Windows XP 64-Bit Customer Preview Program

MBCook writes "I just notice that Microsoft has a new Windows XP 64-Bit Customer Preview Program starting today (February 3rd). If you have a AMD Opteron or Athlon64, you can go to the download page to get your copy. It's a pre-release copy that will expire in 360 days (which probably means the final will be out by then). Now Intel just changed their 64-bit plans, and all of a sudden this appears. Speculate away!"

Comments

Actual Performance Difference

[neomage86]

I wonder if a 64 bit OS will make any performance difference for the average desktop user. Since its not like any normal people have more than a gig of ram anyways. Is it possible that it could even slow down 32 bit apps?

Re:Actual Performance Difference

[Paladin128]

My guess is that the initial rev will, in fact, be slower on 64-bit. Microsofts compilers are new to 64-bitness, and a reasonable amount of memory bandwidth will be wasted on larger ints. On the other hand, in 64-bit mode on the Opteron, there are twice as many GPR's, so it could wind up being faster. My bets are still on slower largely due to immaturity of the platform.

Re:Actual Performance Difference

[Vlad_the_Inhaler]

There were linux benchmarks which made /. recently, comparing the speed of 32-bit code and 64-bit code on the Opteron. 32-bit code ran measurably faster.

Re:Actual Performance Difference

[Neon Spiral Injector]

Microsoft has had a 64-bit version of NT since almost the start. Also sizeof (int) == 4 on AMD64, same as IA32. Just the pointers are larger.

Re:Actual Performance Difference

[cubicledrone]

I wonder if a 64 bit OS will make any performance difference for the average desktop user.

Of course! We need an operating system that will support a pointer to all the RAM on the planet.

Re:Actual Performance Difference

[Zeinfeld]

I wonder if a 64 bit OS will make any performance difference for the average desktop user. Since its not like any normal people have more than a gig of ram anyways. Is it possible that it could even slow down 32 bit apps?

Its been a long time since I had performance issues due to CPU bottlenecks. My twin processor 650MHz box works just as fast as my 2.4GHz single processor box in practice.

The big issue for me are the cases where the stupid machine just locks up and does fuck all for 20 seconds or so. CPU meter shows 3% utilization, no disk activity. What is the stupid thing doing?

Same goes for UNIX systems, its not the processing thats the issues, or even the legitimate I/O delays, its the cretinous delays built into broken device drivers and applications.

I would like Windows to have a meter built in that would show which processes were waiting and the resources they were waiting on.

My other pet peeve is what the cretins at Checkpoint think is an acceptable VPN client. Every time the credentials time out a box appears for me to re-enter my credentials. Only I use cert based credentials stored in CAPI so all I am doing is hitting OK. Even so the box locks the user interface for about 90 seconds while it does something. Oh @$#(& it did it again.

Re:Actual Performance Difference

[Cuthalion]

In ANY halfway modern architecture each fetch from memory pulls in a whole row of cache. Which is more than 8 bytes. The fact that more data needs to be moved for small operations to be completed is NOT a benefit.

Re:Actual Performance Difference

[Anonymous Coward]

Go their benchmarks aslo show a marked improvement for Linux on 64bit.

Some tests scored about as much as a 20% improvement in performance, athough it's more realistic to expect a 8-10% average improvement across the board.

Of course windows likely will run slower since it's so optimized for the older 32bit platform.

Linux is just a much more mature platform for 64bit computers. I mean we've had Suse 64bit (aviable for free from Suse's ftp install stuff) for almost a year now.

Re:Actual Performance Difference

[myg]

Actually, there are a few tasks that benefit from 64-bit code: Encryption, Databases, and file-systems. Most modern filesystems allow files >4GB and with digital video the norm now files much larger than 4GB are common.

So the processor spends less clock ticks doing 64-bit arithmatic for file offsets. Cryptography can benefit too. In particular the DH key agreement protocol and RSA public-private key cipher both require the use of "big numbers" (as in 1024 bits and up). And doing these operations 64-bits at a time rather than 32-bits can result in a performance impact. SSL session negotiation may be improved significantly, so that could be a boon to SSL webservers (not that IIS makes a great webserver).

Databases also need 64-bit numbers. Even low-end databases frequently have ID numbers that are 64-bits. For large databases the ability to compute page offsets (which are typically 64-bits) more efficiently can help.

