ppc: load 64 bit constant 5 instructions = 20 bytes to load an 8 byte value
x86_64: load 64 bit constant 1 instruction = 10 bytes to load an 8 byte value
not only do you have considerably more scheduling issues on ppc, but you also have to burn far more memory bandwidth to do it. suddenly ppc doesn't look so hot anymore.
not to mention ppc has some rather silly design misfeatures which make position independent code not only slower than other processors, but ridiculously difficult to debug as well. the blame for these misfeatures lie solely with the hardware guys.
the problem is that for itanium to be efficient, compilers have to be incredibly sophisticated. they have to get branch prediction near perfect (prediction miss is _expensive_ on itanium). they have to be excellent with huge register files. they have to be good with reordering, and they have to be good with VLIW packing.
no compiler today is good at virtually any of this, even intel's own compilers.:-)
in the end, VLIW turned out to be a very, very hard nut to crack. intel made a huge gamble on it and lost.
sparc is no longer competetive. its major design win (register windows) turned out to be a liability long term. memory got faster, caches got better. register windows became obsolete, and sparc stopped scaling competetively.
powerpc is an overall better (faster, cheaper, more scalable) design than sparc. it doesnt make the same kind of assumptions as sparc did and thus powerpc was able to scale along with the rest of the industry as technology developed -- powerpc was able to take advantage of new developments in silicon without having to lug around old cruft; powerpc had few fundamental dependencies on the underlying technologies. sparc has baggage (register windows) which are no longer architectural wins.
powerpc assembler is really unpleasant to read and debug though. load/store is a pita. two instructions to load a single 32 bit value and _five_ for a 64 bit value(!). now consider that a compiler can completely reorder those instructions. not nice for readability.
Itaniums get such a bad rep here on Slashdot because its cool to do so. Itaniums are made by the "big guy", Intel. If they were made by AMD they would not get the same rap as they do.
bullshit. itaniums get a bad rep on slashdot for any number of reasons, and they cannot all be distilled down to "because it's hip and trendy to bash itanium".
slashdot would still be bashing itanium if it were from amd.
few people like paying $1000+ for a cpu alone, for example.
itanium is a niche processor filling a tiny tiny tiny market. and it is already hitting scaling issues.
itanium also has yet to deliver on most of its performance promises. just about the only one it's delivered on so far is memory bandwidth:-)
intel gambled itanium's future on its dependency of a number of risky and unproven technologies (eg VLIW). in order for itanium to succeed, ALL of these technologies had to succeed. instead what happened is virtually NONE of them did.
it's quite telling when a lot of the intel engineers and scientists involved with itanium are calling it a huge mistake. the p4 guys aren't impressed either:-)
itanium is doomed longterm. most of intel's itanium partners have long since bailed on the architecture, most projects for itanium have been killed off (including windows), which guarantees itanium has no longterm future.
some lessons from itanium may be rolled into other intel mainstream products, but as a product itself itanium's days are numbered. itanium has been a huge black hole sucking billions of r&d from intel while amd has been constantly chipping away at intel's market share with x86_64. itanium has never turned a profit, in over a decade of development on the damned thing. it's only a matter of time before stockholders demand itanium be hauled out to the barn and given both barrels.
most people who have studied itanium closely conclude itanium is an r&d project that should have remained in the labs as pure r&d, and never turned into a product.
However the latest oh-so-fashionable fad is to blame the victim, at least for computer viruses. It's hip, it's trendy.
It's like in the 1980s it was fashionable to blame rape victims because they "were just asking for it". Eventually people realized how silly it was, but it took a couple years.
By creating an incredibly hostile atmosphere for software developers (you never know if the next release of Windows is going to bundle some application that will make your product worthless), they sabotage their own employment market.
Such predatory behavior kills interest in CS, which may partly explain the decline in number of applicants.
cringely predicted an entirely different thing. he predicted microsoft wouldn't be able to keep up with virus signatures. he didn't predict that microsoft would deliberately sabotage the product.
when you consider the tens of millions of infected computers, it would certainly be a deterrent to would be future 1337 hax0rz. he'd be paying off the fines for the rest of his life.
While Microsoft is the application coder, Sysadmins are end users. (every microsoft windows 2000, linux, bsd, osx, etc. user is a sysadmin)
So you most definitely are blaming the victim.
Also, lots of computer attacks are bruteforce (and are on the increase). Going to start prosecuting lock manufacturers because someone used an acetelyne torch to cut through your lock or C4 to blow it up?
Slashdot Mirror
ok, lets try this again.
ppc: load 64 bit constant
5 instructions = 20 bytes to load an 8 byte value
x86_64: load 64 bit constant
1 instruction = 10 bytes to load an 8 byte value
not only do you have considerably more scheduling issues on ppc, but you also have to burn far more memory bandwidth to do it. suddenly ppc doesn't look so hot anymore.
not to mention ppc has some rather silly design misfeatures which make position independent code not only slower than other processors, but ridiculously difficult to debug as well. the blame for these misfeatures lie solely with the hardware guys.
doom3 was incredibly, eye-wateringly repetetive. great graphics and sound solely do not a great game make.
doom3 was a tech demo, not a game. a $55 tech demo, yes. but a game, no.
hopefully the doom3 licensees will make real games with the engine. (et:qw, quake4, etc)
So epic is better at marketing the unreal name than id is. Nothing new here, move along...
i offer my deepest condolences on both counts.
the problem is that for itanium to be efficient, compilers have to be incredibly sophisticated. they have to get branch prediction near perfect (prediction miss is _expensive_ on itanium). they have to be excellent with huge register files. they have to be good with reordering, and they have to be good with VLIW packing.
