Slashdot Mirror
Sun Moves Into Commodity Silicon
Samrobb writes "According to Sun Microsystems CEO Jonathan Schwartz, Sun has decided to release its UltraSPARC T2 processor under the GPL. Schwartz writes, 'We're announcing the fastest microprocessor we've ever shipped this week — delivering 89.6 Ghz of parallel computing power on a single chip — running standard Java applications and open source OS's. Simultaneously, we've said we're entering the commodity marketplace, and opening the chip up to our competition... To add fuel to the fire, the blueprints for our UltraSPARC T2... the core design files and test suites, will be available to the open source community, via its most popular license: the GPL.'" Sun is still working on getting these released; early materials are up on OpenSPARC.net.
Finally a chip that you can run Java on.
That's what I got the first time I tried loading this article on /.
But seriously, what's the real point? Are the means to actually make one of these processors beyond 99% of companies and pretty much 99.99% of the people on the planet? What about the patenting of the process or equipment to actually make the processor?
My Suburban burns less gasoline than your Prius.
Clicking the OpenSPARC.net link returned the message: "This Account Has Exceeded Its CPU Quota"
...when you are talking about a market with massive investment related non-recoverable expenses & high barriers to entry, such as processor fabrication.
Although I submit it would be really cool to just manufacture these things in my garage.
Go look at the CPU cycles per watt that the UltraSPARC T1 delivers.
Now, figure the UltraSPARC T2 is better than that.
I can't wait for somebody to design a new generation of desktop PCs that have lower power consumption than that of previous generations but without sacrificing performance and graphics. Anybody know how much power typical UltraSPARC based desktop PCs consume compared to Intel or AMD based desktop PCs?
Nothing that it matter... just interested, but does anybody know if it is released under GNU GPL 2 or 3?
Many FPGA houses provide free ARM cores etc for inclusion on their FPGAs. You can build an ARM-based (or other core based) device using free download tools and run it on an FPGA that costs a few bucks. To do this the licensee need to pay a heft licencing fee to ARm or whomever. Now they can also distribute GPL cores.
But is this really useful? To use a GPL core would mean that all the rest of the chip design would have to be released too. Very few hardware builders will be prepared to release their silicon source code because that is often the only way they have of preventing mass knock-offs etc.
Engineering is the art of compromise.
Apparently Sun will sell the chips to you already manufactured if you want.
I always wanted a Tech 2 processor.
Oh wait...
Depending upon how the patents (are there patents?) are handled. China has been researching it's own chip design in the past. This could be a huge push for Sun if China abandoned trying to re-invent the wheel and just started cranking out UltraSPARC's.
Not to mention Windows not running on such, but Linux will.
And China would have a home source of chips for their IT industry and would not have to import Intel or AMD.
How many cores does this chip have?
8 cores, 8 threads each.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
One would be to build a simulator that is accurate at the level of silicon, so that you can cross-compile and run binaries for this CPU on a non-native architecture. Another would be to look at some specific module within the core and re-use the code within an OpenCores project. A third would be to reverse this - take OpenCores code (or write your own) and generate a module that would work within the T2 and would provide functionality the developers might want. A fourth would be to produce a specialized version of the chip (rad-hardened, for example) without paying license costs. And so on.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
We're announcing the fastest microprocessor we've ever shipped this week
"We shipped a faster one last week!"
(Apologies to MST3K and the classic flick Diabolik.)
Comment of the year
What workloads does sparc excel on? Is there any gains from running one as an add-in on PCIe and could an existing VM solution be hacked to take advantage of it?
This is one of those moves where some abandonware is being open sourced. Usually this happens with software, but here it's happening for hardware. The SPARC line is in decline; Sun is moving to x86 machines. Sun's hardware business is on the same trajectory as SGI's, but about five years behind. (Remember SGI, the MIPS processors, the overpriced x86 workstations, the bankruptcy?)
As Wikipedia points out, Sun already did this for the UltraSparc T1 in 2006. Nobody cared. Now they're doing it for the UltreSparc T2.
This might be useful if someone needed to emulate a SPARC CPU twenty or fifty years from now. So it's good to have the details of the CPU design on the record for historical purposes. But nobody is going to manufacture the things.
Nothing that it matter... [is it] GNU GPL 2 or 3?
It actually matters a lot because Sun probably owns a lot of patents.
Too true.
