The legal spat between Itanium system maker HP and software and now systems maker Oracle has heated back up, with Oracle filing papers in California's Superior Court alleging that HP was paying Intel to keep the Itanium processor alive. This is sure to surprise readers of El Reg. Or, rather, not. The only way HP would ever …
Sounds like Oracle's trying to pick a fight with Intel too, dumb move
HP vs Oracle aside (and who cares really about that besides DB vendors, Oracle's locked into proprietary shit and has been irrelevant since the Sun acquisition, much less the tit for tat Mark Hurd saga) it sounds like Oracle is trying their level best to try and drag Intel into a fight that they probably really don't want Intel in. They can win against HP, HP and Intel vs Oracle is a fight they more than likely cannot win and if they do, they'd better hope ALL of their customers want Sparc architecture. Nothing says Intel has to sell to them, nor license the Intel Architecture to them. It may cost Intel some market cap, but as long as Apple and Windows vendors default x86 and x86-64 which for the forseeable future they will, Intel does not need Oracle one bit.
Intel may not be as big of a beast as they used to be, but again, I seriously doubt they want Intel involved. All Intel has to do is claim that the Itanic is part of a strategic roadmap which HP has no input on, which may or not be the case (all they have to do is show some email between managers at Rio Rancho and Santa Clara, the burden of proof is on Oracle for making stupid claims) and Oracle's point is moot.
As long as it smacks that self-assured yacht piloting smirk off Larry's face, Im happy.
Shareholders want profits and they do elect the board. So would Intel turn away a customer just to make a point and lose money in the process. Sun can easily just go over to AMD and that is something that Intel doesn't want.
Intel couldn't just lie and state that Intanic is part of a strategic move. Shareholders could sue if that turned out to be false and shareholder was lost due to that false announcement.
I have been seeing a bit of a pattern
emerge on El Reg lately.
Whenever anybody comments negatively regarding Oracle they receive a whole bunch of down votes.
Methinks there are a couple of shills on the case because surely there cant be *that* many Ellison fanboys around?
God, I hope not.
Not necessarily - I downvote people who are simply misinformed, or deliberately spouting shit. See Lance's post for the reasons why he got downvoted.
@amanfromMars 2 & @Goat Jam
For enterprise/datacenter products, up/downvotes are much lower than consumer products. This AC got 6 downvotes - thats quite high (even though it was a somewhat dumb post), seen on iphone/android kind of discussions, where everyone has a strong opinion on the smallest of issues. (Granted a pure iFan can easily draw downvotes in 3 digits... but 6 downvotes is still quite high here).
And Goat Jam got 2 downvotes merely for posting a question.... which sounds like s/he might onto something!! So I am tending towards Goat Jam's point here; though probably we need to watch some more before drawing absolute conclusions.
Then shirley you would be better off posting an informative response rather than churlishly hammering the down vote button?
I use the up vote button a bit but I rarely use the downvote button.
Nothing says "I'm a fanboy" more than downvoting someone without having the guts to confront their view to their face (so to speak)
You quit making touchpads which were decent (but overpriced initially) but continue to subsidize the money losing Itanium which are pure garbage. No wonder HP has a CEO of the month club.
What's so terrible about Itanium? It's faster than one of its two major competitors and continues to sell.
Actually I beg to differ
Itanic _WAS_ garbage initially and I was equally vocal to you and many others saying so.
It is not garbage any more. Its IO bandwidth exceeds anything AMD or Intel's Xeon lineup can muster by 2+ times and it has very good raw compute power. The compilers have also picked up and are now generating code that can use the CPU properly.
There is a whole raft of tasks out there which are presently done on extortionally overpriced specialized processing units and ASICs which can be done on Itanic at the same perf and fraction of the price. There are also people who are looking at using it for this so it is not just HP.
@asdf; That's a little harsh...
"the money losing Itanium which are pure garbage."
Well, it depends on your point of view. They do boot and run, and as Turing taught us that is good enough to do anything. And I'm sure there are some workloads that are well suited to their deeply unfashionable instruction set. The philosophy that instruction pipelining and parallel execution should be under the control of the compiler instead of the processor is fine. It means that you need fewer transistors on a chip to achieve a set level of performance *provided* you get the compiler right. That is the part Intel got wrong, at least initially.
It doesn't bode well for Intel. So far their only attempt at replacing the creaking x86 has been unsuccessful. Now that the world has discovered the benefits of low power the need for Intel to get rid of x86 is paramount, but they almost dare not try. X86 is really going to struggle to be as efficient as ARM, and Intel need to change ISA to properly compete. There's an even chance they will be forced to change to ARM...
I know this is silly question, but isn't IA64 architecture more sane and (potentially) more efficient than IA32?
If so it should be possible to manufacture "consumer" version of IA64 with smaller cache and fewer cores (and smaller power envelope). Well this is big "if" and presumably not before Paulson is out of the door. I'd like see Ellison's face then ;)
"HP was paying Intel to keep the Itanium processor alive" Shock horror. Company A is paying Company B to supply it with the goods it makes...
In other news, it is discovered that many PC manufacturers are paying Intel to make Pentium processors...
HP is a customer of Intel. If HP agrees to buy a service (or product) from Intel for the next X years, then that is a normal commercial agreement. If HP agree Intel can continue to sell said product to other people, so be it. I'm fairly sure there are MANY custom parts in Itanium servers, as well as in SPARC servers. Maybe not the core processor, but what about the mobo, the chipset, firmware, etc.
However, if Oracle signed an agreement to supply and support software on the platform, then they should do that. If they didn't sign, then HP might be struggling. They could always talk to Microsoft to make sure SQLServer is up to the job for Itanium. Perhaps even pay MS to develop it, in the same way they've contracted Intel to supply the chip.
"They could always talk to Microsoft to make sure SQLServer is up to the job for Itanium."
Somehow, I think that Windows isn't the primary platform we're talking about here. We've purchased a number of Itanium boxes running VMS. They can run UNIX as well. Oracle has databases for both OS's. How much effort would it take to port SQL Server to VMS? Probably more than MS would want to exert.
"talk to Microsoft to make sure SQLServer is up to the job for Itanium. "
SQL Server? That runs on Windows, right? Does Windows run on Itanium any more? I don't think so, not after Server 2008 anyway, and why would it?
IA64 is only of interest when the workload requires ultra-massive memory and/or ultra-massive single system image SMP (and maybe ultra-massive IO) above and beyond what AMD64 can do, or when the application suite requires HP-UX, NSK, or VMS which (as I write this) are targeted at IA64 systems (and some legacy boxes). Anything else has a perfectly adequate choice of other non-IA64 platforms (hardware and software) to run on.
How long before it goes the way of the iAPX432, the i860, the i960, and other non-x86 Intel ventures?
the end [of IA64] is nigh?
The arrival of x86-based Superdomes (big boxes for lots of blades, with shared memory on the Quickpath) has been announced.