Cache uses very little power, all things considered. This is why the Pentium M is so cache heavy; it's the lowest power way to increase performance, although not the fewest-transistor way. (Also note that the power architecture of the Pentium M's cache is seriously cool.. at the cost of an additional cycle of latency for L2 access, three quarters of the cache uses basically no power at all times. I'd give my left foot for IBM to license/use that in the 750GX.) Also, keep in mind that the G5 is meant for dual processor machines, meaning ~95 processor watts high end. The P4 is for single processor boxes, giving ~105 watts high end. This is not so big a difference, and the power-hungry chipset on the G5 probably comes close to making up for it. The G5 is a much better engineered machine, in my personal opinion, and seems to get a lot more for its watts... but the power consumption of a high end G5 and a high end Prescott should be pretty damn close to identical.
Re:Performance doesn't come directly from 64 bits
on
Is Prescott 64-bit?
·
· Score: 2, Funny
Stick with it! It took me a while, but I rescued the princess in the end.
Re:Now I know we love apple and hate the RIAA, but
on
Beatles Bite Apple
·
· Score: 1
Apple Computers has more money in the bank than the value of Apple Records. In fact, the value of the Apple Computer brand towards selling music at the iTMS is probably more than the value of Apple Records. It would make more sense for Apple C. to do what they did, with the fallback plan of buying Apple Records wholesale if necessary, than to create a new brand for their music stuff.
Dual core Power5 with 32MB L3 cache on board takes 500 watts. And makes the Itanics look light weight.
Re:currently available configurations
on
G5s Start Shipping
·
· Score: 1
Okay, let's try this again. Get the 800MHz bus, so the Dell has some chance, however small, of competing with the G5. Add Gigabit ethernet, the modem (well, you're better off subtracting it from the Mac, but let's be fair), a decent keyboard, XP Professional (still not nearly as good as OSX, but at least usable), Microsoft Works (I won't say that Appleworks is as good as Word, but many people I know use it regularly), Microsoft Digital Media Edition Plus! Pack (it's shit, but it represents iPhoto, iMovie, and the other great OS X software free)... and you get $2748 (no monitor), compared to $2399 for the fast single-processor mac...
Um, I think just about anyone on slashdot agrees with calling stealing music theft. The issue is that it is impossible to steal music; you can call it theft, but it doesn't happen, so you have little to talk about. It is possible to steal media containing music, and it is possible to infringe upon copyrighted music, but those are both distinct from the hypothetical stealing of music itself. Again, no one says that they should rename "How the Grinch Stole Christmas"; it's just understood that we're using terms that refer to a fantastical world where such things can be.
Think of the difference between Toshiba's jog wheel and Sony's jog wheel (both on their handhelds). The sony one is a real wheel, like a mouse wheel. The toshiba one doesn't actually go in circles; you pull it up or down and, while it's in that position, it scrolls. When you let go, a spring pulls it back to standard position. I actually find the toshiba one much more usable for scrolling, and wish they made a mouse with it. The only real downside is that you can't comfortable click the wheel button while scrolling on the toshiba, and your scroll speed is determined by software.
Re:Former perl, python, java geek gone to Ruby
on
Ruby 1.8.0 Released
·
· Score: 1
While you do mention that this is not an advantage over Smalltalk, it should be emphasized how unexcited smalltalkers will be by the reinvention of internal iterators.
Maybe it's changed since then, but when Iomega started selling Microdrives, they had lower power usage, a lower rotation rate, higher latency, and lower transfer speed.
In order of best to worst: 1) Code that works 2) Code that doesn't work 3) Code that endangers my computer If Java can eliminate the bottom class, more power to it. If a program doesn't work, you don't use it. If a program is unsafe, you may get screwed before you figure out not to use it. Again, which is the real danger here?
If you really care about what the original means, why not read the original? You hear about these people who spend an hour a night studying the bible, their whole life, and they can't be bothered to learn greek or hebrew? I just don't get it. Not a troll.
