Windows Reaches 64-Bits, For OEMs 365
thatrez writes: "Microsoft 's Windows Advanced Server, Limited Edition, is now
available for computers based on Intel's 64-bit Itanium chip. The
Itanium chip supports greater amounts of system memory and offers
stronger floating-point, or mathematical, capabilities than current
32-bit desktop processors. The extra memory support and the
floating-point capabilities increase the performance of Web hosting,
data warehousing and other applications." Now available in this case means that certain OEMs will soon be selling systems loaded with 64-bit Advanced Server, and later other manufacturers will join in. 64-bit versions of XP are expected sometime next year as well.
wow this is great news... (Score:1, Funny)
Re:wow this is great news... (Score:2, Insightful)
Wow...you didn't even need to read the article to figure this one out - it was in the title, of all places:
Windows Reaches 64-Bits, For OEMs
OEM stands for Original Equipment Manufacturer. That leads me to believe that they'll be shipping 64-bit Windows on 64-bit machines. But then, you can't flame someone for doing that...
Re:NT 4 ran on digital alpha 64 bit cpu in 1996 (Score:2)
Re:NT 4 ran on digital alpha 64 bit cpu in 1996 (Score:2)
Re:NT 4 ran on digital alpha 64 bit cpu in 1996 (Score:3)
originally did.
SGI was the first "mainstream" vendor to go
with a 64-bit OS (and it still has 32-bit-mode
and 64-bit-mode executables. DEC was next with
OSF/1 (later renamed Digital UNIX), and eventually VMS. IBM and Sun came later -- about the same
time as Linux (for Alpha and then for MIPS).
IMNHO, there was a very good reason Intel made
such an investment into the IA-64 port of Linux:
so that they could be sure there would be an OS
for it by the time it came on the market!
MS is a latecomer for that (currently have Linux
and two other UNIX ports that I know of with 64-bit support for IA-64).
windows is finally catching up to linux... (Score:3, Insightful)
Re:windows is finally catching up to linux... (Score:4, Offtopic)
IIRC, didn't SGI used to have IRIX running on 64 bit systems? Didn't SGI make a move to Linux? Didn't SGI assist with some 64 bit code?
If I do remember that right, then Linux had a leg up thanks to SGI--a company that use to have a little pocket change itself
Re:windows is finally catching up to linux... (Score:2, Insightful)
If i do remember that right then MS has had a reasonable amount of experience with 64bit too.
Although note that IRIX was running on MIPS. MIPS and Alpha were both 64bit RISC chips. Whereas Itanium is VLIW. No one has had that much experience porting anything sizeable to a VLIW architecture, with linux it is fairly easy, a few Kernel and compiler mods, and your sorted for a fully working system...windows on the other hand...euuughhhh...I feel dirty ;-)
Re:windows is finally catching up to linux... (Score:2)
sgi??? (Score:2)
btw, it seems they are reconsidering their decision to close it down...
Re:windows is finally catching up to linux... (Score:2)
If I do remember that right, then Linux had a leg up thanks to SGI
Ermmm... no. Linux has been 64 bit for ages thanks to it Alpha and Sparc64 ports. Waaaay before Linux was a blip on SGI's radar. Hell, Linux was even fully 64-bit on Sparc64 before Solaris was. Yes, SGI has helped with the IA64 port (along with many other companies and individuals), but the fact that the codebase was already 64-bit clean has made the task considerably easier than you're implying. SGI helped port Linux to a new architecture, not to get it to 64-bits.
Re:windows is finally catching up to linux... (Score:3, Insightful)
Aparantly the 64bit architecture helps out when doing things like accessing large disks, large amounts of memory, etc. But the instructions are basically the same, so in theory.... ImageMagick will convert images at the same performance at 32bit on Irix as compiled at 64bit on Irix. (assuming that they are both running on a 64bit platform.. ie Origin 200).
Anyway, this was from an SGI employee... who knows, maybe he is wrong... but I haven't seen any performance gains or losses to disprove what he has said.
Re:windows is finally catching up to linux... (Score:2, Interesting)
The only functional difference between n32 and 64 is the size of the pointer. Under n32, sizeof(char*) is 32 bits, and under 64 it's 64 bits. That means you can address more than 2 GB of memory with the 64 ABI.
