Mandrake To Support AMD's Hammer

ruiner writes "Mandrake has announced their intention to support AMD's Hammer with a 64 bit version optimized for the new CPU. Redhat is also rumored to be following Suse's lead. 'This new generation of AMD Athlon and AMD Opteron processors is extremely exciting. A version of Mandrake Linux dedicated to these powerful 64-bit processors can certainly accelerate MandrakeSoft's growing adoption in the Linux corporate market' said Jacques Le Marois, CEO of MandrakeSoft."

Comments

That's OPTERON!

[Axe]

Or, no - That's Athlon 2. Or no.. Infinium 36..

I am confused..

What will be the Intel spin on 64-bit is better then 32? He he..

Re:That's OPTERON!

[JPriest]

AMD's roadmap is here

dept.

[El Pollo Loco]

from the hammer-time-joke-goes-here dept.

Followed by a press release from AMD and mandrake saying "can't touch this!"

Re:dept.

[bilbobuggins]

Followed by a press release from AMD and mandrake saying "can't touch this!"

you're talking about when you run an athlon without a fan right?

Re:dept.

[Thing 1]

from the hammer-time-joke-goes-here dept.

Followed by a press release from AMD and mandrake saying "can't touch this!"

"The password's hammer time."

"The Simpsons already did that."

Re:dept.

[salmo]

Despite the humor factor of MC Hammer, I think a real suggestion for a theme song would be the Breeder's "Divine Hammer". It would also work well since they just released their new album Title TK. Although, I admit I haven't heard it yet.

Re:dept.

[Sj0]

Everybody should just grow up. Do you really expect a 1.6 GHZ processor to run as cool as a 486? Perhaps you have slightly unrealistic expectations of an excessively powerful, bleeding edge chip?

Next thing I'm bound to hear from the same vein: "ha ha ha! My twin cylinder dirtbike engine runs cooler and uses less gas than that V12 supercharged DOHC motor! the V12 sux0rz!"

regardless of whether you were being humorous, I've seen it too many times, and decided to say something about it.

OT: Mandrake Forum the last avenue?

[twilight30]

Right, this is definitely offtopic, but why is MandrakeForum *never* the first to reveal this kind of thing? I switched to Debian for different reasons, but it has always amazed me that the community site for Mdk users is seemingly the last to have this information.

Given that Mdk is an avowedly newbie-oriented distro, one would think that the company would have a clear interest in getting this out to its 'channels', to the party faithful, *first*.

Re:OT: Mandrake Forum the last avenue?

[justsomebody]

"June 27th, 2002 - MandrakeSoft cooperates with AMD to support x86-64 architecture - Mandrake Linux 9.x for the upcoming 64-bit AMD Athlon and AMD Opteron processors is expected to be commercially available by the beginning of 2003."

This news are published on the main www.mandrake.com page????? You don't even have to log into channels. Also there was a rumor long time ago. But news are officialy published after signing cooperation. News != Rumors.

I hope your reasons to switch to Debian were better than this bashing.
I don't really use MDK, except on my notebook.

why is this such a big deal?

[wowbagger]

I'm not knocking Hammer, but why does everyone act like the Itanium and Hammer versions of Linux are the first 64 bit versions? I was running 64 bit Linux several years ago on my Multia!

Once you have GCC that will compile for the target arch, and you have the needed changes to Linux to support that arch, why is it more than bunch of builds to get a 64 bit version? Many (perhaps even most) apps are now 64 bit clean (unlike certain other criminal OS's).

Why does everyone ignore the MIPS and Alpha versions?

(and OT: When will a MIPS version of Linux with full support for the extra hardware in an Indy come out?)

Re:why is this such a big deal?

[quasi_steller]

I don't think that the issue is "The first 64 bit Linux distro". We all know that there have already are 64 bit Linux distro's, and that they are great. The issue is that companies who have a major investmen in x86 will now have a linux option (without having to compile everything themselves).

Re:why is this such a big deal?

[Anonymous Coward]

Dude, you are a sad little manwhore. You have no clue about Solaris, do you? It's dying and for good cause. They haven't done anything to improve that OS in over a decade.

Re:why is this such a big deal?

[Rob Kaper]

I'm not knocking Hammer, but why does everyone act like the Itanium and Hammer versions of Linux are the first 64 bit versions?

Noone does. Everybody acts like the Itanium and Hammer will be the first 64 bit versions that will get substantial market share in both server and desktop environments. Like it or not, x86 is pretty big and important, so x86-64 and i64 will be pretty big and important.

Re:why is this such a big deal?

[joestar]

There have been Mandrakes for SPARC & Alpha but apparently the market for these machines is very limited. I can remember a Linux expo with an Internet Café full of UltraSPARC running Mandrake 7.0, that was impressing!

Anyway I think it's interesting to notice the move of Mandrake towards the corporate market. They started as a desktop distribution, and now, as they are known as a desktop distribution, they start to release products targeted to corporates (see the Mandrake Prosuite, the Single Firewall, or the very recent Advantech Firewall Plus (apparently only in Europe). I think it's an excellent strategy because as they have an "easy to use" image, it a big advantage to convince corporations to use their products.

Re:why is this such a big deal?

[guacamole]

They haven't done anything to improve that OS in over a decade.

You are truly an idiot if you think so. As for Linux, no matter how much it has improved, it still has a long way to go to match Solaris what solaris has had for years. now.

Re:why is this such a big deal?

[Afrosheen]

I can remember a Linux expo with an Internet Café full of UltraSPARC running Mandrake 7.0, that was impressing!(sp)

Especially impressive to me considering I can't get SILO straightened out on my Ultrasparc with Mandrake 7.1...or X running..or anything else that'd make it a usable box. Damn sparcs.

Re:why is this such a big deal?

[p3d0]

Like it or not, x86 is pretty big and important, so x86-64 and i64 will be pretty big and important.

IA64 has no similarities whatsoever with x86. The only things that could make ia64 "big and important" are (1) its own merits, or (2) Intel's clout.

Re:why is this such a big deal?

[Rob Kaper]

You just admitted yourself that Intel is the big similarity between x86 and ia64.

Suse?

[Spyky]

For those who don't know, because its very unclear from the article, Suse was the first (or at least before Mandrake) linux distro to announce Hammer support.

Check it out here

-Spyky

Exciting...

[javilon]

Why do they allways use the word "exciting". Do they copy and paste from each other?

Re:Exciting...

[quasi_steller]

Exciting!

Re:Exciting...

[JanneM]

Well, what do you expect? Ninety percent of press-releases are overhyped; they only exist to make something look big and significant. You might want something like this:

"In a totally pedestrian move, Mandrake - as utterly expected - announced a totally unsurprising future port to the Hammer architecture. As everybody have realized for the past year, Linux will be running on the platform, and this of course includes all the major distributions. Spokesmen for the company added: 'of course, no end-users will see any actual results from this announcement for another year or so, when actual systems will be available'. Added the AMD spokesperson: 'As far as novelty value goes, this announcement is the equivalent of watching paint dry.' Both did add that they do get a little free press out of the announcement, making it more valuable than the paper it's printed on."

/Janne

Re:Exciting...

[Ed Avis]

Yeah, it's strange the way the 'press release' has mutated into something no self-respecting journalist would ever use. It used to be that the art in writing press releases was to make them sound like an interesing story, not mentioning your own company too obviously, so that lazy journalists would more or less copy the whole thing.

Even with Internet standards of journalism, I don't think any news site would take much notice of press releases like this one. So what is the intended audience?

Re:Exciting...

[JanneM]

The point of press releases today are more often than not just another way of showing that there is activity. There is often little or no intention of actually inform anybody; instead the point is in showing "interested parties", that includes investors and shareholders, that Things Are Happening. That use unfortunately debases the use of press releases for 'real' news, of course.

Now, I'm not saying this is not interesting, or 'real' news. It might be, or might not; I'm not sufficiently knowledgeable about this aspect of the business to decide, but it the very existence of 'look at me!' press releases does debase the value of the announcement.

/Janne

Re:Exciting...

[m_evanchik]

Speaking as someone who has worked in the technology press, I can tel you that writers love press releases.

Haven't you noticed that most tech-reporting is crap.

Some of the most successful tech writers just regurgitate press releases. Then the company they shill for rewards them by buying advertising in their magazine, or even better, buying the rights to reproduce the article for publicity purposes.

Tech writers are mostly lazy scum of the earth, who never met a freebie they didn't like.

Re:Exciting...

[eddy]

Some of the most successful tech writers just regurgitate press releases.

Ted Bridis of the AP dabbled with that, which of course means all his and the rest of APs tech news goes unread over here. (He may or may not be with the AP still, I have no idea)

Re:Exciting...

[Anonymous Coward]

Purely on principle, I would buy a product advertised thus. Once. I'd buy the upgrade / next model if the ads quoted unbiased performance numbers. On a specification that's of general concern. I'd buy a Mac if "twice as fast as a P4" wasn't followed by the invisible subtext "on 3 seldom-used and highly-optimized Photoshop filters which happen to make heavy use of instruction sequences the G4 executes efficiently." I bought a Seagate X15-36lp because the press releases said it kicked every other drive's ass. And they were right. And it does. A year later, it's finally in jeopardy of losing the pole position. Regrettably, well-marketed but poorly-executed products succeed time and time again so my purchase of good-quality parts is likely lost in the froth...

64-bit Mandrake? Download it now.

[User 956]

You could just download it now for Itanium processors.

"MandrakeSoft announces the availability of the Release Candidate of Mandrake Linux 8.1 for Intel Itanium Architecture. The Itanium 64-bit architecture is used for servers and computers which run highly demanding business applications, such as large database, ERM/CRM applications or advanced computing."

Re:64-bit Mandrake? Download it now.

[ReNeGaDe75]

x86-64 is different from the Itanium's IA64.

x86-64 is an extension to the regular x86 instruction set while the Itanium's instruction set is totally different.

Re:64-bit Mandrake? Download it now.

[User 956]

Correct, however it's still useful if you have an Itanium. (Though, if you have an Itanium, you probably know enought that you won't be running Mandrake)

Re:64-bit Mandrake? Download it now.

[Afrosheen]

Mandrake isn't just for newbies. There have been lots of polls before and the majority of apache sites are running the Advanced Extranet Server tuned version from Mandrake. Actually I think it's either the second or third most popular server distro. Mandrakeforum has talked about this alot and other sites/polls have confirmed it.

Mandrake has alot going for it actually. Decent security out of the box, the advanced apache server, good samba, etc. etc. plus it runs on damn near anything.

Re:64-bit Mandrake? Download it now.

[Billly Gates]

"..Mandrake has alot going for it actually. Decent security out of the box, the advanced apache server, good samba, etc. etc. plus it runs on damn near anything."

Please replace Mandrake with OpenBSD in your post. Thank you.

