Beta

Slashdot: News for Nerds

×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Tilera Releases 64-Way Chip Dev Tools

ScuttleMonkey posted more than 6 years ago | from the another-win-for-the-good-guys dept.

Hardware 72

eldavojohn writes to tell us that Tilera has released a Linux-based development kit for their 64-core system on a chip. "The Tile64 is based on a proprietary VLIW (very long instruction word) architecture, on which a MIPS-like RISC architecture is implemented in microcode. A hypervisor enables each core to run its own instance of Linux, or alternatively the whole chip can run Tilera's 64-way SMP (symmetrical multiprocessing) Linux implementation. An 'iMesh' switching interconnect, developed by Tilera's founder, MIT professor and serial entrepreneur Dr. Anant Agarwal, is said to eliminate the centralized bus intersection that limited scalability in previous multicore designs."

cancel ×

72 comments

64 ways (0)

Anonymous Coward | more than 6 years ago | (#23259694)

I used the 64 way chip to get first post!

Re:64 ways (0)

Anonymous Coward | more than 6 years ago | (#23260010)

Oh really, because I can't find your other 63 posts.

Yummy chips (1)

mrbluze (1034940) | more than 6 years ago | (#23259774)

I used the 64 way chip to get first post!
Which of the 64 ways did you use? And did you have fish with the chips?

Hope it wasn't released under the GPL (-1, Troll)

Anonymous Coward | more than 6 years ago | (#23259780)

As a consultant for several large companies, I'd always done my work on
Windows. Recently however, a top online investment firm asked us to do
some work using Linux. The concept of having access to source code was
very appealing to us, as we'd be able to modify the kernel to meet our
exacting standards which we're unable to do with Microsoft's products.

Although we met several technical challenges along the way
(specifically, Linux's lack of Token Ring support and the fact that we
were unable to defrag its ext2 file system), all in all the process
went smoothly. Everyone was very pleased with Linux, and we were
considering using it for a great deal of future internal projects.

So you can imagine our suprise when we were informed by a lawyer that
we would be required to publish our source code for others to use. It
was brought to our attention that Linux is copyrighted under something
called the GPL, or the Gnu Protective License. Part of this license
states that any changes to the kernel are to be made freely available.
Unfortunately for us, this meant that the great deal of time and money
we spent "touching up" Linux to work for this investment firm would
now be available at no cost to our competitors.

Furthermore, after reviewing this GPL our lawyers advised us that any
products compiled with GPL'ed tools - such as gcc - would also have to
its source code released. This was simply unacceptable.

Although we had planned for no one outside of this company to ever
use, let alone see the source code, we were now put in a difficult
position. We could either give away our hard work, or come up with
another solution. Although it was tought to do, there really was no
option: We had to rewrite the code, from scratch, for Windows 2000.

I think the biggest thing keeping Linux from being truly competitive
with Microsoft is this GPL. Its draconian requirements virtually
guarentee that no business will ever be able to use it. After my
experience with Linux, I won't be recommending it to any of my
associates. I may reconsider if Linux switches its license to
something a little more fair, such as Microsoft's "Shared Source".
Until then its attempts to socialize the software market will insure
it remains only a bit player.

Thank you for your time.

Re:Hope it wasn't released under the GPL (0)

Anonymous Coward | more than 6 years ago | (#23259814)

WTF is this? Smells like a Microsoft marketing asshat. Mod this shit down.

Re:Hope it wasn't released under the GPL (0)

Anonymous Coward | more than 6 years ago | (#23259832)

You and your lawyers are idiots.

EOF

Re:Hope it wasn't released under the GPL (1)

mrbluze (1034940) | more than 6 years ago | (#23259834)

I might be a little inaccurate about this, but AFAIK nobody forces you to GPL license your kernel patches, but you are forced to tell people that you used a GPL'd kernel in the first place and tell them where they can get it for free.

Serves you right if you are trying to make a buck off other people's hard work. Running software (including kernel drivers) on Linux DOES NOT require you to release your source code. Where's the source code for Googleearth, or VMWare, or NXClient/Server or the nVidia proprietary drivers? Where's the source code for your router? Exactly.