Under both ABIs, you have access to the 128-bit "long double" type. Under n32, a "long" is an "int" (32 bits), while under 64 a "long" is a "long long" (64 bits).
What this means is this: if you're careless, and use the "long" type in your source code, the same program compiled for the n32 ABI will probably be faster than the 64 ABI version, because you'll be able to fit more "longs" into cache.
The moral of the story? Don't use the "long" type. If you need 32 bits, use an int, and if you need 64, use a long long.
Blame Intel. (Score:1)
Re:windows is finally catching up to linux... (Score:1)
Bollocks. People have been able to get 64-bit Windows since last year, the difference being that it was in beta and not fully supported. Just because RedHat et al. slapped together a 64-bit Linux distribution before doing any real QA/QC checking on it doesn't mean that they're ahead of anybody. If you think that RedHat's release was anything above beta quality itself, you're kidding yourself.
Re:windows is finally catching up to linux... (Score:1)
What's more impressive is that it'll run most existing Windows software *without needing a recompile*.
Re:windows is finally catching up to linux... (Score:2)
What's more impressive is that it'll run most existing Windows software *without needing a recompile*.
Well, it was impressive when Apple [apple.com] pulled a similar stunt years ago when they switched architectures. Now it's sorta expected.
Re:windows is finally catching up to linux... (Score:2)
Re:windows is finally catching up to linux... (Score:3, Informative)
None. In fact, they get slower. Check out this article [pcworld.com] (link pilfered from poster above).
MS's (correct) mantra about it being all about the apps is gonna bite 'em in the ass on this one. Until SQL server, IIS, and the rest of the back office stuff is also native 64 bit (along with all thier dev tools) it ain't gonna be anything but an expensive, slow box.
Re:windows is finally catching up to linux... (Score:2)
ostiguy
Re:windows is finally catching up to linux... (Score:2)
http://athena.tweakers.net/reviews.dsp?Document
Re:windows is finally catching up to linux... (Score:2, Interesting)
Not exactly. Remember that MS had the Alpha port of WinNT4 going, and lot of 2k/XP is based on NT (most..).
And I really don't think that they started the port only after Intel "Announced" ia64 being available. I bet that MS has been on this port for 2-3 years minimum. Compare that to Linux then, It makes a nicer graph =)
What I really don't understand is why MS fucked Alpha down. In my experience Alpha is STILL pretty nice player in server level (my uni runs mostly on alphas/ x86+linux)
With alpha support it'd be much easier to support amd athlon+x86-64 (it's ev6 style bus, right?)
Re:windows is finally catching up to linux... (Score:2)
Ah, that's because DEC did all the porting. Yes, you heard it right: DEC did Microsoft's job for them in order to get NT on their own hardware. (Similarly, SGI ported NT to MIPS). A year or two after DEC was bought out by Compaq, they told Microsoft they are not going to do porting for them any more. Microsoft decided to kill Alpha port and made the spin that it was all Compaq's fault. There was a story on The Register about it maybe a year ago.
Re:windows is finally catching up to linux... (Score:2)
No, the NT kernel was originally written as a RISC kernel, and the development team hated the Intel RISC chip of the late eighties (i860?) so they wrote it for the R4000. Then they ported it to the i386 architecture, but they said it sucked. I think Microsoft did all the porting work themselves, including PowerPC and Alpha, but they never ported the apps.
Re:windows is finally catching up to linux... (Score:2)
Re:windows is finally catching up to linux... (Score:2)
The i860 never really caught on as a general purpose CPU, but was used as a specialized processor in many instances because of its floating point performance. IIRC, SGI used several of these on their first Reality Engines.
Re:windows is finally catching up to linux... (Score:2)
MS wasn't happy when Compaq pulled the plug on Alpha/NT, but with this level of support from MS, it's easy to see why Compaq made the move.
this means new books (Score:3, Funny)
I can't wait.
