Slashdot Mirror
One Processor, 128 32-bit Cores
Max Entropy writes: "EETimes reports that a German company named Pact GmbH has developed a chip containing 128 microprocessor cores as part of the company's 'Extreme Processor Platform' (XPP). 'Each of the XPP's 128 processor cores sports its own 32-bit fixed-point multiplier, yielding a theoretical output of 12.8 billion multiply-accumulate operations per second at an expected clock frequency of 100 MHz. Pact claims the architecture will scale to produce devices capable of more than 400 giga operations/s in 2002 and into the peta-ops range within a decade.' The transistor budget for this behemoth is 30M, fabricated on a 0.21-micron process." Of course, each one of those processor nodes is completely proprietary and requires some peculiar programming.
'Each of the XPP's 128 processor cores sports its own 32-bit fixed-point multiplier, yielding a theoretical output of 12.8 billion multiply-accumulate operations per second at an expected clock frequency of 100 MHz. Pact claims the architecture will scale to produce devices capable of more than 400 giga operations/s in 2002 and into the peta-ops range within a decade.'
Mother of God! The first time some fool runs Quake IV Slaughter on a Beowulf cluster of these puppies (you knew that had to get in here somewhere), it'll instantly self-evolve into Quake X^100 and wipe out the human race!
A truly excellent pizza parlor is a delight unto the heavens. Treasure the sauce and the toppings!
Yeah, but if you can check and discard possible methods of simulating the protein folding at a decent speed, you'll have more possibilities of finding the right algorithm
I love it! Computers can NEVER be too powerful! Those Sun E10K's over here are already starting to get old, IMHO.
I always keep ten-year-old computer magazines around too; just to see what was state-of-the-art back then. I wonder how many years it would take to get people to laugh at the present day technology.
I'm tellin' ya!
A spokesperson for Pact GmbH claimed the Extreme Processor Platform was "Most bitchin, radical and gnarly". The spokesperson continued "We've totally maxed those proc cores, and out backplane is decidedly off the planet. Our 32 bit multipliers open one extra deluxe can of whupass, and our combination back-flip most tiny fabrication process cuts it it so nice, you wanna hear it twice - they can do 40 k fakie axel grinds a sec, and the shiny future one's gonna do a k of vicious ops".
When approached about the controversy surrounding the redundant second posting of an article concerning the chips to popular news-for-nerds site Slashdot.org, the spokesperson responded: "Its all good".
It's not Rob's job to write new pieces of slash right now. Slash is open source, and as Rob stated over and over again in #forum last week, if someone adds a feature to the code, he will consider it.
I suspect that if someone added a method of moderating articles, and defining user thresholds, it would make its way from slash to slashdot.org.
So, in short, if you want something done, do it yourself.
I'm waiting for a system that can churn out one SETI@Home unit (about 300K of data from Arecibo) per hour. Currently, it takes a nicely equipped P3 about 12 hours to process a unit.
I know, I know, SETI@Home isn't everything. But it is a good way to get a glimpse at the level of FPU performance
"Ancillary does not mean you get to rule the world." --U.S. Circuit Judge Harry Edwards, speaking to the FCC's lawyer
The area of Palestine reserved for a Jewish state by Balfour et al. was less than 50% of the current area of Palestine at the time. Currently the Israelis are controlling over 100% of that area.
That's the same as The United states of America expanding into Canada and Mexico.
The US couldn't do that though, because they give their best assault weapons away - to the Israelis.
That little gem from a reasonably highly ranked officerin the USAF. (who was complaining about how his men weren't well enough supplied)
FatPhil
(Who's lived in the part of Palestine which was never supposed to be anything but Palestine).
Also FatPhil on SoylentNews, id 863
This post and the first post are both funny. Lighten up.
${YEAR+1} is going to be the year of Linux on the desktop!
As already mentioned, as Moore's law runs out of steam, something else will be needed. Besides, it really doesn't matter how long Moore's law still holds; for any extremely powerful single processor it still holds true that for certain problems X of them will be X times more powerful still. While the theoretical processing power of some as-yet-unimplemented molecuar or quantum computing devices could be quite high, it will still be very finite and fall short of the requirements of some types of processing--such as nuclear reaction simulations, for that next generation of smart atomic weapons that can single you out and kill you based on certain profiles.