Re:64-bit Mandrake? Download it now.

[Afrosheen]

Really? OpenBSD ships with Apache Advanced Extranet Server with Thawte enhancements just like Mandrake? Wow I learn something new every day.

redhat rumours

[scotfl]

I'm sorry if I'm being a bit dense here, but nowhere in this message can I find anything suggesting RedHat is going forward with x86-64.

All I see is a (good) suggestion for altering the current set of RedHat install CDs to allow for processor-specific optimisations in the kernel/system binaries. There is mention of how this would carry through into x86-64 support, but this is just a suggestion from someone outside RedHat.

Manufacture support by any chanche?

[GdoL]

Linux is mature enough so that manufactures like those, (AMD, INTEL and other bigs) should themselfs start to buld support and not wait to the Linux boudles companies to do so. Organize teams like they have with MS to build the processor focused on the software and not the opposite.

We had so manu troubles with ATHALON and the INTEL similar to instal RH amd Mandrake that it was almost tempting to change to a less powerfull machine.

I don't understand this

[Anonymous Coward]

Why does Mandrkae have to crow about something that is actually being developed by Linus, Ingo and others?

I'm absolutely sure that Debian will also support AMD's Sledgehammer and so will Slackware and Redhat and SuSE.

What also bugs me is the security advisories from all these companies....it sounds like they are doing the work (for example in the case of Apache's fix, it came from Apache and all the distro's just need to have one consolidated announcement saying LINUX rather than Mandrake/SuSE/etc)

Hmm...

[GreyWolf3000]

I'm going to get modded down for this, but I'll go ahead and say it anyways:

In the past, kernel development has quickly caught on to newer, more powerful processors and features...MTRR (I know what MTRR is thank you very much...), 3D Now instructions, etc. Why the need for official support? And AMD is also going along with this Palladium thing? If you ask me, there's a lot of companies grandiloquently talking up promised "corporate partnerships" that I don't see happening. I mean come on, folks, all AMD has to do is release some rudamentary specs and the kernel team can get right on it.

I'm wary of "features" that require support from the entire OS and not the kernel...

Re:Hmm...

[Thing 1]

(I know what MTRR is thank you very much...)

I don't. By the power of Google, is this it?

(I'm joking of course. I'm pretty sure it's "Mean Time To Recovery" or "Mean Time To Repair", from other Google results.)

Re:Hmm...

[Hoser McMoose]

I think you're missing a few things here.

x86-64 is not just a new feature of x86, it's completely extending the architecture to 64-bits. The Hammer chips will boot a standard i386 install just fine, however there's an entire new architecture being added to the kernel to support these processors! This is a LOT more involved then just adding support for MTRRs, 3DNow, SSE and the like.

Subtle troll

[zulux]

The entire reason for this post is to see how many people gripe when I mis-spell Atheleon.

Re:Subtle troll

[MrSloth]

thats Athlon, gripe gripe gripe gripe

Re:Subtle troll

[Anonymous Coward]

Hey, jackass, you misspelt Athalon.

Re:Subtle troll

[shepd]

You misspelt misspell. ;)

Worth noting

[Anonymous Coward]

FreeBSD already announced report for x86-64 architecture a long time ago.

define "support"...

[spoon42]

Ok, so they're going to support this new platform. That's great, but:

Does "support" mean "put out a press release and then recompile all our packages once the kernel and gcc people do all the work"?

If they're not contributing this is just a bullshit attention-grabbing publicity stunt.

Re:define "support"...

[justsomebody]

"If they're not contributing this is just a bullshit attention-grabbing publicity stunt."??????

Isn't ability to dowload their distribution with sources freely a contribution?

Re:define "support"...

[Anonymous Coward]

That's what it looks like to me.

Organizing binary CDs for x86 chip optimizations

[Anonymous Coward]

- Optimizing select packages for i686 and Athlon in a NFS install

For the past 2 years, I have maintained my own NFS-based install of
RedHat. It is very easy to update, add to and re-mix an NFS-based
install in a matter of minutes, hence why I do it. More recently I have
discovered that I can rebuild RPM packages from source to be optimized
for particular architectures, drop those .i686.rpm and .athlon.rpm
packages into my NFS directory, and the Anaconda installer (after
re-running genhdlist) will choose them based on what chip is in use in
the system where RedHat is being installed. E.g., while RedHat includes
different kernels for i386, 586, 686 and Athlon, and an i686 optimized
GLibC to complement the standard i386 one, I also drop in an Athlon
optimized GLibC package as well as i686 and Athlon rebuilt SDL and
multimedia RPMs to complement the standard i386 ones (because MMX/SSE
and MMX/3DNow! optimizations make a _big_difference_ in performance for
multimedia).

- Issues with distributors x86 chip-specific packages on CD

While this works great for NFS-based installs where storage is,
virtually, limitless (based on your HD capacity), CD is not. So RedHat
often find itself in a crunch when it comes to choosing what packages to
optimize for which x86 architectures. So far, it has limited its
optimizations to the kernel, only recently adding Athlon as an option in
7.1+, and just an i686 optimized GLibC. Other distros, namely Mandrake,
come with all their packages i586 built (Pentium optimized) by default,
which means its performance is often higher than RedHat's
out-of-the-box. As all of the desktop and most server systems I have
assembled and/or installed over the past 18 months have been Athlon, I
find myself yearning for a distro that is at least partially Athlon
optimized in CD form. At the same time, I'm sure many others are
looking for an i686 optimized distro too, so this only adds to the
possible "waste" in accommodating options.

- Possible solution? Make binary CD #1 optimized for specific chips.

Although it is far from perfect, I think I have at least a "partial
solution" to my, among others', lust for x86 chip-specific optimizations
in distro binary packages. A simple solution would be to release
different binary CD #1s for each x86 optimized version. E.g., release a
i386, i586+i686 and Athlon binary CD #1 for the next major version of
RedHat. This x86 chip-specific binary CD #1 would including the
approximately 25-40% of packages that would affect system performance
the most by being specific x86 chip optimized. Those packages that are
not affected by binary optimization then would go on binary CD #2, much
like the current 2 binary CD combination we have now in RedHat 7.x. And
if the default RedHat install should bloat to requiring 3 binary CDs "as
standard" to accomodate additional packages RedHat makes "standard"
(which is very likely), this approach still only requires 1 CD change
for each x86 chip-optimization wanted. And it would most likely result
in higher performance, especially for multimedia on i686/Athlon, than
Mandrake's "everything's i586" approach.

- Addressing the x86-64 future ...

Furthermore, the introduction of x86-64 (AMD's "Hammer", possibly
Intel's "Yamhill" as well) is going to force the requirement of x86
optimized binaries for performance/addressing considerations. As if
optimizing kernels and GLibC for different x86 chips already amounts to
over 100MB of RPMs on current RedHat binary CDs, x86-64 is going to
require at least an additional 100-200MB of other core libraries to
optimized for x86-64. So this distribution approach also has the added
benefit of already addressing the issue -- in a way that could already
be well understood by RedHat users if adopted prior. Or if RedHat 8.0
is not released until x86-64 starts appearing in end-user products
(e.g., if RedHat 7.3 is the next release, which I would NOT suggest
changing the binary CD layout for being a "revision" release), just use
this approach to address the issues with properly supporting x86-64 on
the same CD as an Athlon optimized binary set.

- Distribution impact and other considerations ...

Probably the most direct impact, other than the small, additional
rebuilding time to produce the added binary CDs, is the addition of more
.iso files to RedHat's servers and mirrors. Again, I feel this is
already going to be an issue with accommodating x86-64 anyway, so why
not address it with i586+i686 and Athlon+x86-64 CD options? If RedHat
is looking to minimize the number of downloads and/or create a "default"
binary CD #1 that addresses "all systems," it could package the
"normal/default" binary CD #1 with just uniprocessor versions of the
kernel for the i386, i586, i686, Athlon and x86-64 plus i386, i686 and
x86-64 optimized GLibC. That way, it wouldn't differ from the current
binary CD approach today in any way other than the fact that the
multiprocessor and enterprise kernels wouldn't be on it. Since those
kernels are not used by 99% of systems anyway, the "default binary CD"
strategy would be, more or less, nothing different than now. But power
users, IT professionals and x86-64 adopters would have a clean, easy way
of installing and supporting Pentium, Athlon and x86-64 systems with
optimized packages "out-of-the-box."

- Reviewing the approach ...

To conclude, here is a more "visual" representation to the approach I
suggest:

Today: Tomorrow (Suggestion):

Binary CD #1 -> Binary CD #1 "Default"
Binary CD #2 -> Binary CD #2
Binary CD #3+ (if included) -> Binary CD #3+ (if included)
Doc/Src/Other CD ... -> Doc/Src/Other CD ...
NEW! Binary CD #1 "Pentium/II/III/4"
NEW! Binary CD #1 "Athlon/x86-64"

As you can see, two new Binary CD #1s are introduced.

"Today" Details ...

"Today" Binary CD #1 "Today" Binary CD #2
- i3/5/686/K7 uP/MP/Ent Kernels - kernel source, support
- i3/686 GLibC - lesser binary libraries
- common binary libraries - lesser binary programs
- common binary programs - devel libraries
- common noarch packages - lesser noarch packages

"Today" Binary CD #3+ (if incl) "Today" Doc/Src/Other CDs
- additional libs/prog/devels - as optional
- non-redistributable software

"Tomorrows" Details ...

"New" Binary CD #1 "Default" "New" Binary CD #2
- i3/5/686/K7/8 uP-ONLY Kernels - kernel source, support
- i3/686/K8 GLibC - non-opt'able libs/prog (i386)
- optimizable binary libs (i386) - lesser libs/prog (i386)
- optimizable binary prog (i386) - devel libraries
- noarch packages

"New" Binary CD #3+ (if incl) "New" Doc/Src/Other CDs
- no change from "today" - no change from "today"

Which includes these 2 new CD #1 "options" ...

Binary CD #1 "Pentium/II/III/4" Binary CD #1 "Athlon/x86-64"
- i5/686 uP/MP/Ent Kernels - K7/8 uP/MP/Ent Kernels
- i5/686 GLibC - K7/8 GLibC
- i586-opt binary libraries - K7-opt binary libraries
- i586-opt binary programs - K7-opt binary programs
- select i686-opt bins/libs - select K8-opt bins/libs

[ NOTE: i686 = P2, P3 or P4; K7 = Athlon; K8 = x86-64 ]

Mandrake

[Anonymous Coward]

Kicks Ass!

Because MIPS and Alpha processors suck!

[benzapp]

I think the simple answer to this question is that no one cares about MIPS or the Alpha. Both platforms are now dead. They died because they were too expensive, and didn't provide sufficient performance increases over current 32-bit processors. Hell, as far as I understand it, Only IBM latest PowerPC processor outperforms the fastest Pentium 4's and Athlon processors.

