Slashdot Mirror

Programming tools are one of the areas that hasn't gone to the cloud.

distcc and (sun) grid engine don't count? Both of them can be used for distributed compilation, etc.

Well, you'll probably see something along this: PID user .... CPU% MEM% xxx user1 ... 100% 0.1 xxx user2 ... 100% 0.1 xxx user4 ... 100% 0.1 about 3000 times... I was working for a research center for genomes and chromozones, and I created a high performance cluster with 3 high end servers with SGE. Nothing HUGE, but with tweking of those batch files and memory usage, they where getting results from 10% to 600% faster.

I use gccfss and it's great. Handles GNU-ish code, and yields highly-optimized binaries. Unfortunately (for me), the last available version for Sol9 has a nasty bug in it that the team has no intention of fixing.

Sun has some very cool hardware and software, not to mention an open source friendly attitude...

Sun has a Machiavellian attitude towards open source. They are open-source friendly when it suits their needs (public image). They are certainly not friendly to all open source projects. Need an example? Look at GCC. They feel free to rip off the GCC front-end for GCCFSS. They contribute virtually nothing to the GCC or binutils software in return.

Want to know how the GCC devs feel about it? Read the mailing list thread.

All their customers want is a fast and portable compiler for Sun hardware. Sun's own compiler sucks for C++ development. You have to use a non-standard, obsolete and soon to be deprecated C++ standard library just to have a slight chance of compiling any modern C++ software.

If you want to see just how screwed up their C++ development environment is going to get, you just have to read this thread: http://opensolaris.org/jive/thread.jspa?messageID=289188&tstart=0

Now imagine if Sun started shipping fully supported SAMP (Solaris Apache MySQL PHP) software distributions branded with "High Performance*, 64-bit Sun MySQL".

Note - OpenSPARC T2 currently does not include PCI-Express and 10Gigabit
Ethernet design implementation due to current legal restrictions. Equivalent models
may be available in the subsequent releases of OpenSPARC T2.

The statistics the T1 are available here:

http://fpga.sunsource.net/

The most recent release of the T1 code has a few options for removing functionality (dropping to 1 core and 1 thread) such that it will fit on some of the larger available FPGAs.

The OpenMoko software is more important than the hardware. The Neo1973 is open hardware designed to a specification, but won't ever be a successful commercial product because no phone company is going to subsidise it. There is a bullet to be bitten, that there are few smartphone platforms that are open enough to match the aspirations of OpenMoko, but I can see it being ported to other smartphone chipsets such as those used by HTC or indeed non-dedicated chipsets like the OpenSparc S1. It wouldn't surprise me at all if was running on an iPhone in the next year either. That's where the software's strength will lie: while phones are not PCs, and are probably harder to develop drivers for (one of the reasons, I believe, why the Nokia N8x0 doesn't have a phone module), there will be a group of people who want to get it out there and onto as many phones as they can.

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.

What I find interesting is that I seem to be able to re-license it under various other licenses at my discretion.

How very open of Sun. I approve. ;)

What I find interesting is that I seem to be able to re-license it under various other licenses at my discretion.

How very open of Sun. I approve. ;)

What I find interesting is that I seem to be able to re-license it under various other licenses at my discretion.

How very open of Sun. I approve. ;)

It is GPLv2:

http://opensparc-t2.sunsource.net/

I know you're trying to be funny, but Apple publishes a guide for doing precisely that, without replacing OSX. If you're creative, you can always skip XGrid and use SGE (GridEngine) instead.

So, quit imagining, and just run!

Most Beowulf clusters run parallel codes written to use the Message_Passing_Interface (MPI). MPI programs really don't want to be migrated to different nodes while they're running, so load management is achived through schedulers such as Grid Engine, TORQUE, and others. These schedulers avoid the need for process migration by preallocating the resources (compute nodes) in advance, and prevent the load imbalance from happening in the first place. openMosix waits for the imbalance to slow down the computation before it migrates a process to relieve the problem.

If you check the archives of the Beowulf mailing list, you'll see that while the Beowulf community knows about openMosix, very few Beowulfers use it.

Its a one of the several neat ideas being lifted from the Fortress language.

For the unitiated, Guy Steele (of Scheme fame) is building a new language for scientific computing called Fortress. It has some nice ideas that really should have been there by now. The language would have saved countless headaches in not just scientific but probably all mainstream software development projects.

Of course, its just one of the pet projects in SUN Labs ;)

Performance isn't everything, but then again, when you are 400 times slower than Java..