Besides, this 128 processor devices isn't really any different from most other multi-processing systems. The holy grail is still the development of smart compilers and algorithms that can allow even dumb programmers to write effective multi-processing code.
If you wish for your post to be taken seriously, then please talk sense.
As far as I know the Athlon (note the spelling) only has three pipelines which can perform operations like multiplies. Where does your figure of 6 come from?
Also FatPhil on SoylentNews, id 863
You obviously don't know about the Palestinians flying bat-like through the streets and countrysides terrorizing children and sucking the very life from their veins, BITING GREAT CHUNKS OF FLESH FROM THEIR BODIES, impaling their heads on their car radio antennae and driving past their homes singing ribald songs using the dead child's name.
Did I win?
--
As a matter of fact, I am a lawyer. But I play an actor on TV.
Indeed. The interconnections are apparently modelled on FPGA connectivity (according to the story), and such chips have similar problems: routing delays can easily be 4 or 5 times logic delays. It's due in part to all the switches, which really slow the electrons down from the notional 2/3 c.
On the other hand, you could fit hundreds of simple CPUs on a top of the range 2 million gate FPGA.
If you're a jock, inflict some pain / If you're a nerd then use your brain - DAPHNE AND CELESTE
It might interest you to know that I don't live there. I was pointing out that one writer appears to have the advantage of first-hand observation while the other tells a horrific story with no validating information. That doesn't mean the story is false, but I hear so much garbage typed in ALL CAPS that turns out to be patently untrue that I have an aversion to unsubstantiated claims.
(And you might be interested to know that I know about the region's history back a few thousand years, including the Bible (both OT and NT) and the Koran)
Do you?
While I don't boast an extensive knowledge of the Koran, yes, I do know the basics of the history of the region.
I think you should believe someone from the outside, who has no real reason to prefer one sider over the other, and no religious reasons either ...
And who would that be? You certainly don't sound like a neutral party. Just because the other party lives there now, they grew up in the US and likely have a perspective different than those who grew up in the region. Additionally, just because someone is an Israeli, does not mean that they are religiously inclined at all. The most balanced accounts seem to indicate there is fault on both sides. This is a conflict that has lasted for centuries and I doubt either side has a valid reason to continue beyond the obsessive retaliation for the latest retaliation for the last.... well, you get the picture.
--
As a matter of fact, I am a lawyer. But I play an actor on TV.
I want to delete my account but Slashdot doesn't allow it.
Silly troll, trips are for tweekers.
Name a time that Jews didn't inhabit that region since the Maccabinian revolt in that area in 400bc (was it 400bc?) When were all the Jews kicked out (every one of them?) Its my understanding that THEIR country had Jewish settlements even at the time that they were recognized as Palestine. So why should he/she get out, oh person knowledgable in THEIR history?
I think undisclosed would be a better term than proprietary in this case.
I propose we give the palestinians atomic weapons on par with Israel, and then let nature take its course. They will either be forced to finally make peace, or at least the mideast will then be very quiet, peaceful and still for the next thousand years.
Bork! Bork! Bork!
WWJD -- What Would Jimi Do?
WWJD -- What Would Jimi Do?
(Smash amp, burn guitar, take home the groupies)
The Mercurian processor SB-1250 mentioned yesterday has a peak rate of 16 GFLOPs (with only two processors on the chip) vs 12.8 Gig Fixed point operations. This 128 way fixed point stuff is much less impressive/useful in comparison.
Classic mistake to make. Never assume that any news is any less biased than any other. Don't trust anything without questioning it, especially what is reported as "news". People are people the world over. If you see someone from country X do something shitty, you can bet it can happen in *your* country.
Rich
HOW the hell is this off-topic? it was a question about there hardware set-up! The moderators need to get there dang act together!
You should care. The x86 has had so much bolted onto it and the clock pushed up so high it's not that much of a surprise the 1.13GHz PIII ran out of steam recently. I program VLIW machines and there's some mileage left there, but I can see the writing on the wall; we need something else.
