An Interesting Boot Log On Alpha
Here is an interesting Boot log on an Alpha. What is so interesting about this boot log? Nothing special, just that this Alpha's got 31 Processors, 256GB RAM -- looks VERY impressive. I wouldn't mind having one of those beasts at work *drooling all over*. Oh, and it compiles the kernel very fast :)
But its definately not the same system as the new 12,000 cpu 30 TeraOp ASCI EV68-based dream machine.
Now what would be _cool_ is if during checkout someone were to try out 2.4.0-test103 and find out that it actually outperforms Tru64 for certain classes of problems.
The reference to 2.4.0-test103 is based on the rate of test kernels, and the projected delivery time of this new machine. I hope that 2.4.0 final is available much sooner than that.
About 3 stories ago. I wish Compaq would commit to keeping the line running, at least for a few years. It'd make my 64 bit system purchasing decision a lot easier...
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
Oh, well.
And it so nice to see the Digital logo in this story. For all of us old VMS hackers - it'll never be Compaq.
Yes, Linux is cool and performs fairly well. But Tru64 UNIX is optimised for Alpha hardware and far more scaleable. Yes, Linux is more scaleable than ever before, but still not as scaleable as most UNIX systems. It seems like a waste to spend all this money on excellent hardware and then stick Linux on it, degrading the performance.
But I am drooling, dammit!
:-)
I got the bootlog by email a three weeks ago and had to stay seated for half an hour.
Of course geeks lust after cool hardware, just like car guys lust after cool cars.
:-) My question is why aren't you drooling? Are you some sorta software nancy boy or something?
I browse with my threshold at 2 so I can't read my own comments :-)
/. really needs to start notifying sites when they print up a story about them, this one has already been slashdotted and the story has only been up for a little while. If we could tell them beforehand, perhaps they could brace for the onslaught.
.sigless
Enigma
Enigma
yeah, you all shut up!...
oh wait... that doesn't look right...
Please, let's not draw irrelevant political comparisons; I think this one is, not because you fail to make a point with it, but because it simplifies the British political situation.
He who fights and runs away,
In a related story, it was recently discovered that it IS possible to run the Enlightenment window manager without annoying slowdowns...
Politics, Culture, Food?
I AM, therefore I THINK!
Would we notice if they were kidnapped? Of course we would! The stories would get better!
/. is a commercial entity. goto slashdot.com
"The only caveats are that one of the CPUs was out of
the system at the time (hence 31 CPUs, not 32)"
But check out that floppy, it's a 2.88MB! Even the floppy drive is better than what I've got in the beowulf cluster at work. I feel so inferior.
I bet my keyboard is better than theirs, at least I have that.
Despite the fact I'll probably get struck by lightning, I've always been a fan of "Sweet Zombie Jesus!"
ACHTUNG! ALLES LOOKENSPEEPERS!
Das computermachine ist nicht fuer gefingerpoken und mittengrabben.
Ist easy schnappen der springenwerk, blowenfusen und poppencorken mit spitzensparken.
Ist nicht fuer gewerken bei das dumpkopfen.
Das rubbernecken sichtseeren keepen das cotten-pickenen hans in das pockets muss; relaxen und watchen das blinkenlichten.
Oh, wait, front panels went out in the 70's. nevermind.
"Linux pthreads" is really just a library implemented on top of the Linux kernel clone() system call. The clone API scales just fine. The pthreads library on the other hand has "issues."
A car that can do 200 can generally reach 70 mph a great deal faster than a car that can only do 70. I think the analogy holds for the 32 processor 256 GB RAM Linux box as well.
"Mmmmmm, beer." Homer Simpson
A quick search on Google revealed that an 8x Pentium III (Xeon) at 500 MHz, SMP would run at 3996.06 BogoMips. Compared to 46170.90 BogoMIPS. Or about 8% of the box. Of course, I wonder how the price comparison point would be. :)
Ok, I finally tracked down the Alpha pricing. But I'll be danged if I can get it to work. Can someone else? yeeshk.
