Fitting A Linux Box On A PCI Card

An Anonymous Coward writes: "Running on Newsforge/Linux.com is a hardware review where Slashdot's Krow took a couple of OmniCluster's Slotservers and and built a cluster configuation inside of a singe host computer (and even had DB2 running on one of the card's inside of the host). Could something like this be the future of computing where for additional processing power you just kept adding additional computers inside of a host?"

Comments

!!!!

[Unanonymous Coward]

:)

I guess

[Anonymous Coward]

not!

Frist p0st

[Yebyen]

I H4x0r3d J00!

Ob Beowulf comment

[anticypher]

Imagine...

It would be cool to have completely separate processors in a box, so that as long as there is power, each card can run on its own. Then you could network them together into a beowulf cluster, and then make clusters of clusters

the AC

Re:Ob Beowulf comment

[Khalid]

There has been a project a while ago, which aim was to implement a beowulf of separate StrongArm cards to be plugged in a box, they have even managed to build some prototypes.

http://www.dnaco.net/~kragen/sa-beowulf/

Alas I think the project seems to be dead for some time now.

Re:Ob Beowulf comment

[Anonymous Coward]

Seperate cards that have their own CPU? Well, start by getting a VME or XVI chasis, get CPU cards, write your own code to load the OS on each card, and presto! Beowolf in a box.

Re:Ob Beowulf comment

[BlueMonk]

Can this be extrapolated to a point where computers start to look like they were modeled off of brains? Very simple processors combining their efforts into a larger group effort, but retaining individuality from other similar groups. These groups are further combined (based on task/area of functionality) into larger groups which merge again into a simplified single I/O pipe for the whole super-group. Groups could be arbitrarily nested to any number of levels. Each level would be responsible for multiplexing I/O between the single I/O pipe to the higher level and the multiple sub-groups at the current level. At the top you end up with huge groups operating somewhat similar to a left brain and a right-brain with a fat I/O pipe between them acting like a corpus colossum (sp?). Talk about a huge bus width. What would that be -- a few hundred gigabit bus :)

Just some pie in the sky amusement and speculation.

Hmmm

[Anonymous Coward]

Imagine a Beo... oh. Hang on.

Damnit

[Yebyen]

Fr00th P0s7. :(((((((((

*lowercaseforslashdot lamenessfilter*

Thats like having a shit filter on your ass. WTF!

Too bad...

[NoMoreNicksLeft]

That slots are considered a bad thing nowdays. The trend is to manufacture boards with less expandability, not more. So let's see... Soundblaster 1024 Ultra, or another CPU board... but not both. Then again, I've never been accused of buying crappy consumer motherboards...

Re:Too bad...

[Anonymous Coward]

I purpose that an arrogant filter should be developed to work in conjuction with the lameness filter, because there is truly nothing more lame than being an arrogant computer nerd.

Re:Too bad...

[guroove]

that your DNS doesn't work for your own site when you're offering DNS to others. Learn how to set up BIND. It should take 5 minutes.

Re:Too bad...

[NoMoreNicksLeft]

Well, this comes as a shock, you know. Those 27 friends, fellow experimenters and beta testers must have been lying to me...

I've learned how to set up bind. Maybe you didn't bother to do a proper host command? This isn't an ICANN sanctioned TLD, you know. So until you're willing to cough up the $50,000 application fee, that's not a realistic possibility... why don't you try my homepage instead? http://24.30.242.100:8000 Until you set up your _OWN_ bind properly, it isn't going to work for you.

Seen these for a long time

[ackthpt]

I've seen these around for ages, variety of manufacturers, but usually they're priced significantly higher than just buying several cheap PC's, granted you have a fast bus between cards/PC's, unless you have a redundant powersupply, one failure brings your whole cluster down, whereas networked mobos should be tolerant of one system failing. As for future, eh, they've been around long enough, but I expect the use has been rather specialized.

Re:Seen these for a long time

[pruneau]

Well as for the single-point-of-failure for the host computer, you _are_ right.

Another problem of course is the PCI bus speed, as someone already mentionned : if you are using some !gb/s link between the machines, that will allow you to deliver data much faster.

But... wait ! if that's going through a PCI bus anyway... Hey, can some hardware people invent a _true_ bus, because we _are_ lacking something there.

But that kind of solution might interest people wants to do more with less space... If they are ready to pay the price.

All in all I'm not sure it's that interesting. Do someone have some benchmarking about that ???

Re:Seen these for a long time

[Anonymous Coward]

No different than a room full of machines where the main power gets lost than a box full of cards running off of one power supply, so you have two (or more) power supplies running, where one can take the sudden increase of load if the other dies.

But...They do that already in higher-end servers.

There will always be "single points of failure". So make sure that some of them have redundancy or give you a soft landing.

Re:Seen these for a long time

[Anonymous Coward]

And another...

So how do you do things like disk space and network access? Somewhere you're gonna have a lot of redundant stuff. Think IBM mainframes with their virtual machines and virtual hardware.

At what point do you blur the line between your box of PC-on-a-card w/ virtual network hardware and virtual disks and an IBM mainframe with one cpu core, but with virtual machines (with virtual network, virtual disk, virtual printing, etc.)?
The advantage with the IBM mainframe is that it's all internal, and you could probably assume after 20 or so years of doing it that it's pretty well optimized, and not dependent on a non-existant multi-GB/s bandwidth low-latency internal networking scheme.

Besides, there are still lots of rack mount systems w/ passive backplanes that work this way (isn't this VMEBus and S-Bus hardware?).

It's an old idea that has fallen out of favor, but seems to make a certain amount of sense.

Re:Seen these for a long time

[ironfroggy]

I've seen cards that do this for sale in at a local computer shop these old people run.

Get lost!

[Unanonymous Coward]

This is news for nerds, If you want trolly news like the stuff your posting then go some where else!

Re:Let me get this straight...

[MrPerfekt]

If we do not carry on, they will have won.

Performance

[Anonymous Coward]

I wonder what sort of performance you get out of one of these cards?
The processor is a:

x86 Compatible National Semicondutor pxGeode

Re:Performance

[kzanol]

I wonder what sort of performance you get out of one of these cards?
Since I've got a router/firewall box using the same CPU I should be able to answer that one (300Mhz):
RC5: Summary: 4 packets (25.00 stats units)
1.14:10:07.62 - [48,778 keys/s]
As you can see, number crunching is rigtht out of the question. Still, easily fast enough to push packets around.