We've got at least a decade of Moore's Law left and we have to find some way of really using huge chip complexities. Putting many processors on a die is simple enough for the hardware guys (not to underestimate what they do at all). Just bloating a processor to make use of a whole chip is do-able, but what do you suggest other than tons of cache?
Figuring out how to use parallel processors is a big issue for the future IMHO. Maybe this one will bomb, but we should support their innovation.
Also, programming weird architectures is fun and teaches you stuff - as an example I went to a lecture on optimising code ar Siggraph, people liked it, the content was good, but some of the stuff was already second nature to us VLIW programmers.
Since you are quick to point out IBM deserves credit also give Infineon credit where credit's due.
What do you call a "proprietary" processor ?
As opposed to a "standard" one ? How do you define a "proprietary" processor ? This statement is simply ridiculous, and has probably been expressed by someone who doesn't know what a processor is.
-- javaDragon is an instance of JavaDragon.
And here before that even...
"don't fall into the fallacy of believing that Perl can solve social problems. Maybe Perl 6 can, but that's a ways off"
Am I correct when I read that you are comparing the 100 watts of a complete chip to the 50 picowatts of one single molecular transistor? In that case, it doesn't make sense :-) To compare fairly you have to multiply it with 25 million or so (dunno exactly how many transistors fit on current chips)
The target of this product is telecommunications. (for more info see EE Times, www.eet.com) The clock speeds and some features of the architectures are in line of embedded DSP (digital signal processors). Currently, the telecommunications industry employs racks and racks full of shelves of cards. Each card has several dozen processor cores, mostly DSP. Each core handles up to a few data streams (voice calls). Depending on where the rack is in the voice network (which is the worlds largest and most complicated computer network), it may do GSM-AMR decoding, mu-law conversion, and diversity handling, etc. As for the lower clock speeds- power consumption and cost are a big deal. When you can have a 200 Mhz DSP do the job, why got any faster?
Please read the following to understand EXACTLY what I was trying to say:
SMP (Symmetrical Multi-Processing) is typically associated with high-end operating systems such as UNIX and NT running on high-end servers. These large monolithic systems tend to be quite complex, the result of many man-years of development. Since these large kernels contain the bulk of all OS services, the changes to support SMP are extensive, usually requiring large numbers of modifications and the use of specialized spinlocks throughout the code.
Neutrino, on the other hand, contains a very small microkernel surrounded by processes that act as resource mangers, providing services such as filesystems, character I/O, and networking. By modifying the microkernel alone, all other OS services will gain full advantage of SMP without the need for coding changes. If these service-providing processes are multi-threaded, their many threads will be scheduled among the available processors. Even a single-threaded server would also benefit from an SMP system, because its thread would be scheduled on the available processors beside other servers and client processes.
As a testament to this microkernel approach, the SMP version of Neutrino adds only a few kilobytes of additional code. This version, called procnto-smp, will boot on any system that conforms to the Intel MultiProcessor Specification (MP Spec) with up to 8 Pentium or Pentium Pro processors. The procnto-smp manager will also function on a single non-SMP system. With the cost of building a dual-processor Pentium motherboard very nearly the same as a single-processor motherboard, it's possible to deliver cost-effective solutions that can be scaled in the field by the simple addition of a second CPU. The fact that the OS itself is only a few kilobytes larger also allows SMP to be seriously considered for small CPU-intensive embedded systems, not just high-end servers.
And Please visit the link http://qdn.qnx.co m/s upport/docs/neutrino_qrp/sys_arch/smp.html
This would refute EVERYTHING that you said.
Thank You.
Is there not an Alpha out there that uses this 3 cores on the chip or something?
How every version of MICROS~1 Windows(TM) comes to exist.
Do the following really mean anything? SCSA MCP CCSA CCNA
--I'm not actually after an answer!
Would it fit on my pentiummainboard???? Mark
- if you love something, set it free; if it doesn't come back, hunt it down and kill it
I thought the article looked like Deja 'vu or something like that!
Maybe Slashdot should give him a 1 week vacation to think about what he's done.