The IA64 is great...but.... (Score:1)
Re:The IA64 is great...but.... (Score:2)
With IA64 many IT managers would love to switch their programming departments from c++/solaris systems to VB.net/W2k systems in the name of standards of course. Bussiness users are extremely conservative and they are the ones who are keeping intel processors in the majority of pc's sold. The Athlon is totally foreign in any corporate environment. They will continue to want to use intel based products and the IA64 will give them oppurtunity to do this. Now which OS will come default with any IA64 server purcahse? You guessed it, Windows 2k! Now if they can have an intel based server with a microsoft based operating system running their microsoft based VB.net or c#.net apps that can handle a solaris load, then they will have it made. %100 conformity.
Windows is getting alot more stable and with clustering and switches, downtime problems are going away. This will create more headaches for unix since corporations like to buy computer equipment from the same company. Which will include servers with Windows pre-installed of course.
2 Questions... (Score:1)
What kind of functionality limitations will be placed on these machines (its ms... there's always limitations...)??
This release no SledgeHammer. (Score:4, Informative)
Microsoft is also considering an x86-64 port of Windows XP, but they have not announced their decision yet.
Re:This release no SledgeHammer. (Score:2)
Win95 Nightmares (Score:1)
Well... (Score:1)
Next thing you know, they'll take the money out of your bank account for you, without you even having to ask, and then personally deliver the porn to your door.
yikes ! (Score:1)
Please, just leave it in my 10G inbasket.
tia
Just like microsoft (Score:2, Interesting)
Re:Just like microsoft (Score:1)
More information on the release (Score:1)
Limited Edition (Score:1)
C:\ONGRTLNS.W2K (Score:1)
C:\ONGRTLNS.W95
Similarly, perhaps a coalition of vendors -- Sun, IBM, Compaq, Apple and the gurus of the 64-bit Linux kernel ports -- should run a congratulatory ad for this momentous event: Microsoft finally goes 64-bit.
Re:C:\ONGRTLNS.W2K (Score:2)
Re:C:\ONGRTLNS.W2K (Score:2, Informative)
greater fp perf == better web serving? (Score:3, Interesting)
More memory == better web serving? (Score:2)
"Limited Edition"? (Score:1)
Wonder what they mean by "Limited Edition"?
M$ is only going to sell a limited number of copies, or it's 'limited' in features?
Let's take look (Score:1)
Microsoft 's Windows Advanced Server, Limited Edition
Nomen est omen
In other news... (Score:3, Funny)
ObLinuxComment: Let's make Linux 128-bit clean, just for the hell of it, so it's ready for when someone makes a 128-bit processor to run it on.
Re:In other news... (Score:2)
echo 0 >
cat $MAIN_ARTICLE
cat $JD_POST
/usr/sbin/laugh --force
Wait for WinAS/LE service pack 5, SE, whatever (Score:2, Funny)
Limited use (Score:2, Informative)
All you would get for your trouble is a crippling licence fee (courtesy of MS), a dearth of 64-bit applications & drivers, slower 32-bit execution and double the memory and disk requirements. These are hardly compelling reasons to "upgrade".
Re:Limited use (Score:5, Interesting)
I was going to say the same thing, that, as far as current customers are concerned, this product seems to fill an incredible Non-Niche.
It's exclusively for IA-64, which can't compete head to head with established RISC hardware yet. Also, given that the OS's for the competitive RISC hardware have been around longer, had more bugs shaken out, had more apps (eg, Oracle) developed for them, Advanced Server won't provide any kind of revenue for MS. It's all written off for the sake of future revenue.
Like anything, they're willing to let it slog slowly up through the ranks for a few years until it gains credibility (eg, the first 2 versions of Windows and of NT). Eventually, though, all this beta testing will pay off so that in 2005 they can argue convincingly that they can provide an alternative to the big iron from IBM, Sun, HP, SGI and Compaq (DEC).
The other benefit of this move for MS is to provide a testing ground for their code base so that if IA-64 ever does develop into something so desirable that it begins to appear in desktops, they'll have some experience for it. With the recent boost that Intel gets from killing off the Alpha competitor and from using the Alpha's carcass to improve the sickly Itanium, the IA-64 will eventually become something to be reckoned with, even if through the sheer brute force of the dollars behind it.