The SETI version

[Wire Tap]

Does anyone here remember a while back when that "fake" company tried to sell us SETI @ Home PCI cards? I was about to place my order, until the word came to me that they were a fraud. Kind of a funny joke at the time, though. At any rate, here is the old /. story on it:

http://slashdot.org/article.pl?sid=00/07/23/2158 22 6&mode=thread

It would have been GREAT to have an improvement in CPU speed on a PCI card, as I always have at least two free in every system I own. What I wonder, though, is what instructional speed would the PCI card "CPUs" give us?

Re:Let me get this straight...

[fshalor]

1. We're NOT in a war against Islam.
2. If we think we're in a war against Islam, then we have lost.
3. We are in a war against Microsoft and almost all that she stands for.
4. This would be a major blow in the final battle of that war.
5. Any "linux" based application/hardware that takes even a miniscule share in any market is worth the time spent to develop it, as long as it's intentions are pure.
6. Without inventions like these (when can I get mine?) we regress into nothing.
7. (and last one.) The souls of the victims of every war and conflict in history would like you to get your information straight. And would like us to get on with our lives.

No one can tell anyone else who they don't know how to get on with our lives. I'm not telling you to get over it, am I?

And to struggle for that on-topic (1), think about the heat issue, folks. These cards have got to be smokkin'

-=fshalor

Impractical

[atrowe]

I don't see these things taking off for most uses because the PCI bus is limited to a measly 133 MB/S. Even newer 64 bit PCI slots found in some servers have insignifigant bandwidth to keep the data flowing fast enough to make full use of these things. I can see where they may come in handy for heavy number crunching applications such as SETI, but for web serving and DB applications, the throughput between the card and the host system is simply unacceptable.

Also, I would imagine that the RF interference generated by having several of these in one box would be quite signifigant. PCI slots are only an inch or so apart on most motherboards, and without any sort of RF shielding between multiple cards, I can't imagine they'd function properly. It's a good idea on paper, but in reality, I'd think a few 1U rackmount servers would do the job much better. And for $499 a piece, you could get a decent single processor rackmount server for around the same price.

Re:Impractical

[man_ls]

The way I understood it, the host system's motherboard was just a backplane supplying power to the computer, which was contained on the PCI card. IIRC, several years back, when Pentium IIs came out, lots of people wanted a way to upgrade their Pentium-I based systems. The easy answer was - make the motherboard into a holder for a very compact computer. It had, I think, a 333 Celeron, an SODIMM slot for memory, a single IDE channel and floppy controller, and onboard sound and video. Not too impressive, but the entire workings of the computer onto a single PCI card.

Sun or SGI also has something like this, to allow SparcStation users to run Windows applications natively. Basically, a card with a 450MHz Pentium II, some RAM, video (no sound though), and the other necessities of a computer.

I agree about the RF interference, however. I ran several computers, even in their shielded cases, in my room for a while, and it was a deadzone for our cordless phone. It would be only worse with inches, instead of feet, between the systems. Not all people have room for a rack to mount things on, however.

Re:Impractical

[Anonymous Coward]

That's why they use ethernet for communications and just use the PCI bus for the power supply.

The PCI bus is just an outdated fancy parallel port.

Re:Impractical

[Anonymous Coward]

PCI bus is only constrained to 133MB/S at 33Mhz/32bbit - not only can you widen the bus (as you mentioned) but the freq can also be increased to 66Mhz

Re:Impractical

[Sleuth]

> -atrowe: Card-carrying Mensa member. I have
> no toleranse for stupidity.
>

Cool. Now that you've taken the Mensa test you can spend some time with a dictionary.

lol.

Re:Impractical

[Peter Dyck]

the throughput between the card and the host system is simply unacceptable.

I'd suspect a radar system requires much high a throughput than web or DB serving. Here's an example of such a system. "160Mb/sec, 32 bit parallel synchronous interface" doesn't sound that high to me.

Re:Impractical

[bad-badtz-maru]

I would suspect that the sig is a joke and the misspelling is the "punch line"...

maru
www.mp3.com/pixal

Re:Impractical

[morcheeba]

RF Interference:
I don't think there will be a problem with interference. Check out these computers. They use a similar system, but instead of being on a pidly motherboard, they use the ubiquitous VME format. They really pack in the processors -- 4 G4 PPC's per daughter card, and 4 daughter cards per single 9U VME card, and then 16 9U cards per chassis, and then three chassis. (4*4*16*3=48 TFLOPS) The pitch spacing on PCI is comprable to that on VME.

Also, I wondered about the connector on the tops of these boards. It looks like another PCI card edge. I wonder if this is a duplicate of the host PCI interface (for debug purposes), if it's a new "slot" to connect to the server's internal bus, or if it's a way to connect server cards bypassing the main PCI bus (for better performance).

Re:Impractical

[Indomitus]

One would hope. You never know though. I like that you put quotes around "punch line" because it isn't much of one.

Re:Impractical

[Sleuth]

Deliberate stupidity? At least we can assume [s]he's not ignorant, right? (I'm hoping here.)

Re:Impractical

[Knobby]

don't see these things taking off for most uses because the PCI bus is limited to a measly 133 MB/S. Even newer 64 bit PCI slots found in some servers have insignifigant bandwidth to keep the data flowing fast enough to make full use of these things.

You've heard of Beowolf clusters, right?

Let's imagine I'm running some large routine to model some physical phenomena.. Depending on the problem, it is often possible to split the computational domain into small chunks and then pass only the elements along the interfaces between nodes.. So, how does that impact this discussion? Well, let's assume I can break up an NxM grid onto four subdomains. The communication from each node will consist of N+M elements (not NxM).. Now, let's take a look at our options. I can either purchase 4 machines with gigabit (~1000Mb/s) ethernet, Myranet (~200Mb/s) cards, or maybe I can use ip-over-firewire (~400Mb/s) to communicate between machines.. Gigabit ethernet has some latency problems that are answered by Myranet, but if we just look at the bandwidth issue, then ~1000Mb/s is roughly 125MB/s. That's slower than the 133MB/s you quoted above for a 32bit, 33MHz PCI bus.. Of course there are motherboards out there that support 64bit, 66MHz PCI cards (such as these from TotalImpact)..

You're right that the PCI bus is not as fast as the data io approaches use by IBM, Sun, SGI, etc to feed their processors. BUT, if I'm deciding between one machine sitting in the corner crunching numbers, or 4 machines sitting in the corner talking slowly to each other through an expensive gigabit ethernet switch, guess which system I'm going to look at?