I guess the article begs the point: Why "benchmark" fixed point only? IIRC, the only major uses for super large systems is for FP operations. Decryption? I guess those are integer ops.
The other matter is that 32 processor systems are probably best off left to 64 bit systems because of the 4GB limitation.
A quick check of the other thread didn't seem to turn up these issues.
Am I the only one disappointed by people's use of "kph" instead of "km/h"? I really think the m deserves to be included. Otherwise it's just kilos per hour.
--
Fuck the system? Nah, you might catch something.
(And you might be interested to know that I know about the region's history back a few thousand years, including the Bible (both OT and NT) and the Koran)
Do you?
I think you should believe someone from the outside, who has no real reason to prefer one sider over the other, and no religious reasons either ...
it's in my head
First of all, this is not an existent configuration. It's barely even an existent plan for a configuration. Secondly, it refers to a computing device that none of us will likely ever see or have any reason to see.
You know guys, it's okay to let a few hours go by without a news story, if there's nothing really going on. I know the trolls will get lonely, but what do you care?
-in a fast german car im amazed that i survived... an airbag saved my life!-
Will Taco flame me on IRC for this? Damn, I hope so!
To this day, I cannot understand why the powers that be feel that it is beneath them to participate in the discussion threads.
I just don't get it. Oh well, another item for the list.
A now defunct company called Masspar of Santa Clara California developed a massively parallel computer based on putting dozens of CPUs on a single chip. They were trying to beat Thinking Machines, a defense department funded massively parallel company, that was looking good at the time. Masspar had a nice mchine and several dozen customers. However, as with most of the 1980s and early 90s "mini-super" business, the people who made custom CPUs and ASICs could not keep up with the commodity CPU super-clusters (ironically pioneered by Thinking Machines). At best a custom company could engineer a new generation every three years, while Intel (Sun, IBM and MIPs) come out with a new chip on an annual basis or faster. These mini-supers were often obsolete by the time they shipped.
You don't need anything so fancy.
Just analyze the responses to the editor's articles and rank them by the cumulative karma gains of all respondents. Editors whose articles generate lots of interesting, insightful and funny quotes score high, editors whose articles generate lots of flamebait score low.
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
BTW, the other reply you got was correct - you don't know the truth. You apparently don't know about Palestinian kids (age Leave Palestine to the Palestinians - they lived there before you did.
it's in my head
The ever-sexy Connection Machine used bit serial processors. At least in the first generation, get-me-a-thesis version. Later versions that had to be applicable to more than fluid dynamics used commodity cpus, I believe.
Can anyone summarize why they went belly up? Too hard to program? What was their value-added, after they moved to Centinode commodity CPU systems. Doesn't everyone and their brother have huge systems? Or was CM special in that it had a good architecture for Shared memory busses (based on the hypercube, at least in mk 1)?
Right, and neither does anybody else. Care to make up another one. Halloween is coming up.
The probable reason why this chip and most massively reconfigurable chips have low clock rates is that they have huge tangles of internal routing to allow signals to flow from any internal point to any other point. Delay in modern processors is much more dependant on interconnect (wires) than devices (transistors), so all of this interconnect they need to achieve full reconfigurability causes long critical paths and low clock rates. The trick, as pointed out by the designers of the Remove and PipeRench architectures (both purely academic research efforts so far) is to provide enough local interconnect to allow a wide array of applications to run while limitting the interconnect to a allow the clock to run more quickly.
Ben
You moderate an editor below a negative number, they are booted off the site. :)
- I don't care if they globalize against free speech. All my best free thoughts are done in my head.
the hell there won't be beowulf references!!
Wow! Could you imagine a beowulf cluster of these?!?
Lets see. There is a mention of this that is still on the front page. Uhm... you guys really need to get together and talk, or at least read Slashdot before you post new articles.
it's in my head
Overclock it. ;]
Despite the proprietary nature of this beast, the concept has not been patented (as far as I could determine in US or Europe). Hopefully, this concept will be expanded to general platforms in short order. The possibilities of load balancing low-level operations is quite interesting.
What's the name of their seti@home team? This thing could crunch data in a matter of seconds...