For current customers, though, this OS release is a yawning opportunity to be part of MS beta test program. As with the Linux IA-64 release, it is mildly interesting, with genuine interest deferred until the point that the hardware is competitive with the established RISC vendors.
Anyone care to compare and contrast their 64 bit foray to their first foray into the 32 bit world?
Limited Edition! (Score:3, Funny)
Does anyone know if it comes in a metal lunchbox [red5interactive.com], or tin case [amazon.com]?
Re:Limited Edition! (Score:2)
i've used lots of different versions of windows, and i thought they were all pretty limited.
64 bit Windows (Score:4, Interesting)
Undoubtedly parts of this version of Windows 2000 has to remain 32 bit for compatibility. Or is Microsoft going to port Microsoft Office to 64 bit Windows as well? Unless Microsoft has implemented some type of FX!32 (DECs 32 -> 64 bit layer which "learns" and accellerates), this release of Windows may potentially be quite useless. One of the reasons people use Windows is the availability of applications.
I can't for the life of me think that this is anything different from a marketing release where Microsoft can say "We're in the future, we're 64 bit". But it's nevertheless interesting that Microsoft has gotten something out the door that is 64 bit. Let's see how well Microsoft entrenches itself in the datacenters. My guess is that the 64 bit x86 (Intel or AMD) will become far cheaper than the Sun counterparts and thus taking over a lot, but not everything. But will Windows be the preferred platform or is Linux going to hit Microsoft where it hurts? Or perhaps Microsoft will make this
In either case, from a technical standpoint I will observe how Windows 64 bit is going. Very interesting indeed.
Alex
Re:64 bit Windows (Score:2, Informative)
My understanding is that everything in the core OS has been recompiled as a 64-bit binary, this would include the kernel (obviously), shell (including support DLL's and EXPLORER.EXE, for example), and most likely server components such as IIS (this is my speculation here, if IIS is still a 32-bit binary, someone speak up, because that really WOULD make the entire release almost pointless).
But I've read on MSDN and elsewhere that Explorer and other basic (ie: integrated) components of Windows 2000 were ported to IA-64 for this release.
Re:64 bit Windows (Score:2, Insightful)
Re:64 bit Windows (Score:2)
Windows on it's own useless? You don't say? I'm not sure how useful any MS stuff has ever been, but don't worry YOUR SOFTWARE INVESTMENT WILL BE PROTECTED WITH THE USUAL MS CARE AND CONCERN. No one would inflict needless waste for the sake of their bottom line, would they? Why does this Word 6 document look like poop?
Connections.... (Score:2)
Itanium is aimed squarely at the hideously overpriced Sun e450 and up lines of hardware. If you are Intel, you're not going to get hardcore Solaris shops moving to Windows Adv Server, and Windows DataCenter. This brings us to....
Linux. For the record, I found TurboLinux was the company to produce a usable OS for Itanium, followed by Suse, and then Redhat, this has resulted in:
Resources. At this point. Redhat's Distro seems to be the best on Itanium, giving them a leg up on the Real Prize, McKinley. However, porting apps to the Itanium hasn't been as easy as just treating it as yet another 64bit CPU.
real reason behind the new roll-out (Score:2)
http://homepage.mac.com/jcarusone/iMovieTheater2.
_f
In my day... (Score:2, Funny)
we could crash Windows with only 16 bits. None of this namby-pamby 64 bit stuff. No sir. We used to say, "look at me, I'm crashing with a segmented memory model!" And you know what? It was good enough for us and we liked it.
This new generation doesn't know how to crash Windows. No sir. They say, "oh, I couldn't possibly crash Windows without 64 bits, oh no." Wimps. All of 'em. Think bus bits grow on trees or something.
64 Bit computing just now becoming affordable. (Score:2)
Databases, for those of you who don't know are extreme memory hogs and 64 bit memory space is necessary for most large systems. NT or Linux this is a nice feature.
Video editing/rendering - having this much memory is nice + floating point = fast!
I know exxon will love them, geophysics isn't easy on 32 bit systems when utilizing holographic imagery and trying to produce maps of oil.
