So, here's what I'm trying to figure out... What's the idea behind having 1 SPU "reserved for redundancy"? Wouldn't it be better to have all the SPUs available, and then them for redundancy as needed? Thoughts? Ideas?
"It is unclear how many processing units will be incorporated into either the PlayStation 3 or workstations." is one change.
Also
"The PE is not the primary processor for the system, but acts as a controller for the other eight SPUs" is a bit different than what I've read in the past. This means that the SPUs are more general-purpose (And thus the amount of parallelism is more flexible) than I recall being told.
I think you made my point for me... It's all marketing hype. It really is easier said than done. Trust me, at E3, they're going to say an awful lot. They have a whole year to "go back on" what they said at the show and make "changes" to what they're planning on implementing now.
I think you may be a little confused... The PS3 has 3 cells. Each cell has a primary core, plus SPEs. That means each cell is designed to both serial and parallel code.
Also, gaming is a specific arena where parallel processing is an absolute godsend. Having an army of SPEs doing a transformation on every pixel in an image is vastly more efficient than having CPUs do it. When your system architecture is built and your code is specifically written for that kind of parallelism, software that has intensive parallel-friendly tasks should perform very admarably.
Considering that the PS3 is better than the 360 on every front you listed, I'm guessing the sony marketing guys just watched the xbox release, and upped all their numbers for E3.
The problem is, you're trying to recieve a consise message from an advertisement that does not have one. There is no underlying message, and they are suggesting very little.
The whole idea of this kind of advertising is to get your attention, and then show you a brand name. It's called brand exposure, and it's surprisingly effective. That's it, nothing more.
The game code itself doesn't read the file, it relies on some form of BIOS to do that. So all you really need to do is have the system read from the memory stick instead of the UMD. For these guys, that's no problem.
I disagree with your comments about the API. In this example SDL_GetTicks() is not a very clear function name. A well placed comment saying what the function is really doing will let me know if I need to go to the documentation or not.
For example:
last_update = SDL_GetTicks();// time in milisec
Tells me enough about what the function does, and what kind of data it returns, that I don't have to go all the way to the API documentation.
So, do you think this is really about preventing the child from having any form of contact with the game under any circumstances?
No, it's about putting reasonable bounds in place, like locking away liquor. If the child still gains access to it while they know it's out-of-bounds, then there are concequences.
I'm sorry, but I don't trust the process. Ok, so they're not buying or selling items directly. They still control the origination of the commodities and can control the value of the items by making artifical changes to the supply of the market.
So, basically, Sony is able to create their own economy by creating, manipulating, or removing (virtual) commodities at a whim, and then get/give real money for them?
Here's the leasing rule of thumb that an accounting teacher at my college once gave me:
Every product you buy has an estimated replacement lifespan. Computers, for example, get to be about 3 years old before people tend to replace them. The cost of leases are usually calculated using this estimated lifespan.
If you are going to use a product for LESS THAN 75% of it's replacement lifespan (switch computers every 2 years), you are better off leasing. Anything longer than that, and you're probably better off buying.
Please note that the replacement lifespan is quite different than the actual product lifespan (Computers working 8+ years, etc).
Part of the difference we're talking about here is audience. To scientists flawed concepts still have merit... Much the way buggy code can still provide insight for programmers.
However, outside of that audience, that value is lost. A conceptually great program that is poorly executed would be considered garbage to the layperson. They're told that they'll be able to do XYZ, but they really can't.
In this case, they think they'll have better understanding about ZYX, and get upset when they don't.
Like it or not, people expect science to be exact and correct even though it often isn't.
One day I'm reading an article about how dark matter doesn't need to exist to explain the Universe. A couple days later, an article comes out saying how black holes don't exist, and they're starts made up of dark matter.
I'm sorry, but to the layperson, it looks like these bozos don't know what they're talking about.
I'm not sure if this is supposed to be FUNNY, a TROLL or someone just being STUPID.
It's April Fools, but I have met some people that actually believe this works. Those are also the same people that wrote the worst code I've seen in my life.
Slashdot Mirror
A bookstore with the most flexible return policy ever! I'm in!
~D
Nope. But that's what you get when you have two stories on almost the same subject right in a row.
~D
So, here's what I'm trying to figure out... What's the idea behind having 1 SPU "reserved for redundancy"? Wouldn't it be better to have all the SPUs available, and then them for redundancy as needed? Thoughts? Ideas?
~D
Pardon me, you're right. I'm using old data.
The Wiki on the Cell has also been changed...
"It is unclear how many processing units will be incorporated into either the PlayStation 3 or workstations." is one change.
Also
"The PE is not the primary processor for the system, but acts as a controller for the other eight SPUs" is a bit different than what I've read in the past. This means that the SPUs are more general-purpose (And thus the amount of parallelism is more flexible) than I recall being told.