Hammer of Truth
it's in my head
Seriously, the editors need to start reading Slashdot more often. I know timothy posted this at 5:23 AM and he probably wasn't thinking clearly, but the quality of the articles is becoming a joke.
I have lost count of the number of duplicate articles that appeared on the same day. Or is this a side effect of Slashdot getting hacked (rouge processes non-deterministically posting articles)
Whilst I'm sure that this processor offers a lot of horsepower, the face that it will be so difficult to program is already making it look like next year's big failure. It's never easy to break into the processor market which is saturated at all levels by established products from well-known companies, and making a chip that requires new programming techniques is hardly going to make this a cost-effective solution for 99% of hardware vendors.
How fast does it chomp on a SETI Block? :)
it's in my head
I'm posting this twice.
What the company needs to do, IMHO, is find a way to smooth the programming interface so that people don't have to learn obscure propriety languages to have a usable compiler. If a mainframe can make multiple processors look like a single one to a programmer, maybe they can too!
As for process and clock speed, their low specs are sensible. It is much easier to fab a processor with low specs. The advantage this thing has should come from its massive parallelism, not from trying to beat Intel and AMD at their own game.
Now for the obligatory idiotic comment: IMAGINE a BEOWULF cluster of THESE!!!
hee hee....
"Any connection between your reality and mine is purely coincidental." -Slashdot
I'm posting this twice.
I think we should have a new moderation option - and that it should apply to stories:
Score -1, BEOWULF BAIT
So how fast can you get data on and off this thing? How big is the cache per processor? What is the bandwidth and latency? The bottleneck in most parallel tasks is inter-processor communication. Beowulf (drat - I vowed I'd never say it) is cheap because compromises have been made w.r.t. bandwidth and latency between processors - that's great, but it makes it a poor choice for some tasks. Presumeably, this thing will have other compromises, and, IMHO, until we know what they are, it's not really possible to make any kind of reasoned judgement about what it can be used for...
This is specialized hardware. NASA used to have similar beasts with 2^16 16 bit ALUs for satellite image processing. Not on one chip of course. On multiple.
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
Moore's Law happened, but it is no good for predicting the future.
Intel seem to think so - I quote from the Intel page I provide a link to:
Moore's observation, now known as Moore's Law, described a trend that has continued and is still remarkably accurate. It is the basis for many planners' performance forecasts.
26 years of evidence show the number of transistors increasing over time. It won't continue for ever, most people think we have another decade. So I'm saying we need to come up with ideas on how to best use those transistors. If we are uninspired over the next decade, there's a danger that the computing horsepower (eg MIPS or some such vague metric) won't scale.
Or are we arguing in Philosophy 101? I guarantee that the sun will rise on Friday 13th October 2000. Sigh.
Each ALU runs at 100Mhz. Why so slow ? It makes the chip much less impressive than it seems. I think that an Athalon can theoretically perform 6 integer multiply-accumulates per clock cycle. A 1Ghz Athalon then can theoreticaly perform 6 giga multiply-accumulates per second. The XPU128 theoretically perform 12Giga multiply-accumulates per second. Twice as fast. Big deal. So why is the XPU128 clock rate so slow ????????? Athalon info: http://www.azillionmonkeys.com/qed/cpujihad.shtml
Support the organizations that make up the Global Internet Liberty Campaign http://www.gilc.org/
What happened to russian merced-killer E2K?
Relinquish
Am I the only one annoyed by "Extreme" this and "Extreme" that?
"Crunch your numbers with the [bad rock starts playing, strobing colored lights] Extreme Processor Platform!!! D00D!!"
Maybe it's just me.
--
If I were a Troll Brigadier, I might seriously think about posting stories straight into comments, and then having them "voted on" by moderators. Hence (once again!) you've hijacked Slashdot.
Will Taco flame me on IRC for this? Damn, I hope so! Think of the Dark Side Geek Aura!
be well;
JC.
--
"Don't declare a revolution unless you are prepared to be guillotined." - Anon.
Classical Liberalism: All your base are belong to you.
I posted this in the last story about this chip, but I think I was too late for anyone to notice, so I'll post it again.