I don't know who you have for a lawyer, but I can't see how your case is real. Nobody else is seeming to be having this problem.

Re:Hope it wasn't released under the GPL (1)

zippthorne (748122) | more than 6 years ago | (#23259898)

Clearly some kind of troll.

Note the use of the words:

Recently, ext2, defrag, and.. token ring??

Was that even still around when a student named Linus messaged some newsgroups describing his latest little hobby?

Re:Hope it wasn't released under the GPL (1)

mrbluze (1034940) | more than 6 years ago | (#23259974)

Note the use of the words: Recently, ext2, defrag, and.. token ring??
At first when I saw the formatting, I thought the self-inflicted linebreaks were there because the poster was about to write a poem or something beautiful, but it really was diarrhoea with a dash of sulphur. A total waste of time.

Re:Hope it wasn't released under the GPL (1)

The Analog Kid (565327) | more than 6 years ago | (#23259998)

Was that even still around when a student named Linus messaged some newsgroups describing his latest little hobby?
Well when my brother went to Marist college in 1998, they were still using token ring, so I'd say yes.

Re:Hope it wasn't released under the GPL (1)

Thumper_SVX (239525) | more than 6 years ago | (#23262582)

Given that I was ripping out token ring from banks as recently as 2005... yeah. Big organizations like that don't move quickly, and TR was the default (and at one point only) option available for connecting AS/400 and Mainframe systems to a LAN. It was only in the mid 90's that it started to all switch to Ethernet, to the sound of much wailing and gnashing of teeth from the Mainframe and 400 guys and gals.

Also bear in mind, many mainframes and 400's installed at the end of this TR era may actually still be in use today in many locations... some of them so loaded up with cards that they are incapable of accepting an Ethernet interface. In our client/server-centric world we sometimes forget that there are systems out there with a much longer shelf-life than the 3-5 year x86 system running Windows or Linux.

How did this work with client machines that only had Ethernet? Quite well. TCP/IP is a very versatile protocol, quite happy to go across a set of routers from one architecture to another. As a result, even up until 2005 I touched on Token Ring systems in datacenters that used routers and gateways to get to their Ethernet-connected clients.

Re:Hope it wasn't released under the GPL (1)

aliquis (678370) | more than 6 years ago | (#23261362)

Not mine, but for some peoples routers it's here:
http://openwrt.org/ [openwrt.org] ;D

Re:Hope it wasn't released under the GPL (1)

Profane MuthaFucka (574406) | more than 6 years ago | (#23259892)

An investment firm that obviously has no intelligence? OMFG this is a dream come true!

Gimme a dollar!

Re:Hope it wasn't released under the GPL (1)

budword (680846) | more than 6 years ago | (#23259904)

OK troll. First, you only have to give the source you wrote to people you distribute binaries to. No one else. Of course, they can give them to anyone they want to. Second, you don't have to give the source out to anyone at all IF you don't distribute binaries. So if it's just for in house use, you are free and clear, you don't have to give back to the people you took from. You are lucky you posted as an AC. Who the hell would hire you if you take code without knowing even the basics of the license you got that code under ? Then you hire a lawyer who gets the basics wrong too ? Hey buddy, it would have been cheaper just to read the GPL yourself. Oh, that's the General Public License. Not "Protective". Get the basics right, then complain.

Re:Hope it wasn't released under the GPL (1)

XaXXon (202882) | more than 6 years ago | (#23259986)

Not necessarily. Only if you distribute ALL your binaries with source do you not have to provide source for ANYONE on demand. If you allow written requests for source, you must allow the source to be obtained by ANYONE.

http://www.gnu.org/licenses/gpl-faq.html#WhatDoesWrittenOfferValid [gnu.org]

Re:Hope it wasn't released under the GPL (1)

McCarrum (446375) | more than 6 years ago | (#23259920)

Would you like a little FUD with your whine?

In the odd, small, tiny chance that you're not some idiot with an agenda (the wwworld is fast filling up with posts like this), then your lawyers are the ones with an agenda and/or incompetence.