Now, 64-bit processors will be affordable for everyone but the super rich, and they will natively run all the programs used by 98% of computer users. Thats sweet.

I mean, talking about the Alpha? Who the hell uses that. That has been dead for years!! It was dead as soon as Compaq bought DEC. MIPS?? Hmm, outside of my Casio PDA and my Nintendo 64, I don't think I have ever even seen a MIPS machine. If you had said you were running an RS/6000 in your crib I would feel a little sympathy. Of course, AIX doesn't cut it for you.

Re:Because MIPS and Alpha processors suck!

[Anonymous Coward]

> MIPS?? Hmm, outside of my Casio PDA and my
> Nintendo 64, I don't think I have ever even seen
> a MIPS machine.

Silicon Graphics, anyone? I admit that his is hardly an argument for the non-deadness of MIPS, though.

Re:Because MIPS and Alpha processors suck!

[skyhawker]

I mean, talking about the Alpha? Who the hell uses that. That has been dead for years!!

How about the third and fourth fastest supercomputers in the world? Including the fastest in Europe.

Not to forget that Titanic, by far the highest grossing motion picture of all time, was rendered with the help of an AlphaPC server farm running Red Hat Linux. Admittedly, that was five years ago. But it was still a 64 bit OS.

I must confess my bias, because I have two Alphas faithfully running Red Hat Linux at home. So far as performance is concerned, I'll be surprised if the Intel Itanium or even the AMD Hammer chips will compare favorably to comparable Alphas on floating point performance, which is very important in many high end applications.

Of course, I could be wrong....

Re:Because MIPS and Alpha processors suck!

[benzapp]

To my knowledge, SGI was the only major company to use them... and as far as I know, that tradition is over. SGI is either using sparc or powerpc processors. Further, I seem to recall thinking that the reason this was going was because the MIPS processors were no longer available.

Re:Because MIPS and Alpha processors suck!

[benzapp]

First of all, supercomputers are a little bit different... Having 1024 processors, no matter what the speed kicks ass. Further, I am not saying MIPS processors don't have the performance to be used at all.

I am saying that MIPS processors are no longer being used in the manufacturing of new computers, nor are Alpha processors. Further, cost is a big factor here. I personally never had the money for an SGI or Alpha based system. You are one of the fortunate few, the elite, to be so privilaged.

Good for you. But that doesn't change the fact that Alpha and MIPS processors are dead, and they were never used by the mainstream. I seem to recall reading somewhere that fewer than 10,000 Alpha systems were ever sold.

Also, The hammer and Itanium processors will far exceed the Alpha on floating point operations. I mean, if they can't beat a four year old processor, they shouldn't be in business. I believe the IBM Power4, the latest PowerPC chip is still fastest.

but thats a cool link on the supercomputers... cheers.

Re:Because MIPS and Alpha processors suck!

[cwebster]

>Hell, as far as I understand it, Only IBM latest PowerPC processor outperforms the fastest Pentium 4's and Athlon processors.

that would be IBM's POWER4. PowerPC chips only went into very low ned RS/6000's (and i dont think its even used in the RS/6k line anymore) and into mac's (before they went G*).

Re:Because MIPS and Alpha processors suck!

[linzeal]

I bought 10 indys for 1000 bucks 3 years ago. I still have 4 of them, the rest I gave to a high school a while back.

Re:Because MIPS and Alpha processors suck!

[FattMattP]

To my knowledge, SGI was the only major company to use them... and as far as I know, that tradition is over. SGI is either using sparc or powerpc processors. Further, I seem to recall thinking that the reason this was going was because the MIPS processors were no longer available.

SGI machines still use MIPS processors not PowerPC or SPARC processors. Up until two years ago MIPS Technologies was a subsidiary of SGI at which point they were spun off into an independent company. MIPS processors are clearly still available and many companies are using them.

Re:Because MIPS and Alpha processors suck!

[The Axe]

Uhm, the Alpha CPU still outperforms a lot of CPUs today. I don't know about the latest chips from Intel and AMD compared with the Alpha, but about a year ago there was a comparison between 64-bit chips and a few 32-bit chips, the Alpha won by a huge margin.

MIPS is still being used by SGI, and you can not tell me SGI is dead. They provide a big chunk of the workstations and renderfarms used to render special effects in movies (yes yes, I know, Linux is taking over in many place :).

If I'm not mistaken, Samsung still produces the Alpha chips.

Re:Because MIPS and Alpha processors suck!

[Fnord]

You're right about SGI not using Sparc or PowerPC processors but SGI is attempting to depreciate the MIPS. The latest Origins can have either MIPS or Itanium modules in them, and the Itanium configuration is described by them as the "high performance" option while MIPS is for "legacy configurations".

Re:Because MIPS and Alpha processors suck!

[Anonymous Coward]

Nobody has taken the Alpha seriously since DEC went under. Sure, it's fast, but it's also a dead end, and everyone with a brain has known it for the last 3 years.

And yeah, I'll tell you that SGI is pretty near death. They're limping along with MIPS after a failed attempt to switch to Intel. God knows what their plan is now.

Re:Because MIPS and Alpha processors suck!

[Anonymous Coward]

This is just plain wrong. Why do you post stuff if you do not have any clue? SGI uses MIPS exclusively. There will be an Itanium-2 based ccNUMA machine/architecture in the future, but this will NOT deprecate the MIPS line. Also there are new MIPS processors coming along.

The Itanium workstation that was sold by SGI was the OEM intel box, the same as HP or Dell shipped...crappy stuff!

BTW the most important reason to stick with MIPS is probably Irix - Linux is just not an alternative for it, and I doubt it will be anytime soon!

Re:Because MIPS and Alpha processors suck!

[Hoser McMoose]

I'll be surprised if the Intel Itanium or even the AMD Hammer chips will compare favorably to comparable Alphas on floating point performance, which is very important in many high end applications.

At the moment, only the new 1.0GHz Alpha EV68 chips are faster then the current P4 processors, and the Itanium and Athlon are right up there as well. The rather substatial lead in FP performance that the Alpha used to have has virtually disapeared these days when compared to x86 chips. The very fastest EV67 chips are slower then the fastest x86 chips (note: I'm using Spec CFP2000 for comparison here, if you know of any FP benchmarks that run on both platforms I'd like to hear about them).

As for the future, Alpha's time on this planet is very limited. EV7 is still supposed to come out, and I've heard from reliable sources that it should post some very impressive scores for floating point due to it's HUGE memory bandwidth. However Intel's Itanium 2 is also supposed to post some rather impressive scores (they're talking about 1300-1350 in Spec CFP2000, which would put it ahead of the current champion Power4 processors from IBM). AMD's Hammer won't be any slouch either, as it's on-chip memory controller should boost it's score quite nicely.

Like it or not, Alpha is dead. It's been sold many times and basically salavaged for scrap (Intel now owns most of the old Alpha technology). You're mentioning Titanic is actually quite approriate, because that was the last gasp for Alpha. If you look at new movie production, they're moving to x86, just like all the rest of the world. MIPS might have a future in the embedded space, where it's currently second to ARM. Alpha technology might have a bit of a future in Intel's IA-64 (though I'm skeptical as to how well they'll be able to integrate the pure-RISC Alpha technology into the VLIW IA-64 technology), but as a product on it's own, stick a fork in it.

Re:Because MIPS and Alpha processors suck!

[skyhawker]

I am saying that MIPS processors are no longer being used in the manufacturing of new computers, nor are Alpha processors.

The Alpha based supercomputers that are numbers 3 and 4 on the list were installed in 2001 and the one that's number 6 was installed this year. Clearly, the Alpha is nearly dead, but it's not quite dead yet.

I seem to recall reading somewhere that fewer than 10,000 Alpha systems were ever sold

The three supercomputers I just mentioned (nos 3, 4, and 6) have a combined count of 7600 CPU's. Admittedly, that's only three systems, but your statement is just ridiculous.

Also, The hammer and Itanium processors will far exceed the Alpha on floating point operations. I mean, if they can't beat a four year old processor, they shouldn't be in business.

Floating point operations have never been Intel's strong suit. For similar clock speeds, their CPU's get roundly trounced by Alpha, PPC, and others as well. I found an interesting page here that contains some enlightening comparisons. An 833 MHz Alpha beats a Pentium 4/2000 on the Specfp2000 benchmark, and it remains generally comparable throughout the tests. I suspect that Intel will beef up their results with the Itanium, but they have had the benefit of all the Dec/Compaq experience now that they own it.

One thing I love about the Alpha is how incredibly cool it runs. It's amazing to open the case of my 533 MHz AlphaPC to look at relatively sparsely populated motherboard and note that the CPU has but a modest heat sink and no fan.

Re:Because MIPS and Alpha processors suck!

[Strog]

If you have followed SGI much then you would notice that the Itanium systems were recently moved to the legacy pages and the only systems currently available are MIPS. No more Intel Linux Origins any more now.

Re:Because MIPS and Alpha processors suck!

[Anonymous Coward]

MIPS systems are also used extensively in high end graphics applications. I don't think they'll be overtaken by x86's very soon.

Re:Because MIPS and Alpha processors suck!

[ToLu the Happy Furby]

At the moment, only the new 1.0GHz Alpha EV68 chips are faster then the current P4 processors, and the Itanium and Athlon are right up there as well. The rather substatial lead in FP performance that the Alpha used to have has virtually disapeared these days when compared to x86 chips. The very fastest EV67 chips are slower then the fastest x86 chips (note: I'm using Spec CFP2000 for comparison here, if you know of any FP benchmarks that run on both platforms I'd like to hear about them).

As for the future, Alpha's time on this planet is very limited. EV7 is still supposed to come out, and I've heard from reliable sources that it should post some very impressive scores for floating point due to it's HUGE memory bandwidth. However Intel's Itanium 2 is also supposed to post some rather impressive scores (they're talking about 1300-1350 in Spec CFP2000, which would put it ahead of the current champion Power4 processors from IBM). AMD's Hammer won't be any slouch either, as it's on-chip memory controller should boost it's score quite nicely.

This is a good summary, except that ignorant slashdotters reading this might not realize how far ahead P4, Alpha, Power4 and Athlon are in front of all other CPU architectures (i.e. PA-RISC, US III, Itanium1, MIPS, and most especially G4*) in single-CPU SPEC performance. Just to be clear: Alpha's performance is still at or near the top.

* no, there's no official G4 SPEC entry (because Apple is too chicken), but c't benchmarked it and boy does it suck. SPECint performance on par with a 933 MHz PIII, and SPECfp on par with a 500 MHz one.

Second, you probably haven't seen the leaked slide of SPEC scores for the 1250 MHz EV68 (they should be official at spec.org real soon now), which put it almost precisely equal to 1.3 GHz Power4 SPEC scores, despite not having the dubious advantage of 128MB L2 cache (under normal operation the Power4 cache is shared amongst 4 or 8 cores, but for SPEC one core gets it all).