You quitting proves that the karma kap worked. The most annoying of the whores shut up. --CmdrTaco
I'm working on an open source project, and porting to anything beyond *Linux (i.e. AlphaLinux, IntelLinux, etc.) is a pain in the ass. Header files aren't where you think they are, they don't necessarily do what you think they do, standard library routines don't do what you think they do (try getcwd(NULL, 0) on a solaris) or what they're supposed to do, etc.
When companies like Oracle employ teams of porting engineers and have to substantially modify their coding practices (last time I read them, the Oracle coding standards were over 100 pages) to support the plethora of operating systems they have to support, you start to see the advantage of having one unified platform. One of the reasons why corporate developers have flocked to Java in droves was the original goal of Write Once, Run Everywhere.
While there are specific hardware differences between platforms that porters should be aware of (Alignment is the most significant and obvious), being able to have the same operating system on many different platforms gives developers more of an advantage to writing to that platform. And it makes more esoteric hardware more attractive (becuase they get to exploit all the different applications ported to that platform).
Overall, the more things that Linux runs (and runs well) on, the better it is for applications developers to support the Linux development model and encourage customers to run it.
The solution isn't to just say "Tru64 is better, run it" but to say "Linux needs work, here's the help," which appears to be what Compaq/DEC is doing here.
Well... the S/390 port runs on a VM, so Linux doesn't know the kind of power that is actually down there..
Nope, the S/390 port runs either in a VM or natively. You can boot up a 390 with Linux as your only OS, and then it definitely knows about the whole machine.
It's just that far more people are likely to have access to a VM running on a 390 than there are that have a whole 390 to play with.
No, no, no. It ain't ME babe,
It ain't ME you're looking for.
-- Alastair
As taken from yet another Kernel Traffic Post:
Linux is king-of-the-hill in SpecWeb99 tests on One, Two, Four, and *8-way* systems. :-)
Insanity is the last line of defence for the master diplomat. But you have to lay the groundwork early.
cant be afforded? two words my man: drug dealing
we do not condone the actions being condoned by the condoners condoning the condoned actions.
"sex on tv is bad, you might fall off..."
I lost my concept of community when my community lost all concept of me.
You'll still benefit from sneaky writes (I confess I don't know, does Linux do them?) to swap, so should have TotalPeakMemoryNeeded swap.
The sundown machines are behind a firewall, so pinging them would be futile. In addition to that, this test occured about two weeks ago, so you couldn't do anything with it. Compaq (seeing as Digital is owned by them, not on their own) doesn't let their machines that are as big as three refridgerators (sp?) sit running all the time (something about power consumption comes to mind...). An intresting thing about these machines is that they are prototypes. They are for sale (they model type), but they cost somewhere in the 2 Million Dollar range. As far as the Linux Scalability issue, here is a really intresting report from a UofM alumni, now a worker for Compaq on their zk3 campus (practically Hemos's back yard, it's in New Hampsire I belive). Yes it is only a coincidence.
-Mr. Macx
Moof!
******
Great, now I'm not going to be able to sleep at night. It makes my Dual Pentium II machine look bad.
It's either on the beat or off the beat, it's that easy.
I moderate therefore I rule!
--
yeah... same here, ever since upgrading to 2.4.0-test5 I've been having that problem. couldn't even create the file. best suggestion? ignore it :) i think ip_always_defrag is just used by servers?
---
dd if=/dev/random of=~/.ssh/authorized_keys bs=1 count=1024
Starting system logger: Warning: /boot/System.map has an incorrect kernel version.
/proc/sys/net/ipv4/ip_always_defrag is missing -- cannot control IP defragmentation
Now which processor is doing that? Is it the RAM strips? Oh wait.... it's the modules...
Ok, I don't mean to troll, but if I want to show-off my machine, I'd like to fix those little warnings first before sending the dmesg.
