... is who actually owns the UNIX copyrights after the sale. I'm not talking about patents, which is what keeps getting brought up - who actually owns the copyright to the latest version of UNIX System V? Is it SCO, who developed it, or Novell, who supposedly owns UNIX, or someone else?
They already do. You realize that until recently, something like 60% of all processors sold were 8-bit microcontrollers, right? User-visible "computers" are a niche.
The US military presence is what allows Japan to maintain the illusion and appearance of pacifism, and relieves them of the need to develop nuclear weapons. A number of prominent Japanese politicians have said that if the US's capability or willingness to defend Japan was in doubt, that Japan would drop the "SDF" fiction and develop long-range offensive weapons, including nuclear arms.
Yes... because all algorithms that anyone actually uses are 100% parallel.
Throwing a shitton of SIMD units on a chip isn't that cool anymore. DSP's have been doing it forever. Real workloads require fast sequential code performance, and a GPU will have truly embarrassing results on such workloads.
Yeah. That'll work great. A chip with a peak of
I/O and memory bandwidth matter more than ever. ARM doesn't have competitive numbers in this area. Combined with their absolutely pathetic core speed (look up the SPECint2000 numbers on a Cortex A9 sometime), they aren't enough for high-performance use, and won't be for a while.
Technically, the monarch has a veto power - the Queen is required to grant Royal Assent for laws to take effect, and has a number of other "reserve powers" that are almost never used (the power to appoint her own choice of Prime Minister, for instance.) In that sense, the UK is less de jure democratic than the US is by quite a bit.
I'd bet x64. IBM doesn't have any good high-end embedded core right now - there's the 440/460, which has decent floating-point and is very low-power but won't cut it as a general-purpose CPU, and then there's the Power7, which is the fastest general-purpose processor in the world, but is thirsty and obscenely expensive.
People keep talking about ARM in the next-gen consoles, but I think the chances of that are slim to none. As far as I know, no currently-existing or roadmapped ARM core is even fast enough to replace a triple-core 970 or a 970 with SIMD units bolted on, much less to emulate one.
That's obsolete. As soon as the Itanium 2 came out (~2003) the processor itself was fast enough to get reasonable performance in software mode, through a subsystem called IA-32 EL. The hardware decoder got scrapped since at that point it was useless.
On Windows, that subsystem was totally transparent. Running a Win32 x86 binary on IA64 would happen just like running one on any other Windows system, with no special steps.
Alpha and IA64 both had compatibility modes. IIRC PPC did as well - I don't remember, not having used it, but I've spent a lot of time with Windows on IA64 and its software compatibility is both very good and very fast.
Every version of Windows since NT 3.1 has run on architectures other than x86. 3.1 ran on MIPS, Alpha, and x86. 3.5 added PPC. 2000 killed those except Alpha (which was internal-only) and added IA64. XP added AMD64. Win8 is killing IA64 and adding ARM.
Actually, MIcrosoft pushed for a MIPS reference architecture (ARCS) to be the successor to the PC architecture. They had some substantial support onboard, but it ended up breaking up due to DEC and a couple of other manufacturers pushing for Alpha to be the processor to be used, and then Compaq leaving and returning support to PC-compatibles.
A declaration by some wing of a splintered popular uprising in an African country, reported by a news source with zero credibility? Is this what Slashdot has come to?:(
IIRC, an ISA slot isn't required for any Windows version. I've run Windows 1.0 on an ISA-less machine before. At least in theory, any modern PC with a floppy drive should be able to run all Windows from 1.0 to 7.
The G5's sucked. They did great at the specfp benchmark, but in real world usage, they were a lot like the Pentium 4 - never lived up to anything close to the hype. The G4 actually had superior IPC at everything except floating point, and if they had been clocked as high as the G5, they would have been generally faster for consumer usage. Apple should have done a higher-clocked G4 for consumer systems, and licensed straight Power4 for Xserve.
Slashdot Mirror
... is who actually owns the UNIX copyrights after the sale. I'm not talking about patents, which is what keeps getting brought up - who actually owns the copyright to the latest version of UNIX System V? Is it SCO, who developed it, or Novell, who supposedly owns UNIX, or someone else?
They already do. You realize that until recently, something like 60% of all processors sold were 8-bit microcontrollers, right? User-visible "computers" are a niche.
But let me guess... you're an Android fanboy.
The US military presence is what allows Japan to maintain the illusion and appearance of pacifism, and relieves them of the need to develop nuclear weapons. A number of prominent Japanese politicians have said that if the US's capability or willingness to defend Japan was in doubt, that Japan would drop the "SDF" fiction and develop long-range offensive weapons, including nuclear arms.