Microsoft just has an easy interface for how things work. 64 bit in some form or fashion has been aroiund for a while in solaris, again mainly for a server os. Irix has had it, and again they're for the graphics/producing bit.
SO NT will just fill it's niche.
I have no idea what good linux on itanium is. can only get mysql to go so quick, pgsql doesn't support 64 bit as far as i know and not much for high end graphics. May be good for a rendering farm i guess?
Atleast with SGI & NT/XP/2k you got lots of visualization, data manipulation, mapping and extrapolation type applications. Sun has its share.
So i don't know why people dis it. You aren't going to run Office on the sucker and if you do who cares if its on par with a PIII, a PIII is still fast emulated or not.
Its just nice to know 64 bits is around the corner. With memory and CPU prices falling through the roof its only a matter of time before consumers (gamers / coders and tech heads) upgrade to 64 bit systems.
Be it linux, solaris, irix or not.
Re:64 Bit computing just now becoming affordable. (Score:2)
Just change the 192.168.1.2 to dedserius.com and you'll see the content. That machine is one of several hot-swaps on a private LAN. Sorry. I need to update the httpd.conf file, but it's the least likely swap to go and as an Alpha I don't spend much time keeping it current except for security.
The Link has helped quite a few people get their Apache/PHP/FrontPage server on Linux implementations working so hope that helps you out.
Twice as many users in the ditch (Score:2)
Holy single fucking point of failure Batman!
An Intel box as big as a huge ass Sun or a RS/6k-S80 will cost at least as much to support and twice as much to harden for security.
Not a critcism just a fact - we'll have to reshape all of our SLAs to reflect the unreliability inherent with consolidation.
Re:Twice as many users in the ditch (Score:2)
Unreliability isn't inherent with consolidation, its inherent with ineptness.
If 100% uptime and reliability is what you need then having multiple servers doesn't server you anybetter. You would need duplicate datacenters, redundnat storage arrays, redundant power, redundant network connections, redundant routers, redundant switches and offsite support. When your talking about something like that, its cheaper to Buy an E10k maxes out that gets 99.99999 percent uptime and lease recovery center hardware & provisions from experts in the field like Sungard and such.
Hardware is different from software, no? (Score:2)
It's the software that's a problem. A kernel panic on one of 12 NT servers is less of a problem than a kernel panic on 1 of 1 NT servers. A problematic security hole caused by yet another ubiquitous IIS or Active Server glitch is more of a problem from a change control perspective if all your users are on box. A big box that still uses NetBIOS over TCP and blasts a ton a crud thorough port 139 is probably easier to manage from a firewall perspective if what you want is to filter traffic from-to by address but attempting to mount a network IDS on the box will present correlation engine problems in the shear volume of false positives that an NT based solution will generate.
And so on.
Rendering/Raytracing (Score:2)
Future looks cool. As long as it's not too far
Bah (Score:2)
Random Thoughts on IA-64 (Score:2)
This leads to a couple of side question: How are the Linux and BSD IA-64 ports doing? I heard something about both of them awhile back. Both camps reported stuff is going well with kernels and compilers running but then the news just died away. Anything new an interesting to report? I would be interested in how much bloat going to a pure 64 bit kernel actually is.
Didn't Intel claim that the Pentium is a 64 bit processor? Where are the 64 bit ports? *shock* Does this mean that Intel wasn't exactly truthful?
Re:Random Thoughts on IA-64 (Score:2)
as for linux, it's as far from vapor as it can be. buy an ia64, download any stable 2.4.x kernel and go on.
See your local linux sources or
http://bitmover.com:8888//home/ppc/linux_2_4/sr
Obligatory Blue Screen Post (Score:2)
What happens when the two blue screens of death toggling mechanism breaks? Will we get the magenta screen of tormented afterlife?
64==2*32 (Score:2)
2*.dll
2*reboot
2*hot fix
2*service packs
2*the bloat
2*the suck
?
Get M$ OFF any other architecture but the x86. (Score:2)
Re:FlOps assist serving web pages? (Score:1)
Re:FlOps assist serving web pages? (Score:3, Interesting)
Itanium, and in fact the entire IA64 architecture, is a disaster. They bet the company on the wrong technology*. Intel will probably survive just because their marketing machine will persuade clueless corporate buyers to take the chip.
