Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
AIX6OSX ?

just a speculation...

The POSIX issues won't go away, and IBM has already showed a willingness to modify its software marketing model.

What would really be neato-keen is if (as they did with AIX5L..the
linux version of AIX)..is if IBM began to allow experimentation in UNIX flavors of other types of OS's..including Inferno (Plan9 derivative) and others..

Think about it.

We could have a 256 base OS for multiple processors ..run on AIX master code..with a.i. sub systems OS's like z/VX..or even an OPEN CYC 64 bit variant.

OSX would then be the preferred media os of choice for the scientific GUI..or bio-medical interface. PowerPC is already the preferred processor in the embedded area over custom DSP applications because of its flexiibility in floating point apps.

Most genuine Artificial Intelligence apps (a.i.) need serious fp transfer data rates/ bus speeds for neural net and compression requirements.

Sun Microsystems sure has been real quiet since Steve announced the G5. Maybe McNeally is in Canberra with a 64 processor system already.

<---needs a blonde space maiden right now..
 
for what it's worth...

anyone notice that apple is selling xservers as cluster nodes? Diskless and everything, just optimized for computing. I think Apple works on a good clutering mechanism rather than building real big machines.
It would be easier to target the existing market with that as well as scalability for CPU intensive tasks is given - and you can grow as you need it.
 
The market is obvious and everybody is missing it

Apple wins big when it creates markets. It won big when it introduced the legacy free iMac, the iPod revolutionized portable music in its form factor, and the iTMS revolutionized downloadable music for pay.

A larger server that runs anything from $12k-$500k would be perfect for a home server in a McMansion. If you're spending $5M for a house, $100k seems reasonable for a wireless system that will control your home, give you great terminal services via native Mac OS or virtualized Windows sessions and, especially in Windows settings, being able to throw away and start over when you get virused, trojaned, etc. is a great feature that will sell such systems to the upper middle class and wealthy.

The partnership would be with home builders who would install these monsters in at construction and wire up access points throughout the house and yard (perhaps in parallel with the sprinkler lines?) and the cost gets rolled into the mortgage.

It would be an ultimate reaffirmation of Apple as a high-class status symbol and Apple's scarcity of virus software would ultimately push it ahead of its Windows competitors. Linux would also sag here as the appeal of tweaking your house would be rather limited (and accomplished just as easily on PPC Linux as on x86 variants).
 
Re: The market is obvious and everybody is missing it

Originally posted by dbrutus
Apple wins big when it creates markets. It won big when it introduced the legacy free iMac, the iPod revolutionized portable music in its form factor, and the iTMS revolutionized downloadable music for pay.

A larger server that runs anything from $12k-$500k would be perfect for a home server in a McMansion. If you're spending $5M for a house, $100k seems reasonable for a wireless system that will control your home, give you great terminal services via native Mac OS or virtualized Windows sessions and, especially in Windows settings, being able to throw away and start over when you get virused, trojaned, etc. is a great feature that will sell such systems to the upper middle class and wealthy.

The partnership would be with home builders who would install these monsters in at construction and wire up access points throughout the house and yard (perhaps in parallel with the sprinkler lines?) and the cost gets rolled into the mortgage.

It would be an ultimate reaffirmation of Apple as a high-class status symbol and Apple's scarcity of virus software would ultimately push it ahead of its Windows competitors. Linux would also sag here as the appeal of tweaking your house would be rather limited (and accomplished just as easily on PPC Linux as on x86 variants).
No.
An xServe could run the NAT, handle the web site, personal email, and all the automation systems for a dozen of "McMansions".
You don't buy 4, 8, or 64 cpu computer systems to run your house. You buy them to run the customer database at BankOne, or to model the Big Bang (though a cluster would be more cost effective at the latter)
This is not the market that apple is going to try and create.
Anyway, I've dealt with enough well off people to know that many are extremely cheap (unless it's something that is really status). I don't see a big computer hidden in the bowels of your house being a status symbol unless you are Bill G. and all your friends are in the tech industry.
 
Creating Markets

I agree with the previous post.

Apple does have cluster boxes. Apple does have XRaid. Apple does have XServe, but it is *still* nowhere near enough for my bosses to consider implimenting Apple technology here.

Apple is just starting in the right direction, and still has a long way to go. The market Apple needs to go after is not just the Windows Professional Server market, but low end UNIX environments as well.

For a company our size, we have a lot of high tech equipment and a lot of servers. Just standing in the machine room is amazing. However, this company is still small, all the employees fit on a web page with pictures, and our machine room has plenty of space for expansion. We are not an ideal customer for Sun, just another semi-small business, though a really profitable one.

So we have top notch equipment.

Yet the owners love macs. Why are we using Sun then?

Macs can't do what we need yet.

I want apple to deliver some outstanding products for our market, which is largely seen as the leftover scrap of the larger enterprise market. We're too big to qualify for small business, but too small to be considered enterprise. But we need enterprise quality and enterprise level equipment. Not workgroup servers.

