Slashdot Mirror

It's a proprietary design. If you want open hardware you can have a look for example here:

http://opensparc.net/

Who needs another proprietary closed source crap?

http://www.opensparc.net/

Hatred towards open source on Slashdot?

http://www.opensparc.net/

OK. Just buy proprietary crap from Intel, AMD and ARM.

Fabbing the CPU is the easiest part of the game. Designing it is what is really expensive. Fortunately, we have the source code for a good CPU to get started, should we ever need to re-bootstrap the IT industry.

You can go download openSPARC here complete with testbench and synthesis scripts. Of course, you still need roughly >$200,000 worth of EDA tools to bring it all the way to tape-out.

Almost all hardware is proprietary. Perhaps you mean uncommon? Anyways the designs of the T1 and T2 have been publicly released. http://www.opensparc.net/about.html

Both Sparc and Power now have open specifications that anyone can use to implement their own microprocessor and sell it in the market for any targeted applications. Which is pretty much the goal of open standards. The closed RISC standards that were there - Clipper, PA-RISC and Alpha (Alpha actually less so) are all dead, as are i860 and i960.

Incidentally, the latter Alpha and Power architectures, as well as the MAJC processors all borrowed some VLIW concepts such as concatenating multiple instructions into a single word to enhance their SIMD capabilities, so it's not like VLIW is a complete failure. Itanium managed to, on a PR front, knock down PA-RISC, MIPS and later (after HP bought Compaq) Alpha, but ironically, failed to do much against Am64, w/ the result that it's not made a dent in the marketplace, and Microsoft, Oracle, RedHat and Canonical have all dropped support for it. Even Intel's latest C++ & Fortran compilers don't support Itanium: support is referred to earlier versions. Given that factoid, Intel's announcement reiterating support for Itanium sounds hollow. And w/ the Itanium's list price of $700-$4000, one can't support that CPU even if one wants to.

The game is over - the only CPUs that matter are x64, Power, Sparc and MIPS (I'm not counting ARM here, since it's so far unsuitable for server apps).Intel can forget about dethroning either IBM or Oracle in that arena.

You mean openSPARC ( http://www.opensparc.net/ ) and openRISC ( http://openrisc.net/ ). I thought there was a MIPS and a Power-based open hardware project too but I could not find it right now.

http://www.opensparc.net/

On the CPU front, you can make an equal case for Intel supporting open source.

On the CPU front you can put your money where your mouth is and buy Niagara.

http://www.opensparc.net/

Sorry Oracle. I have to do this without you. http://www.opensparc.net/ Now, to make something awsome

It's a sad day when most know so little about architectures that PPC and Sparc are re-flagged as mere 'alternatives' rather than being recognized for the areas they excel in. The Wintel ideal is a ratio of worse that 2:1 of hardware to services. Each box burning tens if not hundreds of watts. Sparc is an open architecture and handles many threads per core, so for most things you should be able to replace a rack of Wintel boxes with a single Sparc. We'll see how long Oracle allows you to access that Sun paper.

Fujitsu also provides info sparc architecture because it also sells server hardware.

More info at:

http://www.opensparc.net/

Includes the definition for the original UltraSPARC T2 processor

As much as Sun's.

SPARC is open. I don't see open source z/Arch or POWER.

But you have a choice of different OS's supported on their proprietary hardware, unlike Sun.

So? If you want to run Linux or Windows, Sun will sell x86-64 hardware.

Sun sells many proprietary products.

Many of Sun's software products are available for free if not open source. Java and Solaris are examples of Sun products that have been open sourced.

Oracle also makes contributions (check the stats for recent kernels, Oracle is normally somewhere around number 3-5 corporate contributor, behind Red Hat, Novell, IBM.

And Sun has open sourced an entire kernel and operating system.

If you wanted to build a car or whatever and sell it, whether or not it's based on an open source design it would still be required to meet the same standards.

If you want to mod your own car and make it illegal then you could do that but that would only last one year (assuming you don't get caught) until the inspection.

We do have open source hardware already, one big name being Sparc http://www.opensparc.net/ but I suspect the reason it won't ever take off in a big way has to come down to the fact it's probably harder to recoup your R&D costs if someone comes in with dirt cheap chinese labour to build an exact copy. At least with software everyone is more or less on the same terms with distribution costs on the net.

guess who won't be buying any more sparc servers?

Someone who is not interested in supporting Open Source processors?

SPARC is an organization that licenses processor designs (Scalable Processor ARChitecture), provides docs, and even licenses a keyboard interface design (this part is low cost). I'm not sure what they charge, it may be just a fee for making sure your processor conforms to the sparcv9 spec or similar and you can put the SPARC stamp on your box. I have not heard of them being unreasonable before. It's an independent body. Sun and Fujitsu/Siemens happen to be a couple of the companies that use the processor design (modified for thier use). Probably similar to ARM in this case, but I will freely admit I do not know many details about the subject. I do know that one can download the designs and source code for the OpenSPARC T1 and T2 series processors from http://www.opensparc.net./

I hear that someone has a design that runs an OpenSPARC on a FPGA (granted, single core, but still cool).