Re:Impractical

[Anonymous Coward]

Or, in the case of PCI-X, 133 MHz. Let's see, 133 MHz * 8 bytes (64 bits) = about 1 GByte/s + overhead.

Re:Impractical

[Charles Farley]

Please tell me you're not trying to argue that ethernet is faster than PCI. If that's true, your PCI bus would plug into your ethernet card and not the other way about. Do the math. 100/8 is a whole lot smaller than 133.

Re:Impractical

[Shanep]

Moderators need to mod Charles UP and the AC down.

That's why they use ethernet for communications and just use the PCI bus for the power supply.

There is nothing informative about this! I was supporting products like these from Cubix back in '94. Back then, those products were even using the ISA and EISA buses to carry ethernet between other Cubix cards on those buses.

The PCI bus is just an outdated fancy parallel port.

ROFL. ISA can carry data at 8MByte/S (8bit 8MHz = 64Mbit/S) to 32MByte/S (16bit 16MHz = 256Mbit/S) which provided a far better solution in these setups than a 10base-* NIC that was going to be plugged into ISA or EISA *anyway*!

And PCI can carry data at 133MBytes/S (32bit 33.333MHz = 1Gbit/S)

These cards are usually slaves to the host motherboard, not the other way around. This way they're easier to make and the assumption can be that whatever they're plugged into will be the master of the PCI bus, so no need to fiddle with master/slave config. For usage with a dumb PCI backplane, PCI master cards (one per backplane please) can also be purchased. Though I have'nt looked at this companies offerings.

A Linux machine set up as a web server, accelerated with khttpd and one of these cards running FreeBSD serving the db would be an awesome setup. Nice and real fast. Especially with a server mobo with multiple PCI buses (buses != slots) to seperate the 100/1000Mb NIC interfaces and the db card.

Re:Impractical

[ryanr]

You know that that's megaBYTES per second, yes? Or just over a gigbit per second? If that's not fast enough for you, what is? Pretty much any solution to connect something external to the box is going to have to go through the same bottleneck. Really, the only faster buses you have on a PC are the RAM sockets, and the AGP socket. I seem to recall a special high-speed networking solution that goes through AGP, but we're talking a little bit different class of hardware.

Re:Impractical

[crisco]

Hmm, methinks this is a very good troll.

Aside from the nick and the sig, calling 133MB/s 'measly' is absurd. Sure, compared to servers that cost an order of magnitude higher than these do, it is a little slow. But comparted to 100 Mb Ethernet, it is pretty fast. For specific applications it is definately useable.

And RF problems? How about24 CPUs in a 3u package, using a similar concept?

But for a troll, its nicely done. Several detailed replies, even I couldn't resist!

Re:Impractical

[Luminous Coward]

I can either purchase 4 machines with gigabit (~1000Mb/s) ethernet, Myranet (~200Mb/s) cards [...]

You meant Myrinet. And you meant 200MB/s not 200Mb/s. Actually it's almost 2 Gbps (2000 Mbps).

Re:Impractical

[Anonymous Coward]

WTF???? 133 MB/S is insufficient for web serving?!? How exactly do you connect to the internet, and how much does a bag of it cost?

Re:Impractical

[Anonymous Coward]

I would not think this is a troll exactly.
I do wonder how useful they are. At only 300mhz they are not powerhouses. I do wonder if a bunch of 1U rack servers might not be a better deal.
How ever I can see how it could be used for a small web Cluster. Us the Host machine to be a database server and then two or more of the cards as Webservers.
However for many situartions several 1U boxes or one of the new blade servers might be better.

Geode Specs

[Ron Harwood]

Here are the Geode specs... "Speeds offered up to 266 MHz"

Re:Geode Specs

[Anonymous Coward]

The card has a 300 Mhz x86

G4 processor cards

[Peter Dyck]

I've been wondering how expensive/difficult it would be to build a multiprocessor computer for computational physics applications based on G4 PowerPC cards.

I'd just love the idea of having a host PC (or a Beowulf cluster of them ;-) with all the PCI slots filled with G4 7400 boards crushing numbers...

Re:G4 processor cards

[Anonymous Coward]

I remember that I saw some REALLY nice G3 or was it G4 cards... they were a little to expensive.. but really cool.
This crap then... eh.... National Geode cpu.. what a sick joke... my old sparc station 4 performs better than that crap....

Re:G4 processor cards

[statusbar]

'Crushing numbers' is the right term, as g4's altivec is only single precision.

But it would be cool.

--jeff

not that new

[Anonymous Coward]

thats how quite a few of the "old" HPs and other servers worked... just slotted in extra CPU "cards" into the mainboard... sorta same idea as adding memory to routers, etc...

However, surely using something like a PCI slot would kill the data rate drastically... Maybe we may see the same thing that happened to ISA ports happening to PCI as AGP gains new uses... Who knows...

PCI card computers

[hattig]

You have to remember from a certain aspect, you can add a PCI card to a motherboard which made the motherboard the PCI slave.

PCI = PCI = PCI = CPU = PCI = PCI
I I I I
IDE CPU CPU CPU
I I I
USB PCIs PCIs
I I
IDE ..
I
USB

I have left out memory controllers, northbridge, etc, and modern fancy chip interconnects because they are just fluff (no, not fluffers, that is another industry). In the above diagram, what is the host CPU? Is there actually such a thing as a host? The PCI bus is arguably the center of a modern PC, with CPUs and controllers hanging off of it.

Modern motherboards are just a restriction on what you can do in reality. Reality is a PCI backplane on a case, maybe with a couple of PCI-PCI bridges. You can then add anything into any PCI card that you want - normal PCI cards, or CPUs (NB, Memory, CPU, etc).

That is why you can configure these cards to use the 'host' IDE drive. It is just a device on every 'computer' within the case...

I can't post a diagram though, because I must use "fewer junk characters". Bloody lameness filter - affects the real users, the people it is meant to trap just work around it. Would you call this a "lame post"?

Re:PCI card computers

[Anonymous Coward]

you betcha, lamer.

Re:PCI card computers

[Anonymous Coward]

well, lame in a karma whoring way

Re:PCI card computers

[castlan]

At least for AMD based multiproc systems, the Northbridge seems to be the hub around which the CPUs are gathered. AFAIK, the RAM doesn't have a direct connection to either CPU, it has a dedcated bus to the Northbridge. Why is the Northbridge "fluff"? Isn't is the closest thing to a host on PC systems? It is what makes the "Motherboard" the mainboard. Where is the BIOS located? If it is part of the Northbridge, then that would close the argument for me. If it is discrete, then it is a good candidate for the center of a modern PC, if you allow that once a system is booted, "central" functions (like "basic Input and Output") can be migrated to other parts of the system.

