Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

HP's NonStop Servers Go x86, Countdown To Itanium Extinction Begins

Unknown Lamer posted about a year ago | from the itanic-cannot-sink dept.

HP 243

An anonymous reader writes "HP has been the sole holdout on the Itanium, mostly because so much of the PA-RISC architecture lives on in that chip. However, the company recently began migration of Integrity Superdome servers from Itanium to Xeon, and now it has announced that the top of its server line, the NonStop series, will migrate to x86 as well, presumably the 15-core E7 V2 Intel will release next year. So while no one has said it, this likely seems the end of the Itanium experiment, one that went on a lot longer than it should have, given its failure out of the gate."

cancel ×

243 comments

Sorry! There are no comments related to the filter you selected.

Once you go x86 (-1, Troll)

Anonymous Coward | about a year ago | (#45344875)

Once you go x86, you can sux0rs my dicks!

EPIC failure (0)

Anonymous Coward | about a year ago | (#45344881)

Also, Itanic

Re:EPIC failure (1)

Megane (129182) | about a year ago | (#45345231)

I am extremely disappointed that this Register article [theregister.co.uk] does not use the word Itanic. Not even the (currently 8) reader comments. I expected higher standards from such an organisation as El Reg.

Meh (0)

Anonymous Coward | about a year ago | (#45344885)

Meh. This pretty much describes IA64/Itanic, as well as any "news" related to it.

OTOH, you will be able to get those in eBay even cheaper, Debian runs on it just fine, and shellcode for IA64 is pretty much nonexistant.

I suspect it is bcos of HP's TCPA connection (2, Interesting)

jkrise (535370) | about a year ago | (#45344895)

Not a single major hardware or device maker seems ready to support Linux on non-Intel architectures. Intel, MS, HP, Cisco etc. are part of the TCPA alliance; even Linux on ARM based servers have taken a very long time to arrive.

Re:I suspect it is bcos of HP's TCPA connection (4, Insightful)

kthreadd (1558445) | about a year ago | (#45344917)

IBM supports Linux on their Power based systems, and I don't think they have any plans to stop that.

Re:I suspect it is bcos of HP's TCPA connection (0)

Anonymous Coward | about a year ago | (#45345161)

Not to mention s390x...

Re:I suspect it is bcos of HP's TCPA connection (1)

Anonymous Coward | about a year ago | (#45345201)

And its zSeries mainframes too...

Re:I suspect it is bcos of HP's TCPA connection (1)

cyber-vandal (148830) | about a year ago | (#45345047)

Are you saying Itanium is a non-Intel architecture?

Re:I suspect it is bcos of HP's TCPA connection (1)

disposable60 (735022) | about a year ago | (#45345241)

I think it was a sloppy phrasing of non-x86 arch. Itanium _can_ do x86, but not as efficiently ... IIRC (it's been over a decade since I've touched one)

Re:I suspect it is bcos of HP's TCPA connection (2)

cant_get_a_good_nick (172131) | about a year ago | (#45345861)

(it's been over a decade since I've touched one)

And your fingertips are still burnt. (Itanium as a marshmallow and hotdog heater joke)

Re:I suspect it is bcos of HP's TCPA connection (3, Interesting)

armanox (826486) | about a year ago | (#45345101)

More importantly, major Linux vendors (Red Hat and Canonical in particular, I think Novell is the odd ball on this one) don't release for Itanium. Power is still supported by many, and ARM is a rising star, but IA64 seems to be heading the way MIPS went....

Re:I suspect it is bcos of HP's TCPA connection (5, Interesting)

lowen (10529) | about a year ago | (#45345321)

Red Hat Enterprise Linx 5 is still available and supported for IA64. At least at the moment; this will give IA64 users a Linux soure base at least until 2017.

I have personally rebuilt CentOS 5 from source for SGI Altix, which is an IA64 box, and am running a smallish Altix (30 CPU's, 54GB of RAM) in production for data analysis. (NASA's Columbia supercomputer was an IA64 Altix with 10,240 CPU's.....)

But RHEL 6 is indeed not available for IA64.

Re:I suspect it is bcos of HP's TCPA connection (0)

Anonymous Coward | about a year ago | (#45346665)

Doubtful, as MIPS went from servers to high-end embedded.

Re:I suspect it is bcos of HP's TCPA connection (2)

Lumpy (12016) | about a year ago | (#45345163)

Because it requires paying competitive wages to skilled people. They dont want to do that.

Re:I suspect it is bcos of HP's TCPA connection (1)

Anonymous Coward | about a year ago | (#45345491)

Not a single major hardware or device maker seems ready to support Linux on non-Intel architectures

LOL ANDROID

Re:I suspect it is bcos of HP's TCPA connection (3, Informative)

cant_get_a_good_nick (172131) | about a year ago | (#45345829)

Err, I know Slashdot doesn't allow editing of comments, but Itanium is definitely an Intel architecture. I assume you really meant to put "non-Intel-x86 compatible". It's still significant since HP helped designed the chips.

