Miscellaneous Operating Systems/Hardware

I'd say the big delay to market kind of killed the Itanium before it was even born...

Maybe it's me but I have always thought it was the #1 reason that SGI got tripped up, cutting off it's MIPS stuff, and getting ready for the switch to Itanium, to only find out that it wasn't ready, and then it wouldn't be ready for YEARS while they had to scramble to try to keep the company afloat with some kind of faster MIPS.

On the Windows front, we always kept on hearing more and more about how great this 64bit computing would be, but again such massive delays just made it all feel like another vapor ware special from MS & Intel.

Not to mention when the Merced finally came out they were next to impossible to get, and we as customers were told to sit this round out as they frankly were not ready for production. Heck even the first version of Windows 2000 for the Itanium was lovingly called "Windows 2000 Advanced Server Limited Edition" which may sound great if you want to build an exclusive clubhouse, but it just said to us that this thing was more of a proof of concept then a real shipping product.

And then of course in 2003 the x64 stuff launched. It was really cool for Linux & BSD users as we suddenly moved all of our firewalls and internal stuff to the x64 and it was great. The custom in house stuff we did we luckily still have the source to, so we re-built and away we went. The big 'deal' was price, and part availability. And with the AMD stuff we really felt like that we had the ultimate path forwards & backwards. If we needed to run something old on any new machine we got, we could. Hell we can even run our MS-DOS product natively and all is well.

There is a lot to be said for hardware level comparability.

The whole time though we had kept on hearing from our tech contacts at Microsoft that they had a x64 version of Windows "in the wings" and they were just waiting for the right moment to release the product. The basic impression we had gotten from them was that if you didn't barge into the itanium market around 2001-2002 that you would be better off waiting for the x64 as they were already getting ready to scale back and drop the Itanium... Which is pretty evident in that windows 2003 was a 'full' release while the 2008 version is pretty scaled back. Hell you can't even have an Itanium dhcp server or active directory server. They clearly were phasing the product out. This is not even counting how they killed the workstation versions.

Which I guess now is no big surprise, esp once Windows on the x64 shipped, the Itanium version was dead.

It's just like how the MIPS/PPC/Alpha versions of Windows NT all died. Just show someone a RISC version of Windows, and the first thing they notice is that it looks *JUST LIKE THE i386*... But the hardware will cost a lot more. And they can do the exact thing. And that native applications are pretty dammed hard to get.

I mean even the whole 'native application' thing is moot for the Itanium as MS FINALLY SAW THE LIGHT OF CROSS COMPILERS... Mostly because they knew most devs would never buy Itaniums. But even that couldn't save the platform.

Instead it's always been, unless you need xyz app to run FASTER today, and you don't care how much it costs, because you need speed there hasn't been a compelling reason to buy a non x86 version of Windows. And the kicker is that Intel just kept on cranking out faster CPU's that you could constantly upgrade CPU's and get faster machines, as Intel had lept over the CISC performance "wall".

So I think the larger question about the Itanium's future is how much money is it making, vs how much it cost to make? I know they'll keep on making CPU's because of obligations to people that switched over to it, killing their own RISC cpu's, but eventually Intel will give up and move on. Even RedHat has given up on the Itanium.

I bet there is some pretty pissed off people at HP, as they have basically screwed themselves out of control of their own destiny... So where will they be in 5-10 years from now? Trying to fab their own Itaniums? Or porting to the x64..?

I just found out that IBM bought transitive, effectively killing that escape route. I had thought it would be kind of funny how they got all those great contracts getting people ONTO the Itanium (or apple from the PPC to the x64) but their next big thing would be getting people OFF the Itanium.

I guess someone ought to start funding Qemu.......

Ok I think that's enough rambling.

Speaking of Itaniums running Windows, does anyone have one?

I built neko & quake for it, but have no way to test it.....


http://vpsland.superglobalmegacorp.com/install/neko-ia64.zip
http://vpsland.superglobalmegacorp.com/install/ezquake.zip

I can load it on mine but it might take a while lol, I have to set it up on my netboot server. I think I just have 2008 and not 2003.

That'd be cool, but I don't know if I'd kill myself to have an animated cat chase the mouse around... lol
Not to mention, now that I think about it, the market of people to play games on Itaniums is probably around 0.....

Well I wouldn't mind playing games on mine, but I agree the market for games on windows on ia64 is 0 lol, hpux or linux might be more than 0

The first download requires authentication, probably a permissions issue; either way, download won't go. Speaking of games, with D OOM for IRIX and even VMS, who needs anything else?!

I want something with OpenGL so I can test out my ATI FireGL X1/Z1 card lol.

Maybe it's time for a community game development project? (And then only releasing it for non—x86/-64 platforms).