If I've got this right: Under GPL3 anybody with foundry access could make the chip or a derivative, with no more patent issues than Sun itself would have. But under GPL2 they might have to enter separate license agreements to actually implement it.
= = = =
Presuming this release does make the chip open to anybody absent further licensing, it will be interesting to see how it affects Sun's future.
On one hand it means any company that wants to could build the chip and sell it in competition with Sun (which has borne the development costs on the SPARC series - but recouped much of them already).
On the other hand, they have a number of advantages: Already up and fabbing, deep understanding of the chip, etc.
Further, one big source of resistance to adoption of their chips is the concern for what happens if Sun abandons the line, stops developing it, goes belly-up, or closes up again. With a perpetual license to others to build this chip and make improvements on it, that's no longer an issue. Even if Sun went belly-up and left them with no other sources, a big enough company with a product based on this chip could even commission the fabrication of its own chips, rather than twisting in the wind for lack of supplies. So such a company can design this chip into their product line and buy it from Sun without betting their own company on a possibly weak supplier.
Let's see Intel or AMD compete with that that. B-)
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Oh, yes:
If Sun's open-sourcing of this chip leads to a big boost for them, just IMAGINE what an argument that will be against the utility of the government-enforced monopoly in the patent reform debate. B-)
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Because wrapping an open source CPU around street furniture is great advertising. I think you have the wrong article.
Sam ty sig.
I am actually hoping that AMD or Intel decide that there is useful technology they can use in their own chips.
Especially AMD who needs whatever they can get at the moment. It is really far fetched, but possible we see AMD respond with a GPL chip that uses parts of Sun's tech they find useful. If they can get ahead of Intel for another generation or two it could be worth it to them.
Wow, sent an e-mail as suggested when clicking on "use classic" banner, and got a fast response that addressed my msg
As far as I knew, GPL only covered copyright. If this is the case, then Sun would still be protected by patent law. Unless GPL covers patents too.... Anyone know?
Anyone else find it ironic that the processor itself is acclaimed for being open-sourced, among other things, while the podcast of the announcement is only available in Realplayer format?
Now if they'd get S-Bus/hardware specs opened up on a "hobbyist RAND" basis, then you could bury the sun4m specific bits for good. Otherwise, to not aim to this crowd in some form would be stretching the "commodity silicon" term, as well as insisting on sun4m be buried and gone.
Commodity silicon exists, and it's not done on SPARC.
Twitter supports and protects racists - by smearing their critics with the "Hate Speech" label.
That comparison is between a 1.0 GHz UltraSPARC T1 - a three-year-old chip. There are 1.4 GHz chips currently available. How many months has the Intel Q6600 been on the market?
How well will that Intel architecture scale to over 4 CPUs, anyway. At least AMD can do that.
There is a successful FPGA project for the T1. The OpenSPARC engineering people were at FCRC/ISCA this summer demoing the T1 running "adventure" on an FPGA board.
/ \
\ / ASCII ribbon campaign for peace
x
/ \
If the design is GPL then those people will have to release their own touches etc under GPL too. They'd rather have LGPL or BSD licensed cores.
Fully commoditised hardware is going to be a very difficult thing to get hardware companies to sign up to.
As for FPGAs... You can get a few ARM7 cores onto a single FPGA that costs less than $10 and those prices are dropping. I have no idea how complex an OpenSPARC is, but I assume it is something equivalent to an ARM9 or so and will fit in a $10-or-so FPGA.
The hurdles are not technology, but political. Sure people want free-as-in-beer cores, but they don't want GPL cores that force them to release their design.
Engineering is the art of compromise.
FTFS: 89.6 GHz
When did we get close to 100GHz for processors?
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
Another advantage is that no business in their right mind would enter into the chip market as a direct competitor, but embedded or other technologies will be encouraged by the GPL availability, spreading Sun technology.
Returned Peace Corps IT Volunteer
Not hindered by any actual knowledge about the processor, but this was my first thought:
Considering it's Sun, this to me seems to have all the hallmarks of a farewell bid by "going open source" and hoping to hop onto the momentum it generates, if any.
Five years from now, sparc will be history. Not for being bad or outdated, but because nobody really cares. And I think they know it.
I might be wrong but I think that i386, and perhaps more specifically AMD/64 ultimately are just not possible to compete with, with specs going up (relatively) and costs going down (relatively).
It's quite possible that sparc is much better but like I said, no one will care (also relatively).