Yes... because all algorithms that anyone actually uses are 100% parallel.
Throwing a shitton of SIMD units on a chip isn't that cool anymore. DSP's have been doing it forever. Real workloads require fast sequential code performance, and a GPU will have truly embarrassing results on such workloads.
Damn you, Slashdot.
"A chip with a peak of below-10GB/s I/O and memory bandwidth will work great with 64GB of RAM and an SSD."
Yeah. That'll work great. A chip with a peak of
I/O and memory bandwidth matter more than ever. ARM doesn't have competitive numbers in this area. Combined with their absolutely pathetic core speed (look up the SPECint2000 numbers on a Cortex A9 sometime), they aren't enough for high-performance use, and won't be for a while.
Technically, the monarch has a veto power - the Queen is required to grant Royal Assent for laws to take effect, and has a number of other "reserve powers" that are almost never used (the power to appoint her own choice of Prime Minister, for instance.) In that sense, the UK is less de jure democratic than the US is by quite a bit.
We elected Clinton with 43% of the vote - 57% of the population voted against him. How is that not "the few ruling the many"?
Or '96. He got roughly 43% in 92, and 49% in 96.
Also, the Bulldozer isn't even a multi-socket chip, except in its Opteron configs, which cost even more.
I'd bet x64. IBM doesn't have any good high-end embedded core right now - there's the 440/460, which has decent floating-point and is very low-power but won't cut it as a general-purpose CPU, and then there's the Power7, which is the fastest general-purpose processor in the world, but is thirsty and obscenely expensive.
People keep talking about ARM in the next-gen consoles, but I think the chances of that are slim to none. As far as I know, no currently-existing or roadmapped ARM core is even fast enough to replace a triple-core 970 or a 970 with SIMD units bolted on, much less to emulate one.
They aren't going to put $1500 worth of CPU's in the next Xbox. Microsoft is willing to sell at a loss, but not that much of a loss.
That's obsolete. As soon as the Itanium 2 came out (~2003) the processor itself was fast enough to get reasonable performance in software mode, through a subsystem called IA-32 EL. The hardware decoder got scrapped since at that point it was useless.
On Windows, that subsystem was totally transparent. Running a Win32 x86 binary on IA64 would happen just like running one on any other Windows system, with no special steps.
Alpha and IA64 both had compatibility modes. IIRC PPC did as well - I don't remember, not having used it, but I've spent a lot of time with Windows on IA64 and its software compatibility is both very good and very fast.
Every version of Windows since NT 3.1 has run on architectures other than x86. 3.1 ran on MIPS, Alpha, and x86. 3.5 added PPC. 2000 killed those except Alpha (which was internal-only) and added IA64. XP added AMD64. Win8 is killing IA64 and adding ARM.
Android isn't really Linux. Linux is basically a bootloader for it. With minimal porting, you could run it on top of FreeBSD or Mach.
Actually, MIcrosoft pushed for a MIPS reference architecture (ARCS) to be the successor to the PC architecture. They had some substantial support onboard, but it ended up breaking up due to DEC and a couple of other manufacturers pushing for Alpha to be the processor to be used, and then Compaq leaving and returning support to PC-compatibles.
No it isn't. Way to lose the whole point, which is predictability of how many cycles an instruction takes.
Not that it matters at this point. VLIW, like in high-performance DSP's and certain niche processors, is the future.
This is right on.
A declaration by some wing of a splintered popular uprising in an African country, reported by a news source with zero credibility? Is this what Slashdot has come to? :(
IIRC, an ISA slot isn't required for any Windows version. I've run Windows 1.0 on an ISA-less machine before. At least in theory, any modern PC with a floppy drive should be able to run all Windows from 1.0 to 7.
Unisys's MCP operating system, the oldest OS still in production, has binary compatibility back to 1961. That is damn impressive.
The absence of Yandex, the #1 search engine in Russia, is also strange.
By most phone OS's, you mean the trendy ones with all the press? Maemo, Meego, Symbian, and webOS are plenty open.
The G5's sucked. They did great at the specfp benchmark, but in real world usage, they were a lot like the Pentium 4 - never lived up to anything close to the hype. The G4 actually had superior IPC at everything except floating point, and if they had been clocked as high as the G5, they would have been generally faster for consumer usage. Apple should have done a higher-clocked G4 for consumer systems, and licensed straight Power4 for Xserve.