According to your incorrect interpretation, a vast majority of internaly held code that has been written the world over is illegal. Lets take one simple example, Google. Do you honestly think that all of Googles code is released into the open?

This post is mostly pointless. I seriously doubt you'll ever be back to this thread, AC. However, various web archives will be around for a long time, and for each example of FUD that's shoveled down the collective pipe needs a rational "WTF" post after it to point out just what it is.

Enjoy your Windows 2000 solution.

Re:Hope it wasn't released under the GPL (1)

nycguy (892403) | more than 6 years ago | (#23261894)

According to your incorrect interpretation, a vast majority of internaly held code that has been written the world over is illegal. Lets take one simple example, Google. Do you honestly think that all of Googles code is released into the open?

Actually, that may well be coming in the next version of the GPL. For now, it's a separate license [fsf.org] , but I suspect future versions of the GPL will be such that if you write any application based on GPL'd code which provides a service to anyone but yourself, you must provide the users of your service with the source code for that service. This particularly hits Google, with its web services, but I suspect it will be expanded to include any service (e.g., a mail server).

Perhaps more interesting will be when the FSF guys start to address the "content loophole." I.e., if you create content with GPL'd software (or with software derived from GPL'd code), you must provide the source code for the software when you distribute the created content. If I put on my "freedom" hat, that seems like a perfectly rational thing. (If, for example, I wrote a GPL'd 3-D modeling program and some game company modifies it and uses it to help create their next blockbuster, I'd at least like to get the source code for their improvements, even if I don't get a cut of the revenue from the game.) If I put on any other hat (or no hat at all), it seems bonkers, though.

Re:Hope it wasn't released under the GPL (1)

snl2587 (1177409) | more than 6 years ago | (#23259962)

Furthermore, after reviewing this GPL our lawyers advised us that any products compiled with GPL'ed tools - such as gcc - would also have to its source code released. This was simply unacceptable.

Now that is an outright lie. Do you realize how many commercial programs were compiled with gcc? The only time you must release your source is if you were linking GPL'd libraries...and at that, the legal issues are still unresolved.

Re:Hope it wasn't released under the GPL (1)

zeroduck (691015) | more than 6 years ago | (#23259992)

Although we met several technical challenges along the way (specifically, Linux's lack of Token Ring support and the fact that we were unable to defrag its ext2 file system), all in all the process went smoothly. Everyone was very pleased with Linux, and we were considering using it for a great deal of future internal projects.

Come on guys, this isn't a troll. This is just hilarious.

Re:Hope it wasn't released under the GPL (1)

Tangamandapiano (1087091) | more than 6 years ago | (#23260038)

Furthermore, after reviewing this GPL our lawyers advised us that any products compiled with GPL'ed tools - such as gcc - would also have to its source code released. This was simply unacceptable.
Just to exorcise this insane statement, please, take a look at this [gnu.org] .

Re:Hope it wasn't released under the GPL (2, Informative)

Mr. Roadkill (731328) | more than 6 years ago | (#23260096)

A quick giggle for "or the Gnu Protective License" barfed up the following:

http://www.news.com/5208-1030_3-0.html?forumID=1&threadID=2246&messageID=11919&start=-1 [news.com]
http://linux.slashdot.org/comments.pl?sid=389856&cid=21705136 [slashdot.org]
http://yro.slashdot.org/comments.pl?sid=67877&no_d2=1&cid=6220788 [slashdot.org]
http://slashdot.org/comments.pl?sid=159323&cid=13343214 [slashdot.org]
http://www.kuro5hin.org/comments/2003/2/13/8422/16656/11#11 [kuro5hin.org]

I think the biggest thing keeping this troll from being truly informative is the lack of understanding of the licence, and the deliberate mis-statement of its effects. Its fictitious and incorrect pronouncements virtually guarentee that nobody with even rudimentary analytical skills will believe it. After my experience with these beliefs, I won't be recommending them any of my associates. I may reconsider if it switches to something a little more believable, like the HIV-protective benefits of nailing your head to the floor. Until then its attempts to deliberately distort the facts about what you can and cannot do with in-house software that's not for external distribution shall continue to attract such a flurry of indignant responses that it's easy to believe that Mother Henrietta Hickey's day job is posting anti-GPL FUD.