---
dd if=/dev/random of=~/.ssh/authorized_keys bs=1 count=1024
I'm glad I'm not the only one. :)
Free Mac Mini. Yes, I'm
Bite my shiny metal ass.
Just because the West Edmonton Mall has more submarines than the Canadian navy doesn't give you the right to laugh at our miliary.
J
Anyone that doesn't believe that Linux scales, is mistaken. Anyone that professes the same is either ignorant of what's been going down for the past two years or is FUDing.
While I'll admit that the S/390 port is some proof of scalability- every little drop helps dispell myths.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
w00t, i can run 15 copies of quake and frag myself as various players!
I personally find the memory cooler. Why not start up the OS, mount / to a 10-100gb ramdisk, copy the harddisk to the ramdisk, and then finish loading up. *that* would be true speed.
No sig here...
He who fights and runs away,
I'd sooner run Solaris or Digital Unix on their native hardware - at least that way, all that power will be put to use! Linux just doesn't scale well beyond the workstation market yet.
While this is cool, I'm sorry - Linux isn't quite "up there" yet. Show me some hardcore applications that use this (nevermind Linux itself) under Linux...
get a life
mmmm Compaq really is selling some interesting HW.. not only do they sell a 31 way Alpha server, they also sell the Unisys 32 Way PIII 1GHz Win2000 DataCenter box with very similar IO capcity.. maybe its time to go work for them :)
banananananana
Insanity is the last line of defence for the master diplomat. But you have to lay the groundwork early.
To bad I had to put down Fabius, my war-monkey... he'd be a much better fight. Those bio-mechanical weapons just didn't agree with his nervous system, alas.
BTW, the monkeys, paraphrasing Nepos, say that if orthodoxy is your doxy then hetrodoxy is simply their doxy... I don't know how they picked up on Byzantine religous history, but there you go.
Insanity is the last line of defence for the master diplomat. But you have to lay the groundwork early.
Well, not really. With 32 processors, you aren't going to see much, if any, performance gain over a 1-2 processor box, if you are just doing things like playing Quake. And if you're only running, say, emacs, a couple of shells, and Netscape, you're definitely not going to see a performance gain. If anything, you'd see a performance decrease.
The only real use of a 32 CPU machine, is, obviously, an application that requires lots of processing. Scientists need machines like that; normal users don't, unless they are way to into distributed.net.
umm, why is everyone so concerned about bogomips?
doesn't that just mean it can do nothing really really really really fast?
how about some actual benchmarks?
-sean
That machine appears to have 8 100mbit ethernets, 4 scsi adapters, and 8 fiber channel controlers. Thats an IO monster. (Plus a floppy, for when their aboot gets pooched. :-)
Any speculation on what they are planning to do with this machine? (I rather doubt Quake.)
*HUG*
It's an alpha with 31 processors. That does NOT mean that it will help you play quake any faster than a single processor alpha, because quake is only programmed to take advantage of one processor (yes, i know, the kernal supports SMP, but that still doesn't mean quake does).
For those inclined to "imagine a beowulf cluster", this functions in the same way as a beowulf cluster, only processor to processor comm. is handled on the motherboard/daughterboards and not over a network.
====
Crudely Drawn Games
"
I show 35.84 BogoMIPS on an old 90MHz Pentium server here at work... so I would say >41 times is a decent increase ;)
I have come to a conclusion that one useless man is a shame, two is a law firm, and three or more is a congress -J Adams
I worked with the previous High End Alpha servers They were VERY fast...
:-)
Wonder how many Seti at home an hour this sucker swallows??
I really HAD another userid
Do you?
IMHO, it seems that when it comes to linux, there are two groups of people, those that want to show how well Linux performs on the bleeding edge, and those that want to show how well Linux performs on their i386 circa 1991. The middle ground (read: Average User, 99% of computer owners) are somewhat ignored.
