Slashdot Mirror
Own a Piece of An Apple-Based Supercomputer
Graff writes "Now that Apple has come out with the Xserve G5, Virginia Tech has been swapping out parts of their 'System X' supercomputer for the more compact 1U Xserves. MacMall is selling some of those System X component G5 systems with an approximate $200 savings and an extra 512 megs of RAM over a normal G5. You can read more about it at MacCentral."
I wish there was some kind of engraving on the aluminium casing stating something like "Virginia Tech, Supercomputer node #758 - 2003" Then they could definitively sell it at a Premium. I mean I can get this kind of computer off ebay for more or less the smae price. I need some kind of souvenir that it's from Virginia. How about sending it through the IPOD engraving shop?
Artificial intelligence is no match for natural stupidity
As someone else noted, if they were engraved or etched or something that would make them special.
Why couldn't VT hold their horses? ...)
my bet is... they could.
no one likes building any cluster (not to mention a supercomputer) out of desktops, esp. ones configured like desktops (gfx, no ecc,
but apple really wanted the PR of having the computer cluster, and perhaps to list the revenue in 4Q2003.
so i can't blame them - looks like a fair deal.
I believe that the Apple recommended restore procedure for an XServe involves booking from a copy of the OS installed on your iPod (which connects via FireWire).
You're more likely to back up onto tape.
There's no reason why you can't plug the tape drive in over FireWire. FireWire is basically a serial variant SCSI (okay, I'm oversimplifying a bit here) and with speeds of up to 800Mb/s it's fast enough for most things. You probably wouldn't want to connect your RAID array via FireWire, but for backups it's plenty fast enough. Many tape drives only let you write at Oh, and by the way the G5 units they are selling are intended as workstations not as servers (hence the digital audio out and the Radeon 9600 Pro, neither of which is really required for a server).
I am TheRaven on Soylent News
So they can make money when people forget to send the form on time, or fill it in incorrectly.
Several companies in the UK do the same for extended warranties. They say "pay lots of money and if your machine doesn't break, we'll refund it after five years". You typically get 30 days after the five years to get your money back, and most people will just forget.
Should that be a concern? Do these 6 month old computers already have 2-3 years of typical mileage on some of their components?
This is an educational and research establishment, not a commercial enterprise.
/.) that Virginia Tech would have sold the units to students, freshmen, whatever, at a knock down rate. Or even used a ton of them within the university itself.
You'd have thought (as some students were hinting here at
Yet more profiteering from a supposedly educational institution.
VT isn't the one selling these systems... Apple is. VT is tossing them back at Apple for shiny new Xserves. Apparently being the 3rd fastest computer in the world isn't good enough anymore.
8==8 Bones 8==8
I was a bit confused by the decision to include Firewire in this machine. I know it is an apple kind of penchant, but surely a server won't need firewire.
Actually there are lots of reasons to include firewire on a server.
- You can hang a firewire mass storage device off of it to backup (tape, disk, etc), boot from (recovery, etc), add extra storage in a pinch, etc.
- You can create various types of clusters using firewire. One product is the sancube.
- It's cheaper to design in a feature that may not be used in one incarnation of a product, but may be usable in others. Case in point your comment about Audigy sound cards on high end Intel servers, those very same motherboards are probably used in both servers and high end workstations, no point in having two different motherboards just to save a few pennies off of a $500+ mobo.
I'm going to have to disagree with you concerning the amount of wear and tear.
Most computers fail either in the first month or so of use or after many years of good use. In the first case it's usually a bad component that slipped by quality control. In the latter it is simply the ravages of time. Longer quality control "burn-in" times would eliminate many of those first month failures, but the vendor really doesn't have the time/space for long burn-ins.
Now, the Number One way to shorten the life of your computer is to turn it on and off frequently. The computer heats up when you run it and cools when it's off. The expansion and contraction of components associated with these temperature changes stresses every solder joint on every component -- and may even stress the chip-level components themselves. To lengthen the life of your hardware (at the cost of extra electricity), leave your system on unless you aren't going to be using it for a significant length of time (i.e.: don't power cycle more than once a day).
These G5's have been on for approximately six months straight in a very well-controlled temperature environment. This is a burn-in that virtually guarantees that there were no manufacturing defects. However, since they weren't power-cycling on a regular basis, it was actually a VERY low-stress environment.
