No one else. A Mac is still a Mac, even when it's got an Intel processor in it. The only effect of going to x86 is that Apple won't fall behind on performance.
The assumption seems to be that a Mac is somehow more attractive to developers with an Intel processor in it than with a PowerPC. Why!? What difference does it make? If you're developing for Linux and decide to also support the Mac, you still have to go through the same porting process to support OSX that you had to go through before. Sure, on those rare occasions when someone has some x86 assembly they would need to port, it helps, but otherwise, it's a non-issue.
Porting to the Mac is porting to OSX. OSX is the same on both architectures. Apple switching to Intel doesn't make OSX any more like Linux.
Something others haven't mentioned: With an Intel processor and an appropriate syscall translation layer, they can run Windows programs at native speed. Being able to install arbitrary Windows programs makes the available software titles an order of magnitude larger and makes it just that much easier to get people to switch plaforms.
My one question: Why Intel? AMD processors are even faster than Intel per watt. So why Intel?
Oh, no! They're going to drill down and accidentally stop the earth's magnetic field, and then something else bad will happen, and we'll all have to move to Mars!
OSS produces some of the coolest stuff!!
on
McVoy Strikes Back
·
· Score: 1
Mind you, it's mostly very technical, but I joined the Linux Kernel Mailing List just to watch the discussions on the process and I/O schedulers, and it was a lot of fun watching people come up with neat ideas on making the schedulers more scalable and improve interactivity, etc. Ok, so maybe things like GIMP are playing catchup, but there are some FOSS tools that have caught up and surpassed the competition, and the trend is that once they support all of the basic features, they just keep going, adding new stuff. And I also notice that they tend to do the old stuff in novel ways, learning lessons from how people didn't do it so well in the past.
Quantum-mechanics, crosstalk, brain-death, etc.
on
Download Your Brain
·
· Score: 1
As well as complex neural cross-talk that goes on, some scientists have speculated that there are quantum-mechanical processes that go on in the brain that can't be reproduced by a traditional computer. If this is true, then only a quantum computer has a hope of being able to emulate a human brain.
And if your brain is dying, why is it dying? And what makes you so sure that it wouldn't die in emulation if you emulate it correctly?
As technology has moved back to the organic in some areas, I think there's going to be a convergence of the biological and the technological. That being said, I think the better approach is not to move your consciousness into a computer that may be ill-suited to it, but rather to use technology to repair the cellular mechanisms you already have in place.
Thinking of the human brain as a "legacy system" that needs to be replaced seems counterproductive to me. Is it possible to develop computers smarter than us? Sure, but they'll be an entirely new sort of intelligent system, very different from us. Preserving humans and advancing AI seem productive to me, while trying to convert one to the other seems like an unnecessary challenge.
The threat to fork also seems a bit heavy-handed. Yes, it's legal and allowed under the GPL, but it's certainly not a friendly thing to do to threaten to take over (basically) someone's project like that. That sort of thing insults people who have an emotional investment in it, and threatens to undermine the project by alienating the developers.
Yes, there are times and places for this sort of thing, but it must not be done coldheartedly.
You should see the flames the Open Graphics Project (opengraphics.org) is getting for wanting to keep their hardware proprietary, despite the fact that they already provide an unprecedented level of documentation on how the hardware works.
It's interesting that OOo has become so influential that the FSF has decided to poke their noses into it. I for one think that making OOo's java code work with GCJ is a great idea and makes it easier on those of us who don't want to have to download Sun's VM. But I also find it a big disturbing that the FSF has the gall to make demands on projects they don't own.
While DSPs would certainly be faster for graphics than a regular CPU, you just can't beat a special-purpose pipeline, designed specifically for graphics.
Don't believe me? Read their mail list archives. This strategy shines through over and over again. Their "open" focus is a fully documented register set, not fully GPL released "hackable" design. There's even a mimi flamefest between on the project leads and someone who really wants to see more focus on making the hackage FPGA prototypes more of a goal for the project.