That probably makes it the fastest Linux machine in the known universe.
Well, at least at no-ops. :-)
CVS is teh suck. Use Vesta instead.
The entire point of this story is that this guy has a sweet box? Or is there something akin to "news" in here?
Help me out here...
Well, I can't match those processors. Or the RAM. Or the drive space. Or... hey, wait a minute, only 139 swap? Those losers, I can top that!
=repartitions=
"Yeah, I saw that system on Slashdot. What a piece of crap, my P150 has more swap than that...!"
I'm surprised they used tulip (21x40 chipset) NICs. IIRC, they are slow on platforms that can't do unaligned accesses, like the Alpha. (They kick ass on IA32, which can access odd bytes :). I was talking to Don Becker about this at the Ottawa Linux Symposium. He said the best card to use for an Alpha machine was a 3Com NIC.
#define X(x,y) x##y
#define X(x,y) x##y
Peter Cordes ; e-mail: X(peter@cordes ,
Someone please benchmark this box ruuning Linux / TUX against IIS 5.0 running on W2k DCS - I could use a good laugh.
It can play an hour's worth of MP3s in 78 seconds.
(and execute an infinite loop in less than 3.5 minutes)
Duct tape + WD40 => DevOps
an alternative to k5 and /., not yet clogged up by weenies... although the siteops are kinda weeny. =P
From a motherboard manual, error beep codes: S-L-L-L-SS: Speaker Error
I was the one that reported that result. Using the bogo.c file compiled using cc65 and run under VICE. The results were the same whether I set the processor speed to "100%" or "1400%" so it might be fairly close to the real result....
What does RH have to do with anything ? Those CDs where pressed about 2 months ago. Are they going to redo things for a new kerenl ??
until (succeed) try { again(); }
until (succeed) try { again(); }
ttyS00 at 0x03f8 (irq = 4) is a 16550A
All I need is another 255.8725 GB, another 29 processors, another 7 LAN cards, and I'm right there!
Those submarines were made in Port Moody, British Columbia. Just another factoid. :o)
ARRRGH!! you used the words "Java" and "Oracle" in the same paragraph.
Curses! as of Oracle 8.1.6.1.2 - you're stuck with JServer - you can delete it synonyms, shrink its shared pool, but its still in there (the word inoperable comes to mind).
SQL*Plus: Release 8.1.6.0.0 - Production on Wed Sep 27 20:27:49 2000
(c) Copyright 1999 Oracle Corporation. All rights reserved.
Connected to:
Oracle8i Release 8.1.6.1.0 - Production
JServer Release 8.1.6.0.0 - Production
That's great that Oracle supports Java, but they shouldn't force it upon you (yes, I have to run Oracle). I think that Java is great for making body parts move rhythmically in a browser, but think that it does not belong in a database.
.
It depends upon the task being performed of course. Some tasks can never be parallelized at all, regardless of how many CPUs you have and what operating system you are running.
Sure, but assume a task that is 100% parallelized for the purpose of the question. What is the OS overhead as you increase processors?
--
1990: You could easily read everything in comp.*. You bitched about all the weenies clogging up the alt hierarchy.
1995: You could easily read everything in comp.lang.perl. You bitched about all the weenies clogging up the comp.* hierarchy.
1998: You could easily read everything on slashdot. You bitched about all the weenies clogging up Usenet.
July 2000: You could easily read everything on kuro5hin. You bitched about all the weenies clogging up slashdot.
September 2000: Bloody weenies clog up kuro5hin. End of universe as we know it. Film at eleven.
NOTE! EVERYTHING LOOKENSPEEPERS!
The computermachine is not for gefingerpoken and mittengrabben. Easy snatches if that is branching factory, blowenfusen and poppencorken with sharpen-deactivate. Is not for trades with the dumpkopfen. Rubbernecken sichtseeren keepen the that cotten pickenen Hans in pockets must; relaxen and watchen that flash-light.