I forgot the code names, but the first Itanium was Intel designed. Had really bad performance, landed with a thud. HP (back when they had engineers and not marketers) designed the second set, which actually was a decent chip. HP had a lot vested in this, HP slowly moving away from Itanium is very very big.

Re:I suspect it is bcos of HP's TCPA connection (1)

jkrise (535370) | about a year ago | (#45345993)

As another poster pointed out; I intended to say non-x86; not non-Intel. In 2001; I was buying about 250 desktops from HP(Vectra VE5 if I recall right); and the marketing folks from HP sang loud praises of the 'Merced' project which later morphed into the Itanium range. The guys claimed that the architecture was entirely done by HP (primarily to support HP-UX) so I presume even though Intel did the design and production of the chips; they could not implement TCPA in that.

Re:I suspect it is bcos of HP's TCPA connection (0)

Anonymous Coward | about a year ago | (#45346259)

This site is full of some seriously confused (borderline mentally ill?) people. The vendor politics around Itanium have nothing to do with "trusted computing".

Re:I suspect it is bcos of HP's TCPA connection (1)

fuzzyfuzzyfungus (1223518) | about a year ago | (#45346251)

Nobody supports 'Linux on ARM based servers' at the size you appear to be thinking of, because there aren't any, more or less.

Now, at the scale where actually-available-now-at-reasonable-prices-and-with-suitable-peripherals ARM cores are available, everybody supports Linux on ARM based servers, they're just called 'NAS'es, because they are small and feeble and typically just configured for light file-serving duty.

given its failure out of the gate. (4, Interesting)

serviscope_minor (664417) | about a year ago | (#45344909)

given its failure out of the gate.

For a multibillion dollar industry, "failure" is a rather strong term. It may be declining, but it topped over $4.4bn a year at one point. That's probably bigger than AMD.

Re:given its failure out of the gate. (2)

gsnedders (928327) | about a year ago | (#45345067)

Depends on the year, assuming we're talking revenue. AMD topped that in 2000, and then in every year from 2004 onwards.

Re:given its failure out of the gate. (3, Insightful)

Big Hairy Ian (1155547) | about a year ago | (#45345075)

Faliure??? I know of firms and organisations that still haven't retired their old IBM3000 mainframes!! Itanium will be around for a while to come.

Re:given its failure out of the gate. (0)

Anonymous Coward | about a year ago | (#45345319)

The company I'm at is still using an IBM 7094 to maintain the database. You wouldn't believe how hard it is to maintain code for that thing.

Re:given its failure out of the gate. (1)

ZorinLynx (31751) | about a year ago | (#45345369)

Fuck the code, what about the hardware? This machine was designed in the early 60s. :)

Re:given its failure out of the gate. (0)

Anonymous Coward | about a year ago | (#45345789)

Which means it was engineered to last, because I dunno, they knew how to do things then.

John Titor through Stein's gate. (0)

Anonymous Coward | about a year ago | (#45345567)

> How hard it is to maintain code for that thing.

It's that hard. That company (hint: US gov't) actually had to hire a time-travelling confederate soldier of fortune to fetch an 1975 era IBM 5100 desktop debug system of that platform for them them. No kidding!

Re:given its failure out of the gate. (1)

Tore S B (711705) | about a year ago | (#45346663)

That's a lie. There are no running 7094s in the world.

Re:given its failure out of the gate. (0)

Anonymous Coward | about a year ago | (#45345209)

given its failure out of the gate.

For a multibillion dollar industry, "failure" is a rather strong term. It may be declining, but it topped over $4.4bn a year at one point. That's probably bigger than AMD.

Hey, it's not Linux on x86, so it must be a non-existent failure.

Re:given its failure out of the gate. (2)

jedidiah (1196) | about a year ago | (#45346587)

> Hey, it's not Linux on x86, so it must be a non-existent failure.

Linux on x86 is a big part of the reason that it is a non-existent failure. It also didn't help that Itanic was an engineering disaster in it's own right.

Re:given its failure out of the gate. (0)

Anonymous Coward | about a year ago | (#45345401)

Failure isn't defined by numbers alone. Not too surprising this got by you since you didn't do the research on the numbers in the first place.

itanic was supposed to take off big, and they kept talking it up with obviously wildly over-optimistic "estimates" (you know the graphs I'm talking about), only, you know, it didn't deliver, and kept on failing to deliver. It lagged orders of magnitude behind "predictions", giving it a rather unmistakable "Baghdad Bob" flavour. Thus making it a good example of a failure.

It got so bad that intel had to bow to AMD and implement AMD's x86_64 extensions (oh irony! and then double irony on you) just to keep up.

So I don't know by what definition it wasn't a failure, but you're welcome to try and come up with one.

Personally I don't mind the demise of this monster. Now if only we could get over it and get some new architectures to see the light. Even bringing back PARISC and Alpha (updated designs, newer processes, higher clock speeds) looks good at this point, but I'll take MIPS or something yet different again too. In their day those were very viable, insofar as that the x86 dominance allowed it, something itanic never was. Yes, indeed, very good, Alanis.

Re:given its failure out of the gate. (4, Informative)

dj245 (732906) | about a year ago | (#45345461)

given its failure out of the gate.