This is an interesting architecture, but I don't think you'll see it in a personal computer any time soon. You have to program in one of two funky dataflow oriented languages to use this beast. The company says they will have a C compiler out by next year, but the job of mapping a C program onto an architecture like this efficiently is incredibly compilcated. I would be surprised if they managed to get 1/4 of the efficiency out of a complex algorithm written in C, vs. the same algorithm written in their own programming language.
I am currently involved in some research in which we are trying to solve the same problem, roughly. However, we are trying to make the change as transparent to users (i.e. compiler writers & assembly code writers) as possible by making small tweaks to the standard RISC concept that will allow our chip to extract large amounts of parallelism. It is clear that with these reconfiguration based architectures it is possible to have huge performance gains at the expense of programming complexity. Hopefully our architecture will be to these reconfigurable systems what superscalar was to VLIW. VLIW has been around for a long time, but it wasn't popularized until architects came up with ways of making a VLIW core look like a scalar processor to the outside world.
We'll see.
Ben
He eventually went to work around Beaverton Oregon for one of those silicon foundries, but I think he got more interested in parallel, hardware regular expression evaluators.
Seastead this.
...aww, fuck it.
I don't understand the people who moderate posts like this down. Isn't Slashdot about news for nerds? And isn't it nerdy to salivate over high-end computing hardware? And if we built a Beowulf Cluster out of these bad boys with Linux as the OS, wouldn't that be extremely nerdy? Hell, wouldn't it be cool?!
I'm imagining a Beowulf Cluster of these. And I'm imagining running dozens of instances of any distributed.net client. Oh hellz yeah =)
--
Peace,
Lord Omlette
ICQ# 77863057
[o]_O
But not with the benefit of a linked article in English.
--
Chief Frog Inspector
A feeling of having made the same mistake before: Deja Foobar
Maybe I should post this as a separate thread, but your post made me think about it, so here goes.
An old saying is that if you find a way to make an O(n^2) operation into an O(n lg n) one, the world will beat a path to your door finding ways to use it. It was originally said about FFT, which is used in any number of situations where you wouldn't expect it.
Anyways, I've been wondering about cellular simulations, like fluid dynamics or nuclear modelling.
How feasable would it be to make a special purpose SIMD chip that takes a simplish formula and applies it to each a large number of cells. The driving insight behind this proposition is that a cell has fairly predicable communication needs, so that you can hard wire efficient communications, and also that a n-way multicell (as would be expressed as a unit of silicon in my proposition) has comm needs that rise as n, but contains n^2 functional units. So the bigger you can make it, the cheaper communication becomes.
So this would be a very specialised peice of computing machinery. My question to you lot is how applicable the "beating a path to your door" would be. Can cellular computing be applied to much appart from the game of life?
I remember a couple of years ago seeing something like this that was not so specialized. It don't remember exactly what it was called. Basically it contained multiple CPU layed out in a grid pattern, and switches were used to direct the flow through the chip. It could allow mapping from several entry points allowing more than one process to be going at a time. Anyone else remember what I am talking about? I can't seem to find the information anymore.
At the next eco-hypocrisy-meeting, count the private jets used to get to the meeting. Should be interesting to see that
Timothy doesn't read Slashdot
Posted by timothy on 04:20 AM April 1st, 2001
from the at-least-read-the-front-page dept.
Frac writes: "It seems rather obvious that timothy doesn't read Slashdot, considering that the an article still on the main page mentions the exact same news." Interesting stuff. And in other news, there are now proton polymer batteries available, results from ICANN elections, and a really interesting article at ZDNET on reverse-engineering.
All the sub-morons who are going to use the a certain massively parallel form of supercomputer to make a lame joke of this?
I dunno why they're using .21 micron and not a more current .18 or .15 micron fab process. It might drive fabrication cost up some, but the price should stay level due to the fact that they'll get more chips off a wafer. That and the chip would have more headroom before requiring yet another die-shrink.
Chas - The one, the only.
THANK GOD!!!
Chas - The one, the only.
THANK GOD!!!
This thing has onboard embedded linux as well as shipping with linux & NT drivers.