400 times slower? Still too efficient. Consider a Ruby implementation on the JVM and multiply inefficiencies! Should be thousands of times slower for those same benchmarks.

About JRuby; Sun recently hired two (both?) of the JRuby developers and progress has accelerated. The promise is a highly capable Ruby implementation running on a JVM. This, coupled with very recent changes to the JVM to facilitate scripting languages could lead to an interesting future. Sun is also leveraging the JVM for other language projects such as Fortress.

Apparently Sun isn't content to let Microsoft's CLR become the de facto standard bytecode runtime platform. I don't know whether it's possible to make Ruby performance on the JVM competitive with native implementations, but I am hoping.

How is this different from their already-available CoolStack, which I'm already running on my T2000?

http://cooltools.sunsource.net/coolstack/ ...they don't really explain it.

see the Sun Fortress project http://fortress.sunsource.net/ for a language with built-in parallelism...It's a beginning, anyway.

...all i've seen Sun do over the last few years is open more things up. Solaris used to cost us money for disks; now it's free (and there's an open source version), same for Java, now it's GPL'd. Anyone like open source chips?

don't get me wrong, i don't think the sun (not much pun intended) shines out of their collective behind. there's still some stuff that grates; service plans just to get the 'recommended' patch clusters. they are moving in the right direction, as parent said.

Hmmm yeah the ultrasparc hdl has that pesky Gnu Public License.

http://opensparc-t1.sunsource.net/download_hw.html

Pretty much everything around the UltraSPARC processors is out and in the open.

The way the summary is written, you'd think that actual site was down or something. But the website and grid itself was fine - it was just the free example (running on separate hardware) that got busy. (I dunno how busy - I accessed it yesterday and it was fine at the time).

I dunno, Slashdot could have reported on something more meaningful - like Sun GPL'ing their latest processor. You can download it here:
http://opensparc-t1.sunsource.net/download_hw.html

There's a decent write-up here:
http://www.itjungle.com/breaking/bn032106-story01. html

Manufacturing fab not included...

It is a bunch of v20z and x2100 style hardware running the Sun Grid Engine which is an open source grid engine that Sun wrote.
http://www.sun.com/servers/entry/v20z/specs.jsp
http://www.sun.com/servers/entry/x2100/specificati ons.jsp
http://www.sun.com/software/gridware/
http://gridengine.sunsource.net/

All your code needs to do is run on Solaris 10 on x86 hardware. It can either be 32 bit or 64 bit.

This isn't a "platform" like a beowulf cluster, it's just a bunch of loosely coupled machines that have a job scheduler in front of them to dispatch jobs. Any of the jobs you submit will only run on a single machine at a time, and it may take 1 or all of the processors available in that machine.

Sun isn't doing anything magical here, by the end of the summer I plan to have a grid with Solaris SPARCv9, RHEL 4 x64, Windows 2004 x64, and MacOS PPC64 clients, and it will have a web based front end gui scheduler (just like Sun's), and it will be easy to use (just like Sun's), and it will be able to do accounting back to users at the 1 hour interval (just like Sun's) because I will be using _Sun's_ grid engine software to do it.

I would also point out that you can use the Sun N1 Grid Engine software for free, and that you only need to pay Sun money when you want support.

The cell looks like it has a real chance of becoming the next big advancement.

It will be interesting to compare the Cell with the UltraSPARC T1 (Niagara). They both have about 8 cores (T1 is 8 cores, Cell is 8+1), but the T1 can do 32 threads of execution simultaneously. The Cell has good floating point performance, but the T1 only has 1 FPU for all 8 cores (it's specifically not designed for FP performance). The T1 has very low power requirements, at about 72 watts (79 peak), while (as far as I can tell from google) the Cell will have high power consumption and they have not disclosed the exact figures yet.

And both companies are working very closely with the open source community. Sun actually went further, and open sourced the entire SPARC architecture. As far as I know, IBM is not opening up their architecture.

They clearly have different markets, but they are similar in the multithreading aspect. Whoever does a better job of the multithreading and makes good compilers that can help the programmers write parallel code will then be able to move into the other company's space (if Sun does it better, they can add FPUs, if IBM does better they can remove them). And that success depends on open source involvement, which depends on an architecture that is easy to code for. If open source programmers get heavily involved in a concurrent compiler for one architecture, it will win in the long term.

So, it's clear why both companies are fighting to get the attention of the open source community, which is becoming (in a lot of ways) the force that drives which technologies are actually used in business. And that's certainly good news.

Opening up? Things to come?