For a multibillion dollar industry, "failure" is a rather strong term. It may be declining, but it topped over $4.4bn a year at one point. That's probably bigger than AMD.

It is a complete and dismal failure if you consider Intel's plan for this architecture. It was supposed to be the next i386, the architecture all processors would use. Instead it was a huge flop in the beginning, and only redeemed itself 2 generations later. AMD snuck in their own 64 bit architecture which became the de-facto standard for all 64 bit laptop/desktop processors. Itanium became the architecture of a few supercomputers, and gained a toehold into some miscellaneous scientific computing niches.

In this respect it is about a big a failure as "new coke". Sure, selling it may have been profitable, but it failed to meet expectations and become the Next Big Thing.

Re:given its failure out of the gate. (3, Informative)

jythie (914043) | about a year ago | (#45345847)

When they speak of 'failure' they are referring to fashion, not business. At least they should be.....

Itanium was not popular, it was not fashionable, it was not sexy, it didn't have geek credit, but it made a lot of money.

Re:given its failure out of the gate. (0)

Anonymous Coward | about a year ago | (#45346895)

If it "didn't have geek credit," it still had way more geek cred than x86. People unfamiliar with any chip than x86 and insecure about it, who know nothing about CPU architecture besides the marketing codename salad Intel uses to promote x86 to people who are already buying it, shouldn't make comments like this because the story of CPU's since the beginning has been "the technically worst architecture wins for a combination of stupid business reasons, both internal and external to the CPU company."

Re:given its failure out of the gate. (1)

Uteck (127534) | about a year ago | (#45345871)

That is nothing compared to the money it took to design and fab the chips, and most of that money was fronted by HP, http://www.wired.com/wiredenterprise/2012/02/hp-itanium/ [wired.com] .
Given the production costs and the number of customers using it, so it was always a high cost chip with mediocre performance.
Intel would rather shut down thoughs fabs and make more x86 ships which make them more money.

Re:given its failure out of the gate. (1)

cant_get_a_good_nick (172131) | about a year ago | (#45345917)

It's all relative. Itanium was slated to be *the only* 64 bit chip [superuser.com] , replacing x86 with a new architecture. It was supposed to be the only server chip as it cleaned up all the RISC chips out of the market. It kind of did the latter - only Sparc and POWER still really exist, MIPS, PA-RISC, and Alpha are gones.

But the goals were high. Destroy all other chips. even x86. Not have a second vendor (no more AMD making x86 chips) meant all the money went to Andy Grove. They never did close to any of that. Based on the money poured in, and the expectations, it is a failure. Maybe not Apple Newton level failure, but it is a failure.

Re:given its failure out of the gate. (0)

Anonymous Coward | about a year ago | (#45346409)

It's really interesting how SPARC keeps staying in business and is even running incredibly strong again. Who would have thought 5 years ago.

Re:given its failure out of the gate. (4, Informative)

itzdandy (183397) | about a year ago | (#45346927)

It kind of did the latter

That's not even a stretch, it's completely false. Commodity x86/x86-64 clearly did the overwhelming bulk of eliminating other architectures by offering drastically better price/performance or maybe even more importantly, bringing the minimum server configuration down sub-$1000. Before the 'Xeon' and X86-64, servers were very much over powered and over engineered for many businesses.

Placing a $20,000 HP-UX/HPPA server in a small business and getting a baseline of 3% usage put these systems out of reach for obvious reasons. A $1000 Xeon box that performed similarly was the obvious choice. Itanium was never in the discussion and had effectively nothing to do with the decline of the MIPS and RISC server market.

--IMHO

Goes along with the VMS announcement (4, Informative)

brausch (51013) | about a year ago | (#45344937)

Earlier this year HP announced the end of the line for VMS. That was certainly connected with the Itanium retirement as well.

Re:Goes along with the VMS announcement (1)

guru42101 (851700) | about a year ago | (#45345089)

I'm surprised it's still around. It was dieing when I was assisting administrating one in college 15 years ago. All they used it for was DNS, DHCP, some web hosting of minor services (DCL Script ugh!!!), and forwarding of employee addresses (first.last@abc.edu to flast1@mail.abc.edu)

Re:Goes along with the VMS announcement (1)

hughk (248126) | about a year ago | (#45345115)

I think the US military has been using VMS for mission critical applications - payroll!!!!

Re:Goes along with the VMS announcement (1)

sinator (7980) | about a year ago | (#45345185)

Also, all the health care: CHCS runs on VMS and will continue to do so through 2018 or even later, depending on the speed of the DHMSM COTS acquisition process.

Re:Goes along with the VMS announcement (1)

Tore S B (711705) | about a year ago | (#45346743)

To put the reliability into perspective. I was speaking with a VMS sysadmin when I was 19 years old, who exclaimed that he had support contracts on cluster with higher uptimes than I'd been alive.

It is a really, really rugged OS. The clustering has an elegance that I miss on Unices.

Re:Goes along with the VMS announcement (1)

operagost (62405) | about a year ago | (#45345869)

Just because your university was phasing it out didn't mean the platform was dead, did it?