So if you care about having an "Open" CPU design in your system, then you'll be missing out by avoiding SPARC.

Alas I do not have many facts available to back this up. www.sparc.org has been blocked by firefox or my antivirus somehow...perhaps someone is attacking it out of badwill or they got owned independently.

Regards

BTW: I've run gentoo on 'sun4u' SPARC processors, and hear that the kernel has lots of support for the 'Open' T1/T2 line and hypervisor, etc.

At this point, it looks pretty much the same as a (DEC) Alpha or Itanium roadmap.

Except for one thing: the SPARC circuitry is entirely open source. This has interesting implications, such as the fact that enthusiasts can build these things as FPGAs or even ASICs as fab costs come down to within the reach of clubs and schools. And emerging economies can fab these things by the bazillions without paying royalties. Not to mention big rich countries too.

Umm, the spec and SOURCE for the design of the processors is already opensourced. http://www.opensparc.net/

This, running on a T1 or T2 machine, running ${FREEOSOFCHOICE}. yum.

Or perhaps he could just run GNU/Linux on, say, an OpenSPARC based system.

It's not like nobody has ever thought of this before.

While Sun may not be the strongest FOSS advocate, they've made many adjustments over the past few years to open up several products.

Stop right there. Sun is one of the biggest corporate contributors to open source. Go ahead, count lines of code. I'm betting Sun will be in the top two if not #1.

Here's a brief list of things Sun has open sourced:
Solaris - Their entire OS, including ZFS and Dtrace
SPARC - Their CPU line
Java - Maybe you've heard of it.
OpenOffice - The office suite that ships with every desktop Linux distribution.
VirtualBox - A GPL desktop virtual machine.
NetBeans IDE - A multi-platform IDE.
OpenDS - LDAP Directory Server
High Availability Cluster

Honorable mention:
NFS - The Network File System
vi - developed by Sun founder Bill Joy
MySQL - Now owned and maintained by Sun-paid engineers

So, next time you say Sun hadn't done much for open source, look again. It would be a shame if Sun was bought by Oracle and all of their valuable contributions were abandoned.

Niagara-generation SPARC CPUs are open source. Anyone with an idle fab laying around can go ahead and start churning out UltraSPARC T1 and T2 processors. Fujitsu has an interest in the architecture not dying, since the SPARC Enterprise server line is actually Sun/Fujitsu and the SPARC64-V PRIMEPOWER is all Fujitsu.

That doesn't get you very far without compatible hardware to plug the CPU into, but Sun being bought and shut down by the likes of MS doesn't necessarily mean SPARC will go bye-bye.

http://www.opensparc.net/

If IBM drops SPARC, Fujitsu will probably become the only supplier of Solaris systems. Both the SPARC architecture and Solaris are available as open source, so IBM cannot easily prevent that.

Being open source, it is possible that other companies emerge using those technologies. Don't count on it: if anyone believed in that, this would already have happened.

I'm also unsure about how much traction Solaris will have without SUN.

Who needs Dell or Intel when you have Open Sparc?

I hope this helps the revolution.

Yours In Socialism,
Kilgore Trout

You don't think the transistor designs in any current commercial chip are open source do you?

Yes, I do think so.

The UltraSPARC T1 and T2 are Open Source.

http://www.opensparc.net/about.html

I felt these links should be in this thread:

OPENCORES.ORG
Open Hardware
OpenSPARC
The Wikipedia article on Open-Source Hardware, with many more links

"AMD is the last bastion of creativity in CPUs."

Have you checked out UltraSPARC T1 and the upcoming UltraSPARC Rock CPU's ? There's still creativity in microprocessors...

By the way, T1 and T2 are also the only commercial open-source processor

They are coerced into behaving nicely by the huge open source community which is not that much interested in what they have to offer anymore and have a lot of influence in the market

Not really. Java still dominates the enterprise application market (the only place it ever made any money for Sun), and its open source status is likely to have little effect on this. Even without ZFS being open-sourced, Solaris would still have a world leading file system. And I don't see where any pressure at all came from for them to open source the design of their UltraSparc T1 and T2 processors. Sun have been progressively opening more and more of their key business IP, and as far as I can see the only reason they have done so is that they really believe in the benefits of open source.

One of these in an OpenSPARC machine?

Oh yes... the future is now

We don't care about suppliers of proprietary solutions because we have OpenSparc. We wouldn't run an open source OS on closed source hardware and firmware.

http://www.opensparc.net/

Why not? Sun's Sparc processors already are.