We'd have to fill too many racks with XRaids and XServes to equal the computing power and storage space we have with these enterprise products. Sure the XServe costs less, but it doesn't fit our needs.

What we need is Enterprise Server Solutions - For the rest of us.

Full blades, hd raids, and they have to be small in size yet big on storage and processing power. Our facilities are state of the art, and will only get moreso.

We also need an operating system that can do what we are currently doing. Right now, we'd have to rewrite *everything* from scratch, and because of the darwin underpinnings, there are many things we may not be able to do at all. The OS Server needs to become more robust and mature as a UNIX server, because we wouldn't even use the GUI.

Here's hoping Apple gets the clue.

Jaedreth
 
Re: Creating Markets

Originally posted by jaedreth
We also need an operating system that can do what we are currently doing. Right now, we'd have to rewrite *everything* from scratch, and because of the darwin underpinnings, there are many things we may not be able to do at all. The OS Server needs to become more robust and mature as a UNIX server, because we wouldn't even use the GUI.

Here's hoping Apple gets the clue.

Jaedreth
What exactly can't you do with the Darwin underpinnings?
Darwin is based off of FreeBSD. I MUCH prefer FreeBSD to Linux... and that preference goes back to before Apple announced Darwin.

Netcraft reported that FreeBSD is running 2 Million web sites. It's a full on BSD Unix.. just free.

Here's a few companies that rely on FreeBSD to host web sites...
  • Yahoo 159,354 active sites hosted
  • NTT/Verio 152,054 active sites hosted
  • Infospace Inc 129,378 active sites hosted
  • Datasync 100,103 active sites hosted
  • Pair.com 72,626 active sites hosted
  • Tierranet 41,811 active sites hosted
  • Global Internet 39,365 active sites hosted
  • Telus 38,525 active sites hosted
You might want to see which servers post the longest up time too...
 
Let them know

jaedreth,

You know that Apple is going b@lls to the wall working on Panther and the G5 server. They want your company's business, as well as other companies like yours - especially the profitable ones!

Let them know that you (and other employees) are a Mac lover, want to see their servers used where you work and what you feel is needed to get them ordered. Like the "Bug" icon on the beta Safari, information from informed Mac user can help moving towards a configuration your boss will sign off on. Maybe not in the next 3 - 4 months, but in the reasonably near future.
 
Blades

I think IBM and Apple should do a cross licensing deal where Apple would sell IBM's blades with OS X and IBM would sell blades and high end severs with OS X! Think Power 5 running OS X... and Photoshop😀
 
Feedback to Apple

Yes, I do agree feedback to Apple is important.

And I do, through my own channels.

😉

I can't give a detailed description of why Apple doesn't meet our needs, but I've seen what we're running, and Apple doesn't have anything anywhere near it.

I will of course inquire further at work, as I know my bosses would be thrilled to be able to have an Apple solution there, as I set up a brand new tower for one of the head haunchos the other day. 23 HD Display. 🙂

It's good being the most experienced mac guy there. 😀

Jaedreth
 
Carpideum

Heya, the nick looks a little familiar for some reason. Perhaps it's nothing...

However IBM has no interest for OS X on its servers.

Right now, it's fighting a deathmatch against Sun and HP. Solaris and HP/UX are the only serious competition for AIX.

And it is serious competition, and that market is waning... I only hope IBM comes out on top, but right now, it doesn't look like it.

IBM is selling lower end servers and blades based on Intel chips (stupid idea imo) yet they're running Linux.

Those boxes they might be persuaded to ship Darwin with, if Darwin without the OS X GUI could do everything that OS X server can do with the gui. But it's different enough, very few people have experience modifying, customizing, and making Darwin do what they want (unless they are mac and unix heads).

I do hope IBM does teach Apple how to make proper blades, and I think the "G5" chip would be great on these machines currently sporting Intel chips.

A Darwin Server product, sold inexpensively on cd, downloadable for free for those with broadband, would be ideal, providing it only runs on Macintosh G4 or G5. Said product would come with no support, but tons of man pages and an actual instructional manual thick enough to be of use.

If they cleaned up the underpinnings, and made it so there is not one thing that can be done in the gui and not done in darwin server, then the gui would be more efficient due to it simply referring to the underlying programs. They could also sell cheap server boxes, again no support, Apple does not want to get into the business of supporting a unix os in the command line, but then a cheap alternative to a top notch G5 or Xserve with a pure unix environment could sway a lot of customers, and cost Apple less money while making more in the long run.

Just an idea...

Jaedreth
 
Re: for what it's worth...

Originally posted by visor
anyone notice that apple is selling xservers as cluster nodes? Diskless and everything, just optimized for computing. I think Apple works on a good clutering mechanism rather than building real big machines.
It would be easier to target the existing market with that as well as scalability for CPU intensive tasks is given - and you can grow as you need it.

THANK YOU! THANK YOU! THANK YOU! I have said the same thing many times for a while.

This way even joe home user can buy a cluster node when he needs more power for editing his home "movies". Any small business could take advantage of this, i mean this is the way to go. All connected via fiber channel.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.