Re:Goes along with the VMS announcement (0)

Anonymous Coward | about a year ago | (#45346189)

VMS still has a strong following. It's an extremely reliable O/S and platform that pretty much defines the gold standard of "Clustered" systems. It has been declining, but much slower than even HP realizes. One of it's strengths is reliability. I know several systems that do not have any HP support, but have been running continuously for over a decade.

While dependency on Integrity (and it's shrinking market share) is crippling OpenVMS, the major factors was Compaq's and HP's lack of ability to really market it, and more recently, the movement of all OpenVMS Engineering to India. There is a good market for emulated VAX and Alpha hardware that provides continued support for OpenVMS, though. In many cases, the emulated hardware (VAX or Alpha emulators on Intel using Windows or Linux) has better performance than many of the lower end systems.

Re:Goes along with the VMS announcement (1)

bpechter (2885) | about a year ago | (#45345325)

Damned shame that they killed Alpha and with that move doomed VMS. The Itanium port didn't help expand the VMS base, since there wasn't enough support to VARs to keep the support for VMS in applications. There are only two viable OS choices now.
Windows and Linux/Unix. (And the Unix part is weakening over time due to costs vs. Linix).

Re:Goes along with the VMS announcement (1)

TheRaven64 (641858) | about a year ago | (#45345597)

The funny thing about the whole story is that the reason i386 has four protection rings was to make a VMS port from VAX easier. Instead, DEC ported it to Alpha, with only two...

Re:Goes along with the VMS announcement (1)

dpilot (134227) | about a year ago | (#45346307)

And have you ever seen any x86 software use more than 2 rings, even though it has 4?

Re:Goes along with the VMS announcement (0)

Anonymous Coward | about a year ago | (#45346763)

Yep -- OS/2 used ring 2 for some drivers. That's one of the reasons is so hard to support under most VM products.

Re:Goes along with the VMS announcement (1)

dpilot (134227) | about a year ago | (#45346917)

Used OS/2 for a while, never knew it put any drivers in ring 2. I know both OS/2 and Windows used ring0 for the kernel, and I seem to remember that one of them initially used ring 1 for userspace, then in a subsequent release moved userspace to ring 3. Aside from that early use of ring 1, I thought pretty much everything since then put the kernel in ring 0 and userspace in ring 3.

Netcraft Confirms It (-1)

Anonymous Coward | about a year ago | (#45345335)

It is now official. Netcraft has confirmed: VMS is dying.

One more crippling bombshell hit the already beleaguered VMS community when IDC confirmed that VMS market share has dropped yet again, now down to less than a fraction of 1 percent of all servers. Coming on the heels of a recent Netcraft survey which plainly states that VMS has lost more market share, this news serves to reinforce what we've known all along. VMS is collapsing in complete disarray.

You don't need to be Criswell to predict VMS's future. The hand writing is on the wall: VMS faces a bleak future. In fact there won't be any future at all for VMS because VMS is dying. Things are looking very bad for VMS. As many of us are already aware, VMS continues to lose market share. Red ink flows like a river of blood.

(For old times sake...)

Re:Netcraft Confirms It (1)

Tridus (79566) | about a year ago | (#45345867)

Ah, classic.

Re:Netcraft Confirms It (0)

Anonymous Coward | about a year ago | (#45346819)

less than a fraction of 1 percent

99999/100000 is a fraction, so are you saying it is now 99.998%, which rounds up to 1%?

Re:Goes along with the VMS announcement (2)

unixisc (2429386) | about a year ago | (#45346967)

Earlier this year HP announced the end of the line for VMS. That was certainly connected with the Itanium retirement as well.

They should have announced an end to VMS/Itanium with it i.e. no new sales on that. On the above subject, I guess it's good that they are finding a home for Tandem customers. Really speaking, once Alpha/MIPS were dead for HPQ, they should have migrated both VAX and Himalayas to x64 instead of Itanic. That way, their customers would have had more options whenever HP decided to discontinue VMS, or transition NonStop.

Anyway, at this point, I think the writing is on the wall for Itanic - it's now an HP/UX only home, and that too, not anywhere near as popular as PA/RISC was. Too bad, since Itanium III was starting to look good. Intel should either reduce its power consumption, increase the cores and explore the supercomputer market w/ it, or failing that, close it down.

Microsoft knows this (4, Interesting)

Dr. Manhattan (29720) | about a year ago | (#45344991)

I work on a product that supports Itanium, and we have a few customer that are still using Itanium servers, who knows why. We just discovered that unless you get the top-tier developer subscription to Microsoft Visual Studio, you don't get Itanium compilers.

Re:Microsoft knows this (1)

fuzzyfuzzyfungus (1223518) | about a year ago | (#45346357)

Itanium has long been a signal, a signal that says "I really do cut POs that large. Soak me."

Now that it's niche and dying, the fun will really start. Sure, the number of customers will shrink; but the last ones to go will be the ones who will pay almost anything for one last hit before they have to port...