Sun has been one of the biggest commercial open source supporters for years now. Probably only surpassed by IBM and the Linux companies ( RedHat and Suse, Linux is their core business after all ).

Millions to buy StarOffice, millions to setup and run OO.org and OpenDocument development, marketing, promoting OpenDocument. Releasing packages like GridEngine, etc. http://www.sunsource.net/. Years of shipping and support opensource applications to companies that would never have used it otherwise.

Back when I was a network admin, we got a whole lot of GNU software in the system by first showing superiors that Sun endorsed those packages and actually provided solaris binaries.

Sun's main issue is PR, I suspect. When IBM does something good, it makes sure everyone knows. But that doesn't seem to be McNealy's style...

Rocks has a great system for making high-performance clusters from similar machines. A Rocks cluster consists of a front-end ("master") node and a bunch of compute nodes (and I think special-purpose nodes).

The master gets a full Linux (RedHat-based) install. It's a NFS/DHCP/Kickstart server for the compute nodes, and runs whatever other services you want the compute nodes to use. The master has two network cards and acts as a firewall (NAT optional).

The compute nodes boot via DHCP and Kickstart, downloading their kernel and whatever other OS files you want to their local disk. You decide how much NFS or local disk to use.

Job queueing is handled by, e.g., Sun Grid Engine (an Open Source queueing package) or some other queueing software.

Here's the neat thing: to make a change to a compute node setup, you change the Kickstart config and reboot all the compute nodes (as they finish whatever queued work they're doing, or immediately if you want). That makes the sysadmin's life easy, while still maintaining the speed of having the OS on the local disk.

What do you mean by "no"??

If you spend a minute or two more to find out more before jumping into the conclusion, you will find:

http://gridengine.sunsource.net/servlets/ProjectSo urce

http://gridengine.sunsource.net/

Free and opensource, runs on almost all operating systems.

So how does Sun compare to these two? While numbers, 32, does not compare, what about the scope of some of the projects. Open Office and Open Solaris just to name a couple.

IMHO sun is not a friend of free software or even plain open source.

Yeah, I looked at that and went with SGE (http://gridengine.sunsource.net/) at the time, mainly for political reasons. SGE gave me extra buy in from a couple of other departments. Works nicely and relatively transparently even for stuff like OpenOffice.org, Netscape and GIMP which you would normally run locally.

For fairly heavyweight apps we have the machines grouped e.g. There are a bunch of OO machines, GIMP, Mozilla machines etc. It takes advantage of shared libraries; OO is about 90Mb resident, but about 85Mb of that is shared. Along with large CPU caches and pre-loaded shared libraries there are some *huge* performance benefits to be had.

Leasing assumes you can't find a use for obsolete kit. Plus it's a right royal pain in the arse to administer.

Well, if you design your systems correctly your kit will still be in use long after it has depreciated.

That means *don't* put the power on the desktop where it will be obsolete in 18 months. It also means make use of *all* of your computing power. It isn't difficult, there's loads of (free) software out there to help. e.g. http://gridengine.sunsource.net/ . Oh oh oh. Look! It has that "Grid" buzzword! Don't worry it's just a load balancing network queue system, been around for decades.

You can save a *bundle* by designing your network of servers & services properly.

www.sunsource.net
www.opensolaris.org
www.openoffice.org
Sun supports Linux, too

What were you saying?

Gridengine just added Windows support:

- Windows XP and 2000 (December 2004 availability)

http://www.sun.com/software/gridware/

Gridengine's source can be downloaded from:

http://gridengine.sunsource.net/

IBM, the big blue company, decided a long time ago that Open Source isn't so bad.
Sun, the UltraSPARC Processors maker, decided that Open Source isn't so bad.
Intel, the 8086 Processor maker, decided that Open Source isn't so bad.
Munich, Germany's third-largest city, decided that Open Source isn't so bad.

"Microsoft decides Open Source isn't so bad" will be news.

Extrapolating from this, I'm not worried about OSS Solaris killing off anything -- it looks nice on the outside, but has no substance inside. ;)

How about 50 engineers working on Mozilla?

Not to mention Gnome, Linux, etc.

www.sunsource.net

Further, they've involved in several smaller projects. Check out http://www.sunsource.net/ for more information. Oh, and they're a member of the Open Source Development Lab.

Is that good enough for you?

Further, Sun has developed several technologies which have been widely adopted by other Unix vendors, such as NFS and PAM.

