It completely depends on the algorithm involved and the bottlenecks faced by the GPU in performing those algorithms. In general, ASICs are about 2-3x as fast as an FPGA, and can contain about 4 times as much logic. They also cost a LOT of money to make; think 5-10 million USD for a mask set of a large ASIC in modern technology. Depending on the algorithm, you may be able to run many pipelines in parallel, creating a huge calculation capacity, and end up with your PCIexpress bus being the limiting factor. ASICs capable of handling TBs/s of data really are no exception, and a modern FPGA can certainly handle 400 Gb/s interfaces (we built 400 Gb/s AES-GCM encryption in an FPGA a few years ago).

As for your second question: yes. If you have (and understand) the VHDL/Verilog source code of a chip, you can generally make a very accurate estimation of its performance.

http://www.imdb.com/title/tt0364343/
That one you mean?

See, this is where we disagree. I don't understand how anyone in their right mind may think that DRM is a good idea for the simple reason that DRM doesn't accomplish anything useful. When you ask people why they use DRM, you often get the answer: I need to protect my work. Since DRM is incapable of doing that, why would that be a valid reason to use it?

Also: copyright infringers are not your customers, and thus not your concern. Yes, they are using your work without paying for it, but if you somehow manage to stop them from doing that, they're still not buying your products, so in the end getting rid of copyright infringement does not bring any extra money in your pocket. In the mean time, fighting against copyright infringement costs a lot of money, and many of your real customers are pissed off in the process of your futile attempts to stop it. This is not a sound business strategy, no matter how you look at it.

Actually, I think we're more in opposition than you think. :) I may be a musician and software programmer next to my work as a chip designer, but my views on copyright seem to differ quite a bit from yours.

For starters: I don't believe in the phrase "deserve to get paid". Nobody deserves to get paid just because they put a lot of work into something. They deserve a fair chance to make a product that people want to buy, but if they make something nobody's interested in they deserve nothing, no matter how much time they put into it.

I also despise DRM, and not just because there's no such thing as an Open Source DRM implementation. DRM serves no purpose whatsoever. It cannot protect content, it costs a fortune to develop, and the only thing it does is piss off your customers. Why anyone would be foolish enough to invest in that is beyond me.

I do believe in copyright however, but not the travesty that it has become in the last 5 decades. Copyright was meant as an incentive to create, not as a protection for a certain business model. Copyright protects my work from being used by greedy corporations without my permission, but if I go out to sue my fans I'll soon be left without any. Sure that means that some will listen to/use my work without my permission, but they would not have been customers anyway, so I have lost nothing. As an artist, you should focus on those people who WILL be customers.

In a digital age, certain products cease to be scarce. This means you should either go for volume sales and/or focus on those things that ARE scarce. Several artists have already been amazed at how much fans are willing to pay for something that's really rare, like signed copies, limited edition dvds, life performances, etc.

The worst thing for an artist however is to be unknown. Here you have a medium (internet) that will allow you to get your work seen/heared by everybody, practically distributing for free, and yet you're fighting it rather than embracing it. As an artist you don't need to be internet savvy; you just need to find someone who is to do it for you. Besides: setting up a YouTube and MySpace page really isn't that difficult. You just need to invest some time to really get in touch with your fans.

Your idea has a major flaw: DRM does not work. It does not protect your work, will cost a lot of money to develop, and piss off your customers. DRM's not only not the best business strategy, it's the worst! Why on Earth would you want to use it? You would put your faith in Microsoft, a company that can't even protect its own software from being copied, and yet you believe they can keep yours safe?

As for there being no reasonable business strategy for the artist: You're being proven wrong on a daily basis by artists who have managed to make the internet work for them. Learn to sell what's scarce by using that which is not, and you may yet be surprised by mankind.

In other words: you're going to spend a lot of time (and thus money) developing some protection scheme that will be cracked within a week by a college kid with too much time on his hands, which will do nothing to protect your game from being copied illegally, but will at the same time annoy the customers you might get? Sounds like a business strategy to me...

I agree with the above post, though I personally prefer VHDL. That might however have something to do with me having designed ASIC/FPGAs for about 11 years now using VHDL though. :) Both are very powerful languages these days, and I see no problem in teaching a course using both languages, showing how to create the same hardware using different language constructs.

Why would this be Google's problem? Considering the QQ I see from PRS over YouTube dropping their music, it appears that PRS needs YouTube a lot more than YouTube needs PRS' music. If there's no deal, and Google's dropping PRS' music, then what's the problem?

I'm sorry, but that's not at issue here:
Authors: pay us X or we won't allow you to play our music.
Google: X is too much considering how much we're making from your music, but we'll be willing to pay you Y.
Authors: Y is not enough, we don't have a deal.
Google: Ok, then we'll just remove your music since we don't have permission to play your music.
Authors: That's not fair! Why don't you just pay us X like we're asking? We deserve to be paid for our work, and you have enough money anyway.

.....

I'm sorry, but I have to disagree here. I am not condoning copyright infringement, but this is a business deal between two parties. The songwriters wish to get paid for their work being up on YouTube, but don't like the amount Google is willing to pay. Without a deal, Google has to remove said music from YouTube upon request (conform European copyright law), but when they do so the songwriters protest even more. It is clear to me that they have a very different view of what their music is worth from what it is worth to Google, but they can either come to an agreement or have their music removed. They cannot force Google into an amount just because they think it is justified. Failure to understand this is what makes them clueless.

Very simple: they seem to focus on how much money Google is making, and how much money they think their music is worth. The question they SHOULD be asking is: how much money is my music WORTH to Google? How much revenue would Google lose if my music was pulled from YouTube tomorrow, and what % of that money might I fairly claim? They should also ask themselves the question: how much money will I lose/gain if my music was NOT on YouTube? If the payment is not enough for you, then don't complain when Google removes your music.