Slashdot Mirror
Sun Niagara 2 CPU Now Open Source
downix writes "Late last night Sun Microsystems announced the immediate availability of the UltraSPARC T2, also known as the Niagara 2 CPU. While we all might not have a silicon fab in the basement, the access to this source code reaffirms Sun's commitment to open source, and in addition gives us FPGA-lovers something new to play with. The source code can be downloaded (with registration) from OpenSPARC.net. Already the previously open sourced T1 has spawned spin-off projects, such as the Simple RISC S1."
I can remember when the OpenBSD crew was having issues getting sparc specs. My how times have changed.
The RTL code (Verilog) is GPLed:
http://www.opensparc.net/faqs/licensing/
Other people have built and are shipping product with the prior T1 version, the SimpleRISC folks:
http://www.srisc.com/?s1
The licensing pretty much says "Here, have it, have fun!"
See the sibling post below parent . . . this figure is way low for modern processors. There's a reason that there aren't many upstart processor manufacturers. The fabs are expensive and require significant expertise to work out all the fiddly problems that tend to crop up when dealing with a 65 nm process.
Take, for example, the recent $2.5 Billion Intel plant in China.
"There are four boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order." - Ed Howdershelt
Sun Viagra 2 CPU... Ok.. I need glasses
Yes, and I read Viagra. It's time to go to bed !
A billion is low-end fabs. High end cutting edge or even near-cutting edge technology costs much more. Maybe a billion for "old-school" tech like 130nm.
No, your best bet is to just pay the few million to have someone fab it for you - there are very few companies that have their own fabs and can do it inhouse (e.g., Intel, IBM, AMD, Freescale (Motorola), Samsung, Toshiba), at least, cutting edge fabs. Low end fabs can be had for cheap (1um and larger), which is great if you don't particularly care about density (e.g., Gemplus - those smartcards have HUGE silicon for 32k memory and not much more).
Most companies are fabless. They contract out the fab work to places like TSMC (amongst others - they're all well known). These include even heavyweight giants like nVidia, Altera, Xilinx and such. The only real downside is that delays can happen if machinery breaks down, or everyone submits a fab order simultaneously that causes backups at the fab and thus delays shipments. The turnaround time (from tapeout to getting chips back) can be 3 months or more. Luckily, most people test their designs out on FPGAs first to work out their bugs before committing them to silicon. Even places like Intel use computer simulation, discrete circuits, FPGAs, and such before they fab it out to their own fabs just because of the turnaround time.
Of course, what I want to know is what's the smallest FPGA one can put this on and still have something workable. (Where things like bus timings and memory clocks still in the realm of "practical" and "in spec").
Actually, the cost for one-offs is significantly lower than your estimate.
By using a Shuttle run, where the fab batches together a bunch of designs and runs them through using a single mask set, you can get 20 or so instances of a 130 nm design for roughly $100K. Of course, this assumes that you've already done the layout and verification steps yourself...
And the worms ate into his brain.
Dude. There's people who have made derivative works mentioned in the summary. You can't even be bothered reading the summary before posting? Wow.
How we know is more important than what we know.
The vast majority of (bachelors level) computer science degrees don't involve anywhere near enough focus on hardware issues for the "blueprint" of their CPU to be of any real use. The low level source of a CPU is of direct use to a vanishingly small subset of people. But, so is the source of the Linux kernel. I've never submitted a patch to the kernel. I wouldn't know where to start, frankly. And, I'm moderately qualified to do so, having done a fair amount of C, and a bit of embedded programming. I'm certainly more qualified to tinker with the kernel than I am with CPU source.
But, that sort of isn't the point. The fact that you and I wouldn't know where to start with something like that doesn't change the fact that such people do exist. And, there are some people who can't do anything with it, but are really curious to know more about what it is, and this may be the spark that makes them decide to learn. You and I may get the result of one of those guys having access to this. so, even though my own project plans won't be influenced by the availablity, I do expect that you and I will be effected by it indirectly.
It means some other (large) company can take the design, improve it, release a product based on it without paying huge fees, and then contribute their changes back to the public domain. So no, it doesn't help joe user, but it does help industry players save on R&D, which is still a good thing.
Jeremy
How big an FPGA would be required to run this? Can you really download the configs and run it on an FPGA at a reasonable speed? Which Xilinx model?
How about running Linux on that simulated Niagara2, like you can uCLinux on a Microblaze? The exciting part would be replacing parts of the OS, like the TCP/IP stack, with "HW" configs for really high performance, customized per app. None of your processes use some dozen instructions? Drop their microcode in favor of a faster multiplier...
--
make install -not war
It's basically a multi-core barrel processor - it switches threads on every cycle (similar to the old Cray and CDC gear). This gives it time to fetch data while other threads are being serviced to reduce the likelihood of cache misses. This makes it ideal for highly parallel workloads, like web application servers and multi-user database servers. But the workload needs to be highly parallel for it to perform: a quad core Niagara needs at least sixteen threads to be fully utilised. It has weak floating point, too, since web application and database workloads don't usually depend heavily on this. It's pretty much useless for a typical desktop or workstation workload.