#tandem4life (0)

Anonymous Coward | about a year ago | (#45345025)

Tandem on the x86! Intel is making an x86 cell phone. You know what this means? Tandem on a cell phone! Now all we need is a touch interface for TACL.

#tandem4life

IA64 ~~ IPV6 (0, Troll)

CaptSlaq (1491233) | about a year ago | (#45345061)

Now if only the IPV6 community would see the parallels between IA64 and AMD64...

Re:IA64 ~~ IPV6 (0)

Anonymous Coward | about a year ago | (#45345499)

Now if only the IPV6 community would see the parallels between IA64 and AMD64...

What parallels?

The main problem with IPv4 is that it ran out of bits for addressing. Is this like IA-32/x86 running out of bits for (RAM) addressing?

AMD64 added more bits for addresses (in addition to other tweaks). IPv6 added more bits for addresses as well.

Any way you cut it, there was going to be pain when going from IPv4 to "IPng". You had to expand data structures in the packet, and there wasn't a way AFAICT to do that in a backwards compatible way. You had systems that were designed to deal with only 32b and now had to deal with more-than-32b. There was going to be breakage regardless of what design was chosen for "IPng".

The main problem with IPv6 IMHO is that everyone is dragging their feet. I can understand the trepidation from a business/money point of view, but technically rolling out dual-stack sooner rather than later was the best way. Do it slowly over time and you can work through the issues at a leisurely pace. Now we're running against the clock and when (not if) the IPv4 counter goes to zero there's going to be a lot of problems for a lot of people.

Re:IA64 ~~ IPV6 (1)

PlusFiveTroll (754249) | about a year ago | (#45345523)

I'm not really sure where you're going with that one, with dual-stack any system that can run IPv6 can also run IPv4.

Re:IA64 ~~ IPV6 (0)

Anonymous Coward | about a year ago | (#45345547)

And what contrasts with IPv6 in the same way IA64 compared with AMD64?

Ah, Itanium (1)

0123456 (636235) | about a year ago | (#45345073)

Based on the amusing idea that compilers can more easily determine which instructions can be executed simultaneously at compile time than the CPU can at run time...

Years ago one of my friends had the misfortune to have to write code generators for a CPU which required the compiler to determine whether a previous pipelined instruction had completed before reading the result because there were no interlocks to stall the CPU if it hadn't. He could have told Intel a thing or two about trusting software engineers to schedule instructions rather than CPU designers.

Re:Ah, Itanium (1)

KirbyCombat (1142225) | about a year ago | (#45345127)

That would have been the i860.... Ah, the days....

Oh the MEMEs (5, Interesting)

ArhcAngel (247594) | about a year ago | (#45345147)

Let me count the ways
And there was much rejoicing!
And nothing of value was lost.

For those saying it wasn't a failure you must look at what Intel intended Itanium [wikipedia.org] for. If they had succeeded Intel would have owned the 64 bit CPU realm on the desktop with a proprietary architecture effectively eliminating any competition in the space. To succeed they had to get all popular software including Windows to be rewritten for the new processor. This was a daunting task and few were ready at the time to make the switch to 64 bit. AMD introduced the Opteron [wikipedia.org] in 2003 with their 64 bit extensions for the existing x86 architecture which allowed the reuse of the 32 bit code in existence. AMD's x86-64 was well received and Intel ultimately adopted the architecture in their own processors. So yes the Itinaium failed to succeed in its intended task despite lingering for over a decade.

Re:Oh the MEMEs (4, Insightful)

Lluc (703772) | about a year ago | (#45345245)

If they had succeeded Intel would have owned the 64 bit CPU realm on the desktop with a proprietary architecture effectively eliminating any competition in the space.

Realistically, Intel would have licensed the IA64 architecture to AMD or some other third party. Intel would not want to have an absolute CPU monopoly and risk government intervention. It is much better for Intel to have a barely competitive company (currently AMD) operating in the same space but not offering any kind of threat to their market position.

Re:Oh the MEMEs (1)

ArhcAngel (247594) | about a year ago | (#45345955)

It's not a competition if you set the terms for your competitors. Intel doesn't care if they get their money from direct sales or licensing. In fact licensing is a sweet deal since they get paid without having to produce anything. Much like how Apple makes millions just raking in licensing fees for power cords.

Re:Oh the MEMEs (1)

fuzzyfuzzyfungus (1223518) | about a year ago | (#45346385)

Hard to say: having a nice 64-bit address space beats the hell out of PAE; but '32-bit = 4GB of RAM or less' was always only a Windows desktop thing. Server and workstation had PAE support. Had Intel wanted a gimp around, they might have been able to get away with AMD flogging PAE gear...

Re:Oh the MEMEs (2)

dpilot (134227) | about a year ago | (#45346473)

The whole license issue wasn't sufficiently covered in any of this discussion, not in TFA, not even in the Wikipedia article. Reading the Wikipedia article, ia64 seems to have been driven by HP and joined by Intel. I had thought it was the other way around.