Thank you for your time.

Re:Hope it wasn't released under the GPL (1)

Thumper_SVX (239525) | more than 6 years ago | (#23262710)

A quick giggle for "or the Gnu Protective License"...
Methinks someones spell-check database stopped learning around 1998...

Re:Hope it wasn't released under the GPL (1)

Mr. Roadkill (731328) | more than 6 years ago | (#23297756)

Methinks someones spell-check database stopped learning around 1998...
Methinks my fingers went onto autopilot, and followed the posting conventions of one of the Usenet groups I frequent.

Lbh fubhyq or ganaxsgy V gvga'g ebg13 gur jubyr cbfg gb erqhpqr gur evfx bs hajnagrq nggragvba - that would have been really annoying to casual readers, if it made it past the lameness filter.

Re:Hope it wasn't released under the GPL (0)

Anonymous Coward | more than 6 years ago | (#23260108)

Read all the sibling posts above this one. What a fucking brilliant troll & what a bunch of fucking idiots.

Look, folks, a real-live alien! (1)

Jane Q. Public (1010737) | more than 6 years ago | (#23260116)

I don't have a hard time believing that this person consults for large companies, because I have met a number of totally clueless people who made big bucks consulting for large companies. That part is quite believable. But really...

(1) If I were a client, and your company put linux to use for me without researching the licensing well in advance, then your firm lost me money. I don't even need any hard numbers to know that. If you tried to charge me a dime I would mark the contract "loser" and start a lawsuit over payment for services not rendered.

(2) If your lawyers really thought that code compiled with gcc had to be made public, then your lawyers are also from another world. Even if they did not know computers very much, they should understand the basic concept that something made with a tool -- a lathe for example -- does not belong to the tool manufacturer. The same is true of software: someone cannot hand you a tool for "free", then insist that they have legal control over a product you build with that tool. The very idea is ludicrous. If anyone even tried to do that, nobody would use the tool! The majority of linux users are NOT idiots. However, the legal team you describe seems to be a whole pack of them.

(3) Just about anything you wanted to do would be at least POSSIBLE to do without modifying the kernel, though it might not be as easy or perform quite as well. Provision for installable file systems and drivers is built in.

(4) Fuck it. I am tired of explaining why just about everything you say simply ain't so. You are either a troll or one of those annoying people who are well-paid for being clueless. Either way, I would appreciate it very much if you would go away.

about the lathe (0)

Anonymous Coward | more than 6 years ago | (#23261114)

The lathe does not copy itself into what it works on.

There was a case of a pattern-lathe and it was a condition of sale of the patterns that you don't duplicate the patterns!

So while your point is good, the lathe analogy doesn't make it for you.

Re:about the lathe (1)

Jane Q. Public (1010737) | more than 6 years ago | (#23267128)

Linux also does not self-replicate. But in any case, I think you misunderstood me. I was referring to tools like gcc, which are used to compile (build) other programs or tools, not copies of itself.

Re:Look, folks, a real-live alien! (1)

Skjellifetti (561341) | more than 6 years ago | (#23316540)

There are intermediate cases such as GNU bison which inserts a small stub into the output. Because of that stub, it used to be that code that used bison as its yacc was required to be put under the GPL. That changed some time ago, IIRC, but it does show that software tools are not quite as black and white as your lathe example.

Re:Hope it wasn't released under the GPL (0)

Anonymous Coward | more than 6 years ago | (#23260244)

epic troll is epic

Re:Hope it wasn't released under the GPL (0)

Anonymous Coward | more than 6 years ago | (#23261942)

Anyone else notice, Copypasta brand Trolling bait seems to catch a lot of fish lately?

Can you run Windows Vista on it? (0)

Anonymous Coward | more than 6 years ago | (#23259804)

Because if you can't, I'm honestly not interested.