Actually, yes - according to this Sun blogger Ubuntu already runs on it (see at the end).
There used to be some free tools out there long ago that would let you design the actual silicon. Not sure what ever happend to them.
Besides, they said they are releasing the entire thing in this case so that problem is sort of moot.
---- Booth was a patriot ----
There are a couple of companies that are making a living off sparc clones now. sure not many, but id not call it 'nobody'.
And lots of people cared about the T1.
---- Booth was a patriot ----
Just to quell the concerns of "abandonware" and cries of "performance benchmarks"
:-)
Linky on numbers
Summary:
* This puppy comes ahead of Power5 and top-dog (till now) Power6
* Highest single CPU integer and floating point performance
Oh, and it has 2 10G network interfaces on chip... and EIGHT crypto cores to keep them running full throttle too. All this with 8 core each with its own floating point unit and 8 threads.
Oh and BTW, Ubuntu guys just booted their distro on this puppy
So yeah, it runs Linux (too)!
- mritunjai
Can somebody explain what it means to have 8 threads per core?
I understand that each core is like a separate CPU, and AMD and Intel produce lots of 2-core chips.
But what's the hardware meaning of a thread?
I assume that there is some parallelism that is lighter weight than a full core, perhaps with each hardware thread having certain resources like an address counter and a few registers. So how is a SPARC thread unlike an Intel core? What does the thread own vs. what is shared per-core?
Hardware support for thread-level parallelism is what is known as SMT (IBM, others) or hyperthreading. Basically, you have per-thread registers, and per-core ALUs. Exactly what units are replicated vs. shared (branch prediction, dispatch queues, etc.) is going to be implementation dependent.
I've had this sig for three days.
Well, that was fun.
---- Booth was a patriot ----
'petered out'?
Opencores.org
leon3
and tons more.
---- Booth was a patriot ----
'We're announcing the fastest microprocessor we've ever shipped this week'
I bet its the only chip they've shipped this week.
Geeze man. Talk about a limited understanding, even for /. They released the SPECIFICATIONS, that is, the information about how to BUILD on. So you can modify the specs and build your own. What, do you expect to be able to recompile your CPU?
Wouldn't that let people modify the blueprints (which are GPL'd) but not actually make the chip (which is presumably patented). Sounds like a PR stunt to me.
Yes. You, the end user, can modify the processor to the extent that is possible for the technology involved. Since a processor is physical hardware, that means the "compilation" phase for modification involves a microprocessor fab. If you don't have one, that sucks - but it's not something that's possible to fix.
Even Richard Stallman, and even the GPLv3, wouldn't complain about you not owning a fab (and therefore not being able to use a modified UltraSparc T2 in practice) as a freedom issue. This isn't like Tivoization, because Sun can't patch your physical hardware either.
-- The act of censorship is always worse than whatever is being censored. Always.
The posters here seem to be complaining that this is worthless because individuals can't make their own processor chips.
That's not the point. Here's the point:
1: Sun's processors are a niche market. People don't use them because they're harder to use than cheap commodity processors from Intel. Why are they harder to use? Because not enough people use them to create the kind of economic ecosystem that drives down the price of using the processors.
2: All over Asia are chip factories that make low-end embedded devices, RAM chips, and so on. Factories that are owned by companies that don't have the cash on hand to do the R&D to design their own processors to compete with Intel.
3: By GPL'ing their chip designs, Sun lets all those Asian factories produce chips that perform like Intels but cost even less. This gives people an extra incentive to switch away from Intel and to create the very economic ecosystem the processor needs.
4. Next, Sun releases enhanced versions of the chip that aren't GPL'ed. Chip consumers can now choose from fast commodity processors or more expensive deluxe models - that are still code compatible.
And Sun can repeat steps #3 and #4 as often as they like, feeding their previous generation designs to the GPL audience as their newest designs hit the market.
Clear, Dark Skies
That theory might hold water, if Sun hasn't seen one of it's best years in recent history.
WebMink:
Oops! Is he really just trolling?Oh, sorry Animats. Didn't mean to step on your trolling line there...
Ahm, yea, right! They're just heading for financial collapse! They're just making lots of money along the way! Yea, Yea, that's the ticket! ;)
Shane
So what's the DigiKey part number? So I can buy a few (or a thousand, or whatever). I just entered "SPARC" in their database, and got "No records match your search criteria."