* The numbers don't lie: for most applications, all the explicit parallelism, speculative operations, compiler engineering and transistor count in the world can't compensate for not having a real speculating, out-of-order core.
Re:FlOps assist serving web pages? (Score:2)
The thing is that problems of the form "multiply matrices fast" are fairly easy to make fast. Vector units, reconfigurable units (including CPU-attached FPGAs and Stanford Imagine-style coprocessors), multiple CPU cores, lots of FP units, IRAM --- they all kick butt on these problems. Even with traditional designs, you'll notice that Athlon and Pentium 4 performance on these kinds of codes has been improving rapidly. Itanium's lead in these tasks can't be considered secure. And Amdahl's law suggests that terrible integer performance (very hard to fix having given up out-of-order execution) will strangle them even for FP-intensive applications, sooner or later.
Re:FlOps assist serving web pages? (Score:2)
OK, it is 64-bit which is slightly useful for some servers. But by the time 64-bit becomes really useful for many people, AMD will have Hammer/x86-64. That is a much better processor design, 64-bit, easy to port to (just a 64-bit stretch of x86 with a few sensible goodies like more registers), and for extra points will run x86 code incredibly fast! The last point ensures it will ship in volume and hence massively undercut any other 64-bit chip on price.
Re:More MCSE resumes to toss into the shredder! (Score:1)
btw, in regards to 64bit windows - will it crash faster now?
Re:More MCSE resumes to toss into the shredder! (Score:2)
What I hate are those cheap so-called "MCSE boot camp" and other computer-training ads in the local newspapers. They promise you 70k a year in a sort of get rich quick scheme. Many of the applicants are suckers who are mainly poor immigrants who currently work at low wage jobs. At least where I am in New York City. They apply for these 6 week programs which only teach you how to answer questions on the mcse exams. No actual administration or support training is provided. Then afterwards they pass and expect to make 60K a year. A few of them actually made it into the IT world because they had a college degree as well. Its these mcse's are the ones who truly know nothing. Most of them who know dick still work at their low wage jobs and their mcse's did not help.
However I went to a professional business institute (paid by my employer of course) and was a pc junkie since middle school and knew my stuff. I learned most of my computer knowledge from experience but I thank my mcse training school for giving me a foot in the door. A mcse might not make a great admin or programmer but would probably be an ideal candidate for an entry level help desk job.
Re:Wasted Power (Score:2)
Who said that this is for desktop users? Last I checked, there are more than a few admins and developers who read Slashdot who actually use hardware like this (and much bigger)....
Re:Wasted Power (Score:4, Insightful)
Two Words: Video Compression
Seriously, while 64 bit processors running at 1.x GHZ will be wasted on desktops, this power is just the sort of thing to beef up existing dual and quad CPU SQL servers.
Re:But Itanium is slower than P4! (Score:2, Informative)
The benchmarks you're referring to (the ones that placed an Itanium at a slower speed than Pentium III's or Pentium 4's) were running 32-bit code, not native 64-bit code. Intel's Itanium processor family (IA-64) are backwards compatible with their Pentium family (IA-32).. in other words, it can execute the code without any on-the-fly emulation or translation.
The problem with this is that apperently Itanium's implementation of the IA-32 execution unit is shoddy (and thus, slow). However, code written to the native IA-64 spec (which is what this release of Windows 2000 Advanced Server is) should perform MUCH faster.
Besides, an important thing to remember about Quake III is that it's not the CPU that matters really, it's the graphics card.
Once more IA-64 binaries are released I think the benefits of the architecture will become clearer. (Hopefully! I'm not saying Intel can do no wrong, but basing your assessment of their processor on it's EMULATION (basically) of IA-32 is totally off-base-- IA-64 is it's native, preferred mode. IA-32 is just there to make transitioning easier.)
Re:But Itanium is slower than P4! (Score:2)
If you look at the Spec numbers, the benefits of the architecture are already pretty clear: runs great on nice predictable toy floating point benchmarks (SpecFP), complete DOG on everything else (SpecInt). Itanium's SpecInt numbers (400) are the worst of any recently released desktop CPU. Any Athlon or Pentium 4 will blow away Itanium's performance for non-FP stuff (which includes servers). And let's not even start talking about PRICE/performance.