Re:Can you run Windows Vista on it? (1)

The Mighty Buzzard (878441) | more than 6 years ago | (#23259956)

Well, I suppose Vista certainly is enough of a strain on hardware to be a fair benchmark.

serial enterpreneur (0)

Anonymous Coward | more than 6 years ago | (#23259830)

An "iMesh" switching interconnect, developed by Tilera's founder, MIT professor and serial entrepreneur Dr. Anant Agarwal, is said to eliminate the centralized bus intersection that limited scalability in previous multicore designs

Thats GREAT! I used to wonder why the bus used to be late always for my cereal breakfast.

CISC? (1, Informative)

friedman101 (618627) | more than 6 years ago | (#23259844)

I'm not sure about this particular chip but doesn't VLIW normally mean no microcode? Microcode is the set of RISC commands that make up a CISC command. VLIW is just a RISC machine in which the compiler does all the optimization (branch prediction, hazard detection, etc). Normally VLIW machines fetch multiple instructions at once and issue them without fear of any hazards because the compiler takes care of it. Very neat idea, but not CISC

You're mixing implementation and architecture (1)

argent (18001) | more than 6 years ago | (#23261320)

CIISC and RISC describe the architecture. Either can be implemented directly, using horizontal or vertical microcode, or via a translator. RISC is similar to vertical microcode, where each micro-instruction controls part of the core, and VLIW is similar to horizontal microcode, where each micro-instruction controls a number of components at once. Whether you call them "microcode" or "RISC/VLIW" is almost more a matter of marketing at this level, like when Intel started talking about the 486 having a "RISC core". Just don't try and blend them or you're setting yourself up for an EPIC fail.

The way they're describing it, the RISC instruction set is implemented using horizontal microcode that they're calling "VLIW" because it's sexier right now.

Re:You're mixing implementation and architecture (0)

Anonymous Coward | more than 6 years ago | (#23290560)

In Soviet VLIW, you will have many subcodes NOPs in the bigger program and you will need bigger instruction caches!!!

EPIC is the best option in this rivalry.

--- Pipelined, Superscalar, Vectorial, Out-of-Order, Speculative ---

No (1)

dreamchaser (49529) | more than 6 years ago | (#23261472)

No. All processors have microcode, including the most infamous VLIW chip the Itanic. You're confusing yourself.

Re:No (1)

drinkypoo (153816) | more than 6 years ago | (#23262514)

All processors have microcode? Is this true? I thought some were pretty mechanistic little devices (moreso in history.) Today, of course, everything is pretty complex and there is almost nothing purely RISC or CISC (not sure about VLIW-land.)

Re:No (1)

dreamchaser (49529) | more than 6 years ago | (#23262804)

Actually you're probably correct. I should have said "all modern general purpose CPU's." My point was that VLIW style chips still use microcode. Different things.

Oblig. (2, Funny)

r_jensen11 (598210) | more than 6 years ago | (#23259894)

Imagine a beow... oh, never mind....

Re:Oblig. (0)

Anonymous Coward | more than 6 years ago | (#23260122)

Does it run Lin.....
Yes, yes it does.

*sigh*

Re:Oblig. (1)

jd (1658) | more than 6 years ago | (#23260154)

Actually you might want to imagine a Beowulf cluster of them. You'd more likely run MOSIX or OpenMOSIX on the chip to handle the migration from logical machine to logical machine on the same chip, but you can't then use that to build a cluster as you'd get a conflict between the cluster and cluster-of-clusters. SMP and virtualization might work, but there's minimal messaging between nodes and SMP technology doesn't scale well beyond 16 nodes, so a clustering technology seems a more logical way to go.

Re:Oblig. (0)

Anonymous Coward | more than 6 years ago | (#23264716)

You can still beowulf multiple systems running this chip and these board together. Nothing preventing that as a way to get high performance. They even give you 2 10G ports or up multiple 1G ports to make it simpler (more ports means less switches needed).