You quitting proves that the karma kap worked. The most annoying of the whores shut up. --CmdrTaco
Just for comparison, my Celeron 2 800 MHz gets 1599 bogoMIPS, which is very close to these Alphas (1578 or something). I expect a Pentium 3 at about 700 MHz would be comparable, so that gives us a 24,800 MHz Celeron or a 21,700 MHz Pentium 3.
It's fast. I wish I had one for Blender rendering...
And yeah, come back when Linux can touch Tru64 or Solaris on their own turfs...
Those machines aren't aimed at the lowend of the market like Linux is. They're not trying for Linux's marketshare. And they're making great business and good money by doing so...
Marketshare might be important on the desktops, but it's not critically so in the server world. Yes, Linux might have 24%. NT might have 26%. Solaris might have 20%. It just doesn't matter...
dude, you stole my post! i was going to say the exact same thing (even the bit about searching on 'bogo'). 1488.98 BogoMIPS... a mere abacus!
aboot: switching to OSF/1 PALcode version 1.75
aboot: booting from device 'SCSI 3 6 0 1 100 0 0'
This joke was aboot as funny as Canada's military presence, eh?
I could not justify my existence if I were a turkey farmer. Would I terminate myself? Undoubtably, yes.
If you're gonna invoke Jesus in your awe-filled gasp, I prefer one I saw somewhere (maybe the onion?) a while back.
"Well I'll be a greased Jesus!"
-LjM
Er, first, nothing stops one from duplicating the .sig of another.
Second, the UID number is perfectly innoncent. If some weak-minded idiots overrate it, well, it's no different than the era where every Bruce Perens comment was instantly modded up to +5.
Steven E. Ehrbar
I think that there needs to be some sort of "intelligence" in the boot up that will spit out messages like "Damn! 31 Processors. I am gonna kick some a**", and "256 freekin' GB of memory!" or at least at the end spit out "holy sh*t"
type "su" ;)
and then enter the password
Before you email me, remember: "There is no god!"
How much software actually scales well to that many processors? I suspect quake would top out long before 31.
Thank you for not thinking.
Ask Jeeves, of course...
Try Rusted Root, or Root Canal - those might help.
--
"It's tough to be bilingual when you get hit in the head."
SuperID
I've also heard from a colleague that the Linux pthreads implementation was having some problems under SMP (like mutexes not really mutually excluding). Since multi-threading is most useful when you have multiple processors, this seems like a serious deficiency. Maybe it's all better in 2.4, I haven't checked. Does anybody know?
CVS is teh suck. Use Vesta instead.
and why is he posting this drivel without at least introducing hisself first?
--
Peace,
Lord Omlette
ICQ# 77863057
[o]_O
I was just making a "big iron" comparison. A "crappy" K6 can best even the fastest Alpha at RC5... The alpha processor (most RISC chips in general) don't have bitwise rotate instructions. So what takes 2 cycles on a 486 takes a lot of cycles on an alpha. I like the alpha; it does alot of things really well -- RC5 isn't one of them.
Nothing is designed to do everything perfectly -- well, Transmeta is trying, but them may be falling short on the "perfectly" part. (They may be able to emulate any processor, but how efficiently they do it has yet to be proven.)
Having mail delivered to root is stupid and unnecessary.
I use a dual P-III/NT4 at work. It is almost never unresponsive. When I start long builds I don't have to go for a coffe for being unable to type properly. I can run the current version of the program while building the next. I can even play Quake while waiting for the build to finish.
Now if anyone could tell me where to get...
A P-II 400MHz to put in the empty slot of my home Linux machine. The damn things are off the market.
A parallel version of make for Visual C++, so the dual NT machines can actually work with both brains.