Time for me to do more research.
~D
I think you made my point for me... It's all marketing hype. It really is easier said than done. Trust me, at E3, they're going to say an awful lot. They have a whole year to "go back on" what they said at the show and make "changes" to what they're planning on implementing now.
~D
Historically, though, being first to market isn't such an advantage in the console market. Look at what happened to Dreamcast.
~D
I think you may be a little confused... The PS3 has 3 cells. Each cell has a primary core, plus SPEs. That means each cell is designed to both serial and parallel code.
Also, gaming is a specific arena where parallel processing is an absolute godsend. Having an army of SPEs doing a transformation on every pixel in an image is vastly more efficient than having CPUs do it. When your system architecture is built and your code is specifically written for that kind of parallelism, software that has intensive parallel-friendly tasks should perform very admarably.
~D
Has anyone noticed how the concave shape of the 360 appears to be a perfect match for the convex shape of the PS3?
Now the question is: Can you stack these two without the generated heat setting something ablaze?
~D
Considering that the PS3 is better than the 360 on every front you listed, I'm guessing the sony marketing guys just watched the xbox release, and upped all their numbers for E3.
~D
The problem is, you're trying to recieve a consise message from an advertisement that does not have one. There is no underlying message, and they are suggesting very little.
The whole idea of this kind of advertising is to get your attention, and then show you a brand name. It's called brand exposure, and it's surprisingly effective. That's it, nothing more.
~D
The game code itself doesn't read the file, it relies on some form of BIOS to do that. So all you really need to do is have the system read from the memory stick instead of the UMD. For these guys, that's no problem.
~D
Am I the only geek in here that's had enough experience with technology that the most advanced gadget I want in my car is the sterio system?
~D
I disagree with your comments about the API. In this example SDL_GetTicks() is not a very clear function name. A well placed comment saying what the function is really doing will let me know if I need to go to the documentation or not.
// time in milisec
For example:
last_update = SDL_GetTicks();
Tells me enough about what the function does, and what kind of data it returns, that I don't have to go all the way to the API documentation.
~D
So, do you think this is really about preventing the child from having any form of contact with the game under any circumstances?
No, it's about putting reasonable bounds in place, like locking away liquor. If the child still gains access to it while they know it's out-of-bounds, then there are concequences.
~D
You missed the simple, elegant, missing word:
"It just barely works"
~D
Well, if they marry each other then the whole thing could be done for 10k. :O
~D
You missed the George Mason incident earlier this year. Maybe that was only tens of thousands of records though.
~D
I'm sorry, but I don't trust the process. Ok, so they're not buying or selling items directly. They still control the origination of the commodities and can control the value of the items by making artifical changes to the supply of the market.
~D
So, basically, Sony is able to create their own economy by creating, manipulating, or removing (virtual) commodities at a whim, and then get/give real money for them?
Is it just me, or is this INSANE?
~D
Here's the leasing rule of thumb that an accounting teacher at my college once gave me:
Every product you buy has an estimated replacement lifespan. Computers, for example, get to be about 3 years old before people tend to replace them. The cost of leases are usually calculated using this estimated lifespan.
If you are going to use a product for LESS THAN 75% of it's replacement lifespan (switch computers every 2 years), you are better off leasing. Anything longer than that, and you're probably better off buying.
Please note that the replacement lifespan is quite different than the actual product lifespan (Computers working 8+ years, etc).
~D
Your argument doesn't apply when talking about objective phenomenon such as celestial bodies.
Part of the difference we're talking about here is audience. To scientists flawed concepts still have merit... Much the way buggy code can still provide insight for programmers.
However, outside of that audience, that value is lost. A conceptually great program that is poorly executed would be considered garbage to the layperson. They're told that they'll be able to do XYZ, but they really can't.
In this case, they think they'll have better understanding about ZYX, and get upset when they don't.
Like it or not, people expect science to be exact and correct even though it often isn't.
~D
I am well aware of the scientific process. Please keep in mind that:
10 Scientists tell you they know how the Universe works.
9 of them, eventually, are disproven, 1 is accepted. Yay! Progress.
Unforunately that means that 9/10ths of what people were told is garbage... No wonder we're a little jaded.
~D
One day I'm reading an article about how dark matter doesn't need to exist to explain the Universe. A couple days later, an article comes out saying how black holes don't exist, and they're starts made up of dark matter.
I'm sorry, but to the layperson, it looks like these bozos don't know what they're talking about.
~D
I'm not sure if this is supposed to be FUNNY, a TROLL or someone just being STUPID.
It's April Fools, but I have met some people that actually believe this works. Those are also the same people that wrote the worst code I've seen in my life.
~D