For me, anyway, the PCI-PCI bridge seems to be a pretty good negation of the "PCI bus as host" viewpoint. If anything, the PCI bus is just an extension to the PCI controller, which would seem to fall under the "Northbridge chipset as host" perspective.

As we migrate from a single CPU paradyme to multiple CPU architectures, it seems the view of "Primary CPU controlling auxilliary cpus" is vestigal, and we will be moving away from it. This seems apparent if you follow the Locking mechanisms used by Linux migrating from large per-cpu locks, to finer grained locks. It is not very useful to have a CPU centric system when CPUs are commoditized. The Chipset seems to be the lowest common denominator for the forseeable future.

Re:PCI card computers

[hattig]

It is fluff in the context of the fact that you can connect multiple Northbridges (with CPUs, memory, PCI-PCI bridge possibly) to a PCI bus, and the PCI bus will be fine.

The BIOS is located off an LPC device connected to the southbridge.

A modern PC is a subset of what a PC could be. As I said.

You can view a PC any way you like. But you can connect PPC computers on PCI cards to PCs, and they can access any resource on that PCI bus just like the host can. Because, it is simply another host on the PCI bus.

Hence, PCI backplanes work. PCI-PCI bridges are there so you can have more than 6 PCI slots!

How original

[Anonymous Coward]

"Could something like this be the future of computing where for additional processing power you just kept adding additional computers inside of a host?"

Yeah - because Sun never thought of that. Can you say S/390?

CPU Speed

[CBoy]

I read through the site and I could not find ANYTHING vs. relative x86 cpu speed. Anyone find anything? Sure it's great to have a PC, but at least give us some hint of how it performs compared to an x86 cpu.

Re:CPU Speed

[Anonymous Coward]

It is a x86 CPU, a NS Geode which I believe is based on a 486. I got all excited until I realised this thing has practically no use as it does n't have enough grunt. Even if you put 6 cards into one machine and each card had 266MHz thats still only 1596MHz and remember the IPC of a 486 is nowhere near the IPC of an Athlon/P3 or any other modern processor.

Linux on PCI cards is the way forward.

[Anton Anatopopov]

Perhaps this is the way to get over the anti Linux brigade when they say 'Linux is difficult to install'.

Just hand them a PCI card and let them get on with it. I can't help thinking it would be better on a USB device though. Then you wouldn't even need to open the case !

Re:Linux on PCI cards is the way forward.

[Ron Harwood]

Actually, in my LUG we've given the newbies an eval copy of VMware, and a pre-installed linux image... let's them play for a month before they have to think about installing.

Re:Linux on PCI cards is the way forward.

[Anton Anatopopov]

VMware is another way, but its a bit expensive. I would rather spend my $300 and get some hardware to show for it and effectively 2PCs, than spend it on vmware because it will run slower.

I did have a copy of VMware which I paid for, but I lost interest when they went all 'enterprise' on it and the prices got stupid.

Still, theres always plex86, but I want to run it under Windows ME :-(

Re:Linux on PCI cards is the way forward.

[Anonymous Coward]

Just hand them a PCI card and let them get on with it. I can't help thinking it would be better on a USB device though. Then you wouldn't even need to open the case !

Just use a Knoppix CD.

You don't understand...

[Ron Harwood]

This doesn't increase the speed of your existing computer... it adds another computer on-board...

Re:You don't understand...

[krow]

Right. And the cool part (which makes them a bit different then the typical solution) is the loopback ethernet on them.

Isn't that the course we've been on?

[JAZ]

Follow me here:
A computer used to take up a room.
Then, computers were large cabinets in a computer room.
Now, they are boxes in a computer cabinet in a computer room.
So we can extrapolate the next step for computers is to be cardss in computer box in a computer cabinet in a computer room.

It's a natural (obvious) progression really.

Re:Isn't that the course we've been on?

[Anonymous Coward]

Yup. Indeedy. The step after that is:

Chips on a card in a box in a cabinet in a room.

And when I say chip, I mean a chip literally... this is probably at elast 10 years away though :)

Re:Isn't that the course we've been on?

[Anonymous Coward]

Good form! But what's after that? I suggest that we'll be applying Moore's law in the form of "The number of CPUs in a chip doubles.."

Re:Isn't that the course we've been on?

[dillon_rinker]

Computer chips on computer cards in computer boxes in computer cabinets in the computer room...

Re:Isn't that the course we've been on?

[rocjoe]

Congratulations, you seem to be the first beginning-of-thread poster to "get it".

Re:Isn't that the course we've been on?

[CarrotLord]

Then chips on cards in boxes in cabinets in rooms.
and then... ?

salt on chips on cards in boxes in cabinets in rooms?

hmmm

Not that News

[Anonymous Coward]

I worked at Commodore on the Amiga Bridgeboard over 10 years ago, and they had a whole Crawl^h^H^H^H^HBlazingly fast 80286 on the board.
This approach is decidedly not new, the real problem is they don't adverise the speed limit of the processor (if they have a GHz or faster processor on those boards I'll be VERY impressed), nore do they indicate if there is an upgrade path.

Still this is kind of cool, if the boards can be broght up to normal processor speeds, you could make a small (relatively) cheap and fast 6 or 8 way SMP node if you had available slots.

future? yes, but it's here today...

[mysticbob]

sgi has been doing this for a long time. their
newest systems are almost this exactly, but instead of slow, thin pci, they use large, fast
interconnects:

http://www.sgi.com/origin/300/

Wait! What about Beowulf?

[friedmud]

He didn't even try to do any parrallel processing with these things! That was the first thought that came into my head.

Here we have four or five cpus all in one machine, talking to eachother over a native PCI bus. It seems to me this would be a great way to run a Beowulf cluster In a machine.

Anyone care to comment on why he might not have done this?

If you do have storage issues...

[pvera]

Commercial rent is expensive, so the least space you need to dedicate out of your office to store servers the more cost effective they are.

These cards have been around for ages with various degrees of complexity. There used to be (don't know if they are still around) some of these cards that were designed to plug into a Mac so the card would do all the hard work if you wanted to emulate a PC.

I don't see the value for the home user. I can't see why a true home user (not the very small percenteage of hardcore enthusiasts or people that run a business from home) would need so much power that the solution is to get a box, plug a few of these babies and cluster them.