While Sun doesn't get a lot of media attention for their open-source work, they do contribute a lot.

ok I only have experance with a custom T&L rendering platform but the grid engine worked very well I have used it at a couple of hardware design places and it's CPU blancing was very nice indeed
it supports

Apple Mac OS/X
Compaq Tru64 Unix 5.0, 5.1
Hewlett Packard HP-UX 11.x
IBM AIX 4.3, 5.1
Linux x86, kernel 2.4, glibc >= 2.2
Linux AMD64 (Opteron), kernel 2.4, glibc >= 2.2
Silicon Graphics IRIX 6.5
Sun Microsystems Solaris (Sparc) 7 and higher 32-bit
Sun Microsystems Solaris (Sparc) 7 and higher 64-bit
Sun Microsystems Solaris (x86) 8 and higher

see http://gridengine.sunsource.net/
I would look at what software you want to run BEFORE buying hardware

regards

John jones

Apple, in contrast, has embraced Open Source and is delivering a better consumer experience."

What is up with you people and Apple?!!

My God! Give it a rest.... Please. You're killing us here!

I can't get away from the Apple worship even if I block apple stories. It's everyway.

Don't get me wrong, I have nothing against Apple. But Apple is just another Corporation who's goal, as with all other corporations is to, *gasp*, maximize profits for its shareholders.

Ironically Sun ( http://sunsource.net ) and IBM has done orders of magnitude more for Open source than Apple. And at least Sun gets beaten up everyday here. Apple though is worshipped to the point that it is frickin' nauseating to the rest of us.

Come on guys, fanboys just aint cool.

With sun you have the obvious OpenOffice and NFS. They've also open sourced a lot of other software and have provided resources for other projects. Have a look at http://www.sunsource.net/ to see how much Sun contributes to open source projects. People don't like some of the licenses Sun uses because it gives Sun too much control still. Having something GPL'd doesn't make that any better. Just look at what happend with Emacs and XEmacs when a company started paying the FSF to make enhancements to Emacs that they needed but had an uphill battle with RMS who had final say into what happened with Emacs. That's when XEmacs was forked out of it.

Sun has contributed a lot of code to Gnome (accessibility api, work on sawfish, improved usability, tons of documentation and help). They do provide kenel patches, Tim Hockins used to be very active on the linux kernel mailing lists when Sun was working on Sun Linux and still supporting Cobalt servers.

Also Sun is pushing a linux desktop, JDS. And it's pushing hard in different areas. Where is IBM in this? IBM's take on linux, provide it with our servers since we can run websphere on it so that websphere seems cheaper because you don't "have" to buy a Windows Server OS.

Also do a search on the kernel mailing lists. You'll see more references to Solaris than to AIX. Sun had published a lot of papers regarding how they did things and these served as a good guide for many linux kernel hackers. You'll see lots of comparisons to how sun does things. Not just how well it performs to solaris but actually details on how it was implemented in solaris, especially in the case where solaris performed better.

Not saying that linux is a rip off of solaris in case anyone misreads that. I'm saying that Sun has a history for supporting open standards and shares a lot of what it knows and people could benefit from that. Tanenbaum

Everyone needs to remember Open Source is not Linux. Sun does a lot with the Apache Software Foundation.

Sun even provides "scholarships" for open source projects and non profit entities to pay for licensing of some of it's technology that for profit entities have to pay for.

Pointing to a list of kernel changes made in one version to indicate that IBM is the better open source participant is a limited view of open source.

I don't listen to much that RMS has to say. Only so many people in this world can be college proffessors, develop software for free and eventually get a few 100k every so often in awards and money in speaking engagments. The majority of software developers need to be able to make money developing software, they don't have the luxury of clinging to such lofty ideals. How far would all of this gone if RMS had a family to support? Maybe this is why RMS has no family to support? According to him, his child is the GNU project. How many of you can do something like that? Nothing against RMS, it takes a lot of dedication to do what he's doing but it's not very practical for everyone to be doing that.

The Java Community Process Sun set up is pretty good. Individual membership is free. You can help guide the direction of Java. It helps keep things from really going to far astray the way Sun set it up. Which is good for the people that build apps on Java.

Open sourcing Java doesn't really do much for the developer community as most developers build on top of Java, not in it. The people that would benefit would be people like IBM, BEA and Oracle as well as OS companies. The majority of the developer community is made up of the ones building their apps on top of j2se and j2ee. Open Source some great tools and then you're talking. Sun opensourced NetBeans. There's a lot of debate over NetBeans vs IBM's Eclipse. I'v

So let's say I have a massive computing need for, oh say, chip design. Chip designers, like Texas Instruments, Cirrus Logic, or General Semiconductor, require massive amounts of CPU time and even more memory. Sun's ultra-high-end offerings are worthless, since you simply can't cram enough RAM into their higher-end Enterprise servers.