Re:Oblig. (0)

Anonymous Coward | more than 6 years ago | (#23265426)

You can scale shared-memory technology up pretty well -- Altix scales to >1k CPUs, for example. But then you lose the "symmetric" part of "SMP" (it's NUMA with a vengeance), and the best-performance coding style looks more like distributed-memory programming.

"Cluster of SMP" is a standard supercomputer architecture. This would be just an extreme version.

Released? (1)

Wesley Felter (138342) | more than 6 years ago | (#23259918)

I thought the chip has been out for a few months; I feel bad for anyone who has been trying to use the processor with no development tools.

Also, it looks like the tools have been released to Tilera's customers, not to the public. A shame, since I'm sure a lot of Slashdotters would like to at least gawk at the docs for this chip.

How much for one? (0, Offtopic)

rossz (67331) | more than 6 years ago | (#23259942)

I hate press releases like this. "$435 in 10,000 quantities". How much for just one of the damn things with the software I need to fool around with it? Is that too much to ask? It would be cool to have one of these, but I'm sure it won't be worth the effort or the expense (I figure a single unit will be at least double the quantity price).

Re:How much for one? (0)

Anonymous Coward | more than 6 years ago | (#23259970)

I dislike price announcements like this too. However, when you hear things like this it is because the company does not sell directly to the public. There is no way for the company to reliably know how much its distribution channels are marking-up the product by the time it gets to an end-user, so they can only announce the volume-prices they know.

Re:How much for one? (2)

cheater512 (783349) | more than 6 years ago | (#23260034)

Wouldnt you be looking for a development board for it?

Re:How much for one? (2)

42forty-two42 (532340) | more than 6 years ago | (#23260168)

It *is* a dev board. The whole thing is a PCIe riser card, so you can upload firmware from the host machine etc.

Re:How much for one? (1)

quarrel (194077) | more than 6 years ago | (#23260070)

I hear what you're saying - it is sort of annoying.

However, a VC funded startup like this has to be focused. They're going to have a list of customers probably about 10 companies long that they want to sell to, and everyone else at this stage is a distraction. Your $1000 does nothing for them, you and your 2000 friends barely helps.

They're going to want Cisco, Juniper, Nortel, Lucent-Alcatel, 3Com, Huaweii, maybe 1 or 2 of the big telcos and that's about it (ok, there are a couple of other big-iron vendors, then the odd niche players like an F5 or so, but it really is limited). They get designed in to core routers at Cisco and they know they succeed.

The interesting thing here will be if they can dislodge the cutting edge in multi-core MIPS CPUs that are currently going in to the current/next-gen products of networking vendors from Cavium (their Octeon line) and RMI (their XLR).

--Q

Re:How much for one? (1)

emjay88 (1178161) | more than 6 years ago | (#23260452)

This may be a stupid question, but what does $435 in 10,000 quantities even mean?
I either get 10,000 of them for $435 (working out to about 4 cents each) or they cost $435 each with a minimum buy of 10,000 (total spend: $4,350,000). These seem incredibly cheap and incredibly expensive respectively. Am I interpreting it wrong?

Re:How much for one? (1)

Random Destruction (866027) | more than 6 years ago | (#23261622)

they cost $435 each with a minimum buy of 10,000 (total spend: $4,350,000).
that one

Re:How much for one? (1)

jandrese (485) | more than 6 years ago | (#23269480)

$435 each, and that's not horribly expensive when you consider where they're coming from. In fact for a first gen device like this from a startup it's downright inexpensive.

Licensing containment barrier?? (2, Interesting)

sweet_petunias_full_ (1091547) | more than 6 years ago | (#23259954)

I went on to read their in-depth article (linked to the main article) at http://linuxdevices.com/news/NS8981295285.html [linuxdevices.com] and I found this:

"Another touted benefit is the ability to consolidate control- and data-plane functions on a single device, with "solid-wall" processor boundaries reinforcing security and licensing containment barrier. In this regard, the Tile64 chip resembles another heavily multicore MIPS64 chip, Cavium's 16-way Octeon."

Does anyone know what the heck a "licensing containment barrier" is? It definitely sounds like a performance hit if it's turned on. And if it's forced to have it on then this design just lost a lot of its sex-appeal.