Re:But Itanium is slower than P4! (Score:2)
Re:But Itanium is slower than P4! (Score:2)
Still, this is their first iteration of the new architecture-- I'm sure it'll take time to re-engineer it so they can squeeze the performance out of it that we expect.
Re:But Itanium is slower than P4! (Score:3, Informative)
The fundamental problem is that IA64 is an in-order design. Your code hums along, does a load from memory, misses in the cache --- and everything stops until that value comes in. In an out-of-order machine (every other high performance CPU since the Pentium Pro in 1995), while it's waiting for that value to come in from memory, it will be executing other instructions that technically are supposed to happen AFTER the load, but don't actually depend on its value.
The IA64 was supposed to get around this problem by providing speculative loads with alias masks and other tricks so the compiler could hoist the load and perform it super-early, long before the value was needed, so any delay due to cache miss would not impact execution. Intel's big bet was that this would make out-of-order execution unnecessary. They lost the bet, for two reasons: 1) real programs (as opposed to toy benchmarks) are too unpredictable. There just isn't enough information available at compile time to decide what can and should be loaded early, which order instructions should go in, etc. The decisions have to be made at run time by the CPU itself. 2) The cost of out-of-order execution was overestimated. In the last five years we've been able to build really big horrible superscalar out of order cores with really fast cycle times. OTOH, the Itanium's supposedly "simpler" core has a crap clock speed. Go figure.
95% of what you read about the IA64 architecture is marketing hype. Just because it's "new" and "different" doesn't make it better. In fact, on the numbers, it appears to be worse in most cases.
Re:But Itanium is slower than P4! (Score:2)
Out of order execution is HORRIBLE on performance-- you're right, CPU designers and compiler writers (and low level assembly developers) have found ways to eek out every bit of performance from current generation processors, but in an IN ORDER execution unit, all of this SPECULATION isn't needed.
Your example of a load from memory being missed in the cache killing everything else is outright wrong. It's the COMPILER's responsibility to order the instructions such that a cache miss shouldn't ever occur. One of the hallmarks of the design is in fact that instead of doing branch prediction, EVERY branch is taken (fail or pass), and when the result of the the comparison that affects that branch is known, the code is either already in the pipe or in fact already executing (since IA-64 binaries are output in such a way that as more pipelines are added, more code can be executed SIMULTANEOUSLY).
Your notion that there "isn't enough information available at compile time" is horseshit. If you know compiler writers that think along those lines, could you ask them when Visual Basic will quit sucking for me? Compilers can (and should) do multiple passes to gather all of the information they need, then write the most optimized and streamlined binary they can. This is in fact in the Intel documentation-- and even Intel admitted it will take time to get compilers to work as desired.
And if you knew anything about CPU's, you'd know clock speed means crap-- AMD's Athlon has a lower clock speed, but out-performs Intel's higher-clocked Pentium 4 CPU's. If the pipeline is shorter than any CPU on the market (and thus, instructions execute MUCH faster), then the cycle count is going to be irrelevent (or inaccurate) for comparing one CPU against another.
Again.. this is their FIRST release of the processor. You may be some anti-Intel / pro-AMD zealot, but take your preachy attitude elsewhere. I've pointed out that their current processor definately does not meet the performance people were expecting or hoping for-- but it takes TIME for an architecture to mature. Were you expecting it to out-perform everything on it's first day?
Re:But Itanium is slower than P4! (Score:2)
> commitment to this architecture.
Intel actually kept the details of the design secret until very late in the game. But anyway, when I wrote "most of what you read" I meant "most of what you read IN THE PRESS". Not a lot of people look at the programming manuals.
> It's the COMPILER's responsibility to order the instructions such that a cache > miss shouldn't ever occur.
You should go to college and take a computer architecture class. In that class you will learn why this is fundamentally impossible. It's like a driver trying to predict which traffic lights he will have to stop at before he even leaves the house.
