It's certainly going to be good for real world apps (I just read a review that pointed out it's essentially like having an L3 cache in the chipset), but I'd have thought it'd show up in a synthetic streaming memory benchmark also - unless the CPU was already achieving it's theoretical max (which seems unlikely).
The Tom's article says that the chipset does speculative data prefetch to improve memory-CPU throughput by up to 20%...cool... but why are they bothering to do this when Athlon is about to be replaced by Palomino that has hardware prefetch built in?
I don't know what my ISP's TOS are, but in practice I don't seem to be getting disconnected based on connection time, although I've had a few download sites that will do that. The way I workaround that one is to telnet into my ISP (Solaris shell account is for free) and download into the/tmp directory where there's no file quota limits, then ftp from the ISP back into my Linux box and transfer it. I normally leave an "at" job running to delete the big temp file a few hours later just in case they'd get upset - never had any complaints about doing this yet, which surprises me!
You're lucky to even get that 48Kbps dial-up connection - I'm too far from the CO too, and only get 28.8Kbps. You probably know it, but you *would* be able to get IDSL (144/144), since the range is considerably further, although it's a lot more expensive than ADSL (figure around $70/mo at the cheapest, maybe more depending on the provider).
Well for some poeople the cost difference between dial-up (anywhere from free to $20-25/mo) and DSL ($40 or so) is going to be an issue, even if it is available.
There's also the fact that even where DSL is "available", you may be too far from the CO (like me) to be able to get it, and there's not much that can be done about that, even if the max distance can be extended - there's always going to be people out of range.
I've just got a 28.8 modem, and if there's anything big (or a bunch of them) I want to download I just add them to a script ("getem"!) using wget that I'll run when I'm done using the net. If anything fails wget can continue from where it stopped if there was any error.
I think most ISP's (like mine) will disconnect you anyway after enough dead time - I can't think I've ever seen the connection still up in the morning.
Yeah, the key is to find a movie reviewer who's taste consistently agrees with your own. I got put off seeeing a lot movies I'd probably have enjoyed before I also settled on Roger Ebert's web site as the only one I pay attention too.
Yeah, I saw "A Knight's Tale" on friday, and it was basically just bland - I didn't get drawn into the story at all, and it really seemed that Heath Ledger's character should have ended up with the blacksmith chick, not the high maintenance society bimbo he ended up with where there was zero chemistry (i.e. lousy acting). I've never read the book - it was probably a whole lot more fun, as Chaucer can be quite raunchy.
Peak work done per clock or price per clock are meaningless. You should simply look at what you can get for your money at a particulat price and see which is faster. If a $1000 PC beats a $1000 Mac running your application, then who cares what clock speed they're running at - it's irrelevant.
Sure he can choose whatever licencing terms he chooses, but I don't think he can *retroactively* change the terms of a licence unless that was a right he originally claimed. If he want's to insist that his existing licence restricts modification and redistribution, that that be something he'd have to fight in court if needed - the licence itself already exists, and IANAL but I doubt a court would agree with the way he now wants it interpreted.
The point is that gcc is going to be going a different amount of work compiling the same program for different targets due to the different work done in the code generator and optimizer for the different targets; therefore, if you want to compage gcc running on different platforms, you should compare them compiling to the same target (i.e doing a cross compile - which is what he did) rather than compiling to the native target.
No... you should test the boxes the way you are expecting to use them. If you expect to write code in C and compile with gcc, then test that. If you are expecting to write hand optimized assembler than test that!
oops - i replied before I even read the article! my comments still stand, but it seems he was doing cross compile not native compilation, so it was a fair benchmark.
I disagree. It's easier for a code generator/optimizer to generate code for a simple/orthogonal (RISC) instruction set than for a CISC one. I agree though that compiling native code isn't a good benchmark since it's doing different things on different machines - better just to compile for, say, x86 on all platforms if you want to use it as a benchmark.
Still I think the browser does give AOL a lot of leverage because if your 100 free hours AOL coaster automatically installs and uses Netscape/Mozilla, then I think there are going to be a lot of AOL users who will simply use it. Apparently the reason so many people use archaic browsers is because they just used whatever came with the OS/whatever, and never consider to switch or upgrade.
I'd really like to see instant messanging and streaming video become standardized. They're really no different than the telephone or television - they're part of our national infrastructure (as, for that matter, are electronic interchange/document formats).