Intel does not own IP on ia64, nor does HP. It was done that way on purpose. The ia64 IP is owned by a separate organization spawned by Intel and HP, and that organization licenses the IP back to Intel and HP. The reason... it's cross-license-proof. Intel and HP use other comanies' patents and vice-versa - that's normal practice. Pretty much everyone has to do some measure of cross-licensing, just to be in the business. Had HP or Intel owned the ia64 IP, sooner or later it would have gotten included into some cross-licensing agreement, the cat would have been out of the bag, and competition from clones would have started. The IP holding company is a NPE, so they have no need to cross-license with anyone. (Is this the only non-troll use of a NPE IP holding company?)

It was all done this way to avoid competition with clones. That's why I thought ia64 originated with Intel - it's an anti-clone scheme. It's also Intel solving Intel's problems, rather than their customers' problems, which was what amd64 did. This is why I really hate seeing AMD fade - it removes market pressure from Intel and lets them wander into left field again. Hopefully ARM will keep them decent.

Also, I'd tend to echo ArhcAngel's peer comment, that it's not really competition if you can set the license terms.

That wasn't Itanium's intended task (0)

Anonymous Coward | about a year ago | (#45345437)

It's intent was to kill DEC's Alpha chip. In that it succeeded.

Re:That wasn't Itanium's intended task (1)

PlusFiveTroll (754249) | about a year ago | (#45345549)

Other then it pushed Alpha in to AMD's hands giving birth to the Opteron, Intel snatched defeat from the jaws of victory there.

Re:That wasn't Itanium's intended task (0)

Anonymous Coward | about a year ago | (#45346171)

Alpha was a benchmark champ, but it was not a commercially successful platform & DEC was going under.

It's real intent was to take on IBM, and in that role it failed.

OpenVMS (0)

Anonymous Coward | about a year ago | (#45345153)

Went on for too long? Itanium's the platform for VMS, as in, only platform if you want new iron that runs VMS w/o emulation. Plenty "legacy" control systems run on VMS, so we'll see where all that goes.

Itanium was a legend (4, Informative)

onyxruby (118189) | about a year ago | (#45345233)

Unfortunately it became a legend for all of the wrong reasons. Billions of dollars have been sunk into it over the years and many lawsuits have been filed over it demise by vendors desperate to get out of it or force another vendor to stay in it.

http://www.eweek.com/servers/hp-to-seek-4-billion-in-damages-from-oracle-over-itanium/ [eweek.com]
http://news.cnet.com/Allies-pledge-10-billion-to-boost-Itanium/2100-1006_3-6031773.html [cnet.com]
http://www.masslive.com/news/index.ssf/2013/09/hudson_intel_plant_closing_wil.html [masslive.com]

Unfortunately sales never came close to the billions of dollars that have been sunk into it, and it has been that way for years:

http://www.theregister.co.uk/2005/02/28/itanium_04_sales/ [theregister.co.uk]
http://www.wired.com/wiredenterprise/2012/02/hpearnings/ [wired.com]
http://www.zdnet.com/photos/charts-mining-itanium/21115 [zdnet.com]

I'm sure someone has a comparison of how much money has been invested compared to how much money has been made in sales. I might be mistaken, but from what I've been reading from the beginning Itanium has never come close to breaking even for hardware or software sales. Certainly companies like HP and Oracle spent millions of dollars on their lawsuit trying to get out Itanium.

Itanium has always been nothing more than a desperate multi-billion dollar effort to break free from the chains of x86.

Re:Itanium was a legend (0)

Anonymous Coward | about a year ago | (#45345361)

multi-billion dollar effort to break free from the chains of x86

To become more beholden to the particular company that invented x86...

The idea was sound. The compilers were never up to the task though.

Re:Itanium was a legend (0)

Anonymous Coward | about a year ago | (#45346581)

> The idea was sound. The compilers were never up to the task though.

Uh, is that the new way of saying "the idea was obviously idiotic from the start"? Compilers are crap, always have been crap and probably for a very long time to come will remain crap, even if you ignore the cases where a compiler just can't do anything because the knowledge will only be available at runtime.

Re:Itanium was a legend (4, Insightful)

green is the enemy (3021751) | about a year ago | (#45345665)

Don't look at Itanium in a completely bad light. It was a good microprocessor architecture experiment, and had the right motivations (break free of the x86 legacy cruft, design a truly scalable architecture). A lot of useful technology was developed along the way. This technology will be incorporated into future chips. Intel is rare among large technology companies to actually take huge long-term risks, and even survive failure. We need more high-risk projects like this to develop truly breakthrough technology.

Re:Itanium was a legend (0)

Anonymous Coward | about a year ago | (#45346749)

If so...how come none of it is being used? We're still basically doing a batch of kludges on top of the stuff from back in the 70's.

I stand by my pet name... (sh)Itanium.

OpenVMS? (1)

Dimwit (36756) | about a year ago | (#45345339)

Looks like it's the end of the line for OpenVMS as well.

I would pay good, American money, to have OpenVMS open-sourced instead of just languishing like other DEC OS's. Why can't RSTS/E or RSX-11 be free? What could that possibly cost HP? Same with OpenVMS at this point. It's a great system, and I would love to see it available to average joes.