Third, EV7 will have a substantial lead in SPECfp upon release (and might briefly take SPECint as well), although there's no doubt that Power4 and P4 will continue to improve, and Itanium2 on SPECfp and Hammer on SPECint will also be contenders.

But the most shocking part of all this is that, unlike Athlon/Hammer, P4, Itanium and Power4, Alpha is achieving all this performance on someone else's standard fab process (rather than tweaking the process to fit the chip design, which is a huge huge help; compare performance of those architectures where the designer owns the fab--P4, Athlon, Power4, Itanium--to those where it doesn't--PA, MIPS, SPARC). And, unlike current Athlons, Hammer, P4s since January, and Power4, it's an old .18um process. Alpha still gives more performance with fewer engineering resources than any other chip, and EV7 will only widen the gap.

Like it or not, Alpha is dead. It's been sold many times and basically salavaged for scrap (Intel now owns most of the old Alpha technology). You're mentioning Titanic is actually quite approriate, because that was the last gasp for Alpha. If you look at new movie production, they're moving to x86, just like all the rest of the world. MIPS might have a future in the embedded space, where it's currently second to ARM. Alpha technology might have a bit of a future in Intel's IA-64 (though I'm skeptical as to how well they'll be able to integrate the pure-RISC Alpha technology into the VLIW IA-64 technology), but as a product on it's own, stick a fork in it.

Again, you're correct. But the death of Alpha is entirely to do with marketing and zero to do with performance. Before Power4, Alpha was the clear single-CPU performance leader in the 64-bit market, and if the Alpha team had gotten more resources and support from Compaq, Alpha's performance lead would still be huge.

That said, all of this seems slightly academic considering that for the past year and the forseeable future, the SPECint leader has been and will be a commodity x86 chip costing ~$600 (against 64-bit competition costing 100x that per CPU). The P4 seems to have taken on the Alpha mantle: world-beating SPEC performance with a high-clocked small-die chip utilizing innovative microarchitectural features and excellent circuit-level design. That it has done so despite being hobbled with the x86 ISA is even more impressive.

Re:Because MIPS and Alpha processors suck!

[Anonymous Coward]

The alpha processor is the most efficient CPU up to 4 processors, MHz by MHz. After that, with 8 CPUs or more other processors are more efficient (even if the alpha still may outperform them by a margin!).

A 16 CPU setup of the IBM Power4 compiled the Linux kenel in 7.5 seconds... The alpha cannot do that.

see www.spec.org for details.

Re:Because MIPS and Alpha processors suck!

[incabulos]

I'm sure someone has probably mentioned this already somewhere, but the Sony PS2 uses MIPS processors, a pair of them IIRC. One to drive the graphics subsystem exclusively (R4000?), and one as the general purpose system processor (R10000?). I'd consider the PS2 to be a fairly widespread device, and as Linux kits are selling out as fast as they hit the shelves, I wouldnt consider Linux/MIPS, nor MIPS as a platform to be considered dead :>

Also, I believe the R14000 and R16000 series MIPS cpus are still in development, handy when MIPS Inc. is just an IP company and doesnt need to kit out expensive fab plants to do their particular magic..

Re:Because MIPS and Alpha processors suck!

[Anonymous Coward]

"One thing I love about the Alpha is how incredibly cool it runs. It's amazing to open the case of my 533 MHz AlphaPC to look at relatively sparsely populated motherboard and note that the CPU has but a modest heat sink and no fan."

I had to scrape my jaw off the floor after reading that.

My 667Mhz 21264 has a heatsink the size of a paperback book, with TWO fans on it. The box has a 400Watt PSU to run the single Alpha, 2 IDE hard drives, and some whacky Number Nine video card.

My older 21164 400 something Mhz box had a 4 inch fan on the front with complicated ductwork to duct it directly onto a heatsink that was _HUGE_ for the time (1997). I nearly burned my finger off once when I didn't wait long enough after powering down.

Suffice to say, my experiences with Alphas have been very different from yours, at least WRT power consumption.

If you're going to "mispel" it...

[ScumSucker]

at least do it right! Its Athalon!

Re:If you're going to "mispel" it...

[Afrosheen]

And a 10 cpu cluster is called...

*drumroll*

A decathlon! Thank you, I'll be here all week. Remember to tip you waitstaff.

MandrakeSoft cooperates with AMD to support x86-64

[Anonymous Coward]

To provide customers with a powerful, scalable 64-bit Linux distribution with native 32-bit compatibility-

Atadena, Paris - June 27, 2002, MandrakeSoft today announced they are cooperating with AMD to port Mandrake Linux to the forthcoming eighth-generation AMD AthlonTM and AMD OpteronTM processor-based platforms.

Delivering on its commitment to innovative development in the Linux arena, the MandrakeSoft research and development team has begun work on adapting the Mandrake Linux operating system to run optimally on the upcoming AMD processors.

AMD is collaborating with MandrakeSoft to ensure rapid porting and commercial availability of Mandrake Linux for 64-bit AMD Athlon and AMD Opteron processors. AMD Athlon and AMD Opteron processors, based on x86-64 technology, have been designed to support large, intensive workloads providing high-performance solutions for desktops, workstations and servers. The new processors can benefit customers by providing 64-bit application support, while also offering optimized, native support for 32-bit Linux applications to provide a smooth transition from a 32-bit to a 64-bit environment.

"This new generation of AMD Athlon and AMD Opteron processors is extremely exciting. A version of Mandrake Linux dedicated to these powerful 64-bit processors can certainly accelerate MandrakeSoft's growing adoption in the Linux corporate market" said Jacques Le Marois, CEO of MandrakeSoft.

"The 64-bit version of Mandrake Linux optimized for the upcoming eighth-generation AMD Athlon and AMD Opteron processors can provide enterprise customers with a powerful, reliable and easy to install and use Linux distribution", said Robert Stead, Director of European Marketing, AMD.

Mandrake Linux 9.x for the upcoming 64-bit AMD Athlon and AMD Opteron processors is expected to be commercially available by the beginning of 2003.

About Mandrake Linux

Mandrake Linux is a powerful operating system that is available for the Intel Pentium®, AMD Athlon®, and PowerPC® processors. Mandrake Linux includes many graphical administration assistants & wizards that make it intuitive and fun to use while providing all the power and robustness of other Linux systems. Hundreds of included applications make it an ideal solution for both enterprises and individual users. Mandrake Linux is seen as the most feature-rich, multi-purpose Linux operating system ever made available.

About MandrakeSoft

MandrakeSoft provides a trusted interface between users of information technology and open source developers. The company offers its enterprise, government and educational customers a set of GNU Linux and Open-Source software and related services, and user-friendly and highly competitive information technologies. In addition, MandrakeSoft offers technologists committed to open software and courseware a trusted channel to offer their services.

The company has technologists in over 20 countries, and is traded on Paris Euronext Marche Libre (Euroclear code: 4477.PA; Reuters code: MAKE.PA) and the US OTC market (stock symbol MDKFF). "Born on the Internet'' in late 1998, MandrakeSoft has established headquarters in the U.S.A., Montreal, England and France. Please visit the Web site, http://www.mandrakesoft.com for more information.

About the AMD OpteronTM Processor

The AMD Opteron processor is based on AMD's eighth-generation processor core which is planned to mark the introduction of the industry's first x86-64 technology. This technology is planned to preserve companies' investments in 32-bit applications while allowing a seamless transition to 64-bit computing as those companies require.

The AMD Opteron processor is designed to deliver high-performance server and workstation solutions for today's most demanding enterprise applications. The processor is designed to be scalable, reliable and compatible, which can result in lower total cost of ownership. Key AMD Opteron processor innovations include an integrated memory controller, which reduces memory bottlenecks, and HyperTransport(TM) technology, which increases overall performance by removing or reducing I/O bottlenecks, increasing bandwidth and reducing latency.

About AMD

AMD is a global supplier of integrated circuits for the personal and networked computer and communications markets with manufacturing facilities in the United States, Europe, Japan, and Asia. AMD, a Fortune 500 and Standard & Poor's 500 company, produces microprocessors, Flash memory devices, and support circuitry for communications and networking applications. Founded in 1969 and based in Sunnyvale, California, AMD had revenues of $3.9 billion in 2001. (NYSE: AMD - News).

NOTE: Linux is a registered trademark of Linus Torvalds. AMD, the AMD Arrow logo, AMD Athlon, AMD Opteron, and combinations thereof, and AMD-8151 are trademarks of Advanced Micro Devices, Inc. HyperTransport is a trademark of the HyperTransport Technology Consortium. Mandrake and Mandrake Linux are trademarks of MandrakeSoft. All other trademarks and copyrights are owned by their respective companies.

Do we really need another layer on x86 PC arch

[aaron_pet]

I've heard lots of bad stuff about the x86 architecture...

hacks upon hacks...

Do any of these new 64 bit processors offer a way out?

Is backward compatability with 8086, 286, 386 (sx and dx), 486(sx and dx), 586, 686 now trivial, or solved with a (long term low significance cost to benifit ratio)?
</flame>

Re:Do we really need another layer on x86 PC arch

[Junks Jerzey]

I've heard lots of bad stuff about the x86 architecture...
hacks upon hacks...

That's overstating things. Go to Intel's site and dowload the PDF file describing the Pentium II instruction set. It's absolutely huge. There are hundreds and hundreds of instructions, and the funny thing is that only minority--maybe 30%--really matter. The rest of them are things like MMX, old instructions that are no longer relevant, and lots of peculiar special purpose instructions that are rarely used. And this is only the Pentium II, so it doesn't include all of the SIMD instructions added with the P3.

So most of the cruft comes from old stuff that was relevant at one time, and now there's no way to get rid of it. It isn't because of hacks, per se.

Re:Do we really need another layer on x86 PC arch

[mikefoley]

If people wanted a clean architecture, they would have bought Alpha's. The thing is, the vast majority don't give a rats ass about computer architectures. They only want to know one thing:

Does it run my application?

x86-64 does, without re-compiling. People can move at their own pace, when they want to move. IA-64 is making the mistake of Alpha. Few apps, high price, force people to change to get on board.

Re:Do we really need another layer on x86 PC arch

[ozzmosis]

All of them do.

thats the whole point of new architecture.

Re:Do we really need another layer on x86 PC arch

[cheezedawg]

Intel's Itanium is different than AMD's 64 bit processor. It is a completely new RISC architecture and Intel and HP developed together. It does not build upon the x86 "hacks". Compatibility with existing x86 code is achieved with an on-chip x86 emulation unit.

Re:Do we really need another layer on x86 PC arch

[Anonymous Coward]

Itanium is not RISC it is VLIW, besides even "RISC" machines are n't RISC. Look at the instruction decode section of the the latest SPARC processor it does n't look very simple to me.