Life is short: void the warranty.
You can do some good high speed networking using firewire.
It's available in Mac OS and Linux.
http://www.homenethelp.com/network/firewire.asp
400mbps isn't to be sneezed at. With repeaters it'd probably make a decent fail-over network in case the main gigabit link failed.
Worst
oo..damn.. it was!
Probably the only actual record the big mac can claim is the shortest time to obsolescence. Not to downplay the achievement though...
Well, it's the fastest supercomputer ever built with off-the-shelf components and the number three fastest machine on the planet -- and that's before the upgrade.
They will probably make quite a decent profit out of this, despite the $200 discount. They must have got pretty decent discount from apple for both bulk buying and promotion. And any self respecting geek will want one of these over a stock G5
As someone else pointed out, these are refurbished by Apple and then sold through MacMall. VT isn't selling them, but traded them back in to Apple for credit towards the Xserves.
Life is short: void the warranty.
Why oh why do companies use mail in rebates?
so they can stick low price tags in big numbers on the shelf. That's gotten me a few times - you see, "Oh, an X for only $19.95!" so you take it up to the counter, and the cashier rings you up for $39.95 - often by then the consumer is already psychologically committed and just pays it. It's a common tactic, rebates are just one methode of exploiting consumer naivety. Bottom line is, it generates more sales.
try { do() || do_not(); } catch (JediException err) { yoda(err); }
The G5 was the clear winner out of all the chips on the market, and Apple was the clear winner of the platforms considered, and they considered *ALL* of them worth considering.
The success of the venture simply proves the superiority of keeping an open mind and not bringing tired old pre-conceptions (Apple's slow, Apple sux, etc.) to your work.
Apple's online store is charging $2399 for refurbed dual G5s, and the student store's price for a new dual G5 is $2699, $100 less than the MacMall refurbs from Tech. You can even take $26 more off for getting rid of the internal mode, which the supercomputer refurbs don't have either.
Even if you add the extra 512 megs of RAM from Apple's site (where prices aren't the best), these Va Tech refurbs are only $100 less than what a student would spend on the same box new. Not to mention these 2 GHz duallies are rated as "Buy only if you need it - Approaching the end of a cycle" on the Macrumors buyers' guide page.
So not a deal at all if you're a student (though I have to think students at Va Tech could get the inside track on the boxes -- anyone know?) and not a great price for a refurb if you're Joe Schmoe. And not a box with great longevity, relatively speaking, either, if Macrumors has the lifecycle pegged.
Wait for a processor speed bump unless you're dying to own a little bit of celebrity.
It's all 0s and 1s. Or it's not.
What would be spiffy if there was a way that they could do SOME of the math on the GPUs. I never saw a product that could do that, but it would be rather fast. No?
Never hit your grandmother with a shovel, for it leaves a bad impression on her mind...
They did need them badly. Without having them so soon there wouldn't currently be a Virginia Tech supercomputer on the TOP500 list and they'd have to wait another year to try to get listed.
mbbac
Actually, FireWire supports transfers of up to 3.2Gbps depending on the interconnect used.
mbbac
Virginia Tech wanted Xserves in the first place. Apple didn't reveal them until January. Look at the benefits. Smaller case, less heat generated, less power consumed, error corecting code memory; all things which would be extemely useful in a supercomputer like this. From what I have read, the upgrade will use 25% less power and take up a third less space. If I were Tech, I would throw the current G5s back too.
It dosnt have to be an engraving, although it would look nice on the metal finish. It could be a sticker at the very least or a plaque mounted on the side....with a free t shirt saying " I paid nearly $3K for a G5 and all I got was a lousy supercomputer"
"Persistance is Fertile" - Me. I can quote myself if I want to.
My guess is that both they and Apple wanted to have a spot on the last Top 500 list, with all the associated press at the Supercomputing 2003 conference. Apple's been trying to convince somebody, anybody to build a large HPC cluster with their hardware since the G3 came out. Until the G5 came out, it made very little sense economically -- the per-system price for Apple kit was 30-40% more than comparable Intel-based stuff, and the memory bandwidth and 64-bit floating point performance was the same or worse. The G5 fixed that, for the most part
Nobody in their right mind wants to build a cluster out of machines in desktop/deskside chasses. We've done it once, with the first generation Itanium systems where there was no rackmount option for a 2-way box, and we'll never do it again -- remote management of those machines was and is actively painful. (Our 1st-gen Itanium cluster is out of production service now, but it's been partitioned up into smaller clusters at universities around the state as part of the Cluster Ohio project, which we still manage.)