And perhaps a subtle improvement may be in the handling of bitmaps. For example doing a bitwise operation 64-bits at a time will process more pixels than a 32-bit operation in the same amount of time.

But all of those things are really just noise for the special case of amd64. The x86 has always been a register-starved architecture. Going to 64-bit mode gives you 8 more general purpose registers. That alone may very well counter any of the performance loss due to 64-bit integers.

Lots of people on Slashdot have said 64-bit code is slower but haven't pointed out why. Probably the most important thing is cache coherency. 64-bit data structures are (surprise) larger than their 32-bit counterparts. That means that the effective rate of the cache is reduced. As an added bummer, address translation on 64-bit values is slower when there are more levels of translation.

Re:Actual Performance Difference

[Miguelito]

Well I did some benchmarks using openssl's built in speed tests, and running in 64 bit vs 32 bit made a HUGE difference. Of course that was running a 64bit openssl as well. The opterons I have access to even beat the Itanium2s I have access to at work. By a lot in the smaller bit key sizes, but still either tied or beat the ia64 in the larger key sizes.

Here are the charts I made in OpenOffice on the data I collected. Even a 2GHz opteron beat a 3.2Ghz Xeon in 32 bit mode. :)

Of course this was just a benchmark, but it does show that things that use openssl would benefit from running under 64bit on an opteron.

Re:Actual Performance Difference

[Soko]

NT for Alpha had no long pointers and the same 4GB memory limit etc. as its IA32 counterpart. The processor was certainly 64 bits, but the OS was 32 bits at it's core. IIRC there were some places where they had to use 64 in order to get the thing to load, but not where it counted.

If Microsoft had actually used the Alpha to it's fullest potential, all of my servers would likely be runnning 21464s, not Xeons.

Yes, I'm still mad at DEC/Compaq/HP for squandering the Alpha tech. *grumble*

Soko

Re:Actual Performance Difference

[jkovach]

It depends on what you're doing. I recently gave myself a 1.8 GHz Opteron system for Christmas and have been running some performance tests comparing 32 and 64 bit versions of the same applications.

Using LAME compiled from source with the default compiler options and "--alt-preset standard" encoding settings, it took 4m20s to encode an 11 minute MP3 with the 32 bit version and 2m51s to encode using the 64 bit version - about 30% faster. However, comparing some of the filters in a 64 bit GIMP 1.3.23 compiled from source with default settings to the GIMP 1.3.23 from the Debian distribution was quite different - running the "Diffraction Patterns" filter with default options on a 512x512 image took 11 seconds with the 32 bit version and 16 seconds with the 64 bit version - about 50% slower!

Of the other tests I ran, MP3 decoding with mpg123, bzip2 compression and AES encryption with OpenSSL were faster in 64 bit mode, and 3DES encryption with OpenSSL was faster in 32 bit mode. Of course, one of the advantages of having a hybrid 32/64 bit processor is that you can run whichever version is faster for a given task, onlike on the Itanium which until recently had only software emulation for 32 bit code.

I have also heard that 64 bit Windows on Opteron is slower at running 32 bit code than 32 bit Windows on Opteron, which has made me want to rerun my 32 bit tests under a 32 bit kernel. I also want to run the same tests on a Mac G5 and a Pentium 4 for comparison. Some other time...

Re:Actual Performance Difference

[vrt3]

No, no: what he meant was that sometimes (quite often, actually) one process or the whole desktop becomes unresponsive, even though almost no CPU time is being used, the hard drive isn't working, no network traffic should be occuring etc. You just have to assume that something is waiting for something, but there is absolutely no indication what it is.

I see such delays almost every day; mostly on Windows (XP and 2000), but occasionaly on Linux too.

Intel will have to follow AMD

[Random BedHead Ed]

At this point there's enough Opteron stuff out there that Intel can't avoid implementing an architecture compatible with amd64. Yes, I think the timing is probably not a coincidence. Though I'm sure Microsoft wasn't waiting for Intel, they probably informed Groves and Co. of the release in advance.

If Intel can't stay compatible with AMD's lineup they could end up behind. That would certainly be a first for Intel.

Simple

[Tenfish]

Intel and Microsoft may or may not have a business deal to promote Intel's 64-bit processors. Microsoft wouldn't do this for AMD, because splitting the processor targets would cost Microsoft money. Developing for two processors is more expensive than developing for one. There's no need to delve into anti-corporate mumbo-jumbo to explain this one. It's simple economics, and probably good business from Bill Gates' point of view.