Re:Do we really need another layer on x86 PC arch

[Hoser McMoose]

You know, everyone complains about how x86 is just "hacks upon hacks", but guess what, it's the cheapest platform out there and it outperforms damn near everything else!

Right now, only Power4 and the Alpha EV68 are the only chips in the same ballpark as x86 when it comes to raw processing power. In Spec CINT2000, x86 (P4 2.53GHz) manages a score of 896, to beat out ALL other chips. The 1.3GHz Power4 scores 839, while the 1.0GHz Alpha EV68 scores 679. In Spec CFP2000, IBM and DEC/Compaq/HP redeem themselves slightly, with scores of 1266 and 960 respectively, as compared to 879 for the Intel chip.

Alpha has been on the chopping block for some time. Sun? They're SPARC chips and especially their systems as a whole have some advantages, but when it comes to raw processing power, $2000+ USIII chips are beaten by $100 x86 chips by a factor of almost 2 to 1 in most tests.

x86 may have it's problems, and it's certainly not the nicest architecture to write assembly for. But guess what, no one writes in assembly anymore, and compilers have been better optimized for x86 then for any other architecture. So why do people always complain about x86? Well let's see:

Complaint #1: Backwards compatibility holding it back.

Answer: As mentioned above, there are only two chips out there that are faster then x86 chips, and both of those cost a LOT more and neither are all that much faster. Backwards compatibility has also been the reason why x86 has sold so well over the years and is the best supported architecture with the widest selection of software.

Complaint #2: Not enough registers.

Answer: x86 is slightly limited by it's registers, but due to it's CISC nature it doesn't need as many registers as a RISC chip does and no where near as many as RISC chips. What's more, register renaming allows for x86 chips to have a lot more registers then it initially seems like it has (you just can't use them all at the same time).

Complaint #3: x87 stack is dumb

Answer: There is no answer to this, except maybe SSE2. The x87 stack is dumb (I'll grant the complainers this one! :> ). However even this has been mostly worked around by the compilers, to the point where x86 chips are now some of the fastest chips out there at floating point.

Complaint #4: Decoding CISC instructions to RISC instructions is a waste of time.

Answer: The decoding front-end of a moern x86 chip is now a rather small proportion of the total chip and can usually keep the backend well fed (assuming that the decoders can get data from the cache fast enough). What's more, perhaps the most interesting feature of the P4 is it's trace cache, which caches already decoded instructions, there by removing the decoders from ~95% of all instructions executed (AMD doesn't have this yet).

So where does AMD's x86-64 fit into all of this? Well, it maintains backwards compatibility with all previous chips (x86's real strength), it should offer rather impressive performance (particularly the Opteron and 64-bit Athlon should improve the I/O performance of x86 systems, which had traditionally been one of the weak points). It also helps on the register front, in that x86-64 doubles the number of available registers. It doesn't do anything to fix the x87 stack, but even that is mostly compensated for by the generally high quality of x86 compilers. Other then that, the real reason for it is to add seamless 64-bit support.

Re:Do we really need another layer on x86 PC arch

[Paul Jakma]

good news:

Hammer in 64bit mode:

- double the number of registers.

- linux ABI will use the SSE registers for fp (apart from double double), AMD are strongly encouraging people not to use x87 for FP in 64bit mode.

Re:finally

[dastrike]

IIRC, it is only the AltiVec unit that is 128 bit in them.

Re:finally

[Anonymous Coward]

which does wonders for your photoshop filters

Re:Sad news ... Stephen King dead at 54

[Anonymous Coward]

is this confirmed true or not?

Too Many Directions?

[shoemakc]

It always appeared to me that Nandrake/Suse was most strongly targeted at the desktop, while redhat maintained the most alignment with the server market.

A serious comitment by Mandrake/Suse to the server market my be pulling mandrake in too many directions at once.

-Chris

Great! But.....WAIT!

[artsygeek]

Mandrake needs to make the 64-bit version a more server-friendly design....and make it less desktop-ish....for now.... I mean, Mandrake's great in that I can get any member of my family to install linux or config their system or whatever with it, but it needs to "cut the fat" that comes with it being an all-inclusive desktop OS.....of course that's just my opinion...I could be wrong...

Re:Great! But.....WAIT!

[foonf]

Mandrake needs to make the 64-bit version a more server-friendly design....and make it less desktop-ish

You know what, I would agree with you that mandrake in general should be less bloated. But I don't think that de-emphasizing the desktop for x86-64 is wise at all--if anything they should make it even more desktop-focused than the current x86 version. Why? Well, its very likely that some kind of Linux (along with NetBSD of course) will be the only usable operating system on the x86-64 platform when it arrives. There will be a brief window, when the hardware is available, but Windows hasn't been ported yet, that running a free *nix will be the only option for anyone who wants to use x86-64 hardware.

Of course it will be used in a lot of servers, but these people won't choose mandrake anyway (I would be surprised if any of the major distributions don't have x86-64 distributions available soon after the hardware is publicly released, as long as a good compiler is available and most common software compiles with minimal trouble). But certainly people will be drawn to this platform for high-end desktop applications also, and the very fact that there will be no full-performance native alternative will render many of the typical arguments against desktop linux moot, and offer an opportunity to draw users in who might not consider it otherwise. So I think there NEEDS to be a good desktop distribution for x86-64 available when the hardware arrives.

Re:Great! But.....WAIT!

[Anonymous Coward]

Don't forget that the 'ClawHammer' desktop variant will be the mainstream replacement for the 32-bit Athlon.

Re:Great! But.....WAIT!

[Afrosheen]

ACk, where are my mod points when I need them. -1 overrated.

Mandrake IS server-friendly. Consider this: alot of companies are running NT and win2k right now. Their admins are toying with linux to see if they can save some cash and add some stability to their networks. These point-and-clicky MSCE's are going to shoot for a distro they can point-and-click their way to success on. Mandrake happens to be one of these distros.

Also there's no gun in the installer that points to your head and says "install ALL the desktop cruft or I'll shoot you". I got a minimum install on my old Compaq rackmount server that weighed about 100 megs and you could make it even smaller. This was DURING install, not post-install. Individual package selection is where it's at, you gotta do an expert install and have some foresight.

Maybe you haven't seen or used or installed Mandrake in awhile. I suggest you try it on a server soon just for fun. There are thousands of people (if not tens or hundreds of thousands) using Mandrake to serve right now regardless of it's desktop orientation. I happen to be one of them.

Re:Sad news ... Stephen King dead at 54

[Anonymous Coward]

Yes, try Reuters.

Re:finally

[Anonymous Coward]

There is no such thing as a widespread general purpose 128bit processor...

And SSE2 has 128bit regs as well.

wrong on just about every account

[g4dget]

The first 64 bit processor in widespread use on the desktop was the Alpha. It even ran Windows NT. How much more "PC world" can you get? It is still being used with Linux.

Furthermore, the G4 is not really a 64 bit processor. The biggest difference between 32 bit and true 64 bit processors is that 64 bit processors use 64 bit pointers.

Apple puts out some nice hardware and software, but they are rarely the first with anything and they don't really do much research anymore either.

Re:wrong on just about every account

[styrotech]

The first 64 bit processor in widespread use on the desktop was the Alpha. It even ran Windows NT.

Although an Alpha running NT is for all intents and purposes a 32bit processor :)

64 Bit... sounds powerful...

[Space Coyote]

... So rumour has it it will start KDE in less than a minute. :) </old joke&gt.

Re:64 Bit... sounds powerful...

[Anonymous Coward]

Wow, that's overstating it. Rumour has it over here that it can possibly start KDE in less than 2 hours.

Face it, Slashdot. X11 and everything built for it is fucking stupid.

Linux framebuffer -- w00t w00t! Fuck you X!

Re:64 Bit... sounds powerful...

[Anonymous Coward]

Aye... but how long will it take for Emacs to boot? :)

Why are we still compiling for the 386?

[frooyo]

Talk about how people say MIPS and Alpha is dead (just a little trolling) - who honestly uses or knows of someone using a 386 anymore.

Why don't all distro companiesstart atleast compiling for 486 and also have at the least a distro that is compiled entirely for, say 586 (like Mandrake).

I don't understand why companies like RedHat (who make a great solid modern distro) don't make available for the more modern processors a distro optimized for it.

Why sacrafice new technology (speed) for the old and thus making the new run at the speeds of the old?

Re:Why are we still compiling for the 386?

[Anonymous Coward]

Most modern mandrake distros are actually 686 (pentium pro) compiled.

Most of the rest are 586 to optimise for the general pentium class, and still be reasonable on 686 processors.

If you look for a 486 distro, you will actually have a hard time, as I found out when I wanted to set up my old 486 as a nice web/ftp server for work. Mandrake 7.0 did the job, after trying slackware, microlinux, etc.

I would like to see in 'instant server' distro for 386/486, though, with Apache, Samba, Perl, PHP and MySQL/Postgres and maybe just linuxconf for the rest. No X server or anything else. Would be nice.

Instant server for modern machines, with the same config, but 686 optimised, with good security defaults, etc. would be great :)

::thinks about sending off a quick mail to mandrake in that regard::

Re:Why are we still compiling for the 386?

[Anonymous Coward]

Yeah and it pisses me the hell off too.

For all the Linux market's bitching about bloatware there's still no good reason that SuSe should require 64 MB just to install.

There's also no good reason for them to ship 686 code. Just make it an easy recompile at installation option, because the guy with the Pentium 8 and 5 gigs of RAM is going to compile a kernel a lot faster than some poor schlob with a 386.

If a distro won't install on the least of my boxen, I'm not going to put it on the best either. I may have to break this rule when the x86-64 comes out though :)

Re:Why are we still compiling for the 386?

[Anonymous Coward]

Why?

I guess so that it will run on practically every PC.

(Probably not the answer you wanted? :) )

Re:Why are we still compiling for the 386?

[Anonymous Coward]

So, you are saying that RedHat compiles the ENTIRE system for 586. I think not, it is 386 with ONLY the kernel has 686 enhancements.

I don't know of any distro other than Mandrake that compile the entire system with 586 or higher.

Re:Why are we still compiling for the 386?

[Anonymous Coward]

I have said to myself many times over 'WTF' over this exact thing. Why doesn't RedHat and others allow a distro to be COMPLETELY compiled for the 586/686 processor.

Other than Mandrake, I don't know of any distro (and Mandrake is a piece of crap newbies shit)

Re:Why are we still compiling for the 386?

[Anonymous Coward]

Alpha is dead

When was it alive?

Re:Why are we still compiling for the 386?

[Anonymous Coward]

Gentoo all the way !!!!!!

Re:Why are we still compiling for the 386?