"My life's work has been to prompt others... and be forgotten." --Cyrano de Bergerac
They used to run all simulations twice to verify the non-ECC RAM was returning the desired result. As a consequence the system will speed up two-fold in real life use. Now that is a performance gain!
I'll admit that a 6-month replacement cycle is pretty short, but it actually makes sense because they're avoiding the worst of the depreciation. I'm not up on used Mac prices, but x86 server hardware depreciates around 50% per year (refurbished 2 year old x86 servers routinely sell for around 20% - 30% of their original price; refurbished 3 year old gear sells for well under 10% of it's original price. You can get a maxed out Quad processor P-III server for well under $5000 which cost $50K when new.
That said, I don't think that this is a good deal. $200 savings on a $3000 box is only a 6.7% discount for 6 month old hardware; a 20% - 25% discount would be more in line with current market.
Why is it that the proponents of "one nation under God" are so eager to get rid of "liberty and justice for all"?
And an original Lisa sold from Apple for $9,999. Hmmm. $1 profit. There's your return on investment.
I'm fairly sure Virginia Tech wanted the 1u cases all along (makes more sense). However, they needed the cluster up in time to make the Top 100 list. Being on that list brings in _lots_ of research money. So yes, they did need them.
Why?
Actually the list is updated every 6 months, but the one in November is the bigger of the two because that is the time for the annual supercomputing conference.
At the risk of the inevitable flaming, I have to aqree with this posting. I'm not questioning Linux or it's proven abilities, but what so many Windows (and Linux) users fail to understand about Apple's business model is that it's -not- the same as the commodity market that they enjoy. Apple makes their products in a "holistic" manner... they tightly control the hardware, operating system and pre-installed applications in order to deliver a seamless experience to the end user.
Bottom line... if you're the kind of person who feels Apple hardware is overpriced because you can't assemble it yourself and run Linux on it, then you're missing the point and Apple probably doesn't want your business anyways. It's not outright stupidity that has kept Apple out low-end market all these years... it's a market they've intentionally chosen not to enter. Case in point.... Steve Jobs killed the whole Macintosh clone market when he returned to power at Apple. UMAX and PowerComputing were offering faster machines for less money, but were totally clueless when it came to delivering any value-add on the end-user experience.
I'm not saying this is in a "snob" context... it's just the reality of their business model. Apple wants to sell a G5 (or iMac or PowerBook) to somebody who -doesn't- want to assemble their machine. The whole point of the Macintosh is in the fact that you don't need to do any of that.
Take heart though... IBM is apparently going to be (or already is) selling 970 (G5) based systems in some form running some flavor of Linux... so if it's only the bad-ass CPU you're after there will be other ways of getting your little flippers on them.
bash-3.00$ uname -a
SunOS panda 5.10 Generic sun4u sparc SUNW,Ultra-2
Well, from what I can see there is barely any difference between the memory controllers on the two systems. It looks like it was just a new revision of the same ASIC. Apple doesn't exactly provide many details on this, but it looks like the new memory/processor controller chip would be a drop-in replacement for the chip used on the original Powermac. Therefore it's possible (even likely) that they will use this new revision on the next revamp of the G5 line. In fact, they could well start slipping them into the current line-up without telling anyone about it.
I don't anticipate that Apple will sell any desktop G5's with ECC memory installed at the factory, but if the memory controller supports ECC you could easily replace the factory memory with third-party ECC memory.
Just called, no need to call.. they're all gone.. shucks.
-Jim
Not every. But most. Solving ODE's usually boils down to iterating a (possibly implicit) linear system. Solving PDE's with finite differences does too. Or with finite elements. Or spectral methods. Lots of statistical computations do too.
Certainly there *are* scientific applications that don't involve multiply-adds, it's just that the vast bulk of scientific computations that are suitable for parallelization really boil down to solving linear systems, some kind of linear iteration, least-squares problems, or some combination. All of which are solved using lots of multiply-adds. So, while linpack isn't the end-all and be-all of hpc benchmarks, i'd say that it's a pretty good guideline; i'd also say that the speed of multiply-adds matters a whole hell of a lot for scientific computing.