We (that's Sun Microsystems) chose the GPL as the license under which to release everything necessary to make an UltraSPARC T1 (and more recently, T2). We placed it all - RTL, tools, the lot - at OpenSPARC.Net. The license choice was for two main reasons:

1. We felt the GPL was widely accepted as a Free license. We hoped that would encourage all kinds of people to feel free to take a look.
2. We wanted those who used the code from OpenSPARC to publish the new work they built from it.

While releasing hardware sources under a Free license is a different deal to software, the GPL seems to encourage the same willingness to examine and use the code as it does in software. The mechanisms for community have to be different because of the capital-intensive nature of the processes to use the code. We've still seen people rework it to fit it on FPGAs, create single-core chips for embedding and run university degree courses on it. I remain pretty happy with the license choice we made.

*cough*

are there other open-source processors ?

Yeah, it's there.

http://www.opensparc.net/opensparc-t2/downloads.html

Be warned, the 233MB file decompresses to about 1.5 GB.

I wonder what they'd do if someone started selling processors based on the information they just released.

Oh, yeah, the whole "open source" thing. Excepting core functionality, some of Mathematica and the majority of Maple is provided in source form.

For now. But since the program is closed source and very expensive, what happens 30 years from now when Wolfram won't give you the version the original result of interest was created on and it's illegal to get it anywhere else? Oh, and the formatting options changed over 30 years so the results look different and you can't tell easily if they're still the same? Not that this is guaranteed to happen, but it might. Open source is a guarantee, and when you're doing research of this sort that guarantee is very good to have.

You can whine about needing peer review of implementation at all levels, but how many of you have inspected your CPU's microcode or circuit diagrams?

A very good point, but the idea is that (in theory) you could if you have to. Indeed, my own interests with Axiom have lead me in those directions - I have downloaded the MIT CADR machine circuit diagrams and have acquired a couple books on Forth (which has the virtue of being "easily" bootstrapped from machine instructions). I'm also aware of things like OpenSPARC and OpenCores. The point being not that I will ever be good enough at understanding them to verify them other than experimentally (unit testing, etc) I COULD do it in principle because it is available. I would very much like to seen a machine built entirely on open hardware even if it would be slower, but commercial realities may make that difficult.

Anyway, the point is you strive to be as open as possible. Even if hardware today isn't verified, someday open code could be ported to a verified open platform. The principle is worthwhile even if the implementation of it isn't perfect from the beginning.

At some point the line is drawn, and you combine a trust in the reputation of your vendor with the fact that usually you're prototyping and modelling.

Indeed that is what must (practically) be done now, but that doesn't mean we shouldn't fight to make sure that things remain open - it's a foundational principle of science that things be reproducible and openness facilitates that. Perfection is not currently possible, but that doesn't mean we give up and don't do what we can.

Verily! Most old SPARC cores, up to UltraSPARC, have their RTL source code open for download for exactly that purpose. So throw that on an FPGA and have at it. Toss linux or minix on their, and you're running open software on open hardware. I can think of nothing sexier.
Although I can't deny that a home-wired CPU is badass...that's a lot of free time...

One of the more interesting developments following the release of the Sun UltraSPARC T1 & T2 chipsets under the GPL has been the S1, a single core implementation of the T1, which combined some other other GPLed hardware can be built as a RISC based system on a chip. It has massive potential as a powerful, low wattage processor that could compete with ARM and Intel in the portable device marketplace. It might be a couple of years in the future but I think it has the potential to be a competitor. It should run Solaris and BSD as well as Linux.

People built Pentium Ms and the like in their garages 20 years ago?

No. But they could at least design such a microprocessor now. Heck, Sun has open sourced its SPARC series of processors to get the community's input on its design.

There are tons of research chips made from the OpenSparc designs and Simply RISC claims to have an embedded processor made from a single core T1 design.

...If they put THESE under the GPL, along with the T1, they'd be getting more press than they could imagine.

http://www.opensparc.net/

They are openly discussing making the Niagara 2 available as open source as well, but note that there are some roadblocks such as the US government's restrictions on crypto technology.

...If they put THESE under the GPL, along with the T1, they'd be getting more press than they could imagine.

http://www.opensparc.net/

They are openly discussing making the Niagara 2 available as open source as well, but note that there are some roadblocks such as the US government's restrictions on crypto technology.

suddenly, everyone's favorite company doesn't look that bad after all.

he'll then ask for the BIOS, and the CPU firmware

SPARC is an open CPU design http://www.opensparc.net/

You should be able to build a machine based entirely on open components, from the CPU, through the firmware to the OS and finally up to the voting software.

Sun released SPARC?

Strange, given OpenBSD folk lack documentation to properly implement UltraSPARC III hardware.

OpenSPARC is available from Sun Microsystems. The SPARC architecture is still highly relevant. Open source hardware projects like this are worth noting.

There is even a modern open source processor. Look at http://opensparc.net/

This isn't FOSS hardware. They will be using Niagara. 8 core @ 72W.

http://www.opensparc.net/