That is nonsense. How about you show me a PC that takes from 96 to 192 Gigs of ram and can use most of that in a single process? You can't. Sun's midrage servers can do that sort of thing.

If you are doing BIG chips, you are almost certainly going to need either an IBM, HP, or Sun Unix box somewhere in the design flow. Linux just isn't there yet to handle the really big stuff.

Your idea about distributed processing is great... if the software supports it. There is still a lot of EDA software that is single threaded doing tasks that are either hard to split, or can't be split. And that is assuming that you can afford the extra seats of software to actually use it in a distributed computing scheme. Since there are tools that cost $750,000+ per seat there don't tend to be a lot those those dedicated to grid computing.

And now that major vendors are offering Linux versions of their design tools, we are no longer tied to Solaris.

I doubt that you would have ever really been tied to Solaris. You could always go to HP or IBM for most vendors. Now you can also do to Linux for the stuff that will fit. Not all of it will fit though if you are doing anything substantial.

Oh, you want those servers to load balance/load share? To be in a cluster? More $$$'s.

Free and open source from Sun.

Want RAID?

Disksuite is free from Sun.

Want some kind of SAN solution? ... But when they're Linux, clustering is free (software-wise.) And while the hardware costs for RAID and SAN remain high, the software to make them work is dirt cheap compared to anything you would have to buy for Solaris.

So, what software would you use on Linux that you wouldn't or couldn't use on a Sun?

... no longer do we stick Ultra 5's and 10's on the designer's desktops, now they're running their tools on Microsoft Windows 2000/XP.

I would guess from this that you aren't doing anything too tough since practically every serious EDA vendor (Synopsys, Cadence, Mentor Graphics, etc) has pretty much bailed from Windows for their tools to do real chips as opposed to FPGAs.

Based on your comments it looks to me like you have been out of touch with what Sun has been doing for quite some time.

Looks pretty Open & Free stuff to me.
including a licence that's FSF and OSI approved

Looks pretty Open & Free stuff to me.
including a licence that's FSF and OSI approved

Try again please.

Personally, I think one of the worst ideas is to tie your software and hardware directly together, as Sun has. Their hardware is way behind the times (too slow, and too expensive). It's better to use Linux as a way to lower overall system cost and say, "hey, now you can get a computer for $199!"

Some people think that Sun does have a future as a hardware manufacturer, but I think I will have to agree with the article, they can't win the fight against being squeezed out of the market by cheap Intel/AMD servers running Linux (or Windows..).

They really have to decide where they are going, and find a new way to earn money. I think Java is their best bet. I HOPE they will do something like IBM, and jump on the Linux bandwagon as the main platform for Java. Still, finding a steady and large revenue stream from that could be difficult. I suspect they get some from Websphere and the other one (forget what its called), and maybe some from selling courses in Java, but that can't be enough. If they started charging money for using Java I think they would discover that their customer loyalty would evaporate pretty quickly.

I suspect some people here on Slashdot will crow about the problems Sun is going through, but consider that Sun has actually been good for the Open Source world. If it wasn't for the fact that it is a cheap Java platform, Linux would not be as widespread as it is in the business world. Also, they gave us Open Office, and participates and even sponsors a number of Open Source projects. Ant, GNOME, Tomcat, GNUlpr, Open Office... Sure, most projects are Java related, but that is understandable and it is still more than most of the big companies have given us.

Well, if they die, it will be interesting to see what happens with Java. Perhaps they will Open Source it completely, if not out of the goodness of their hearts, then at least as a poison pill against Microsoft...

[disclaimer: I'm a Sun paid engineer on a Sun sponsored open source project, JXTA]

You musn't have been paying attention if you haven't seen Sun's contributions to open systems.

See Sun Open Source for info on quite a number of non-dead open source projects either initiated or sponsored by Sun. Many are significant efforts like OpenOffice, Sun Grid Engine, the Gnome contributions, JXTA, etc.

In addition to the dozen or so on that page you can add NFSv4, ChorusOS, and others to the list of Sun involved open source projects.

The comment was about open source and Java, not about GNU/Linux. IBM's record of starting and maintaining open source projects isn't that great, whereas Sun has a whole lot more history - see SunSource.net for all the details. But I'm sure there are plenty of folk here who will argue with me :-)

By the way, Sun bought the source rights to Unix (and thus became immune to SCO's unethical behaviour, which I believe IBM are rightly resisting) many years ago, long before SCO held them.

S.