Someone who isn't as lazy as I am should start an "Open Source OpenVMS!" petition.

Re:OpenVMS? (1)

fuzzyfuzzyfungus (1223518) | about a year ago | (#45346423)

HP's secret, secondary, mission statement requires them to acquire and kill good technologies. It's the rules. In rare cases of dire necessity, they are allowed to spin them off; but that is frowned upon.

Bye Itanium (1)

ErichTheRed (39327) | about a year ago | (#45345343)

I'm sure HP has been staring this one down forever, saying "We sunk all this money into Itanium, there's no way we can abandon it." In fact, if you look at the documents from HP's lawsuit that Oracle helpfully put up on their website, you can see internal discussions of their intention to port HP-UX to x86 and the fact that they're basically paying Intel to keep developing Itanium processors for them.

Itanium was an interesting idea, and the only way to get 64-bit non-Sun, non-IBM hardware until the Opteron came out. But it's a really good example of a technology hanging on way past the point where it's relevant.

I wonder if they've inadvertently sent OpenVMS to the old folks' home by doing this...unless they're planning to port OpenVMS to x86. I know there's plenty of legacy OpenVMS stuff out there, but who knows if those customers would be willing to finance a port by buying machines from HP?

I also wonder if at least some of the ProLiant line is going to get that awesome RAS (Reliability, Availability, Serviceability) that NonStop and the Itanium boxes have. That would be cool.

Re:Bye Itanium (1)

fuzzyfuzzyfungus (1223518) | about a year ago | (#45346443)

I don't doubt that it'll cost you, plenty, to have Intel not laser off their bits and HP not gimp their half of things in firmware; but my understanding is that Intel has every interest in looting Itanium's corpse for the interesting bits and offering those on (high end) Xeon SKUs.

I do often wonder what the point of Itanium was. (1)

ZorinLynx (31751) | about a year ago | (#45345349)

Did Itanium have any performance advantage over x86_64? It certainly didn't have a price advantage, if anything it was horrendously expensive for the performance you got.

We only have a SINGLE Itanium based server here, purchased more out of curiosity than anything else, years ago when the platform was new. There's nothing special about it whatsoever.

I keep thinking the platform should have been declared a failure years ago, unless there was some specific thing it was really good at that I'm not aware of...

Re:I do often wonder what the point of Itanium was (0)

Anonymous Coward | about a year ago | (#45345497)

It's point was that it came out before the AMD x86_64, so if you adopted early how can you even compare.

Re:I do often wonder what the point of Itanium was (1)

afidel (530433) | about a year ago | (#45346339)

HP had Intel put a ton of RAS features that are needed for the Nonstop and Superdome lines into the Itanium chip, in addition it ran modified 64bit code originally targeted at Alpha, MIPS, and NonStop processors, something which couldn't be done with x86. Basically it was a way for HP to jettison all the legacy hardware from their acquisitions over the years and merge them into one "commodity" chip that they could buy from Intel.

Re:I do often wonder what the point of Itanium was (2)

0123456 (636235) | about a year ago | (#45346705)

Itanium would have allowed Intel to dump all the x86 baggage and move the world to a Brave New Shinier CPU that was 64-bit and appeared to offer substantially better performance.

And it would have made them sole supplier for the mainstream CPU market, taking out AMD and the other clone x86 makers.

Unfortunately, the early compilers sucked and x86 emulation really, really, really sucked, so no-one with a big investment in x86 software was going to make the switch. If I remember correctly, it was also years late, so performance that would have been impressive at the initial release date had become 'meh' by the time it actually hit the market.

first pXo5t. (-1, Flamebait)

Anonymous Coward | about a year ago | (#45345467)

distended. All I FrreBSD project, WE ALL KNOW, that sorded, kill myself like disturbinMg. If you very sick and its guys are usually as to which *BSD About a project

And that's what they killed the Alpha for... (1)

Anonymous Coward | about a year ago | (#45345647)

Leaving us with the crummiest architecture of all contenders as the market leader.

But why should hardware have it better than software?

Not so fast. (2)

attemptedgoalie (634133) | about a year ago | (#45345685)

Until there is a supported COBOL environment in Linux, HP-UX on Itanium will be around for a long time.

I work in the power industry, and we use some very specific applications that are only available on HP-UX and AIX. HP-UX is by far their largest install base.

These apps are used by the power plants/coal mines for everything. As you'd expect, there are very few applications that are certified for use by the power industry that meet the regulations. The one we use will begin supporting LDAP instead of NIS next year.

There's no incentive for new players in this software market due to the small number of potential customers and the massive trust curve they'd have to meet to make somebody switch.

We're one of the reasons there's a pretty long road map for Itaniums and HP-UX.

Re:Not so fast. (1)

rubycodez (864176) | about a year ago | (#45346313)

what!!??? The mainstream COBOL compilers have been available on Linux for over a decade: MicroFocus COBOL and Fujitsu NetCOBOL,

Re:Not so fast. (2)

attemptedgoalie (634133) | about a year ago | (#45346971)

And yet ABB won't port their HP-UX app over. We use MicroFocus on UX, so it should be easy.