If I can't buy it from major distributors, it's not "commodity" silicon.
When will it be possible to get a PCI-e board with a load of FPGAs to which I can download a T2 to run code on, when some code I download happens to be T2 opcodes? Or even anywhere close to that, like a wrapper on the FPGA that can use the T2 config as a starting point to emulate the T2?
--
make install -not war
what's the hardware meaning of a thread?
As another poster has pointed out: You build a core with multiple copies of the register set and replicate (or take turns on) the associated instruction-dispatching logic.
But these multiple CPUs share a common set of arithmetic/logic execution units, along with arbitration logic. Different threads will be doing different things at any given instant, and thus using different sections of the ALU.
The arbitration logic decides which thread gets which hunk of ALU at any given moment. And some threads will be stalled waiting for data and won't need any ALU function at all. Of course when more threads want a particular kind of execution unit than are available, one or more of them must stall. But by having the right number of copies of the commonly-used types of execution units you can keep a number of threads running at or near full speed most of the time and the ALU components mostly busy, with much less silicon than if each thread had a full-blown CPU and most of the ALU logic was idle at any given moment. With less silicon logic you can put things closer together and speed it up still more.
This approach has been around since Cray was at Control Data.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
From the GPL (3):
Some devices are designed to deny users access to install or run modified versions of the software inside them. This is fundamentally incompatible with the aim of protecting users' freedom to change the software. The systematic pattern of such abuse occurs in the area of products for individuals to use, which is precisely where it is most unacceptable. Therefore, we have designed this version of the GPL to prohibit the practice for those products. If such problems arise substantially in other domains, we stand ready to extend this provision to those domains in future versions of the GPL, as needed to protect the freedom of users.
Do you even lift?
These aren't the 'roids you're looking for.
* This puppy comes ahead of Power5 and top-dog (till now) Power6
* Highest single CPU integer and floating point performance
In two benchmarks: SPECint_rate and SPECfp_rate. Now let's see some real-world application performance.
From the GPLv3:
If you convey an object code work under this section in, or with, or specifically for use in, a User Product, and the conveying occurs as part of a transaction in which the right of possession and use of the User Product is transferred to the recipient in perpetuity or for a fixed term (regardless of how the transaction is characterized), the Corresponding Source conveyed under this section must be accompanied by the Installation Information. But this requirement does not apply if neither you nor any third party retains the ability to install modified object code on the User Product (for example, the work has been installed in ROM).
I'd say that providing the code in the form of physical hardware would be covered pretty cleanly by that last sentence.
-- The act of censorship is always worse than whatever is being censored. Always.
While it supports 8 threads per core, only 1 thread per core is executed at any given time, 8 in a core aren't executing concurrently. It's targeting business servers which typically have many threads but at any given only a subset can perform calculations, the others are waiting on I/O.
that's contrary to the ideals of free software. if the FSF considers that ok then we need a new license which ensures software (and hardware) freedom.
Do you even lift?
These aren't the 'roids you're looking for.
What is the business outcome and advantage for Sun with this open hardware move?
It looks to me like Sun has figured out that they are in a knowledge business. They are operating on an information theory algorithm. They are creating a much larger pyramid of customers for their particular computer knowledge. There is a new enormous bottom layer of people contemplating using this fascinating powerful chunk of information. They are emitting information, not hardware.
The thing from information theory is: The more high quality information a source emits, the more valuable the source becomes. Sun still has the stable of PhD researchers from U.C. Berkeley and some more wags from Stanford. So the company will continue in the business of emitting information.
If you want the Macintosh of mainframes, they will sell them to you. If you want to boot Solaris or wire up Sparc chips you are still their customer. Sun will be your first publisher, web site and consultant.
It seems to me that this is a business innovation. It has been 60 years since Shannon's information theory paper suggested that the source that emits information increases in entropy. Sun is doing that by making available a uniquely sophisticated design - not hiding it in file cabinets in the basement.
You're clearly just trolling. The assertion that the ideals of free software require the user to be able to reprogram an ASIC or ROM on the fly is absurd. What's next? Rocks steal our freedom because we can't turn them into birds with our psychic powers?
If it gets to the point where, in practice, all computer processors are implemented on user-reprogrammable FPGAs or something, then we'll need to have a discussion about the freedom to modify software-implemented hardware. But, as always, free software is about giving the end users freedoms that are possible - not demanding absurdities.