I'm not sure how this is less than what any other graphics vendor is offering.
There will be a mechanism provided via the PCI controller that will allow you to reprogram the FPGA through a parallel interface. Open source software will be available to take advantage of that. Some people are trying to negotiate with Xilinx to get Linux support for the synthesis tools.
I'm going to say all the bad things I can think of so we don't have to waste time rereading them all day.
Thank you for commenting.
1. The hardware will be underpowered because this group has little experience (if any) designing bleeding edge graphics hardware
Is 6.4 GB/sec memory bandwidth "underpowered"? Perhaps compared to bleeding-edge Windows cards, but not compared to the latest cards FULLY supported by open source drivers. Your typical Linux server board sports a Rage XL. Furthermore, this group has a long history of experience with extremely high-end graphics cards used in air traffic control and medical systems, driving multiple high-res displays at resolutions like 2560x2048 and 3840x2400.
2. The card will be overpriced because this group doesn't have the manufacturing clout of NVidia or ATI
The initial product isn't really a graphics card. It's an FPGA project board that's a quarter the price of the next comparable product. The OGP ASIC-based product will be competitively priced. It will be on par (or better) in performance and price with other embedded solutions, and it will be affordable as a graphics card.
3. The drivers will suck because nobody's going to buy this card and nobody will develop for it.
There are already a good number of driver developers involved in the project, some of whom have gotten funding from their employers to work on it.
4. The drivers will suck MORE because of all the trans-gamers out there who dual boot, they won't get the card because it won't be supported in Windows (or just very weakly).
We fully intend to have the maximum Windows support possible. While the card isn't intended for games, the specs make are sufficient for Quake 3.
5. The company has no financial backing, so they will crash and burn early on and we will be stuck with abandoned hardware.
We've come up with a project plan that doesn't require financial backing, other than a few thousand dollars out of our own pockets. What more could you ask for?
6. This time, effort and money would be better spent harassing the existing graphics card manufacturers into opening up their drivers, as least the non-trade-secret parts so we can do our magic on it.
Harrassing only makes companies mad. Who are you anyhow? You're a Linux user, representing maybe 5% of the graphics market. If ATI or nVidia were to dedicate proper resources to Linux support, it would cost them more money than it makes them. Plus, ATI has a FAQ that states that they CANNOT open source their drivers due to IP licensing issues.
7. (asbestos ON) I still don't think any Linux Distro in its current state should even be considered for desktop or gaming. But that's me being an elitist prick. Come up with a cleaner development model, make it "just work", and redo the whole windowing system into something that is NOT X, and maybe then we can start talking. The reason OSX works so well is because it does fifty backflips to almost completely hide the underlying Unix layer. It's not because I know Linux that I want to put up with its PMS all the time, sometimes it's nice to just click things with your brain switched off.
This is a WHOLE other topic, but in large part, I agree with you.
Merely open-sourcing the drivers isn't enough. XGI and VIA have done that. In order to make open source drivers truly valiable, hackers have to be able to FIX them when there are bugs. That's very difficulty when the vendor doesn't release full specs on their hardware.
Actually, the product is targeted at the mass market. This includes Linux desktops, heavy workstations, and embedded systems. Being open architecture, it can be supported on all other platforms as well. Certainly, this market isn't as large as, say, the Windows market, but lacking another product as OPEN as this one, open source users are likely to prefer this, because the device will be fully supported by open source drivers, and it won't be a stability concern.
Also, a memory bandwidth of 6.4 GB/sec isn't all that slow.
Slashdot Mirror
Before you know it, Microsoft and Intel will start to get (more) hostile towards each other, with the result being that Microsoft has to rely more on AMD.
It's official. There is no longer any difference between 'good' and 'evil'. Just like how successful Democrats and Republicans are mostly just moderates with different names, good and evil have met in the middle in the tech industry.
It's too painful. Apple (good) with Intel (bad) and so Microsoft (bad) with AMD (good). I can't take it anymore!