Re:Licensing containment barrier?? (4, Insightful)

quarrel (194077) | more than 6 years ago | (#23260136)

> It definitely sounds like a performance hit if it's turned on.

No, if anything they pitch it as a performance gain.

The idea is to run Linux (or OS of your choice), with various control plane functions (it can have an IP address, you can do config, stats collection etc etc), on say (making up these numbers!) 4 of the cores, while the other 60 cores are running without an OS (they offer a BIOS like environment with basic functions to get access to the backplane and subsequently the packets) doing data-plane functions, perhaps doing deep-packet inspection for QoS delivery, security functions (IPS?) etc.

The Linux side, particularly the kernel isn't going to contain your real IP, while the data-plane side is all your secret-sauce. It involves embedded style programming without lots of OS support, but you get speed and the networking vendors are used to this sort of model - it sure beats the hell outa doing it in an ASIC on the dataplane side which is what they're used to.

This isn't an attack on open source - it's using it in a sensible fashion IMO. However, for the paranoid types who've seen the fud, they probably pitch this split of operations as a "licensing containment barrier" cause a marketing person thought it might help somewhere.

--Q

Re:Licensing containment barrier?? (1)

jd (1658) | more than 6 years ago | (#23260562)

It's like an antimatter containment field - circular and serves the purpose of being dramatic on television.

Jargon overload! (1)

wvmarle (1070040) | more than 6 years ago | (#23260234)

Can anyone please translate this to Layman? I mean I do know bits and pieces about computing but this is really unintelligible for anyone but maybe hardware engineers.

Re:Jargon overload! (1)

story645 (1278106) | more than 6 years ago | (#23260422)

Dunno-I think it'd be more coherent if I hadn't flaked on my organization & architecture class (where we used MIPS), but still-eek.

The idiots version: (with help from wiki) This chip carries out one instruction [wikipedia.org] that's really filled with tons of different instructions-(basically whatever can fit in 48 bits-and because it's treated as one instruction-instantish parallism) and does so using assembly code similar to MIPS assembly. Then it offers two options for a core running it's own instance of linux- vm [wikipedia.org] or Tilera's custom linux. More chips can be added 'cause of a switch that treats the processors as a mesh and routes data that way-instead of the usual method of all data having to pass some central location before it gets where it needs to. (Reduced traffic flow by switching out a major intersection for lots of side roads.) I have no idea if any of the above is right-pure ad-hoc translation.

Simple version. (4, Informative)

jd (1658) | more than 6 years ago | (#23260628)

They have set up an 8x8 grid of processors, not unlike a chessboard. Each square on this grid can talk only to adjacent squares (up, down, left, right), with the edge squares connecting to I/O devices. They refer to their network as a mesh, but the correct term for this design is a Manhattan Network. This is not significantly different from a processor I dearly loved in the late 80s, the Transputer. That, too, had 4 connections from each processor, but you were not restricted in how you connected the Transputers together. A grid, it transpired, was not efficient, you needed to arrange the connections to form a hypercube. (Yes, it's 2D, so it's actually a 2D representation of a hypercube. Now stop fussing or I won't get you that Beowulf cluster for Christmas.)

I like the idea, I like the idea a lot, but the fact that they opted for a simple but slow topology doesn't fill me with hope. Especially as they suggest running SMP over it. Processors close to the centre of the "mesh" will be resource-starved. There needs to be strong affinity of a given thread to a given core, where the weighting is by the operations expected and where that weighting can (and will) shift as code blocks change or new threads start. In other words, you want something that is semi-static, semi-dynamic according to need. Only the OS is capable of obtaining that kind of information, so it is the OS that needs to do the dividing, NOT the architecture underneath OR a system administrator.