Must be magic numbers. We do multiplication through shifting, to find texture values at a given point... that is, on a 512x512 texture, pixel (17,41) is found as 17+(419), rather than 17+(41*512). The cards don't even have a multiplier in this location in the pipeline. Scaling the texture down to 254x254, and then adding a border, may work, and I can't imagine that there would be much of a quality difference. The actual value of the border pixels is an excersize for the reader.;-) Note that the solution being discussed on ATI uses a centroid-based point selection method, so you can guarantee that all points are on one side of the line; the nvidia solution, using the pixel shader, is probably just a hack to notice and correct the problem by looking at individual pixels, each time, rather than not generating the problem ones in the first place.
You should. For reasons I don't fully comprehend, you don't. I suspect (having written and optimized a polygon-based renderer) that they're cheating a little here, which is far from rare (or inherently bad). Let's say an edge falls through the middle of the pixel. On the right hand side of that edge (say), the textures are well defined. On the left hand side, they're not. Now, how to we pick points for antialiasing? In the naive 4x model, we treat each pixel as four pixels, and things work fine. For each of those four pixels, the line either passes to the left of the center, in which case we use the texture of this object, or the line passes to the right of the center, in which case we use the texture of the underlying object. For this to work, we need to replicate the z-buffer 4x also; that is, we're basically making an image 4x bigger (2x by 2x), and then scaling it down. More advanced FSAA algorithms (like the horridly-named Quincunx, or whatever) use strange numbers of points (5, here), and use different selection criteria. Also, they minimize scaling up of the z-buffer, since it's expensive and should make little difference. That means, though, that for a given point we may not be able to tell if the line is to its left or its right; we guess, and if we guess wrong, there's a chance we take from the undefined pixel. Again, I'm not claiming to know the exact algorithmic details, but I'm pretty sure this is the root problem.
Slashdot Mirror
Dude, it's a laptop hard drive... how hard are you tapping that poor trackpad?
Cache uses very little power, all things considered. This is why the Pentium M is so cache heavy; it's the lowest power way to increase performance, although not the fewest-transistor way. (Also note that the power architecture of the Pentium M's cache is seriously cool.. at the cost of an additional cycle of latency for L2 access, three quarters of the cache uses basically no power at all times. I'd give my left foot for IBM to license/use that in the 750GX.) Also, keep in mind that the G5 is meant for dual processor machines, meaning ~95 processor watts high end. The P4 is for single processor boxes, giving ~105 watts high end. This is not so big a difference, and the power-hungry chipset on the G5 probably comes close to making up for it. The G5 is a much better engineered machine, in my personal opinion, and seems to get a lot more for its watts... but the power consumption of a high end G5 and a high end Prescott should be pretty damn close to identical.
I Suspect Another Is Somewhat Appropriate...
Stick with it! It took me a while, but I rescued the princess in the end.
Apple Computers has more money in the bank than the value of Apple Records. In fact, the value of the Apple Computer brand towards selling music at the iTMS is probably more than the value of Apple Records. It would make more sense for Apple C. to do what they did, with the fallback plan of buying Apple Records wholesale if necessary, than to create a new brand for their music stuff.
Does it come with an off switch?
Dual core Power5 with 32MB L3 cache on board takes 500 watts. And makes the Itanics look light weight.
Okay, let's try this again. Get the 800MHz bus, so the Dell has some chance, however small, of competing with the G5. Add Gigabit ethernet, the modem (well, you're better off subtracting it from the Mac, but let's be fair), a decent keyboard, XP Professional (still not nearly as good as OSX, but at least usable), Microsoft Works (I won't say that Appleworks is as good as Word, but many people I know use it regularly), Microsoft Digital Media Edition Plus! Pack (it's shit, but it represents iPhoto, iMovie, and the other great OS X software free)... and you get $2748 (no monitor), compared to $2399 for the fast single-processor mac...
What was your point again?