Re:Simple

[diamondsw]

Looks like someone needs to double-check that with Bill, then:

"Important: Windows XP 64-Bit Edition for 64-Bit Extended Systems is only compatible with 64-bit AMD Opteron- or Athlon 64-based computers. It cannot be successfully installed on 64-bit Intel Itanium-based systems."

I don't see any mention of an upcoming Intel it will run on, either.

Speculation...

[ackthpt]

Now Intel just changed their 64-bit plans, and all of a sudden this appears. Speculate away!"

I speculate in a couple days Microsoft will deny this release exists, as they suddenly pull it to give their old cartel partner a chance to catch up and save face.

Consider this:

Microsoft has an evaluation operating system for the Yamhill before Intel actually ships. That doesn't just look like they've been playing patty-cake, but that Intel is running to keep up with AMD. How embarrassing. What's Moore's Law got to say about this? "Every 2 years Intel will get a little further behind where they need to be, by an ever increasing margin until operating systems exists for processors they haven't even designed yet."

System requirements

[Faust7]

Windows XP 64-Bit Edition is designed to address the most demanding needs of technical workstation users who require large amounts of memory

System Requirements:
PC with an AMD Athlon64 or Opteron processor
256MB RAM

Yeah, I know they're minimum requirements, but it's amusing to think that anyone who needed this OS would actually have to refer to them.

How's this for speculation

[iminplaya]

Micrososft is doing this to keep us distracted away from Linux until Longhorn comes out. Two years is a long time, so we can always expect bits and pieces of "neato" stuff in the meantime.

Microsoft Sold Me an OS!

[tonyr60]

I went to the site and odered a copy of the CD. First Microsoft OS I have bought for many years, but for $US7.95 (including international delivery) the price is quite reasonable and comparable to the competition.

Re:Microsoft Sold Me an OS!

[skinny.net]

$7.95?! That can't be right! Here in the US, 64 bits is $8.

Intel is not impressing me these days

[AviLazar]

I still recall reading the article in Wired magazine a few months back. A company approached them with an offer to provide them perfect diamond wafers (produced at less then $5/wafer) and Intel did not take the offer because they have not gotten their full investment back on the silicon. So given that they refused to have an easy method of increasing their processor spead by a very big number, it is not surprising that they still haven't gotten 64 bit over AMD... Shame, Intel used to be the best. -A

with 64 bits...

[vpscolo]

I suppose it means windows will take 1/2 the time to crash and at least twice as long to write out dump files

Rus

Free? Which kind of free?

[Call Me Black Cloud]

Damn, I never get it straight. Is this software free as in speech or free as in beer? Since the software is only good for about a year, I'd guess this is free as in beer, because my beer has an expiration date too. But does that mean Linux is free as in speech because it doesn't have an expiry? Of course some really skunky beers don't either. Now, can you have free as in beer in a cathedral? All the cathedrals I've been too serve wine. So is this free Windows free as in wine?

See, it's this kind of thing that is holding up the adoption of Linux. It's too confusing.

I guess...

[Cytlid]

...the submitter didn't RTFA!

Important: Windows XP 64-Bit Edition for 64-Bit Extended Systems is only compatible with 64-bit AMD Opteron- or Athlon 64-based computers. It cannot be successfully installed on 64-bit Intel Itanium-based systems.

I don't read too much into this

[rabtech]

I don't read too much into this; Microsoft already told Intel that they (Intel) will in fact stay compatible with AMD because they will not be releasing three separate 64-bit editions of their operating systems (Itanium, x86-64, ???).

In this case, I think the monopoly is working for us in preventing Intel from fracturing the market, which I am quite sure they would love to do.

So in this respect, I think Microsoft is just having the usual Microsoft delays. One of the biggest is supporting Read vs ReadExecute memory pages on the x86-64 chips. A Lot of code, including microsoft code (ATL anyone?), breaks the standards and is having to be cleaned up to work.

In addition, Itanium lacked WMP, DirectX, and many other features that customers will NOT do without on Athlon64 systems, meaning even more development time.

sizeof on modern OSes

[Anonymous Coward]

char is 8-bit, nearly by definition
short is 16-bit
int is 32-bit (DOS and Xenix are not modern)
long is 32-bit on all Windows OSes
long is the same size as a pointer Linux, BSD, Mac
void* is the natural size for the machine
long long is 64-bit

So 32-bit Windows and 32-bit UNIX have the
same sizes a each other. 64-bit systems differ,
because Windows sets sizeof(long)==4 and a
UNIX system sets sizeof(long)==sizeof(void*).