I haven't seen a price point for it yet but it may solve a few problems for people (and yes .. someone might be able to write an ueber-decent seti@home client for it - rather then using ex-russion missile control chips =) , ey ?)
Jon - TheSpork
I suppose you are right about commercial sucesss. There are only so many cancer treatment centers and national labs. Then again, it's hard to tell what people will do with things you give them. It takes a very special person, like Von Braun, to interst people in their toys otherwise.
Friends don't help friends install M$ junk.
> This would refute EVERYTHING that you said.
It's interesting that in the three lengthy paragraphs preceding your link you don't offer a single valid explanation for why even single-threaded apps would benefit from SMP, only strong assertions that such would be the case. In fact, the three paragraphs smack suspiciously of marketing-speak, conveniently sidestepping anything resembing an argument.
Microkernels are great, not many people would dispute that, but what the hell do they have to do with SMP per se? In order for an efficient MP architecture (be it symmetrical, distributed etc) to be effective, applications have to be coded to take advantage of it. This includes multi-threading it in meaningful ways. Just putting QNX and Neutrino on your system won't make your linear app any more efficient than running it under DOS--except for the GUI being more responsive, possibly. Just because system resources are multi-threaded doesn't mean squat--your app is still sitting around waiting for disk-bound operations to complete, no matter how those are implemented.
Unless you know how to effectively multi-thread your app, neither QNX nor BeOS will do much to improve your code. You can make any assertions and include any links to the opposite you want, but that's a pretty fundamental tenet of current OSs.
I was kind of looking forward to more and easier to find information on the big chip here. Instead, I find all of these complaints which are even less interesting than other chip news.
Friends don't help friends install M$ junk.
This wouldn't necessarily work; consider repeat articles in the past, which have gotten lots of posts and karma. Also consdider that it is the editors that people hate (e.g. John Katz) that consistently get the most posts, and hence karma.
All the cumulative karma would tell you is the number of people interested in the story...it wouldn't give any indication of whether that interest was positive or negative.
Ahh, but do you remember what Transmeta did with thier curusoe chip? You could easyly put a shell arround the mulitple core CPU to make it appear (to the OS and any software) as 1 very fast chip... Just a thought... Microft
Compared to the power consumed by many boxes, a cluster of these things will be a very useful tool. I can imagine 10 of these in a single box doing great work. It will be very nice when they scale up the clock speed.
Friends don't help friends install M$ junk.
It is in the parlance of research one area that I can personally re-assure you as totally solvable and surprisingly simple in certain ways.
Perhaps you will come up with your own elegant solution like the 'layer cake' one Sakharov did after knowing about yet another way to get to the same smashing result?
Yes I have considered going open design and open source with a POST-VLIW architectural solution from my 80's research. Any suggestions on others that have gone that way for me to study is appreciated.
http://www.aisnota.com/slashdot/ Welcome to Logic and the Future
IBM is using a fabrication process that allows them to synthesize fast, highly dense DRAM and fast, highly dense logic on the same die. To date, they are the only company capable of doing this economically and effectively. This results in extremely fast interconnect between the processor and main memory, since they don't have to boost a signal off-chip. It also makes it trivial to create extremely wide main memory busses with high reliability and low cost (1024 bits wide is, iirc, what BlueGene uses). The latency to access a memory page for the first time (the DRAM's internal pages, not the processor's virtualized memory pages) is around 50ns for commodity SDRAMs, but only 10ns for IBM's EDRAM. For a 1GHz processor with reasonable (compiled) ILP, the difference is easily more than 100 instructions' worth.
Furthermore, IBM is using a new, highly dynamic interconnect which virtualizes the processor interface so that processors that go bad are automatically detected and remapped, much like EIDE drives' bad sectors, so that a good processor takes its place. This is very important because managing the upkeep of this many processors is a significant problem. If losing a processor posed any sort of significant loss of work, then even if the MTBF of processors was very, very high, upkeep would become a significant bottleneck on the system's work capacity.
