96 Processors Under Your Desktop

Roland Piquepaille writes "A small Santa Clara-based company, Orion Multisystems, today unveils a new concept in computing, 'cluster workstations.' In October, you'll be able to choose between a 12-processor unit for less than $10,000 or a 96-processor system for less than $100,000. These new systems are powered by Efficeon processor from Transmeta and are running Fedora Linux version 2.6.6. Apparently, this new company has friends in the industry. You already can read articles in CNET News.com ("A renaissance for the workstation?"), the New York Times ("A PC That Packs Real Power, and All Just for Me," free registration, permanent link) and the Wall Street Journal ("Orion Sees Gold in Moribund Workstations," paid registration). The company is targeting engineers, life scientists and movie animators. It's too early to know if the company can be successful, but I would certainly have to get one of these systems under my desk. In this overview, I've picked the essential details from the three stories mentioned above."

Comments

Cooling?

[justinmc]

Any ideas?

Re:Cooling?

[pedestrian crossing]

That was my first thought. I guess that's the point of going with Transmeta...

Re:Cooling?

[evil_one666]

the transmeta chips used are specially developed to run at a low temperature. It is in fact this development alone which has enabled these "mini clusters" to now be manufactured

Re:Cooling?

[AKAImBatman]

Cooling?
Any ideas?

I've got one!

96 processors on the motherboard,
96 processors on the motherboard,
take one down,
pass it around,
95 processors on the motherboard...

Re:Cooling?

[Jeremy Erwin]

The 12 node system has a peak power consumption of 220 watts. Not a lot of heat, in the scheme of things.

Dual 2.5GHZ

[Ziwcam]

Great... by october my brand-new machine will be hopelessly out-of-date. I knew it would happen, but had no idea they'd usurp me by 94 processors.

Friends in the industry

[slutdot]

Apparently, this new company has friends in the industry.

Apparently Slashdot is one of them

Re:Friends in the industry

[Duncan3]

How much does one of these obvious ad posts cost anyway?

Please send pricing.

yeah but

[MalaclypseTheYounger]

Can I run Doom 3 on it in maximum resolution mode?

Re:yeah but

[WormholeFiend]

Can I run Doom 3 on it in maximum resolution mode?

probably, but the main character is still doesnt have enough processing power out-of-the-box to use both a gun and the flashlight simultaneously...

Re:For a moment...

[Anonymous Coward]

64? who can do with ONLY 64 processors?? sheesh..

128? who the fuck needs 128 processors? you have to think in resonable proportions! sheeeeeesh..

Re:For a moment...

[Anonymous Coward]

Seriously, why 96? Why not 64 or 128?

Because 96 reversed is...

64 and 128 reversed don't have the same flavor

Sounds nice, but

[grunt107]

One thing perplexes me:
Chips on the same board communicate using Gigabit Ethernet, while board-to-board communication takes place on 10 Gigabit Ethernet.

Wouldn't same board communication be more frequent, hence needing the faster connection?
Better yet, why not 10GBe for both?

Re:Sounds nice, but

[Vo0k]

More frequent yes. But there are more parts within one board, so each of them separately needs less bandwidth than all of them taken together. So, 1G carefully engineered/switched (so each part has 1G bandwidth, not 1G shared between all) is quite sufficient. But then, say, 100 parts need 1G bandwidth between each other and 100M bandwidth to the other board, each. Makes 10G of throughput between boards easily.

Re:Sounds nice, but

[Dutch_Cap]

"Wouldn't same board communication be more frequent, hence needing the faster connection?"

I guess it depends how you look at things. On the same board you have one processor talking to one other processor. Between boards, however, you have up to twelve processors talking to up to twelve other processors. So to me it makes sense to me to have more bandwidth between boards than internally on a board.

Colin Hunter...

[mantera]

I really admire this guy; although the ventures he took part in haven't gone anywhere financially, they were pretty cool. Transmeta, OQO, and now this! Go Colin Hunter!

Re:Colin Hunter...

[Proc6]

Yea it's a pretty ingenious business philosophy.

Start a company that sells CPUs. When profits are failing, start 2 more companies that can be big customers for the first company. When all 3 fail he can start another company that is built on Transmeta clusters and OQO handhelds.

It's like floating checks around banks, but with venture capital.

Beowolf Cluster

[Eclypser]

This time we don't have to imagine what a cluster would be like. It's already in the box!

Less than $10,000

[Cryogenes]

That'll be $9999, please.

Re:strange

[isaac]

why not get a huge server where more users can benefit from the processing powers?

You could use these systems as such servers. The idea, though, is that these might be cheap enough to allocate to individuals.

and what kind of videocard does this baby pack?

No video card. These are just render/compute clusters in a box.

I'm impressed at the claimed 220W peak power consumption of the 12-node box, but wonder what kind of real computing performance it provides.

-Isaac

Fedora 2.6.6?

[Halo-]

I hate being this guy, but this is a big pet peeve.

Fedora currently is either Core 1 or Core 2. 2.6.6 is a kernel version number.

Kernel version != Distribution

Saying "Fedora 2.6.6" is like saying a car is a "Ford 2.4 liter".

Re:Fedora 2.6.6?

[Zocalo]

This is a problem at the Slashdot end rather than with Orion - either from the original submitter or the editos. Checking out the product descriptions on Orion Multisystems' site reveals the following (and other interesting specs):

• Based on Fedora Core 2
• Linux kernel 2.6.6 with performance optimized Orion drivers

So I think they know the difference at least... :)

Whee!

[thephotoman]

96 processors running Fedora? I want one!