[Anonymous Coward]

Well there are many reasons. First and formost there is very very little performace gain for most apps. If the kernel gets rebuild for your arch and some other super intensive stuff you add the system yourself gets compiles 586+ then your at maximum performace. Xfree86, your word proccessor, spreadsheet, etc just won't see much performace gain nor will most other important apps. What will happen is they will likely use more ram and the binarys will have a larger footprint on disk. They are also very likely depending on gcc versions to be considerably less stable. In general compileing for better then 386 is not gonna help enough to out weigh the costs in other areas.

Re:Why are we still compiling for the 386?

[Anonymous Coward]

Dont forgot 'Linux from Scratch' ...

Re:Why are we still compiling for the 386?

[Anonymous Coward]

Xfree86 ... just won't see much performace gain

The X-server is massive. Now that deserves a WTF.

Re:Why are we still compiling for the 386?

[Anonymous Coward]

I guess so to too. Because you and I know how there are thousands of apache servers out there running on a 386 and all.

Re:Why are we still compiling for the 386?

[joestar]

I'm amazed about such ignorance. It's been _3_ years that Mandrake comes compiled for i586 and more (it comes with gcc optimizations for pentium instructions set), so you can't run a Mandrake on a i386 or a i486 machine (they released a version of Mandrake 7.0 for i486 but it seems that as the demand for these processors is very low, they abandonned it).

Re:Why are we still compiling for the 386?

[MarcQuadra]

I wish Mandrake would produce i586, 1686+mmx, and k7+3dnow distros. It would be a lot easier than recompiling everything myself. GCC is about to have automagic use of 3Dnow! instructions (should be in 3.2, IIRC) which would boost system speeds a bunch.

Re:Why are we still compiling for the 386?

[treke]

glibc is also compiled with processor enhancements.

Re:Why are we still compiling for the 386?

[Anonymous Coward]

Well as a part time developer which would you rather develop for an arch that has documented features/bugs to avoid for the last 2-3 years or an arch that you have no fscking idea whether they are actually improvements or could they be bugs in the way the compiler generates code.

Having the latest and greatest processor or compiler doesn't automagically make the code better.

Remember the bubble sort (the one that we were told to avoid like the plague) using a new processor we could get "acceptable performance" while the original algorithm is "flawed" it still sorts our data at a reasonable speed. So why learn about quicksort/shellsort/heapsort when the processor is "fast enough" to make our jobs "easier".

I will get nostaligic here but does anyone remember programming for an 8 bit system? The requirements were to make it small and as fast as reasonably possible. Now we have programs that take (at a minimum of 1-2Meg that a talented programmer could do in an optimized version that would take 20-30% less memory and have the equivalent speed).

Sure hd space is at an alltime low but there is a reason we have 120Gig drives. People have no idea what constitutes a good program does it matter whether it takes 4M to run your program when you "waste" a whole lot of memory caching something that used to be helpfull back in the days when it took 30 seconds to display a result of a calculation that today would take 3 seconds?

I shared a computer with 2-3 other people back in the early 90's and we never did fill up the 20M hard drive we had on the old "master" but as soon as we got a 120M drive and switched to Win3.X and a 386 we we found that we could fill it up. This is just anecodetal evidence but it still says a lot, Our using habits didn't really change but the software that we were using changed so voila the hard drive space suddenly became scarce that we were constantly shuffling data between the HD and some floppies so we could run the "latest and greatest" update of the program we were already using but took up 3-4 times the space.

Re:Why are we still compiling for the 386?

[cdn-programmer]

Right - demand for i486's is low.

Well - I had a couple of them and found OpenBSD is broken on the AHA1542 SCSI cards, broken on the AHA2842 32bit VL bus SCSI cards and these old boxen won't run big IDE drives.

Mandrake doesn't run at all on them and I don't feel like putting up redhat which will run just fine.

So I gave up and gave in to progress and spent $100 bux and picked up a pair of 200Mhz machines with 64MB ram. These run just fine thank you. My only mistake was not buying 4 of those machines!!!

My point is that high quality pentium class machines are so cheap there is zero reason to use those old 386's and 486's. I have some to give away for free if anyone is interested. Alas, it is sad but they really are at the end of their useful life.

Re:Why are we still compiling for the 386?

[p3d0]

Aside from the fact that I still run a 386, you should know that there is a difference between processor compatibility and processor optimization. Code can be mostly optimized for a Pentium-IV and still run on a 386.

Anyway, do you have any evidence that breaking 386 compatibility would provide any performance benefit?

Re:Why are we still compiling for the 386?

[shepd]

486 code runs slower on all other intel processors than 386 code.

586 code won't run (technically) on 486s, which puts people using linux on old machines for fun or for routers out of the picture.

Of course, that doesn't stop Mandrake, which I believe uses 586 code for everything (I could be wrong).

HTH, and I could be wrong.

Re:Why are we still compiling for the 386?

[Cro Magnon]

The problem is that code optimized for the Pentium might run SLOWER on a K6 than 386 code does.

Re:Alpha does Titanic, Athlon does Episode II

[benzapp]

Also, FYI.. Athlon MP processors were used to make Episode II... not too shabby, for a 32-bit processor.

AMD Technology Used for Production in Star Wars: Episode II Attack of the Clones

SUNNYVALE, Calif., Jun 27, 2002 (BUSINESS WIRE) -- AMD (AMD) today announced its collaboration with JAK Films and Industrial Light & Magic (ILM), a division of Lucas Digital Ltd. LLC, in the production of "Star Wars: Episode II Attack of the Clones".

JAK Films used AMD Athlon(TM) MP processor-based workstations in its labs to advance the art of cinema storyboarding through digital pre-visualization. ILM used a high-performance cluster of AMD Athlon MP processor-based servers designed and manufactured by RackSaver.

"Our goal was to paint an accurate picture of what `Star Wars: Episode II Attack of the Clones' would look like long before filming was even started," said Dan Gregoire, director of the pre-visualization lab, JAK Films. "Our AMD Athlon MP processor-based systems are ideal for running the powerful software we rely upon."

"Given the challenge of producing top quality visual imagery in a compressed time frame, we really appreciate the performance boost we got with our AMD Athlon MP processor-based systems," said Michael Kiernan, Manager of Systems R&D at ILM.

"We at RackSaver, with our expertise in design and manufacturing of high-density, high-performance computing clusters, are excited to have met the critical computing demands required in the production of `Star Wars: Episode II Attack of the Clones,'" said David Driggers, CEO RackSaver.

"George Lucas is admired for the many technical advances he has contributed to the film industry. The innovative AMD technology used by ILM and JAK Films is no exception, " said Ed Ellett, vice-president of product marketing for AMD's Computation Products Group. "JAK Films pre-visualized literally every scene in `Star Wars: Episode II Attack of the Clones,' which helped Lucas translate his vision to the big screen with exceptional accuracy. ILM is widely recognized in the film industry as a standard-bearer in visual effects and digital animation."

Customers using systems based on powerful AMD Athlon MP processors can experience tremendous productivity, and fast turnaround time on a variety of graphically intensive tasks ranging from creating complex special effects sequences to manipulating video, audio content and high resolution still images.

Note: For additional information about AMD's involvement in the production of "Star Wars: Episode II Attack of the Clones" please visit: http://www.amd.com/us-en/Processors/ProductInforma tion/ 0,,30_118_756_759^566~31514,00.html

if I could ask a lamer question...

[Anonymous Coward]

..hmmm, for someone who just surfs, does xmms audio, etc, exactly how "much better" would the computing experience be switching to one of these advanced processors, with the modem remaining a static semi low speed? If you aren't a gamer or a server or do advanced graphics or scientific configuring, are they really worth it?

Honest question. I'm still on a 200 mghz machine, it seems zippy enough with a lot of ram in it for what I do mostly. I understand thast there always has to be newer and better, at one time the PPro I have was the latest and best, but is there really that a significant change in normal home casual surfing? I've used moderately faster machines at the library that are on broadband, the download pipe seems more important than the processor, or am I wrong on that? which is a better deal buck for buck, get broadband if possible, or get the latest and greatest hardware, with the caveat you can't afford *both*?

Re:if I could ask a lamer question...

[Anonymous Coward]

Yeah the average person probably doesn't need this.

I've got a 5x86 100 and 32 MB of RAM and if I'm honest with myself it does most everything the average user really NEEDS.

Of course I've upgraded twice since then, but I still use the old 5x86 box for Dreamcast coding.

I'm pumped about the mathematical possibilities of the 64-bit chip. I'm better my CGH program will be hella impressive with the extra juice.

Re:if I could ask a lamer question...

[Hoser McMoose]

Well, it all depends on just which applications you use and how patient you are.

A while back, I was using a K6-III+ 450MHz system with 192MB of RAM, and this seemed pretty speedy (the K6-III was a very speedy chip for all non-gaming type applications). But then I set up a new system for a friend of mine, and used it for a week before I dropped it off to her. It was only a 750MHz Duron with 256MB of memory, but the difference in performance was VERY noticeable to me. Browsing and reading e-mail was MUCH snappier. I also suddenly realized just how often my usenet reader paused for half a second or so, and how much quicker and easier using this program seemed on the faster system. Programs started up much quicker. Now, some may write this off to just differences in hard drive speed, but really they were using nearly identical drives (the K6-III+ had one of the infamous IBM 75GXP drive (knock on wood, still working great), while the Duron had an IBM 60GXP drive, slightly faster, but only by about 5% or so). Once I gave that system to my friend and went back to my "slow" 450MHz system, I found myself getting rather impatient with a lot of programs (particularly web browsers, mainly Mozilla and occasionally IE or Konqueror). Before long, I was looking into an upgrade.

I eventually got an Athlon 1700+ with 512MB of DDR SDRAM, and let me tell you, it's quick (for today at least! :> ). Unfortunately BOTH of my sticks of memory have gone bad, and guess what, I'm back on the old 450MHz system and boy does it ever seem SLOW! (trust me, I'm checking FedEx's website on an almost hourly basis to see how far my replacement sticks have made it!). Now, I'll be the first to admit, I'm VERY impatient when it comes to computers, however I think that a lot of other people would find that if they actually tried using a somewhat faster system, they would feel that the ugprade is worthwhile. Other people who are more patient or possibly just use less processor/ram intensive applications (the memory access speeds of PPros is quite poor, and the memory access speed of Pentiums is terrible as compared to modern systems, not to mention those also have butt-slow cache, so it's not just procesor-intensive apps that are faster on new systems).

Still, all that being said, if it came down to a choice between broadband and a decently fast computer, I'd take broadband any day. They'll have to pry my cable modem from my cold dead fingers dammit! Fortunately for me, this isn't an issue since broadband is cheaper then any dial-up I could ever find around here. Dial-up is definitely in it's dying days in Canada (more then half of Internet connections are now broadband).

Re:if I could ask a lamer question...