Give IBM due credit. They have lots and lots of money, but the Justice Department has made it clear that if they use that money to make money in the usual ways (eg: buying other companies, defending patents, et al) they'll come down on them like a ton of bricks. IBM's too vast to spend the money in the same way a small company can, and they're already at the point of diminishing returns on advertisement and other marketing, so they're shoving huge volumes of cash through research labs, funding hard-core basic research, much like Xerox in its heyday. They have been analyzing problems and finding solutions that won't be profitable for years, but since they're one of the few entities doing it, they'll be one of the few entities with the research already done to come out with the hot products of the year 2020. The best voice recognition software on the market today is derived from research IBM has been funding since the 1960's. IBM isn't the only one researching quantum computing, or magelectric memory, or nanotechnology, but as far as I know they are the only ones researching all of these technologies, which lends them many financial and scientific advantages.
As I said, give IBM due credit. In some respects they are the epitome of the BigEvilCorporation(TM), but they are also a world-class powerhouse of technological and purely scientific innovation.
-- Guges --
This souns a lot more impressive than it really is. Getting all these processors to actually do something useful isn't easy. The posted maximum number of operations/second is for pure calculations only. A lot of real life applications, even if they are scalable to a large number of processors still need to have access to large data structures.
Having to use a special language is going to shy people away as well. Anybody remember transputers ? Occam ?
These processors, made by small start-ups also have lots of real-life issues that need to be solved. For example, you also need a good motherboard, with appropriate chip sets to access peripherals (network/hard disks) as well as a high-bandwidth memory bus. Who's going to make those ? Also, what's the quality of the development tools, like the compiler? Even if all those obstacles are overcome, users still need to spend time and money to get acquainted with this platform, and they are risking that the manufacturer(s) will be out of business a couple of years later.
For applications requiring more than just pure calculations, it's not going to be easy to offer a solution that offers users more value for money than a bunch of networked SMP machines based on off-the-shelf hardware, and using development tools that they are already used to, and can be assumed to be bug-free.
I'd like about 5000000 of these for a beowulf cluster. Anyone got a couple extra?
Moore's Law states that every few years the number of transistors able to be fit on the same amount of space doubles. Moore's Law is one of the main reasons for technological trend up until now. Unfortunately, this trend will most likely cease by the year 2015 and advancements will be extremely slow and costly. Never fear, there is hope. Through a field called Molecular-Scale Electronics, scientists are attempting to develop chips the size of, well, molecules! Each individual transistor on the cheap is one molecule large but that's not all. The amount of heat outputted by normal chips is around 100 watts. The amount of heat from molecular transistors is about 50 picowatts, which is much much smaller (about 50 millions of a millionth of a watt).
--
I watched C-beams glitter in the dark near the Tannhauser gate.
Before you go off on tirades, maybe analyse what you're about to say first. First, there's nothing magical about QNX or BeOS. Sure, BeOS apps may be automatically multithreaded, but only in a relatively superficial way (by putting the message loop into a separate thread if I'm not mistaken). BeOS certainly doesn't take your linear code and somehow magically extract multiple threads of execuion--you still have to do the leg work. Same with QNX. Same with WinNT/2K, if you multi-thread you apps, they will spread quite nicely across CPUs (though HOW nicely is a matter of debate for the religious). Don't get me wrong, I think what BeOS is doing is perfectly fine and laudable, but it's not exactly what you're implying.
The fact is, in you model the unit of execution is the thread. What you yourself don't know how to de-serialise and pull into separate threads (and properly synchronise), the compiler or OS certainly won't do for you. So even if your app is multithreaded, if your threads are big fat chains of serial code, the app won't benefit any from multiple CPUs. The holy grail of MP is a compiler that could, for a trivial example, look at your loop and be able to unroll it into x smaller loops working on subunits of the data.
One of the big promises of MP is to avoid the end of Moore's law through parallel computing, IN GENERAL PURPOSE COMPUTERS, not just arcane research machines. We already know how to tickle esoteric MP hardware today into doing our bidding, but it's no trivial task and takes a lot of skill. If MP is to give us a mainstream migration path from single processing, it can't expect more from programmers than they can give today. In other words, MP machines will have to deliver even with mediocre programmers, because they form the bulk of the work force. You can't stipulate as a condition for effective MP an overall higher quality work force, because it ain't happening.