Disclosure: I work for Sun in their software division. This is not secret information, but Sun plans on making money in the next century by selling hardware. Lots and lots of hardware. Why buy it from Sun when you can get it cheaper from elsewhere? That's the other part of Sun's super secret master plan: support contracts. Business do tend to buy from Sun if they have already done so. Maybe it's just easier, maybe it makes the original decision to buy from Sun look better, I don't know. But Sun still sells $billions in hardware each year. The software revenues are a whole lot less.
Then I guess one area that open source hardware nerds can work on is to rework the FPU to be faster.
/.'s webserver do?), and Sun is fairly close to production of the Rock, a processor that in some sense is similar to the Niagra but will also have much heavier FP capabilities. (In a development that is pretty exciting for the architecture people down the hall, the Rock will be the first commercial system supporting transactional memory.)
It's not so much that the FPU is slow as, at least on the T1 (Niagra 1), there was only one of them for the whole chip. The applications the Niagra targets don't really need FPU power (how much FP work does
I'd love to know where people like you get this kind of insight from.. Sun is an open source company.. they always have been. They were doing open source before "the revolution" even happened. They're more than "at ease" with it, they have some of the oldest living open source hackers on payroll.
As for this whole "oh noes!! Linux might takeover from Solaris!!" crap, why do people continue to ascribe the Microsoft world view - everything and everyone is a competitor - to other companies? Sun and IBM and every other normal company (read: not a monopoly) has one business strategy: give the customer what they want. If the customer wants Linux, Sun will sell them Linux. If the customer wants Solaris, Sun will sell them Solaris. If the customer wants Intel or AMD or SPARC, Sun will sell them that. IBM will also sell you Linux or Aix or Intel or AMD or SPARC.. if you want to pay them for that, that's what you'll get. It's only Microsoft who seems to think they can dictate the solutions to the customers instead of the other way around.
How we know is more important than what we know.
Hardware. I transitioned all of my company's servers to Sun (when they started selling x86 servers) over the last few years, and I couldn't be happier. The equipment is more expensive, but it is also of a superior quality and features real enterprise management features. Opening up Solaris had a lot to do with my initial decision, and I wouldn't be surprised if we don't transition to the T2 in the future. Since all of our servers are virtualized anyway, it isn't really a bad idea to move lots of small servers to a few big servers, and this open architecture will undoubtedly give some experts the opportunity to publish very in depth analysis of the chips.
You don't make the poor richer by making the rich poorer. - Winston Churchill
Yeah, it's there.
http://www.opensparc.net/opensparc-t2/downloads.html
Be warned, the 233MB file decompresses to about 1.5 GB.
who's the considerate jerk who tagged this story 'thanks'? We don't work that way here at Slashdot, buddy. When a company does something like this, you're supposed to tag it 'whocares' or 'toolittletoolate' or something equally dismissive. Damn noobs...
Dear Slashdot: next time you want to mess with the site, add a rich-text editor for comments.
While software folks may not understand the hardware world, its quite sad that hardware folks rarely understand the software side as well. One of the most challenging jobs, which gets little attention, is software-hardware codesign. Those applications, like Cadence VLSI suite, are quite challenging as they require EE expertise to implement features while software mastery to develop the product. This results in very advanced, but also very cryptic, software stacks.
I don't think open source hardware is too interesting or valuable, but I really hope software developers will feel comfortable enough to begin reading through HDL code just like they do their favorite open source project. Verilog/VHDL are both fairly old languages, though capable of doing some absolutely amazing tasks (I was in awe the first time I compiled my VHDL chip into a VLSI layout). It would be a really great to see language gurus apply the same innovative spirit towards modernizing HDL languages as they currently do in trying out different techniques in software languages (Haskell, Erlang, etc).
"Open Source?" - Press any key to continue
(disclaimer, I work for Sun, but I manage to be completely clueless about many things, including the official names of Sun's products)
Niagara 1, had one FPU per chip. Niagara 2, has more than one.
The way you pose it -- doesn't perform unless you can find the parallelism -- is not the right way. Some clever person found a market where there was parallelism, and that turns the problem around. "Given that I have all this work to do, what's the throughput per watt?" Niagara wins there. And it happened that those people, or a lot of them, didn't have a burning need for floating point.
Or to use a lame car analogy, a schoolbus is no good unless you can find 32 kids to haul, whereas a minicooper is cool and zippy with only two. But if you regularly have 32 kids to haul, and some people do, you want a schoolbus, and a minicooper is not very efficient.
On the T2 there is one FPU per core, the new FPU's are also accessed in 6 cycles vs the 40 of the T1 and the full ISA is done in hardware (some rare media instructions were emulated in the T1 FPU). The biggest win for the typical customer though is the enhanced crypto performance, now you can do SSL at the rate that the T1 could do normal HTTP.
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
I've had this position for about 2 years now.
IMO, Sun is one of the only companies left innovating.
-Google is just rehashing old ideas.(Gmail? come on....I had webmail 10 years ago.)
-Oracle(eh... RDBMS v45.2 anyone?)
-IBM(If I see one more pointless black-and-white commercial about "ideas" I'm going to scream. IBM should listen to their marketing department and instead of telling us to "Stop thinking, start doing" they should create something that isn't AIX)
And, I will be the lone voice and dare to say that Microsoft, yes them, has a few teams that are starting to 'get it'. Apple is doing a great job with human-computer interaction.
Show me new, for I am tired of your old.
Website Hosting