Re:Windows Media Player?

[ikewillis]

Evidently you haven't heard about the Windows-on-Windows (WoW) subsystem utilized on AMD64 processors:

"The Microsoft(R) WOW64 (Windows(R) on Windows) subsystem will allow most 32-bit applications to launch seamlessly on 64-bit Windows," said Brian Marr, Windows product manager. "WOW64 is designed to provide interoperability and great performance on AMD64 processors across the 32/64-bit boundaries. As customers migrate to 64-bit Windows XP and Windows Server, they will have a code base that will support both 32- and 64-bit applications."

While it's wonderful Linux understands multiple ABIs natively, Windows does not, and utilizes WoW to seamlessly launch 32-bit applications on 64-bit builds of Windows.

Terrible...

[Stupid Dog]

After installing this release on my MSI K8T-Neo Board, I have no drivers for scanner, printer (LPT1 unsupported...), ISDN PCI Card, Wireless LAN card and many other things.

I spent nights searching and begging for drivers. Thank you Microsoft for bringing this truly Linux user feeling to the Windows platform.

Re:Thank you!

[higgins]

I was very confused and kept going to this Mike Rowe guy's site instead. I tell ya. There oughta be a law.

sizeof (long long) == 8...

[Ayanami Rei]

and the Opteron can hold it one register. Just prefix the instruction with the OTHER size prefix byte. (for those who don't already know, most intel instructions if operating on a 16-bit short require a prefix byte. On the Opteron, you use a different prefix to get 64-bit ints and the extended regs)

There are plenty of places where it makes sense to use 64-bit regs, especially in the kernel when involving counters, timers, GIDs, and such.

Windows Server 2003 Also Available for Free

[xTK-421x]

Here: http://www.microsoft.com/windowsserver2003/64bit/e xtended/trial/default.mspx

Beta of course.

64 bit lies

[bl8n8r]

Finally, we'll be able to have an integer big enough to count the number of lies SCO has told so far.

Re:Well, on Windows

[DexterX]

On Win2K and higher, you can split out the file system browsing from the desktop instance of Explorer.exe. Run Explorer and select Tools->Folder Options->View->"Launch folder windows in a separate process". It uses a little more memory but prevents a slow device from slowing down the rest of the GUI.

You're right, though; Explorer blocks waiting for I/O way more often than it ought to. Most of the multithreading in GUIs is pretty good on Windows, so I'm not sure why Explorer tends to block on floppy I/O, network I/O, etc. Fortunately, it usually doesn't affect other running apps.

64-bit Windows

[kylef]

Of course windows likely will run slower since it's so optimized for the older 32bit platform.

The last time I checked, NT is built on something called a "Hardware Abstraction Layer" that made it relatively painless to port NT from MIPS to x86 and then to PowerPC. (NT was designed on MIPS R4000 machines which themselves were completely designed internally by Microsoft. This effort was deemed necessary to keep the codebase free of x86-specific assumptions and optimizations since portability was a key NT goal.) The hardest part about getting your system to run on a new 64-bit platform is getting drivers to work; generally you need lots of support from hardware vendors to accomplish this feat. Getting the OS itself to compile is the easy part.

But I doubt seriously that Windows NT is "so optimized for the older 32bit platform." The kernel is clearly portable to other architectures, and was in fact developed FIRST on a non-x86 architecture with different properties (page size, Endian-ness, etc). This leads me to believe that it is emphatically not "optimized" specifically for 32-bit x86. If you have evidence otherwise, I would like to see it.

Linux is just a much more mature platform for 64bit computers.

Much more mature? Perhaps you were unaware of Windows XP 64-bit Edition? Sure, it only runs on Itanium, but do you not honestly think that for Microsoft to have released it in early 2003 that they would probably have been working on it and testing it for at least a couple years prior to that? Also, from Microsoft's website, I notice that they have also implemented a 32-bit emulation layer for Itanium called "Windows On Windows 64" (WOW64) that lets the OS run 32-bit X86 code. Does Suse have this capability built-in?

The other issue which I pointed out earlier is the driver situation. You can't really call a product "much more mature" unless its drivers are more mature. I don't see a clear win either way at the moment.