Still, its not so hard to come up with a home scenario:

1. Send your broadband connection to the basement of your house and spread it to all the rooms in the house with a $80 broadband router, cheap switches and hubs.

2. Put a box in a closet in the basement with different PCI cards to serve a specific purpose. For my own personal needs (I am a Microsoft dot whore, sorry) I would have an exchange server, one dedicated as a network file server, a sql server and a IIS server. A person of the Unix persuasion would have a card with sendmail and some kind of pop server, a file server, mysql or posgres and Apache.

With just a little bit of money the house now packs as much punch inside of that box in a basement closet than what it takes my company to do with a row of bulky servers. Add in a blackbox switch and a cheap 14-in monitor, keyboard and mouse and you are set. Of course Unix people would use some kind of secure shell and save themselves the trip to the basement, and us lazy Microsoft whores will just have to rely on Terminal Services or pcAnywhere.

In a corporate environment the space saving actually pays off (you don't pay your apartment rent or home mortgage by the square foot like most businesses do) as soon as you recover some of the space wasted by the server room. Right now I can see how I could take ours, gut it out, put a couple boxes full of these PCI cards in a good closet with the proper ventilation, and then turn the old equipment room into a telecommuter's lounge.

The home solution would rock because my wife will not bother me anymore about all those weird boxes sitting under my desk in my home office. All the clutter goes away and I just keep my tower case.

Geode?

[tcc]

266mhz max. Their target audience is the firewall/network application.

Too bad a Dual Athlon-based solution (on a full length PCI card) would suck too much juice... at least from the current PCI specs... AMD needs to make a move like intel did with their Low wattage PIII, I'd love to see a 12 processor (5 pci slots plus host) renderfarm in a single box for a decent price. Not only it would be space saving, but imagine that in a plexi-glass case :) a geek's dream.

Re:Geode?

[Anonymous Coward]

You could always supply the additional power via a powersupply connector. I'm pretty sure there was some monster 3DFX graphics card that needed additional power and got it this way.

Re:Geode?

[Anonymous Coward]

And we all know how well that went over....

Wait just a minute...

[Zenjive]

this better not be another Krasnoconv with that hoax SETI-accellerator card!
I don't know if I can take another disappointment like that.

aka transputers

[wganz]

Transputer advertisements were common in the back of the old Byte magazine. They were more popular in the UK than the US. With the newer low power consumption Transmeta & PowerPC CPUs + low RAM prices, this is more viable from a cost/power ratio now than then.

It is not like Transmeta has a shortage of Linux talent to help bring this off. If Transmeta makes such a product and puts an advertisement in something like Linux Journal with Linus's smiling face beside it, it will sale like the proverbial hot cakes. I would buy one with or without his picture.

Just a thought

P3 on a PCI card

[Anonymous Coward]

Advantech carries similar "PC on a PCI" products, much of which are much more powerful (P3 1Ghz anyone?) than the one referenced in the article.

Re:P3 on a PCI card

[Anonymous Coward]

i went and looked, those are SBC's it is a PC but you cant plug them into a server and create an nework across the pci and share a hd

Imagine a new kind of bus

[Anonymous Coward]

Imagine if all the devices in your computer were attached to each other with 100 GB optical cable.

Essentially there would be a switch that allowed about 32 devices to be attached.

The devices could be storage devices, processors, audio/video devices, or communication devices.

Storage devices would be things like memory, hard drives, cdroms and the like.

This bus would allow multiple processors to access the same device at the same time and would allow devices to communicate directly to each other, like allowing a program to be loaded directly from a hard drive into memory, or from a video capture device directly onto a hard drive.

No motherboard, just slots that held different form factor devices with power and optical wires attached.

A networking device would allow the internal protocol to be wrapped in IP and allow the interntal network to be bridged onto ethernet. This would allow the busses on seperate computers to work like a single computer. The processors on all the machines could easily network together, memory could be shared seamlessly, harddrive storage would be shared and kept backedup in real time. Any device in any machine could communicate directly with any other device in any other machine. Security allowing.

Want 20 processors in your machine? Install them.

Want 6 memory devices with 1GB each? Add them.

Want 100 desktop devices with only a network device, display device and input/output device that use the processor and storage out of an application server? No problem.

Want a box that seemlessly runs 20 different OSes each in a virtual machine that are ran across 10 boxes in a redundant failover system? No problem, it's all done in hardware.

Want the hard drives in all the desktop machines to act like one giant raid 5 to store all the companies data on? No problem. (1000 machines with 10 GB each is 10 TB of storage)

This is the future of computing.

Re:Imagine a new kind of bus

[MikeFM]

I think the basic form to use is some simplified base system designed to be upgraded to the extreme. No built-in crap on the motherboard to speak of.. just lots of PCI slots. If they could share harddrive and RAM and provide a keyboard/mouse/monitor switching method similar to KVM switches but all in one box it'd be great. So rather than replacing older computers we could just add to them. Maybe perfect something like MOSIX and drop the whole stupid SMP idea. I've always imagined computers would someday be like legos where you could buy a CPU lego, a RAM lego, a harddrive lego, etc and just plug them together in any order to add to a hot system. No reboot and no case to open. If one burned out just toss it and put a new one in.

Re:Imagine a new kind of bus

[bernz]

it exists already, sweetie. check out the infiniband spec somewhere.

SUN has a similar product..

[Phizzy]

I am actually typing this comment on a Sun Microsystems SUNPCI card.. It's a celeron, I beleive a 466mhz or so, w/ 128m of ram. It has onboard video if you want to use an external monitor or can use the sun's native card if you want to run it windowed, ditto w/ ethernet. I've been using the card for about 3 months now, and other than some instability w/ SunOS 2.6 (which dissapeared in 2.8), I haven't had problems with it.. you can copy/paste between the Sun window and the 'PC' window, which is very helpful.. and though we are running WIN2000 on it (ok.. so shoot me) I don't see any reason why you couldn't run linux on it if you really wanted too.. All-in-all, the card is pretty badass..

//Phizzy

Re:SUN has a similar product..

[Peter Dyck]

I agree.

I'm posting this with Konqueror on Sun Blade 100. Next to the Konq window I have a SunPCI window with W2K/Office2K. As nice as Sun's StarOffice it still doesn't import/export clients' office documents properly.

Re:SUN has a similar product..

[Anonymous Coward]

Apple had the Houdini cards, and Orang Micro had PC cards that were essentially 166 mhz pentium PCs that shared resources with the MAC. Prior to that, Radius had the Radius rocket that was a 68040 processor that was essentially a MAC within a MAC that could run in Parrallel to accelerate processor intensive tasks.