[Billly Gates]

I would happily downgrade to my old p166 for broadband in an instant. If you pirate oops I mean sure mp3's all day long then the connection is the most important bottleneck in regards to performance. Infact many new portable discplayers like the sony I am using now easily has enough cpu power to decode mp3's. Hell, even a 486 could has enough power to do it. However your 200mhz system is useless for video games and quite painfull under linux when you actually need to compile anything. I remember waiting for 45 minutes to compile one of the 2.2 kernels. My current system ( pIII700) can compile the bigger 2.4 kernels in around 20-25 minutes and I played with an athlonXP1800 which could do it in around 5 minutes. My guess is your system would take at least an hour if not more to compile just the more recent 2.4 linux kernels. Unless your happy playind doom1 and doom2 and quake my guess is that your system is quite useless for anything modern besides playing cards. Go to asus's website and find a local reseller for a cheap AMD duron motherboard or an intel celeron motherboard if you are worried your powersupply could cause problems. You can get a great cpu/mobo combo for like $160. Add another $125 for 256 megs of DDram. Then buy a pny el-cheapo geforce3mx for like $70. Whats great about the newer motherboards is that they will fit in your case. So for $350 you could have a brand new computer that will play quakeIII, run kde and gnome fine in Linux and you will still have money left over for broadband. I never buy new computers anymore. I just build them.

I am in an interesting diloma myself. I want to try gentoo Linux which compiles all your apps automatically when you have a dependency problem. This would be a big royal pain in the ass with my old pIII700 when I have gigs and gigs of apps installed. So I am looking for an athlonMP2000 dual cpu system which would make me cream in my pants. The problem? Money. How often would I do a gentoo style "emerge world"? About maybe once a month. But I need to buy a new car and money problems are hurting me. So thinking wisely I will only upgrade modestly and just do an "emerge world" at night. So I know exactly how you feel. Life sucks but we have to make smart choices and use what we have.

Thanks for letting us know, ass-muncher.

[rasactive]

Wanna cookie?

Re:Thanks for letting us know, ass-muncher.

[cha0sadddddddd]

regarding your .sig....lots of people use extacy that way....it's called "booty bumping" if memory serves.
(i have a little sister who goes to lots of raves and her and all her little friends do it this way)

Re:Thanks for letting us know, ass-muncher.

[Anonymous Coward]

Heroin is also taken this way, from time to time. Not like I would know.

Eeerh..

[dmouritsendk]

Why sit around and wait for somebody else to compile the images for you? Use a source based disto dammit, one that grabs the source to your system and then compiles it. As other people commented before, most linux apps are allready 64bit ready. Because most needs to be compileable on MIPS/ALPHA/SPARC platforms.

GCC(and binutils) supports compiling to x86-64 , this is still experimental though. But searching/looking abit in the mailinglist archives @ x86-64.org show that stuff like qt allready are comiling fine, only needing a wee-change in the makefile(Link).I think thats quite impressive, and im willing to bet good money that GCC has production class x86-64 support by the time the processor is actually available to buy.

So, armed with gcc and a version of Gentoo, Linux From Scratch or any other sourcebased disto that supports compiling the entire system from scratch. You will beable to create/compile your very own system, which can be WAY more optimized that anything a vendor does(i cant really see how its possible for a precompiled kernel images cant be optimized to a system).

The the only bad thing about thiese kinda of disto s is that big large packages as x/gnome/openoffice/what-ever takes for ever to compile. But on 64bit processor, who cares =)

clarification (from a debian guy)

[Xtifr]

They're not crowing about the fact that they can compile for these systems, they're crowing about the fact that they are going to compile for these systems, and support them. Since compiling code into working binaries and supporting those binaries is what Mandrake does, I think they're justified in crowing about this. As a big AMD fan, I applaud Mandrake for this, even though I use and support Debian myself.

Slackware and Redhat and SuSE may or may not support this platform directly, I don't know. It's certainly not guaranteed. There are plenty of platforms they don't support, even though they could. It's probably going to depend on whether they think they can make enough money off of it.

And yes, Debian will almost certainly support the Hammer as soon as we get our hands on some. But then we're insane, and support everything we can. Who else still supports m68k and ARM? Who else is _adding_ support for HPPA and Super8? We do it because it's fun, not because we're trying to make money.

(As for the thing about security advisories, that's a bit off-topic, but I will say that Debian's security list is intended for Debian's users, so that they know when officially supported packages are available, and it's not our fault that bugtraq decided to subscribe to our list. Complain to bugtraq if it bothers you that much.)

Nope, Not Gonna Happen.

[mchappee]

>'A version of Mandrake Linux dedicated to these powerful
>64-bit processors can certainly accelerate MandrakeSoft's
>growing adoption in the Linux corporate market' said
>Jacques Le Marois, CEO of MandrakeSoft.

*swat* Bad CEO! Bad, bad! AMD is not doing very well in the corporate market. PHBs just don't like it. Sure, the grunts that take care of the machines like it, but they (we) just don't have the pull when it comes to writing a check. Imagine this conversation:

Me:"Hey boss, if we went with Corp_PC-A instead of Corp_PC-B, we would save about $200 per computer!"
Boss:"Really? What's the difference between the two?"
Me:"Corp_PC-B uses an AMD processor instead of an Intel."
Boss:"Whoa, now. The processor is pretty important, right? I don't think that we want to skimp there. We better 'go long' on the processor and trim the fat on memory and disk space. Oh, and stick a Trident video card in there for good measure! Silly hacker, what do you know about corporate responsibility. Go fix me turkey pot pie!"

OK, a little creative leeway. The point is that Intel has the aura of a rock-solid performer, while AMD has a hacker/gamer hippie case-modder feel.

Besides, AMD lost it's underdog luster to me when they struck a deal with Microsoft to testify on their behalf for an endorsement.

Matthew
In or near St. Louis? Hire me

Re:Nope, Not Gonna Happen.

[Anonymous Coward]

I doubt your boss really cares about Intel versus AMD. What he cares about is Dell (Intel) versus Bubba's MicroShack (AMD). If Dell or HP shipped mainstream corporate desktops with AMD CPUs, people would have no problem buying them.

[OT] Gentoo

[tunah]

Give gentoo a try. You tell it what processor you have and it compiles and optimises every piece of software you install (including glibc, gcc, linux) for your processor.

This, and some other advantages of compiling from source, make this distro much faster than any other I've tried.

The install process is definitely not for the inexperienced, so your point about precompiled distros still applies, but once it is installed, it is very easy to administrate, and one of the first to get new security packages (I woke up after OpenSSH 3.4 was released and typed 'emerge rsync openssh' and it was installed).

Sorry to come across as yet another offtopic smug evangelising gentoo user... I guess I am :P</rant>

I did not know

[Anonymous Coward]

Mandrake was a supporting AMD Hammer? I thought he was working on E 17

3D-GUIs need support too

[Anonymous Coward]

3D-GUIs need support too.

There will soon enough be 3D-GUIs available from Microsoft and Apple. One Parhelia's next generation friends from ATI and nVidia are here the Joe Schmoe family will buy into these new whizzy dizzy 3D GUIs. Linux need a viable 3D GUI too.

Support www.3dwm.org or www.berlin-consortium.org or whatever other options there are. Their development seem to have stalled in mid-air.

thanks, & please consider more thn just proces

[aaron_pet]

I understand that compatability is good for migration...

I've heard the most bad things about BIOS and bus, and other aspects... not just the x86 processors...

When will it be worthwhile to start getting rid of the not used stuff?

Are they keeping the short word instructions doing what they did, or are they making them do more usefull stuff...

(I suppose you could have a single instruction perform multiple operations... one that is old spec and one that is new spec.

Maybe I am considering something to be a problem that isn't...
but as the instruction set increases, don't we have to have more bits to to describe the instruction?... Now I am realising that memory addressing is the big benefit for larger... bus width?

Also, how much faster are natively compiled for transmeta applications than those that go through the instruction conversion caching? (I think thats a way of putting it)

Thank you for your discourse, I am attempting to learn stuff from this conversation.
</cow>

Re:Sad news ... Stephen King dead at 54

[Anonymous Coward]

Post a url, I've checked about 4 news sites and there's no articles about this. I would of thought his official website might of had an announcement about this too.

Mandrake Forum is mainly desktop oriented right?

[ImaLamer]

How many users there are business/IT managers?

Re:MandrakeSoft cooperates with AMD to support x86

[rapidweather]

Didn't I read somewhere that the AMD Opteron will be able to run both 32 and 64 bit applications, whereas the Intel counterpart will only run 64 bit? This is in response to:

"This technology is planned to preserve companies' investments in 32-bit applications while allowing a seamless transition to 64-bit computing as those companies require."

So, does this mean that something like Mandrake 8, that I am running now (with Mozilla build 2002061108, aka 1.1a) will run on the Opteron? I'm asking this in this thread as it appears some of you know quite a bit about these new processors, and might like to give us some information;-)

take the plunge! get a Miata

[aunchaki]

You can easily find DEC Personal Work Stations (PWS) in 433, 500 and 600MHz speeds for cheap on ebay (they're also known as "Miatas"). These machines are common. I got mine for under $600 a few years ago and it's been great as a workstation and a server.

That and my awesome digital Server 3300R (a rack-mount 400MHz hot-swap alpha I got for $600 on ebay) have run SuSE, Red Hat and are now happily running the latest FreeBSD.

Stop complaining you can't affort it. YOU CAN!

Who needs 64bits?

[thogard]

99.99% of what I run will be much more effecent with 32 bit ints than 64 bit ints. At work we use 64 bit sparcs and they can run in 32 bit mode. They are slightly faster for what we do in 32 bit mode than 64 bit mode.

I don't think most businesses will ever go to 64 bit because its just the wrong step. Most code that will get any speed advantage out of 64 bits will get more advanatge out of 256 bit but thats an insane amount of data to push around for task swaps.

About the only thing that hits the 64 bit sweet spot is database access and disk access routines but in reality they do thouse calcualtions so seldom, that they lose the advanatges with task swaping in most cases. You can do an 32 bit add with carry into another 32 bit add about 400 times faster than you can dump the extra bits on the stack.

If you look at the high speed database base benchmarks, most of them are running on systems where the register stacks don't ever get out of the cache if even out of the register windows (as in Sparc)

With the pc server market no longer subsidising that fastest machines, the gamer market (and super computer field) are all thats left and those markets aren't interesting enough for most compaines to spend billions in R&D.

Re:Who needs 64bits?

[Hoser McMoose]

You're actually quite right in general, 32-bits is enough for most applications and, all else being equal, it IS faster then 64-bits (half as much data to read from memory).

Now, that being said, there are a few reasons for going to 64-bits. The real main reason is the native support for more then 4GB of memory. You can hack support for more then 4GB of memory into a 32-bit processor (see Intel's PSE), but it's not pretty. This was a bad kludge when it was done back in the 16-bit days, and it's still a bad kludge.