The switch from PPC to x86 for Apple is an issue only for Mac developers and maybe some of the users. No one else. A Mac is still a Mac, even when it's got an Intel processor in it. The only effect of going to x86 is that Apple won't fall behind on performance.
The assumption seems to be that a Mac is somehow more attractive to developers with an Intel processor in it than with a PowerPC. Why!? What difference does it make? If you're developing for Linux and decide to also support the Mac, you still have to go through the same porting process to support OSX that you had to go through before. Sure, on those rare occasions when someone has some x86 assembly they would need to port, it helps, but otherwise, it's a non-issue.
Porting to the Mac is porting to OSX. OSX is the same on both architectures. Apple switching to Intel doesn't make OSX any more like Linux.
The switch from PPC to x86 for Apple is an issue only for Mac developers and maybe some of the users. No one else. A Mac is still a Mac, even when it's got an Intel processor in it. The only effect of going to x86 is that Apple won't fall behind on performance.
The assumption seems to be that a Mac is somehow more attractive to developers with an Intel processor in it than with a PowerPC. Why!? What difference does it make? If you're developing for Linux and decide to also support the Mac, you still have to go through the same porting process to support OSX that you had to go through before. Sure, on those rare occasions when someone has some x86 assembly they would need to port, it helps, but otherwise, it's a non-issue.
Porting to the Mac is porting to OSX. OSX is the same on both architectures. Apple switching to Intel doesn't make OSX any more like Linux.
Something others haven't mentioned: With an Intel processor and an appropriate syscall translation layer, they can run Windows programs at native speed. Being able to install arbitrary Windows programs makes the available software titles an order of magnitude larger and makes it just that much easier to get people to switch plaforms.
My one question: Why Intel? AMD processors are even faster than Intel per watt. So why Intel?
Oh, no! They're going to drill down and accidentally stop the earth's magnetic field, and then something else bad will happen, and we'll all have to move to Mars!
Mind you, it's mostly very technical, but I joined the Linux Kernel Mailing List just to watch the discussions on the process and I/O schedulers, and it was a lot of fun watching people come up with neat ideas on making the schedulers more scalable and improve interactivity, etc. Ok, so maybe things like GIMP are playing catchup, but there are some FOSS tools that have caught up and surpassed the competition, and the trend is that once they support all of the basic features, they just keep going, adding new stuff. And I also notice that they tend to do the old stuff in novel ways, learning lessons from how people didn't do it so well in the past.
As well as complex neural cross-talk that goes on, some scientists have speculated that there are quantum-mechanical processes that go on in the brain that can't be reproduced by a traditional computer. If this is true, then only a quantum computer has a hope of being able to emulate a human brain.
And if your brain is dying, why is it dying? And what makes you so sure that it wouldn't die in emulation if you emulate it correctly?
As technology has moved back to the organic in some areas, I think there's going to be a convergence of the biological and the technological. That being said, I think the better approach is not to move your consciousness into a computer that may be ill-suited to it, but rather to use technology to repair the cellular mechanisms you already have in place.
Thinking of the human brain as a "legacy system" that needs to be replaced seems counterproductive to me. Is it possible to develop computers smarter than us? Sure, but they'll be an entirely new sort of intelligent system, very different from us. Preserving humans and advancing AI seem productive to me, while trying to convert one to the other seems like an unnecessary challenge.
No reason not to try, though.
Wouldn't it be better to say that it fails for BAD?
For the backward compatibility, does the Xbox 360 have an x86 processor in addition to the PowerPC? Or does it emulate the x86?
The threat to fork also seems a bit heavy-handed. Yes, it's legal and allowed under the GPL, but it's certainly not a friendly thing to do to threaten to take over (basically) someone's project like that. That sort of thing insults people who have an emotional investment in it, and threatens to undermine the project by alienating the developers.
Yes, there are times and places for this sort of thing, but it must not be done coldheartedly.