Re:SUN has a similar product..

[Phizzy]

yeah.. that, and StarOffice eats all your ram, not to mention your desktop when you run it. ;)

The test I've run of SunPCI has convinced our management to do away w/ separate NT/2000 systems when we move to a new building in april, and just outfit everyone w/ Ultra 5s, SunPCI cards and dual-head monitors..

//Phizzy

Re:SUN has a similar product..

[11223]

Actually, Sun is making them now with 733 mhz Celeries in them.

Definitely an awesome product.

Already happening...

[Anonymous Coward]

Most high-end unix servers are already just variations on this theme. Sun's E10k, for example, is just 16 briefcase sized E450's with a domain interconnect.

Re:Let me get this straight...

[Anonymous Coward]

How does this guy do it...

YHBT. HAND.

hmmm..

[ZaneMcAuley]

Brings back memories of Transputer cards :D

How does sharing of the disk between each machine on a card affect the performance ?

Audio Apps -- Digidesigns DSP Farms

[namespan]

I haven't been paying attention to the market... I guess things like this aren't all that rare. Apparently there's a G4 PPC computer-on-a-card as well.

But anyway, it reminds me a quite a bit of what Avid/Digidesign do for their high-end systems.
You see people who've got 6 slot PCI systems and 4 of those slots are filled with extra computing cards (sometimes more... some people get expansion chasis'). You can rely on your computers processor if you're not doing to many complex effects on a track of audio, but at some point (not too hard to reach... throw in a tube amp emulator and a reverb) you run out of CPU. So they have PCI cards which have a couple of DSP chips (Motorola 56xxx series, I think) on them, and the more of these you add, the more audio processing you can do simultaneously.

At some point, perhaps people will think: hey, why add a specialized card? Why not just more general purpose computing power?

Re:Audio Apps -- Digidesigns DSP Farms

[RFC959]

At some point, perhaps people will think: hey, why add a specialized card? Why not just more general purpose computing power?

Thank you for contributing to the turning of the Wheel of Reincarnation.

Supporting other Host OSses?

[Anonymous Coward]

I wonder if this Ethernet simulation and drive sharing is supported for other host operating systems, so that you can have a Linux box-on-a-board in a computer running something evil, like Windows? (Just in case you need to run mainly Windows on your production box since your boss requires you to, but you want a Linux ghost-in-the-machine for fun and/or increased productivity... And - yes! - I read in the review the card itself can run Winoze, but that*s not my point...

Thx

teq

Re:Supporting other Host OSses?

[Anonymous Coward]

the site says you can run linux, bsd and windows all in the same box

No serial connection?

[Anonymous Coward]

I don't mean a USB, but I would have liked to see the option to have a serial connection so to have a tty with which read the system console directly, so to check the system status without having to leave a telnet/ssh service on...

my problem is price

[Eugene]

I'm fairly interested in those devices, but right now the the cost for those boards is not cheap enough for me to get it. at ~$500 a pop, I could put together a cheap system with better specs (not on a board, of course). I know it's targeted for server/commercial applications, but if they are willing to lower the price some, I'm sure there'll be a lot of takers.

my idea setup will be using a CF card with CF-IDE adapter as the boot drive(which eliminate the dependancy of the host OS on powerup and no actual HD required).

Re:my problem is price

[J. Random Software]

Minaturization costs money, at least in the macroscopic world. If you have a square foot of motherboard, relatively straightforward designs and big cheap commodity components become feasible. Smaller systems save a little power and make higher clock rates attainable (in a 1GHz clock cycle of 1ns, an electronic signal only travels about a foot), but until commodity desktop/minitower hits such a performance wall (or simply goes out of mass-market fashion) it'll always be cheaper than embeddable versions or laptops.

uhnnn,yeah, thats it...

[Anonymous Coward]

doesnt SUN already make this? machine of the future indeed....

up tp 104 processors in a box, add as you need...

even HP/paq has pa-risc based machines that are built like utility payments....four (or x) processors built in, only three turned on...ooops need more, crank up the meter...activate another processpr at a higher payment level...

how far back..?

[Anonymous Coward]

I remember my Apple IIe had a Microsoft card with a zilog processor so the Apple could run CP/M. Nice mix, eh? So what is the earliest version of this sort of thing?

G4 PCI cpu

[starman97]

Here's a G4 card that plugs into a PC or anything with a PCI slot for $400
http://www.sonnettech.com/product/crescendo_7200 .h tml#pricing

The Catch: You have to write the device driver for the Motorola MPC107 PCI bridge chip.

Switched Bus, Multipurpose cards

[swb]

I'd like to see a bus that was little more than a switch, with a minimum of logic for management.

For cards, it'd be great if each card had its own CPU and RAM. Ideally the cards would have a few universal connectors, each of which could accomodate an I/O module which would contain just the discrete electronics necessary to drive a specific device or medium (eg, video, audio, disk, network).

Bus-Switch modules would be interconnectable to accomodate more cards, and would have switch-like manegement features for segmentation, isolation and failover type features.

The CPU cards themselves ought to be less complicated than motherboards since there's no bus logic, just interconnect logic to the Switch-Bus and the I/O modules, and RAM.

Since each board has its own RAM and CPU it ought improve system performance because the O/S could offload much more processing to CPU boards dedicated to specific tasks. Instead of the kernal bothering with lower-level filesystem tasks and driving the hardware, a "driver" for filesystem and devices could be loaded on a CPU board dedicated to I/O.

The same could be true of user interfaces -- run the the UI on the board dedicated to video, audio and USB. The kernal could run applications or other jobs on the "processing" CPU board(s).

Networking? Offload the entire IP stack to the networking CPU board.

Re:Switched Bus, Multipurpose cards

[dy_dx]

it doesn't seem like this would work so well...the PCI bus would get very hosed with I/O requests from the RAM on each chip. Instead of limiting the hosage of physical memory access to the IDE or SCSI bus, you take the PCI bus down with you too (making the bandwidth of the PCI bus limit scalability of how many procs you can use).

also, having multiple memories accessing the same data in a distributed program adds plenty of overhead to make sure all the memories maintain validity and access control of the data. thus the chips wouldn't be as simple as CPU, RAM, interface.

Could be cool for games

[Anonymous Coward]

Hm.