Oops it kept the permissions across the shares.... Don't you hate it when NT actually *WORKS* and you are so used to it not mattering?

lol

Anyways I fixed the link.

For the quake thing, I built it for the x86, x64, MIPS, Alpha and Ia64... They all seem to play locally fine.

If you have compilers.... check out quake... The source is out there, and it's all got bindings for OpenGL...

As far as OpenVMS goes, the work that they did porting from Alpha to Itanium essentially removed most of the hardware dependencies of OpenVMS (no more PALcode, even though it really should have been the other way around with other processors picking up the idea), so provided HP keeps VMS around the switch to AMD64 should be pretty smooth. The hardest part will be IEST or whatever they call the Itanium->AMD64 version of DECmigrate.

Er...

...



...

Bwahahahahaha okay, hold it - are you seriously suggesting that they would consider porting UX and VMS to AMD64?

No, seriously. They won't. The big customers of enterprise platforms need levels of error correction that wintel hardware cannot give, and won't be able to give in a long, long time - if ever at all. If they were to move to wintel, the OSes themselves would have little to no added value anymore.

Oh absolutely, and expect the Itanium to be pulled in 5 years or so. The volume isn't there and all those "ecc" features you want are appearing on the xenons. The writing is on the wall..(well it's been for some time now) that the x86 has basically won out.

Most OSes outlive the hardware they run on. HP have had to port HP-UX from M68k to PA-RISC and Itanium in the past, and VMS from VAX to Alpha and then Itanium. I'd be very surprised if internally they haven't at least analysed what would be involved in a move to x64, in case it some day becomes necessary, just as has in the past.



Going back to the original post of this thread, thats actually the reason MS give for killing the IA64 port - reliablity and scalability of x64 is evolving to the mission critical levels needed by industry, therefore making the main selling point of Windows on IA64 somewhat redundant.

That's the MIckeysoft m.o., all right. Tell a big enough lie with enough sincerity and get Ziff-Davis to print iit enough times, all the Bo Ewalds of the world will follow you anywhere

I think this is sort of covering up the point that once you take the innate reliability of Windows into account, you're not giving up much by moving from IA64 to x64.

Fact is, Mickeysoft can't compete with their Windows-suite on the enterprise (IA64) platforms simply because noone asks for them. In that territory the customers don't even care for Windows since all their apps are running on other platforms. Want to run notepad.exe? There are verrrry few apps out there for Win/IA64, so why bother?

Mickeysoft DO however try to target bankings and other larger firms with their SQL-server on IA64 platform, but as much as HP wants to advertise about this, they don't win marketshares in the extent they want to.

Just look at Windows NT/2000 on Alpha, Mickeysofts tryouts in the PPC/MIPS territory, rescent IA64-venture... It all boils down to the volume segment and where they can get enough money for their investments. Just about any company financial director would look at the facts and just axe it, like they've done several times before.

MS had a crazy-sweet deal on Alpha NT. DEC did the porting, DEC did the patches, DEC did the support. MS just took a cut of each copy. It got cancelled because DEC/Compaq realized there was little-to-no point in putting in all that work - people who bought Alphas generally wanted Linux/DUNIX or OVMS.

Well don't forget MS got the DEC team after DEC killed prisim/mica.. But then DEC could never sell something like NT, just look at how they mismanged the Alpha... Even the ex DEC people used to laugh that DEC could have had NT for 'free' but instead DEC went to lay them all off... Dec only ened up 'getting' NT after lawsuits saying that NT's razzle/dazzle was the whole prisim/mica thing stolen.

Compaq just had no vision, nor any care about being a chip maker, so killed the Alpha..

People used the Alpha who had lots of $$$ and needed speed *NOW*... That's one thing for sure, it never was cost effective, or for a long term buy, as intel kept on closing the gap on the Alpha CPU's by 1-2 years...

But with the Itanium, we've seen it lose linux support, and now windows.. Like the money pit that was OS/2, eventually someone at Intel is going to look at the cost of the chip, look at the revenue, and see how much it'd cost to just dump it on HP and abandon the whole thing.....

Intel always is about big volume, and the fact is the Itanium is never going to deliver there.

No, but it's insanely high-margin, as well as being a huge industry in terms of revenue. Something like a third of total server revenue comes from IA64, SPARC, and POWER, despite being only a couple percent of units shipped. Of that, IA64 had a 26% market share of that market in 2008, and it's supposedly grown since due to Sun's implosion. That's serious money, and I don't think Intel wants to just leave it to IBM and Oracle. Windows and RHEL losing IA64 support are probably inconvenient for some people, but really just aren't that big of a deal, since the vast majority of IA64 systems run HP-UX or VMS anyway.

IA64 is going to survive long-term.