*sigh*
A good friend and I were discussing the exact topic of the mainstream and 64-bit processing the other day. Here\'s the problem: it doesn\'t work.
Guru: \"So what do you think of the new 64-bit processors coming out?\"
Newbie: \"64-bit? What are we using now?\"
Guru: \"32-bit processors, and we\'ve been using them for over ten years.\"
Newbie: \"Oh, so these new computers will be twice as fast as the ones we now, right? That\'ll be cool.\"
Guru: \"*sigh* No, 64-bit only describes the size of the internal registers of the CPU. Basically, how big of a number it can work on at once.\"
Newbie: \"So then they can use twice as big numbers? That\'ll still speed things up a bit, right?\"
Guru: \"Only if you\'re using applications that work with really big numbers.\"
Newbie: \"Do I?\"
Guru: \"No.\"
The concept of how 64-bit processing is better than 32-bit processing is something that the layman (I\'m not accusing anybody here, I\'m just saying) cannot understand. It takes a knowledge of how microprocessors function, how they store information internally, how they perform calculations, how they communicate with the rest of the system, and so on. The majority of the population does not understand these things, and therefore, does not understand how 64-bit processing is better. All they can do is \"assume\" it\'s twice as fast because 64 is twice the size of 32.
Now, I\'ll try to answer your questions:
-DSP: Never heard of it, can\'t help you here
-Bus speed: all the components in your computer are connected via a \"bus,\" basically an information highway. Any data travelling from one place to another, be it processor -> graphics card, hard drive -> processor, processor -> memory, will all use the same bus. Because we have 32-bit processors, that means that our bus size is 32-bits (note: I\'m still confused how 64-bit PCI slots can communicate with 32-bit processors, perhaps someone can explain this). With 64-bit processors, we will likely see 64-bit busses, which means that we can push data all around our computer in chunks that are twice as big (communication via bus is parallel, that is, 32/64 lanes each carrying one bit). What this will improve: loading times, because loading is the process of moving a chunk of information from one place to another, which could be the hydlaa plaza textures from your hard drive to memory, and from memory to your video card. What this will not improve: your fps, because this is something that is calculated internally by your graphics processor on your video card, and server performance, the same deal because most of what the servers are doing is computing and shoving data across the Internet, and server performance will only increase with 1) faster processors (I said faster, not wider) and 2) more servers (better client connections, lower ping)
-\"static vs. dynamic\":
Movie Special Effects:
1. Create 3D models
2. Create a specific animation sequence for each model
3. Put all the models together for a single purpose (\"static\")
4. Render the scene (using your system CPU), frame by frame, and save it as a movie that you can use later
3D Games (Planeshift):
1. Create 3D models
2. Create a set of animation sequences for each model, for every action they can perform
3. Utilize game code to make the models animate in real-time, based on user commands (\"dynamic\")
4. Render the scene (using your graphic processor), frame by frame, and display the results in real time
-Why not?
I have already explained the main \"advantages\" of 64-bit: you can perform calculations on 64-bit numbers natively (as opposed to 32-bit processors which have to do something special to work with 64-bit numbers) which is something that is almost exclusively the realm of modeling (whether it be a graphical render of virtual 3D objects, of a scientific model of some real-world phenomenon) AND you will get a larger bus (push data around your computer faster).
So to your question \"Whats the problem...\" I answer: there is no problem with using the advantage; your problem (and everyone else\'s) is that they don\'t understand WHAT the advantages are. If your code uses 64-bit numbers then your code will run a bit faster on 64-bit processors. If your computer pushes a lot of data around, it will do that a bit faster.
And that is it, don\'t tell me I\'m dodging, because I\'m telling you that there is no reason to go out and buy a 64-bit processor the day they come out because you\'re throwing your money away! In the future, 90% of people will use 64-bit processors, but this isn\'t the future, it\'s the present. 90% of the population uses 32-bit processors, and there\'s not a thing wrong with them.
Another thing: awhile back, Intel created the 8086 processor, their first 16-bit. Everyone was all well and happy, but the thing was, no on had 16-bit devices, all people had was 8-bit. They actually made another chip, the 8088, which was identical to the 8086, only it had an 8-bit external bus. There wasn\'t a lot of demand for the 8086 because people didn\'t want to have to go out and pay to buy a bunch of new devices just so they could be cool and have full 16-bit power. In all practicality, the 8088 was a success because it had all the features of 16-bit processing, but was cheaper because you could 8-bit components were a heck of a lot cheaper.
Note to kbilik: I spent a VERY long time typing up this post in response to yours. Make it worth my while.