Over the past few years, more and more CPU
time in games has been allocated to AI
processing. We've got video and audio
hardware cards to offload the significant
demands of those tasks, but AI would seem to
need a more flexible solution than a hw
implementation of 'DirectAI' or somesuch.
One of these might work well as a programmable
AI subproc. That, or perhaps an engine could
be tweaked to offload physics calculations to
one o' these puppies. Of course, keeping
state information consistent would be a major
issue, so the whole undertaking would be
subject to a overhead/payoff analysis.

i dunno about these

[mikus]

Seems pci bus would be a horrific limiting factor considering you now have a series of processors sccessing resources across the pci bus, as well as on-board. As a firewall, unless the nic's are something decent that can handle hardware trunking to a real switch, 1 interface isn't going to take you far. If they could do dot1q trunking, it would definately be nice. Beyond that, I can't see something commercial like checkpoint fw1 running efficiently on the daughter cards with no more than 266mhz and 256mb of ram. Might work ok for something such as ipf, pf, or ip chains on a stripped down linux kernel, but no freakin way no on win2k. I say stick to a cluster of 1u's if space is that big of an issue. I doubt it's any less complicated to set up a cluster 1u's than getting these things to work *correctly*.

Re:i dunno about these

[Anonymous Coward]

It has a 300 Mhz and up to 512 RAM and will run Checkpoint FW-1 . It has 2 MAC's , 1 10/100 external , and 1 internal GB on the PCI bus.

ClearCube has another one

[studboy]

(disclaimer: my bro's wife usta work for this company)

rackmounted PCs with video, etc. They're intended for offices: you run cables to each person's monitor/keyboard/mouse, manage all the actual hardware in one place ~~ ClearCube

Linux on PCI a year old, dudes!

[scaryjohn]

I looked at this and said... wait a minute, hasn't this already been sorta done? Despite not being a full featured box, Firecard is a PCI-card running Linux... for the purposes of supporting a firewall (as you could have guessed from the name if you'd not read the story -- Nov 14 2001... but it's cool that they've taken it to the next level.

Nothing particularly interesting about this

[Anonymous Coward]

How is this anything interesting at all?

You just have a bunch of PCs talking IP over a bus. Who gives a shit. There's nothing that makes it any harder to talk IP over PCI versus Ethernet.

Other companies, especially those in the CompactPCI space, have been doing this for years.

Typical of Slashdot readers to jump up and start screaming any time anyone does anything.

The voices of the dead cry out for vengence in the form of bowls of hot grits.

Radius Rocket

[beerits]

This sounds like the old radius rocket for mac. Each rocket was essentiallly a quadra on a nubus card.
http://lowendmac.com/radius/rocket.shtml

Single card servers

[Capacitor]

Those of you wondering about expandable servers should have a look at Transmeta's homepage (http://www.transmeta.com)- among their featured products is a server using something called Serverblade - single board servers. I think you can fit 24 of them in a 3u rack enclosure.

I want some !

[jooniqzb1tch]

This is really a great system ! of course it's a bit expensive but still, having two more linux boxes inside your system to play around with, without the extra cases/cables/noise sounds awesome. it would also rock to host several small servers .. I'll definitely try to get some.

This idea is actually the standard

[clustermonkey]

In the communications industry, this isn't anything new. The idea of having multiple computers on blades in a chassis is a (the) standard called 'PICMG', based on CompactPCI technology. It's been around a long while. I validate systems like this at work.

You can get 4, 8, 16, or even 24 SBCs (Single Blade Computers) in a chassis, and link these chassis together via switches. Each chassis has a switch that links all the SBCs in the backplane together and has external ports to hook it up to the outside world.

Check this out:
http://www.picmg.org/compactpci.stm
and this:
http://www.intel.com/network/csp/products/cpci_i nd ex.htm

Read the article...

[YuppieScum]

These cards have TWO NICs, one that talks across the PCI bus and one physical RJ45 10/100...

I'm only go to say this once, but I could copy/paste the same response to 20 or 30 posts on here...

Haven't Sun been doing this for a while?

[parryr]

Sun Microsystems have had "PC cards" for a while now. There was a whitepaper they published some time ago on using a small Sun server (say, an E450) and populating it with PC cards.

They demonstrated how an entire Windows NT cluster could be built using this technology, chucked in some Terminal Services under Windows, ran Exchange, and then did all the important stuff (mail, DNS, whatever) on the Sun box itself.

Granted, it's not Linux, and granted, he cost of a Sun box is quite high - but the PC cards are significantly cheaper over here for Sun hardware, and Sun architecture seems to be a bit more robust and scalable than PC stuff.

who cares?

[adolf]

Stuff like this started in the PC world, IIRC, with 386s on 16-bit ISA cards.

Nobody cared then.

Why would anyone care now?

Please explain your point using no more than 100 words.

-

The Briq

[Anonymous Coward]

Something similar to this has been down, though it uses a 5 1/4 inch bay instead of a PCI slot. It called the Briq, it's a self contained PPC based machine created by Yellow Dog. (http://www.yellowdoglinux.com) It was engineerd to run Yellow Dog..so no, it's not a Mac, though you can run MaconLinux.

Re:who cares? lots of possibilities

[Anonymous Coward]

1. anyone with exposed servers that could use an independant firewall or virus screen without adding space, power, cables
2. anyone building multiserver solitions that wants a private network and shared RAID storage in a single box.
3. anyone who wants to run multiple OS's and or images on real isolated HW in the same box with a built-in network.

Sure its possible to do some of this with other unique and expensive HW and networking stuff but this is ready made servers that drop-in industry standard servers and run linux, BSD ( or windoze if req ) . It certainly makes alot of solutions possible without just adding boxes.

Sunpci card

[johnnyp]

I've got an Ultra 5 with a PCI card which has an AMD K6/400 on it so's the Ultra can run a Windoze machine. The K6 shares the HD disk and can either use the Sun display or you can plug a separate monitor into its on-board vid. Also shares the ethernet card. It works OK, runs Win 98 fine (95 is supposed to work though I can't get it to, but I have seen one running NT 4.0) and you can cut and paste between CDE and Win 98. The only real use I find for it is firing up IE to see if webpages I've made look OK.

I think you can pick them up pretty cheap nowadays if you like that sort of thing. I don't imagine much mileage from trying to install a.n.other unless you feel like writing the relevant drivers to get everything to talk to each other.

Re:Sunpci card

[johnnyp]

By which I meant a.n.other OS than Win95/98 or NT 4.0. As far as I know you can put as many of these PCI cards in your Sun box as you can fit in should it take your fancy.