-- The act of censorship is always worse than whatever is being censored. Always.
While impressive, these are not workstation chips. They only thrive in an environment where they run many processes or threads in parallel, like a webserver. If you were to use it as workstation and run Linux and browser and desktop suites, you'd probably find it feels slow!
Just look at the 4-core Mac Pro vs. 8-core benchmarks; not much difference.
The only "workstation" activity where these multi-core multi-thread CPUs shine is video rendering.
Did anybody ever fab the thing and offer it for sale? Simply RISC doesn't seem to be shipping any real hardware. The picture of the chip looks fake. It looks like they put the logo onto a picture of some other chip with Photoshop. Look at the jaggies on the logo, and how the logo doesn't line up with the upper edge of the chip.
Polaris Micro says they're shipping on Q3 2007, but their web site hasn't updated their news section since 2005.
It's neat that you can load their cut-down version of a SPARC CPU into an FPGA, but that's not a partcularly useful product.
I don't think he was trolling either. Sun's CPU design division is in decline in my opinion. I fully expect that at some point in the next 5 to 10 years Sun will be selling off its CPU design resources to either Intel or AMD. I expect they will eventually pull out of the design side of the market just like Compaq and HP did. Just my prediction. I've seen no indication that this newest chip from Sun is technically the best chip out there either. I expect it to be fairly good at the types of workloads that Sun's current T1 customers run most frequently. But expect it will probably be rather mediocre at other tasks. I've no data to back that up, but there's no data being offered by Sun or anyone else to prove otherwise either. So who knows for sure?
In a minute there is time For decisions and revisions which a minute will reverse. -T.S. Eliot
I'm assuming it will be GPLv3 since Sun generally has said they like v3.
Did I miss that anywhere? Did he say it would be GPLv2 or v3
and...somebody get RMS on the phone for comment.
That doesn't mean the end of SPARK. That means that Sun wants to go into the MIPS market, that is licensing the core to custom designs.
But Sun isn't getting money out of this. That probably means that they want to fill that market so their core will be more used and so they have better support and a bigger production scale scale for their servers.
Rethinking email
As fast as this chip is in Ghz, would it be possible to run a single threaded .Net application in a VM (Virtual Machine) on this chip and still have it perform faster than on the fastest Intel chip?
Does Sun have a VM that will do this.
I know about mono but there are many .Net applications that on API's not present in the mono environement.
I know Ghz is no longer the end all/be all measure of computing performance especially when comparing CISC and RISC but would dearly like to know if there's anything like this that could leapfrog Intel by a few generations.
GPLv2 has an implicit patent license in it. It makes it pretty clear that people who receive the code can do whatever they like with it - including make chips in this case. It also makes it pretty clear that they can distribute or sell derived works to others who have the same permissions. If you can't give the code or chips (which are literally "printed" derived works) to someone else then you don't really have the rights the GPL says you do. In case that isn't enough, GPLv3 has explicit patent coverage.
Might be safe to say, that Sun is betting they will be able to improve the design better and faster than brand X (lenevo?), but that brand x's can help establish a broad and diverse culture of software, support, and education - in which Sun can expect to retain the highest margin business, in exchange for some, or most of the low-margin fare.
The gamble loses if sun cannot produce a better chip next year than brand x.
I like sun's odds.
AIK
The gamble loses if sun cannot produce a better chip next year than brand x.
It's even better for sun than that.
Even if Sun doesn't produce a better chip next year, it's still a win if they make more money as a result of opening than they would have if they'd stayed closed.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Closes up again? The SPARC architecture is and pretty much always has been an open standard. Anyone can implement it, and several other companies have, including TI, Cypress Semiconductor, and Fujitsu. The architecture is controlled by a separate organization, which is a non-profit: SPARC International.
Now this only applies to the architecture, not the design of the chips that implement the architecture. Nevertheless, if your concern is to be able to run SPARC code, you don't need to rely on Sun to do that.
What is the situation about the processes used to make the chips? Are there processes or materials (components, machines, etc) that are required to produce it?
it isn't necessarily about the average Joe making one in his garage. It is more than blueprints. I could "GPL" blueprints to a miraculous, functioning, FTL drive but what good would it do SpaceX if it requires some proprietary hardware/processes/material to actually make it? Hence the question about the entire process and requirements.
My Suburban burns less gasoline than your Prius.
I rest my case.
IANAL but write like a drunk one.