Re:Games: XFree86 with DRI, or Linux FBDev?
on
XFree 4.1.0 Out
·
· Score: 2
Probably what would make more sense would be to ditch the kernel fb drivers and instead have XF86 export a low level API to it's drivers that could be used for games and a frame buffer device.
Interesting - thanks. It Microsoft and AOL do come to an agreement over streaming video and IM standards, then I'd like to see the DoJ step in and force them to open these standards. That'd be a tough combination for providers to ignore and continue to support competetive standards. Probably the smartest thing Real could do would be to open up their standard which would widen their user base and possibly sucker Microsoft into supporting it.
Or it could just be that they're feeling the effects of the economic downturn like ever other goddamn company on the planet.
Oh no, CD's are so reasonably priced that they're going to be the last thing people cut back on! Hell, I'd cut back on food before I cut back on mailing my pay checks to the record labels!
Considering the amount of money, rebates and restructurings they've invested in MSN, I'd say it's doing pretty badly. MSN has about 5M users, same as tiny Earthlink, while AOL had 25M+ as of sometime late last year.
If MSMN has the same number of users as AIM, then that still puts them at less than half of AOL's AIM+ICQ user base, which is split about exactly 50/50.
Your AOL browser client 6% market share conveniently forgets that for years AOL has been available as a straight ISP and distributes IE under a previous agreement. With 25M AOL users, and only 50-60% of the 100M US households net connected, I'd say that's pretty damn significant!
You don't think the browser used by AOL is significant here... that what exactly else do you think is the leverage AOL is using, as a direct competitor to MSN, to get on the Windows desktop?!!!
Slashdot may be anti-Microsoft, but if you want to defend them then at least try to use some real fascts!
Funny, but it looked like a classic Wall St. bubble to me. Maybe Microsoft engineered the biotech crash of a few years back, and also have aliens been kept alive in the basement at Redmond.
Slashdot Mirror
/* 2,191 lines of complete and utter shit coming up... */
It's certainly going to be good for real world apps (I just read a review that pointed out it's essentially like having an L3 cache in the chipset), but I'd have thought it'd show up in a synthetic streaming memory benchmark also - unless the CPU was already achieving it's theoretical max (which seems unlikely).
Thanks - good point! They go together rather well. It'll be interesting to see the Sandra scores for this one.
The Tom's article says that the chipset does speculative data prefetch to improve memory-CPU throughput by up to 20%...cool... but why are they bothering to do this when Athlon is about to be replaced by Palomino that has hardware prefetch built in?
I don't know what my ISP's TOS are, but in practice I don't seem to be getting disconnected based on connection time, although I've had a few download sites that will do that. The way I workaround that one is to telnet into my ISP (Solaris shell account is for free) and download into the /tmp directory where there's no file quota limits, then ftp from the ISP back into my Linux box and transfer it. I normally leave an "at" job running to delete the big temp file a few hours later just in case they'd get upset - never had any complaints about doing this yet, which surprises me!
You're lucky to even get that 48Kbps dial-up connection - I'm too far from the CO too, and only get 28.8Kbps. You probably know it, but you *would* be able to get IDSL (144/144), since the range is considerably further, although it's a lot more expensive than ADSL (figure around $70/mo at the cheapest, maybe more depending on the provider).
Well for some poeople the cost difference between dial-up (anywhere from free to $20-25/mo) and DSL ($40 or so) is going to be an issue, even if it is available.
:-(
There's also the fact that even where DSL is "available", you may be too far from the CO (like me) to be able to get it, and there's not much that can be done about that, even if the max distance can be extended - there's always going to be people out of range.
No cable access here either
Who need's to disconnect?
I've just got a 28.8 modem, and if there's anything big (or a bunch of them) I want to download I just add them to a script ("getem"!) using wget that I'll run when I'm done using the net. If anything fails wget can continue from where it stopped if there was any error.
I think most ISP's (like mine) will disconnect you anyway after enough dead time - I can't think I've ever seen the connection still up in the morning.
Yeah, the key is to find a movie reviewer who's taste consistently agrees with your own. I got put off seeeing a lot movies I'd probably have enjoyed before I also settled on Roger Ebert's web site as the only one I pay attention too.
http://www.suntimes.com/ebert/ebert.html
Yeah, I saw "A Knight's Tale" on friday, and it was basically just bland - I didn't get drawn into the story at all, and it really seemed that Heath Ledger's character should have ended up with the blacksmith chick, not the high maintenance society bimbo he ended up with where there was zero chemistry (i.e. lousy acting). I've never read the book - it was probably a whole lot more fun, as Chaucer can be quite raunchy.