But no.

HP/UX still not ported (1)

rubycodez (864176) | about a year ago | (#45345691)

when HP/UX x86-64 port is out, then Itanium2 will be dead

Re:HP/UX still not ported (1)

fuzzyfuzzyfungus (1223518) | about a year ago | (#45346469)

when HP/UX x86-64 port is out, then Itanium2 will be dead

Which happens first: HP/UX gets ported to x86, or Xeons pull far enough ahead of now-stagnant Itanium2s that somebody can release a (probably very expensive) Itanium VM that runs at adequate speed?

15-core E7 v2? (1)

ArcadeMan (2766669) | about a year ago | (#45346183)

What are the people at Intel thinking? Where did they get 15-core from? Are they making a 16-core CPU knowing there's always going to be one bad core? Why not make a 17-core CPU instead? 15-core seems odd, it's always been 1,2,4 or 8 cores AFAIK.

Re:15-core E7 v2? (2)

roothog (635998) | about a year ago | (#45346457)

It's just a grid of cores on the chip layout. Nothing wrong with a grid that's 3x5. A 2-dimensional grid does not force the number of cores to be a power of 2.

Re:15-core E7 v2? (1)

kry73n (2742191) | about a year ago | (#45346481)

Expecting that one core will have a defect and can be disabled is a too long shot IMHO, usually a defect is more likely to occur in the big areas like caches that cannot be easily be deactivated. My take on this is that 15 is probably just the number of cores that'll fit on the DIE.

YOU INSENSITIVE CLOD?! (-1)

Anonymous Coward | about a year ago | (#45346279)

Now how am I... (1)

cephus440 (828210) | about a year ago | (#45346361)

... going to write FORTRAN 77 code in OpenVMS? No seriously, what's OpenVMS going to do?

Design by Comittee (4, Interesting)

kent.dickey (685796) | about a year ago | (#45346487)

IA64 started as an HP Labs project to be a new instruction set to replace HP's PA-RISC. VLIW has a hot topic around 1995. HP Labs was always proposing stuff and the development groups (those making chips/systems) ignored it, but for some reason this had legs.

The HP executive culture is: HP hired mid-level executives from outside. They would then do something big to get a bigger job in another company. A lot of HP's poor decisions in the last 20 years can be directly traced to this culture. And there was no downside--if you failed, you'd go to an equivalent job at another company to try again.

So enterprising HP executives turned HP's VLIW project into a partnership with Intel, and in return HP got access to Intel's fabs. This was not done for technical reasons. Intel wanted a 64-bit architecture with patents to lock out AMD, and would never buy PA-RISC. So it had to be new. HP was behind the CPU performance curve by 1995 due to its own internal fab not keeping up with the industry due to HP not wanting to spend money. So HP could save billions in fab costs if Intel would fab HP's PA-RISC CPU chips until IA64 took off. So, for these non-technical reasons, IA64 was born, and enough executives at both companies became committed enough to guarantee it would ship.

For a while, this worked well for HP. The HP CPUs went from 360MHz to 550MHz in one generation, then pretty quickly up to 750MHz. I thought IA64 would be canceled many times, but then it became clear that Intel was fully committed, and they did get Merced out the door only 2 years late. IA64 was a power struggle inside Intel, with the IA64 group trying to wrest control from the x86 group. That's where the "IA64 will replace x86" was coming from--but even inside Intel many people knew that was unlikely. Large companies easily can do two things at once--try something, but have a backup plan in case it doesn't work.

But IA64 as an architecture is a huge mess. It became full of every performance idea anyone ever had. This just meant there was a lot of complexity to get right, and many of the first implementations made poor implementation choices. It was a bad time for a new architecture--designed for performance, IA64 missed out on the power wall about to hit the industry hard. It also bet too heavily on compiler technology, which again all the engineers knew would be a problem. But see the above non-technical reasons--IA64 was going to happen, performance features had to be put in to make it crush the competition and be successful. The powerpoint presentations looked impressive. It didn't work out--performance features ended up lowering the clock speed and delaying the projects, and hurting overall performance.

Good... Let it die. (0)

Anonymous Coward | about a year ago | (#45346725)

The (sh)Itanium was a boondoggle for Intel, HP, and others- more that it's execution was so flawed from the get-go than anything else.

Time to let it die and pour effort into something with better legs. MIPS and ARM come immediately to mind as good front-runners if you're going the "ditching X86" route. X86 probably ought to go the same route. While it's "fast" it's because all the effort that could've been put into real performance enhacnements went into kludging the beast. If you put the same effort into either of the currently extant usable performance RISCs, you'd get even higher levels of performance with less power consumption (Though more than with the current MIPS and ARM answers...).

NonStop (1)

kbahey (102895) | about a year ago | (#45346797)

For those who don't remember.

NonStop used to be Tandem, whic was acquired by Compaq, which got acquired by HP.

Tandem had proprietary hardware, proprietary operating system, and even proprietary languages. It was big in high availability stuff, like bank networks running ATMs, ...etc.

Load More Comments
Slashdot Login

Need an Account?

Forgot your password?