Actually, I would be willing to bet that the university I'm at could use a few of these things. After all, we've got undergrads doing BLAST database work, just to teach them about it. Having been through that hell myself, it'd be a lot easier if you didn't have to have a cluster to do the work by computer. For those who don't know, BLAST is a genetic sequencing database that allows for comparison with an extracted gene (retrived through polymerase chain reaction) with a known, sequenced gene in their database.

Re:Whee!

[rice0067]

BLAST doesnt take that long anymore.. well at least not for some things.
We use it all the time to compare our DNA products to all known Gnomes. It takes like 30 seconds.
(300 BP search against the whole library takes less than 40 seconds. using http://www.ncbi.nlm.nih.gov/BLAST/)
Maybe its much longer for other things ?

Seems Very steep

[Crashmarik]

The pricing seems quite steep. 800/cpu for 12 configuration, 1000+/cpu for the 96 cpu configuration. I can see why they have friends in the industry the prospect of selling 10 to 100 times the equipment per seat must have marketing departments salivating.

If your'e going to spend that kind of money though theres alllready solutions that will provide that level of processing cheaper.

There is also the utilization isssue, programming tasks hardly require 96 processors except on compile and link. You don't need 96 processors to wait for a keystroke. The same holds true in applications. You don't need cpu's waiting for a user to decide what to do.

Re:Seems Very steep

[timeOday]

I'm sure you're right, these aren't going to sell in huge numbers. As for the utilization issue, though, I don't think it's for programmers. Like the summary says, think animators - nothing is easier to paralellize than rendering frames in an animation.

Personally, I think most of these will still end up as servers for groups of people instead of individual "workstations." But the logistics of a normal 100-workstation cluster are pretty bad - a large server room, enormous air conditioning unit, a massive power supply, and lots of cabling to be done. This new thing can probably share an existing server room with other computers.

Granted, it's probably just a bit smaller and more power efficient than previous "blade" servers, but maybe presenting it as something brand new is a good marketing angle.

Re:For a moment...

[dreamt]

I woiuld guess it is more a physicical issue. Processors are probalby arranged 3x4, and can fit that way into a 2 foot wide case to fit onto an "average" desk. Stacking these boards 8 high gets to a height that can fit under a desk.

Re:For a moment...

[skaffen42]

Could be cost. 10K and 100K are nice round numbers.

I've realized that most strange tech descisions can usually be traced to some guy in sales...

Isn't there a Jay-Z song about this?

[schtum]

If you're having lag problems, I feel bad for you son.
I got 96 processors and you got one!"

a positive Slashdot effect?

[cybergrue]

OK, apparently, someone took the "Imagine a beowulf cluster of these" a bit too seriously.
Way to go!

gentoo

[LousyPhreak]

now thats a system i'd like to install gentoo on :)

96? That's it?

[keiferb]

I've got well over a hundred in the box under my desk. Unfortunately, it's just that. A box of over 100 CPUs, mostly Pentiums/Pentium IIs.

Re:AMD instead of Transmeta?

[QuickFox]

It'd be cool to have 12 high-end AMD processors [...] heat will melt computer case

First you say it'd be cool, then it'd be hot ... Make up your mind!

Re:For a moment...

[Peter La Casse]

Seriously, why 96? Why not 64 or 128?

12 processors fit on one board, and 8 boards fit into the chassis they chose.

Re:Mass storage I/O?

[Zocalo]

To save you having to RTFA and check out the product descriptions, the say "1 to 96 high performance 2.5" disk drives, 30-80GB capacity, 7.8TB max capacity on deskside". You can probably infer from the 2.5" that they are using notebook harddrives which are most likely EIDE - at least I haven't seen any SATA ones yet, although they can't be too far away.

You should probably check out the product description anyway though; there are some quite interesting hardware design decisions in there!

Piquepaille Slashdot spam must stop !

[wsapplegate]

Warning ! Warning ! Warning ! Warning ! Warning !

Attention, a public service announcement follows : do not read the "overview" touted by Mr. Piquepaille. This person constantly spams Slashdot, trying to get traffic to his site on Radio Userland (which I'm not linking to, for obvious reasons). Do NOT go to his overview, you're only giving traffic to a spammer. See these recurring complaints, for instance. Not to mention he steals the images he puts on his blog and sometimes also spews bullshit for lack of knowing better. This must stop. In any way, do not fall for the spam, and do not provide him any more traffic. Please also warn fellow readers when you see one of his self-serving posts.

And now, a personal message (warning : verbal abuse in foreign language follows) : Roland, tu nous les brises. Va te pendre, hé Ducon !

[disclaimer : I'm not commenting on whether the subject is interesting, or not. But the kind of astroturfing the submitter engages in regularly is just wrong]

Warning ! Warning ! Warning ! Warning ! Warning !

Piquepaille Ripping Off Again

[Anonymous Coward]

How many stories a week does this frigging guy get into Slashdot?

His business plan:
1) Sell Ad Space on "News" Website
2) Shovel In Content From Online Articles
3) Submit To Slashdot Daily
4) Tout "Slashdot Coverage" To Advertises
5) Profit!

And looking at his site, it works fine and dandy indeed.

Then again, is he just doing a service to us?

I agree completely. HEMOS, listen to me...

[Ayanami Rei]

Can the editors PLEASE STOP POSTING PRESS RELEASE COPY?!?!

If you're going to post a story announcing a product or discovery, at least link to a weblog or site that actually has a little commentary on the subject, or the original site itself.

Roland "Fuckyfacey" Piquepaille is neither of these.

Thanks.

Re:AMD instead of Transmeta?

[Frohboy]

You can get an 8 proc Opteron cluster for about $10k from Rocketcalc. I can't speak for their new systems, but I am currently running some simulations on one of their older 8 node P3 1Ghz clusters. It absolutely blows my desktop P4 3 Ghz out of the water (even with some rather poor parallelism in my code.)