:-)
no compiler today is good at virtually any of this, even intel's own compilers.
in the end, VLIW turned out to be a very, very hard nut to crack. intel made a huge gamble on it and lost.
thing is, powerpc really was a huge performance win over 68k. it also was not much more expensive than 68k
itanium is not a huge performance win over x86. in many cases it is a huge performance loss. and it is supremely expensive.
intel has been beating the itanium dead horse for over a decade now.
you do remember microsoft cancelled windows xp for itanium, don't you?
sparc is no longer competetive. its major design win (register windows) turned out to be a liability long term. memory got faster, caches got better. register windows became obsolete, and sparc stopped scaling competetively.
powerpc is an overall better (faster, cheaper, more scalable) design than sparc. it doesnt make the same kind of assumptions as sparc did and thus powerpc was able to scale along with the rest of the industry as technology developed -- powerpc was able to take advantage of new developments in silicon without having to lug around old cruft; powerpc had few fundamental dependencies on the underlying technologies. sparc has baggage (register windows) which are no longer architectural wins.
powerpc assembler is really unpleasant to read and debug though. load/store is a pita. two instructions to load a single 32 bit value and _five_ for a 64 bit value(!). now consider that a compiler can completely reorder those instructions. not nice for readability.
itanium is a very pretty architecture, but "pretty" rarely means performance.
Itaniums get such a bad rep here on Slashdot because its cool to do so. Itaniums are made by the "big guy", Intel. If they were made by AMD they would not get the same rap as they do.
:-)
:-)
bullshit. itaniums get a bad rep on slashdot for any number of reasons, and they cannot all be distilled down to "because it's hip and trendy to bash itanium".
slashdot would still be bashing itanium if it were from amd.
few people like paying $1000+ for a cpu alone, for example.
itanium is a niche processor filling a tiny tiny tiny market. and it is already hitting scaling issues.
itanium also has yet to deliver on most of its performance promises. just about the only one it's delivered on so far is memory bandwidth
intel gambled itanium's future on its dependency of a number of risky and unproven technologies (eg VLIW). in order for itanium to succeed, ALL of these technologies had to succeed. instead what happened is virtually NONE of them did.
it's quite telling when a lot of the intel engineers and scientists involved with itanium are calling it a huge mistake. the p4 guys aren't impressed either
itanium is doomed longterm. most of intel's itanium partners have long since bailed on the architecture, most projects for itanium have been killed off (including windows), which guarantees itanium has no longterm future.
some lessons from itanium may be rolled into other intel mainstream products, but as a product itself itanium's days are numbered. itanium has been a huge black hole sucking billions of r&d from intel while amd has been constantly chipping away at intel's market share with x86_64. itanium has never turned a profit, in over a decade of development on the damned thing. it's only a matter of time before stockholders demand itanium be hauled out to the barn and given both barrels.
most people who have studied itanium closely conclude itanium is an r&d project that should have remained in the labs as pure r&d, and never turned into a product.
Why should he trust intel again? They already blew a huge wad on intel's promises for itanium 1, which turned out to be complete bullshit.
FWIW itanium 2 is better than itanium 1, but it's *still* a mediocre processor.
The Grandmother = Lowest Common Denominator
and with OSX, that's pretty damn low. OSX is like computing with a straitjacket on. you do things apple's way or not at all.
they think this is such a clever way of avoiding saying microwaves ...
wonder how long it will take for websites to start collecting and distributing pr0n collected from these scanners.
from tfa:
Simon Stringer, managing director of QinetiQ's security division, said: "We have been asked to deploy some of this equipment.
"It would certainly assist in preventing this sort of thing from happening again.
No, it won't. Sorry. It just won't. Would you be willing to stake your company on it? Yeah. Didn't think so.
and more vehemently hostile towards alternative OSes than any other sort of OS zealot than I've ever seen
You've never seen OSX zealots then. They're far worse than amiga and sgi zealots.
...for $12? righto.
Yeah, it's pretty funny how many punks think 'caller id block' makes them untouchable.
We had the police arrest some local script kiddies by filing a criminal complaint and getting the phone records subpoena'd.
However the latest oh-so-fashionable fad is to blame the victim, at least for computer viruses. It's hip, it's trendy.
It's like in the 1980s it was fashionable to blame rape victims because they "were just asking for it". Eventually people realized how silly it was, but it took a couple years.
Blaming the victim is so 1980s. It didnt work do blame rape victims and it wont work to blame end users.
By creating an incredibly hostile atmosphere for software developers (you never know if the next release of Windows is going to bundle some application that will make your product worthless), they sabotage their own employment market.
Such predatory behavior kills interest in CS, which may partly explain the decline in number of applicants.
cringely predicted an entirely different thing. he predicted microsoft wouldn't be able to keep up with virus signatures. he didn't predict that microsoft would deliberately sabotage the product.
so cringely's prediction FAILS.
However, SCO's real measurable assets have a value >0, which is why the stock is not 0.
If you read the recent SCO filings, you realize this may no longer be the case very soon.
when you consider the tens of millions of infected computers, it would certainly be a deterrent to would be future 1337 hax0rz. he'd be paying off the fines for the rest of his life.
Linus is responsible for kernel holes.
Which puts the blame for being left vulnerable to either the sysadmin (the one who has root access)
Thank you for acknowledging and confirming my statement that all linux users are sysadmins.
every linux user is a sysadmin.
or... i guess you're advocating that it's time to prosecute linus torvalds?
Prosecute application coders and lazy sysadmins
While Microsoft is the application coder, Sysadmins are end users. (every microsoft windows 2000, linux, bsd, osx, etc. user is a sysadmin)
So you most definitely are blaming the victim.
Also, lots of computer attacks are bruteforce (and are on the increase). Going to start prosecuting lock manufacturers because someone used an acetelyne torch to cut through your lock or C4 to blow it up?