> One of the hallmarks of the design is in fact that instead of doing branch\
> prediction, EVERY branch is taken (fail or pass)
This is called predicated execution and Intel did not invent it. It does help avoid some of the penalties of mispredicted branches, but it is not universally applicable. In many situations you still have to use real branches. Furthermore branches are just one of the things a compiler has to predict correctly for in-order execution to work well.
> Compilers can (and should) do multiple passes to gather all of the
> information they need, then write the most optimized and streamlined
> binary they can.
Sure, that is the goal. But in practice there are severe limits to the amount of information a compiler can get at compile time. It's all about predicting the future, and that's always hard and often fundamentally impossible. BTW, I just spent seven years working on my PhD on whole-program static analysis, which is all about this subject
I understand that clock speed isn't everything. I just bought an Athlon for that very reason. But a lot of the design choices in IA64 were made to *increase* the potential clock speed, so the fact that they weren't able to get the clock speed up anyway does not bode well for the architecture.
It will take time for the architecture to mature, but most of the bad decisions are already baked into the ISA (instruction set architecture). And of course AMD (and even the Pentium 4) aren't standing around waiting for IA64 to catch up.
Re:But Itanium is slower than P4! (Score:2)
PS, EPIC is Intel-marketing speak. What you are talking about was called "predicated execution" long before Intel 'invented' it.
Re:But Itanium is slower than P4! (Score:2)
First gen Itanium? Sucks ass, I wouldn't buy one except to experiment on or develop with. But as future generations are fabbed, we'll start to see the performance shine. Ditto for compilers as they mature.
My money is with Intel's Itanium in the long run over AMD's 64-bit offering.
Re:But Itanium is slower than P4! (Score:3, Interesting)
AMD should bring up this point in their marketing, actually.
I'm betting on the AMD Hammer series (2H2002) over the second-generation Itaniums (McKinley?) regardless.
Re:Intel Microsoft bed-buddies (Score:2, Informative)
I don't know what vendors who sell HP-UX based software are doing right now, but Debian GNU/Linux has ported a very large number of packages to IA-64 and is planning on releasing it at some point not to far from now. (you can install it from Testing and Unstable already).
Major developers know what's out there, it's only you who is in the dark.
(may or may not be the opinions of my employer, I don't speak for them)
Re:Actual improvements? (Score:2, Informative)
The big bonus is that you can access a lot more memory. A lot of databases are larger than 2GB, which means they cannot be loaded into RAM on a 32-bit system (well, 3GB if you play some tricks). Since disk access is on the order of 1000 times slower than RAM access, if you can load your 12GB database into RAM and query it from there, you should see significant performance improvements.
For editing WordPerfect documents, it's probably not that significant. :)
Re:Actual improvements? (Score:2)
More directly, a 64 bit processor moves twice as much data around as a 32 bit processor, assuming all bits are significant. It's worth noting that an IEEE double precision number is 8 bytes (that's 64 bits) of strorage, and so a 64 bit procesor can load, add etc a double precision FP number in one operation. That helps a lot.
It's not going to be moticable for most people, but if your doing serious calculations, then it wil show.
Re:Actual improvements? (Score:2, Informative)
Real world benifits? Depends on the Application and OS. Usually Databases and big imaging applications see benifits first. And remember, many machines have been 64 bits for years, Intel and Microsoft are late to this party.
I understand that when the AS400 (basically database machine) line went from 48 bit CISC to 64 bit RISC chips several years back there were significant benifits in uptime, of all things. Seems the AS400/iSeries Single Level Store memory & storage management scheme requres a unique id for each object in memory. Bigger CPU words meant more (2 to the 12th?) id numbers were available. So, reboots to reset the counter for temporary objects were now needed only (all other things equal which they weren't) 1/2^12 as often. That's a big benifit for big iron.
Re:Doesn't the G4 use 128-bit processing? (Score:2, Informative)
Re:NO! IA64 provides WEAKER floating-point capabil (Score:2)
What's your point? (Score:2)
Re:Great (Score:2)
If you look on Redhat's FTP sites, you can see the IA64 subdirectories right next to the i386 ones.
Re:I've been waiting for a chance to share this on (Score:2)