This brings a whole new meaning to "I've got a woody"!
Compiler optimizations patented?! :-(
Do you have any examples?
Surely they can't have patented emitting certain code though - just the algorithms they use to produce it?
Peak work done per clock or price per clock are meaningless. You should simply look at what you can get for your money at a particulat price and see which is faster. If a $1000 PC beats a $1000 Mac running your application, then who cares what clock speed they're running at - it's irrelevant.
Sure he can choose whatever licencing terms he chooses, but I don't think he can *retroactively* change the terms of a licence unless that was a right he originally claimed. If he want's to insist that his existing licence restricts modification and redistribution, that that be something he'd have to fight in court if needed - the licence itself already exists, and IANAL but I doubt a court would agree with the way he now wants it interpreted.
The point is that gcc is going to be going a different amount of work compiling the same program for different targets due to the different work done in the code generator and optimizer for the different targets; therefore, if you want to compage gcc running on different platforms, you should compare them compiling to the same target (i.e doing a cross compile - which is what he did) rather than compiling to the native target.
No... you should test the boxes the way you are expecting to use them. If you expect to write code in C and compile with gcc, then test that. If you are expecting to write hand optimized assembler than test that!
oops - i replied before I even read the article! my comments still stand, but it seems he was doing cross compile not native compilation, so it was a fair benchmark.
I disagree. It's easier for a code generator/optimizer to generate code for a simple/orthogonal (RISC) instruction set than for a CISC one. I agree though that compiling native code isn't a good benchmark since it's doing different things on different machines - better just to compile for, say, x86 on all platforms if you want to use it as a benchmark.
You've got a point there!
Still I think the browser does give AOL a lot of leverage because if your 100 free hours AOL coaster automatically installs and uses Netscape/Mozilla, then I think there are going to be a lot of AOL users who will simply use it. Apparently the reason so many people use archaic browsers is because they just used whatever came with the OS/whatever, and never consider to switch or upgrade.
I'd really like to see instant messanging and streaming video become standardized. They're really no different than the telephone or television - they're part of our national infrastructure (as, for that matter, are electronic interchange/document formats).
Probably what would make more sense would be to ditch the kernel fb drivers and instead have XF86 export a low level API to it's drivers that could be used for games and a frame buffer device.
Interesting - thanks. It Microsoft and AOL do come to an agreement over streaming video and IM standards, then I'd like to see the DoJ step in and force them to open these standards. That'd be a tough combination for providers to ignore and continue to support competetive standards. Probably the smartest thing Real could do would be to open up their standard which would widen their user base and possibly sucker Microsoft into supporting it.
That's scary. That's like what were effectively suicide cleanup crews at Chernobyl running with fuel rods - it'll not hurt you if you're quick!
Poor bastards.
Or it could just be that they're feeling the effects of the economic downturn like ever other goddamn company on the planet.
Oh no, CD's are so reasonably priced that they're going to be the last thing people cut back on! Hell, I'd cut back on food before I cut back on mailing my pay checks to the record labels!
Considering the amount of money, rebates and restructurings they've invested in MSN, I'd say it's doing pretty badly. MSN has about 5M users, same as tiny Earthlink, while AOL had 25M+ as of sometime late last year.
If MSMN has the same number of users as AIM, then that still puts them at less than half of AOL's AIM+ICQ user base, which is split about exactly 50/50.
Your AOL browser client 6% market share conveniently forgets that for years AOL has been available as a straight ISP and distributes IE under a previous agreement. With 25M AOL users, and only 50-60% of the 100M US households net connected, I'd say that's pretty damn significant!
You don't think the browser used by AOL is significant here... that what exactly else do you think is the leverage AOL is using, as a direct competitor to MSN, to get on the Windows desktop?!!!
Slashdot may be anti-Microsoft, but if you want to defend them then at least try to use some real fascts!
Microsoft engineered the dot com crash eh?
Funny, but it looked like a classic Wall St. bubble to me. Maybe Microsoft engineered the biotech crash of a few years back, and also have aliens been kept alive in the basement at Redmond.