The second reason for going to x86-64 in particular instead of standard x86 is that it doubles the number of registers you have available. This is not a 32-bit vs. 64-bit issue in general, but rather an x86 vs. x86-64 in specific issue. AMD estimates that this will boost performance by 5 to 15% on applications, simply by recomiling them for x86-64 (same hardware and same compiler, just x86 vs. x86-64).

As for the people who talk about "why don't we have 256-bit processors yet" though, you're dead-on as to the reason, they would quite simply be SLOW. The "128-bit" processors that you see in modern gaming consoles like the Playstation 2 are talking about VERY different bitness then in the general computing world (what they talk about is more analogous to the 128-bit or 256-bit graphics cards that most PCs use these days).

Re:Who needs 64bits?

[Anonymous Coward]

So what would a 64 bit cpu do for me that a 40 bit cpu wouldn't do?

Debian

[Jagasian]

Yeah, I was just thinking that as I read this article. I thought, "Debian supports every practical platform and then some... whats the big news about some other Linux distro supporting a non x86 platform?"

For those that want to experiment with a auto-hardware detecting graphical user-friendly installer for Debian should try the new (this is a beta release!!! be careful) Progeny Graphical Installer ISO images. This installer is based on the now defunct Progeny Linux installer. Good Linux software doesn't drop dead, but instead gets reincarnated.

Remember, Debian is not commercial and is purely community based. Helping beta test this new installer on all sorts of various systems will help.

-1 for not reading carefully

[vanguard]

You should reread the post that you just called ignorant. He mentions that Mandrake comes compiled for the Pentium and wonders why everybody doesn't do the same thing.

Power4 is a PowerPC

[RabidChipmunk]

PowerPC is the instruction set. Power4 is IBM's most recent implementation. Motorola's G3s and G4s are also PowerPC chips, with an aditional SIMD extention (AltiVec).

Re: Power4 = PowerPC

[benzapp]

There is a common misunderstanding here about PowerPC chips. For a technical paper from the IBM Watson facility on the Power4 architecture, read here.

This research paper clearly indicates that the Power4 processor utilizes the 64-bit extensions of outlined by the original PowerPC consortium. It also indicates that The Power4 refers more to the architecture of processor interconnects than the processor itself. Since the days of IBM writing OS/2, they have always rightly believed multiprocessing and multithreading provides the best performance.

IBM did the vast majority of the work on the PowerPC processor, and owns the name to it. The reason Motorola processors are now called G* is because they do not own the rights to the title. The G3 was to be the PowerPC 670 proccessor as I recall. In all honesty, I amazed that people here think the G3 is not a PowerPC processor. How ludicrous is that? It was a pain apple's ass to get people to dump the 68k series in 1993, could you imagine what would have happened if a whole new archicture was used only five years later with the first G3? Apple would be out of business.

I only wish I had the links on current benchmarks to refute the lunacy that the antiquated DEC Alpha is still the fastest processor but oh well.

SuSE leads the way.

[inc_x]

Of course SuSE supports this platform. Those who follow these kind of things know that it was SuSE who added Hammer support to the official Linux kernel at the start of this year. See here

That's my post referenced! Mod this up please ...

[BitMan]

All -- that LEAP-CF post is mine. That is not a "rumor" but just my "recommendation" to RedHat. Nothing more.

MTRR

[Paul Jakma]

MTRR == Memory Type Range Register

Used to set different policy (uncacheable, write-back, write-combing) to address ranges. Eg, for address ranges that correspond to PCI addresses (ie memory mapped IO addresses), by setting these ranges to write-combining the CPU will try to gather writes up into big writes to make most efficient use of IO bus bandwidth. (ie get higher MB/s out of your AGP or PCI - important for graphics).

see linux/Documentation/mtrr.txt and /proc/mtrr.

Re:MandrakeSoft cooperates with AMD to support x86

[Paul Jakma]

AMD Hammer/Opteron is completely IA32 (ie normal 32bit x86) compatible - all IA32 OSes boot on it, it has a standard IA32 BIOS, applications will run fine on it. If you run a x86-64 OS, then you will be able to run both 32bit and 64bit x86-64 software (side by side).

Ie x86-64 is:

- IA32 (8086 mode et al too - i /guess/)
- standard IA32 BIOS
- additional x86-64 mode

Apparently 32-bit Linux and Windows booted almost first time on early silicon, and they've had absolutely no 32bit compatibility problems - it all works. then it took just a week for AMD to get linux to boot into x86-64 mode (iirc from the talk linked below).

IA64 / Itanium on the other hand is a completely new architecture:

- completely different instruction set
- completely different ABI
- new weird "look it does everything" BIOS (EFI)
- IA32 is /emulated/ in silicon and hence slow

There's a good talk by an AMD engineer on the AMD Hammer arch. given at the recent kernel summit at:

http://ksmp3rep.sf.net/KSMP3s/amd64.mp3

found amongst other kernel summit talks at:

http://linuxkernel.foundries.sourceforge.net/artic le.pl?sid=02/06/26/0116225

Itanium != RISC, == VLIW

[Paul Jakma]

Itanium is /not/ RISC. Go read up on VLIW (Very Long Instruction Word), which is what Itanium is. (as is the Transmeta Crusoe silicon).

Take a look at gentoo for your 486

[Billly Gates]

Try Gentoo. Whats great about this distro is that you can recompile the whole sytem with "emerge world". That also includes all the apps and not just the kernel. I have never used it( yet) but later next month I will have broadband so I can try it out. It uses something called portage which a mix between debian apt-get and FreeBSD's ports, where you can have automatic dependency downloads or you can download the newest libs and recompile not only the kernel but the whole system with the new libs! Great for experimenting with gcc 3.1. Whats ideal about this is that you can recompile the whole system for your 486. Go here for info. Just add the settings to CHOST in /etc/make.conf to "platform=i486" and go to the CFLAGS section in the same file and add the following "o3" to make sure its optimized. After this type "emerge world" and reboot and thats it. Your whole system including all your apps will be 486 optimized. The 2 catches with gentoo is that its not a newbie distro. Make sure you fully read all of the documentation and print it all out before you even start the install. Second it will be painfully slow on your 486 to do a "emerge world". You need at least 64 megs of ram or a big ass swap space to do the install. Expect a literal all nighter for this. If your box is at work do it in the evening and pray it will be done by morning. The pro's are you only have to do the "emerge world" once. The second thing is that it will be fully optimized for your 486 and you will learn a hell of alot of info about unix and linux in general. The third is you will not have the 20 are so bloated daemons installed by defualt like mandrake and redhat. It will be slim which is important on your old computer. I have heard that this makes gentoo faster then any other distro. Even on modern athlon systems. Since Gentoo in my opinion has the best documentation so this shouldn't be too much of a problem.

'instant server' distro

[azimir]

I've got one.

I call it OpenBSD and it runs great. Take a 486 with 16+ megs of ram (32 is nice) and you've got all of that.

No, it's not Linux, but it does have those good security defaults you asked for.

I'm currently running two 486's as firewalls for my DSL line and I feel very confident that I'm locked down and secure here. When I need an Apache, Samba, Perl, PHP, or MySQL (never used Postgres) server I've got another 486 in the DMZ (between the two firewalls) that I muck with and viola! All that on less that 500MB per machine.

It would be nice to see a Linux distro that installs as easily and with as much usability as OpenBSD does. Until that day my servers keep running OpenBSD. If you haven't tried it, do so, you'll like it once you got over the shock that there is no GUI installer and the documentation is actually consistant.

I couldn't recommend diving into the mailing lists until you've read the FAQ on the OpenBSD.org site. It's almost all there.

Happy serving!

Re: Power4 = PowerPC

[Anonymous Coward]

Um, Apple's "G3"'s for the ibook are supplied mainly by... *drumroll* IBM

the G3 is the PPC 750 btw....

Mandrake #1 Linux distribution server

[Anonymous Coward]

Apache-AdvancedExtranetServer = Mandrakesoft

http://www.netcraft.com/Survey/Reports/0205/

- Apache is not linux distribution only
- Microsoft-IIS the name say it all ;-)
- Zeus is not a distribution
- Netscape-Enterprise is not a distribution
- unknown your guess is as good as mine
- Rapidsite is unix
- Webstar Built for Mac OS X and not a distribution
- thttpd ACME-brand Unix software
- Tigershark custom Unix Web server not a linux distribution
- ConcentricHost-Ashurbanipal I dont get info on this but I dont think its a distribution ...

lamer question version 2

[Anonymous Coward]

--second performance issue. I'm the d00d got the 200 mghz PPro system. The mobo (machine is IBM desktop, BTW, quite stout actually) is dual cpu capable, it needs the second chip plus a voltage regulator, fan, etc. This stuff isn't being made new anymore (duh) but I have been occassionaly able to find like an ebay deal or something but haven't popped for the upgrade yet. Is it really worth it for the second cpu in just general over all normal performance? Again, I'm not serving with this machine (might some day though, ya never know). I know some of my desire is just "because" I got that second slot staring at me, it looks lonely. I got this machine brand new in the box salvage, no second processor but was cool to actually get the factory manuals, diagnostics diskm and there was a shrink wrapped copy of NT4 with it I haven't even used yet.

Worth it? Or will it be just "because it's there"?

I do upgrades cheap and as I can afford it, I know that if I start from scrastch with a new mobo , etc, it's better, but for now this is what I got. Ram hasn't been bad, every other month or so I try to afford a new stick of "bigger", eventually all four dimm slots will be filled with 128 sticks. Now they got a blend of sizes, installed correctly.

Also, say I do it, I'm still confused on install issue with SMP. Do I have to reinstall/re compile to get any new dual response? Seems like I have read yes AND no with that question the small amount of research I have seen.

thanks for the tech advice people!

*yawn*

[SaDan]

Ever tried Slackware?

And yes, I have tried OpenBSD.

That's NetBSD, actually.

[fvdl]

It was NetBSD support that was announced a little over a year ago. Should be in the /. archives, and see also this link and more recently this one

A question for Linux users and AMD suits

[vandan]

With AMD's announcement that they will be supporting the DRM push by building in DRM features into CPUs / motherboards, I wonder how long can the Linux community tolerate (and contribute to) their AMD Hammer architecture.
Surely these hardware embedded cyber-cops will require closed-source, patented drivers and will dissolve the usefullness of the GPL and Linux.
http://www.theregister.co.uk/content/54/25905.html >
http://www.theregister.co.uk/content/4/25891.html& gt;
Isn't it time to do something about this now?
Do we really trust AMD to do the right thing, or will they take advantage of Open Source developers right until the last nail is hammered into our coffin and then run all the way to the bank?
HINT: AMD is an American for-profit corporation.