Um, I think just about anyone on slashdot agrees with calling stealing music theft. The issue is that it is impossible to steal music; you can call it theft, but it doesn't happen, so you have little to talk about. It is possible to steal media containing music, and it is possible to infringe upon copyrighted music, but those are both distinct from the hypothetical stealing of music itself. Again, no one says that they should rename "How the Grinch Stole Christmas"; it's just understood that we're using terms that refer to a fantastical world where such things can be.
yes, if you equip it with a hack saw and a roll of duct tape so I can put it in my pocket and reassemble it afterwards.
Think of the difference between Toshiba's jog wheel and Sony's jog wheel (both on their handhelds). The sony one is a real wheel, like a mouse wheel. The toshiba one doesn't actually go in circles; you pull it up or down and, while it's in that position, it scrolls. When you let go, a spring pulls it back to standard position. I actually find the toshiba one much more usable for scrolling, and wish they made a mouse with it. The only real downside is that you can't comfortable click the wheel button while scrolling on the toshiba, and your scroll speed is determined by software.
While you do mention that this is not an advantage over Smalltalk, it should be emphasized how unexcited smalltalkers will be by the reinvention of internal iterators.
Maybe it's changed since then, but when Iomega started selling Microdrives, they had lower power usage, a lower rotation rate, higher latency, and lower transfer speed.
... panama!
Quick summary: Our software is good, thrashing is bad, and exceptions are still screwy.
Anyone know of a similar OS X program?
I'm glad that you manage to find enjoyment in literature even without comprehension of written sarcasm.
Cause it says the same thing in those langauges, that is the definition of translation.
Virgin? Unicorn?
In order of best to worst:
1) Code that works
2) Code that doesn't work
3) Code that endangers my computer
If Java can eliminate the bottom class, more power to it. If a program doesn't work, you don't use it. If a program is unsafe, you may get screwed before you figure out not to use it. Again, which is the real danger here?
www.colorforth.com
I was just in 2050 last week, and I can promise there's nothing to worry about. /beep
If you really care about what the original means, why not read the original? You hear about these people who spend an hour a night studying the bible, their whole life, and they can't be bothered to learn greek or hebrew? I just don't get it. Not a troll.
Must be magic numbers. We do multiplication through shifting, to find texture values at a given point... that is, on a 512x512 texture, pixel (17,41) is found as 17+(419), rather than 17+(41*512). The cards don't even have a multiplier in this location in the pipeline. Scaling the texture down to 254x254, and then adding a border, may work, and I can't imagine that there would be much of a quality difference. The actual value of the border pixels is an excersize for the reader. ;-) Note that the solution being discussed on ATI uses a centroid-based point selection method, so you can guarantee that all points are on one side of the line; the nvidia solution, using the pixel shader, is probably just a hack to notice and correct the problem by looking at individual pixels, each time, rather than not generating the problem ones in the first place.
Swing! Then again, if Swing is the answer, what was the question?
You should. For reasons I don't fully comprehend, you don't. I suspect (having written and optimized a polygon-based renderer) that they're cheating a little here, which is far from rare (or inherently bad). Let's say an edge falls through the middle of the pixel. On the right hand side of that edge (say), the textures are well defined. On the left hand side, they're not. Now, how to we pick points for antialiasing? In the naive 4x model, we treat each pixel as four pixels, and things work fine. For each of those four pixels, the line either passes to the left of the center, in which case we use the texture of this object, or the line passes to the right of the center, in which case we use the texture of the underlying object. For this to work, we need to replicate the z-buffer 4x also; that is, we're basically making an image 4x bigger (2x by 2x), and then scaling it down. More advanced FSAA algorithms (like the horridly-named Quincunx, or whatever) use strange numbers of points (5, here), and use different selection criteria. Also, they minimize scaling up of the z-buffer, since it's expensive and should make little difference. That means, though, that for a given point we may not be able to tell if the line is to its left or its right; we guess, and if we guess wrong, there's a chance we take from the undefined pixel. Again, I'm not claiming to know the exact algorithmic details, but I'm pretty sure this is the root problem.