You should see the flames the Open Graphics Project (opengraphics.org) is getting for wanting to keep their hardware proprietary, despite the fact that they already provide an unprecedented level of documentation on how the hardware works.
It's interesting that OOo has become so influential that the FSF has decided to poke their noses into it. I for one think that making OOo's java code work with GCJ is a great idea and makes it easier on those of us who don't want to have to download Sun's VM. But I also find it a big disturbing that the FSF has the gall to make demands on projects they don't own.
I don't know about you, but I no idea what this guy is saying.
Correct me if I'm wrong, but pandas eat mostly bamboo, but they evolved from carnivores and are still enticed by the smell of meat.
So how is this anything new?
A typical GPU does about 20 gigaflops (more, really, but that's what you can practically get out of them for supercomputing applications).
On the other hand, a DSP that does 1 gigaflop costs hundreds of dollars.
Do you STILL think they're better than specialized logic?
While DSPs would certainly be faster for graphics than a regular CPU, you just can't beat a special-purpose pipeline, designed specifically for graphics.
Don't believe me? Read their mail list archives. This strategy shines through over and over again. Their "open" focus is a fully documented register set, not fully GPL released "hackable" design. There's even a mimi flamefest between on the project leads and someone who really wants to see more focus on making the hackage FPGA prototypes more of a goal for the project.
I'm not sure how this is less than what any other graphics vendor is offering.
I suggest joining the mailing list and opening that discussion.
There will be a mechanism provided via the PCI controller that will allow you to reprogram the FPGA through a parallel interface. Open source software will be available to take advantage of that. Some people are trying to negotiate with Xilinx to get Linux support for the synthesis tools.
You're a bit behind the times. Modern ATC systems are always full-color.
Thank you for commenting.
Is 6.4 GB/sec memory bandwidth "underpowered"? Perhaps compared to bleeding-edge Windows cards, but not compared to the latest cards FULLY supported by open source drivers. Your typical Linux server board sports a Rage XL. Furthermore, this group has a long history of experience with extremely high-end graphics cards used in air traffic control and medical systems, driving multiple high-res displays at resolutions like 2560x2048 and 3840x2400.
The initial product isn't really a graphics card. It's an FPGA project board that's a quarter the price of the next comparable product. The OGP ASIC-based product will be competitively priced. It will be on par (or better) in performance and price with other embedded solutions, and it will be affordable as a graphics card.
There are already a good number of driver developers involved in the project, some of whom have gotten funding from their employers to work on it.
We fully intend to have the maximum Windows support possible. While the card isn't intended for games, the specs make are sufficient for Quake 3.
We've come up with a project plan that doesn't require financial backing, other than a few thousand dollars out of our own pockets. What more could you ask for?
Harrassing only makes companies mad. Who are you anyhow? You're a Linux user, representing maybe 5% of the graphics market. If ATI or nVidia were to dedicate proper resources to Linux support, it would cost them more money than it makes them. Plus, ATI has a FAQ that states that they CANNOT open source their drivers due to IP licensing issues.
This is a WHOLE other topic, but in large part, I agree with you.
nVidia aren't actively hostile to FOSS, but many people believe that it's unacceptable to use binary drivers.
Merely open-sourcing the drivers isn't enough. XGI and VIA have done that. In order to make open source drivers truly valiable, hackers have to be able to FIX them when there are bugs. That's very difficulty when the vendor doesn't release full specs on their hardware.
The OGP is based around open specs.
Actually, the product is targeted at the mass market. This includes Linux desktops, heavy workstations, and embedded systems. Being open architecture, it can be supported on all other platforms as well. Certainly, this market isn't as large as, say, the Windows market, but lacking another product as OPEN as this one, open source users are likely to prefer this, because the device will be fully supported by open source drivers, and it won't be a stability concern.
Also, a memory bandwidth of 6.4 GB/sec isn't all that slow.
The OGP is NOT a hobbyist project.
You're quoting prices for very SMALL FPGAs. What makes you think we could fit something as complex as a GPU into a 3S200?