Re:Sunpci card

[PerfectWorld]

Sun limits the number of these cards which are supported in each system type, due to bandwidth limitations, but obeying the supported rule is up to you. I wouldn't put more than 2 in a desktop box (U5/10,SB100) and I think only 1 is supported. IIRC, Sun supports up to 6 in an E450.

The current model is a 733MHz Celeron with 128MB RAM base, going up to 1GB RAM, onboard Rage something or other graphics. Supports all versions from Windows from 95 to 2K Adv Server.

You can do some interesting things with these. Since Windows is 'installed' in an image file on the UFS filesystem, you can copy them, easily back them up, and bring differant images up on differant cards at differant times. You could have the office apps images running during the day and the Quake servers running at night ... ;-) They are also cheap enough to have a spare in the server should one go tits up.

They won't run Linux unfortunately. They would have to add support for that to the SunPCi software.

- Mark

Re:who cares? lots of possibilities

[adolf]

But is just adding boxes not cheaper, more flexible, and less prone to catastrophic failure?

By my reckoning, half-width 1U rackmount PCs stack more densely than mid-tower cases each with a half-dozen cards in it. Overall reliability increases, as well: you don't need to rely on a single motherboard and power supply to keep things going.

Putting a bunch of rackmount PCs into a single (portable, or not) box is also a very trivial exercise.

Is realestate at the desk really at such a premium that anyone need care about these things?

RLX already does this [was Re:Single card servers]

[Anonymous Coward]

Servers based on Crusoe Blades have been available
for some time:

http://www.rlxtechnologies.com/

[Linux Magazine, for one, has been carrying double
page ads for RLX webservers for a few months.]

The Server Blades are each complete PCs that come
with Linux or Win2K pre-installed, and you can pretty
much adapt them to your needs. The basic
idea is 24 blades per 3U rackmounted box. Here is
a features list with more info:

http://www.rlxtechnologies.com/purchase/salesfaq /

wait a minute

[dakoda]

this is exactly what many good video cards do, but in a specialized manner. same with high end sound etc. the idea of putting powerful cpu's on cards is probably as ancient as cards themselves.

as has been noted before, this would really be useful if the pci bus was extended (faster/wider). of course, making it faster/wider gives you what sgi has been doing for a while too (also mentioned above).

perhaps the most dissapointing thing is that all that power goes to waste on users playing solitare, running windows, aol, and quake, not on something that will actually need the power to perform the tasks. well, maybe quake isnt so bad...

days of future passed

[hatless]

Can this be the, uh, future?

No, not if it's existed for decades. It's what's referred to as a "mainframe". You know. An expandable number of processor boards running under an operating system that can treat them as any number of single-processor or multiprocessor machines, with the ability to reassign processes between CPUs.

The Unix world has had them for a long time, too. Modern examples include Sun's higher-end servers, which support hot-swappable and hot-pluggable processors and memory.

Doing it with x86 processors and standard x86 OSes like x86 Unixes and Windows is less common but I believe Compaq and maybe Unisys can sell you machines that can do it, too, with one or several instances of the OS running at once.

This hatdware approach is not quite the same as VMWare's server products, which do it via software and don't limit you to one OS per processor or block of processors. It in turn mimics other decades-old mainframe operating environments in approach.

Some merit to the design, but this is better.

[The Panther!]

The only two cool things about the PCI card in the article is host-IP connectivity (which essentially is a dual ethernet card then, perfect for firewalls and such), and 10W power draw straight from the host's power supply.

However, Powerleap, ubiquitous for upgrades and socket adapters, also has a card which touts some similar attributes called the Renaissance/370S based on a Socket 370 or FC-PGA chip. It cranks with Celeron, Celeron II, and P3 chips. Quite rockin'.

The main cool thing about this device is it does NOT use the motherboard slot it sits in. It just uses it as a place to mount. That's right, you can put one in an ISA slot and still run the motherboard it sits in and they won't know a thing about each other, because no pins are connected between them. The price is also a lot better (~$250 for a low end model), you can swap out the CPU and it has two DIMM slots, with a max ram per slot of 512mb (1GB combined). The specs are much better and the price is much lower. It's just marketed as an upgrade option rather than a performance enhancement to an existing machine.

I've been looking into this as a solution for my cluster, but haven't gotten up the nerve to buy them yet. From what I can find on the web, they're the best cluster card option, especially if you are handy with soldering. To really maximize the power per box, I'd probably buy a dead 486 motherboard (ISA slots all the way across the board, which this card requires), slam four Renaissance cards in it, link two power supplies in parallel, rig extra power and reset switches for each card separate from the power supply, and there's your mini-cluster. Probably 4 machines per 4U case, which notably isn't a huge space savings over 4 1U pizza boxes, but it costs less than a single 1U server would.

This is nothing new

[mrdisco99]

IBM AS/400's have offered integrated Netfinity adapters for years. These are PCI cards with processors, memory, and console/network connections which share power and storage on the AS/400. You can fit up to 16 of these in a single machine. Check it out

Drivers only work with a stock RedHat kernel

[apayne]

I bought one of these cards back at LinuxWorld. The hardware's nice, they don't work exactly as advertised though.

The bundled kernel module only works with the stock kernel distribution in RedHat 6.2-7.1 (kernel 2.4.2 max). The kernel module sets up a virtual network device that allows the host pc to talk to the slotserver. The kernel module needs to run on both the host and the card to be able to communicate with each other localy. (You can still communicate via the 10/100 interface over the network.)

Another thing they advertize is the ability to have the card boot off of the host computer via a "virtual disk" (rather than having IDE drives hanging off of the card). I have't been able to get this working at all - and the only documentation available tells you that the feature exists.

It would kind of suck to have a PC loaded with 4 cards and 4 additional disks. I could sell a kidney and purchase some disk-on-chips I guess.

-Andy

Duhhhh... I'll take obvious tech directions 4 $500

[Anonymous Coward]

Wow...

Wow...

Strain my credulity to the breaking point

Move along, move along, no sense of wonder to be found here...

Would make a great web development & test stat

[Anonymous Coward]

This could make a great web development and test station. create pages on host system and view in realtime on multiple cards running each with diff OS / browser

Re:This is nothing new BUT...

[Anonymous Coward]

but this only works in an AS/400 doesnt run linux and costs $2800

Computer in a computer

[StikyPad]

I believe Amiga made something similar about 10 years ago. It was called a bridgeboard. It even had onboard CGA graphics..! I didn't have the slightest clue of the existance of Linux at the time, so I can't say whether it would have worked or not. Maybe someone else out there has tried it.