Re:Simple version. (2, Informative)

Anonymous Coward | more than 6 years ago | (#23260986)


I like the idea, I like the idea a lot, but the fact that they opted for a simple but slow topology doesn't fill me with hope.

hypercubes were great in the 80's when everything was multiprocessors and wire lengths didn't kill you. But it turns out that low-dimension networks (ie, a 2d grid) are faster for a network of cores fabbed onto a single processor. while you can decrease the number of jumps with hypercubes, you increase the amount of wiring (and the length of wires) that goes on the chip when you had more dimensions to your network. There are more variables that go into designing a network than just the number of hops a message has to take. So I'd hardly call it a slow topology. That, or Intel, Tilera, and many others are all screwing it up. Keep the hope up! =)

Re:Simple version. (1)

smallfries (601545) | more than 6 years ago | (#23261448)

Very lucid description. The other problem with the design is that you don't get what you expect; using a simple 4-way grid should give predictable latency costs between nodes. Unfortunately their routing algorithm is non-predictable so you can't statically schedule threads at compile-time to feed each other, it all has to use dynamic control-flow. Shame really.

If you liked the transputer then you should look at its other descendant [xmos.com] that is in the process of coming to market. There isn't a wealth of public information yet although they are in the process of releasing dev-tools and simulators. The largest chip only has 4 tiles, instead of the Tilera's 64, but it is aimed at the low power market. They should scale it up to similar levels without the silly amount of power the Tilera draws.

Re:Simple version. (0)

Anonymous Coward | more than 6 years ago | (#23267464)

There actually is a static network on the Tile64. You can use it to get really low (as in, 1-cycle) latency between adjacent tiles.

Re:Simple version. (1)

smallfries (601545) | more than 6 years ago | (#23268784)

I know, I've followed their research. But unfortunately the fact that the network is static does not (in their case) give static routing costs. It can be as low as one cycle - but you cannot rely upon it.

Re:Simple version. (1)

hackstraw (262471) | more than 6 years ago | (#23262448)

I like the idea, I like the idea a lot, but the fact that they opted for a simple but slow topology doesn't fill me with hope. Especially as they suggest running SMP over it. Processors close to the centre of the "mesh" will be resource-starved.

These chips seem to be designed for specific applications, not as a general purpose CPU, especially in the DSP and digital video markets. I found this [design-reuse.com] and this [arstechnica.com] .

I don't see these chips as being that revolutionary or anything. Yes, they are similar to the transputer, and somewhere I read where they have a "revolutionary" shared L3 cache, which AMD ships today, and Intel either has them or are shipping soon as well.

These things seem pretty cool, but I see these as having a limited range of usage.

But, they run Linux!

I've been looking for a new pc (0)

Anonymous Coward | more than 6 years ago | (#23260318)

Wish 1.
It's less than $4000.
Wish 2. It runs games well.

Wish 3. It could plug into my AGP slot 3

Really? Can I buy one unit today? (1)

itsybitsy (149808) | more than 6 years ago | (#23260330)

The press release, oh pardon me, the article linked to in the posting above lists pricing in units of 10,000. I want one or two TILexpress-64 boards please, not 10,000 units. Until the software is built 10,000 units would just sit in my garage doing nothing.

Also, I guess they had to put out a press release to respond to the massive threat that NVidia's new Tesla board represents. At least this is going to be good for competition.

Re:Really? Can I buy one unit today? (0)

Anonymous Coward | more than 6 years ago | (#23260550)

Ever heard of a sample? Cretin.

VLIW (2, Funny)

ravrazor (69324) | more than 6 years ago | (#23263554)

Okay, so I know that VLIW stands for some very long word, but couldn't you have told us what it's an acronym for anyway?

buy pl0x (1)

Sicnarf (529730) | more than 6 years ago | (#23263616)

sounds impressive =) "Tilera claims that the Tile64 outperforms Intel's dual-core Xeon processor by a factor of 10, while offering 30 times better performance per Watt." and "The Tile64 is available now, in three variants differentiated by I/O mix and clock. Pricing starts at $435 in 10,000 quantities".
where can i buy this?

Really? Can I buy one unit today? (1)

clint999 (1277046) | more than 6 years ago | (#23269148)

I used the 64 way chip to get first post!
Check for New Comments
Slashdot Account

Need an Account?

Forgot your password?

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>
Create a Slashdot Account

Loading...