Pavlos
Actually, to be more accurate, this computer was purchased by the DOE for the Sandia National Laboratories to simulate nuclear explosions. What the truth is is that if you read the entire article the computer will actually be running a modified version of RedHat Linux and their proprietary, but soon to be open source software, called CPlant. The article has all the links and it an interesting read. Check it out. More great Linux headlines that should have probably been their own post on Slashdot.
Ok so it boots successfully, how about benchmarks and compile times? specweb scores? etc etc.
Only the State obtains its revenue by coercion. - Murray Rothbard
At one of my jobs, I was in the computer room of a Vast Fianancial Company That Shall Remain Nameless. They had an IBM RS/6000 SP-2 supercomputer that filled most of one wall, 32 POWER3 processors and untold terrabytes of storage.
I was left alone to work my mojo on a much smaller Sun server, but once I was good and certain i was alone, and that there were no cameras monitoring me, I wandered over to the supercomputer. I looked at the gray and blue tower that held the processors and RAM. It was worth $20, easy.
I touched it, caressed the cool metal of the mesh grid over the airvents with my fingertips, feeling the warm air and the low buzz. I'd pay $100 to do that again.
Then I licked it.
Priceless.
SoupIsGood Food
The kernel is no better off than user apps, since it can be interrupted at any time.
Most processors now have hardware test-and-set instructions, or something similar.
In the absence of hardware support, there are a number of algorithms for guaranteeing mutual exclusion. (Any OS textbook should have a few.) These are just kinda slow compared to hardware, and are mainly used for distributed systems.
Oh baby ... oh baby ... OH BABY show me those sweeeet MIPS honey.
We all know that crap is king
Give us dirty laundry!
Linux market share means ZERO if it doesn't do what it's supposed to.
This is a hardware, not an OS issue. And sure enough, Alpha (like Pentia) has an atomic test-and-set instruction! Good thing, too -- you need this sort of instruction to implement mutexes and other locks both in kernel and user space. There is no efficient software workaround for the lack of such an instruction.
Agree,
everybody bragging about his 31-way multiprocessor SMP machine is just so stupid, that can't understant that 8 4-processor machines were (1) much cheaper, (2) much faster (due to faster memory connects that avoid SMP-bottleneck).
Advice them to read classic book "In search for clusters" by Gregory F. Pfister.
MSDOS: 20+ years without remote hole in the default install
...that if the app never or rarely makes system calls, and the hardware provides enough resources (ie 32 fast procs, a dozen NICs, and loads of memory) so that the app(s) are not in contention with each other, then the os is not involved. The hardware architecture becomes much more important, ie does the bus have enough bandwidth, cache snooping, etc.?
The only situation where the os has influence is when there are context switches, or multiple processes contending for limited resources of other types. Even then if the app consumes 99.9999% of the CPU, suppose I come along with an OS that's half as efficient. Now the app gets 99.9998% of the CPU and what does that do to the throughput of my app? Diddly.
If my app spends all of its time in loops calculating matrix values it does not care if the os even exists. It could just as well have been booted straight from the ROM. There is no support required. OTOH, if my app needs such things as virtual memory, swap space, multiple tasks waiting for the same proc, etc. then the efficiency of the os DOES make a difference.
I'm willing to bet that a 32 proc, 256GB, 8 NIC, 8 FC, 4 SCSI chain box was built with a particular purpose in mind, not just on a whim as in "Oh, I'll take the 128MB SDRAM since it only costs $1 per meg...". Try 31 or 32 simultaneously active processes or threads, each with a working data set size of ~4GB or so. Throw in a net across all processes box-to-box bandwidth that needs some combo of 8 fiberchannel devices, some 600-800Mb over 8 NICs, and disk i/o needs that are just too much for 3 chains, right around 4, and not enough to justify 5 hosts. Otherwise there would be fewer or more procs, more or less RAM, NICs, FCs and SCSI cards. Even rich people don't throw around >$500K on a whim. They also will not toss Linux in there just for the fun of it. The gov might, but then they know it's not their money that they're spending.
Good judgement comes from experience, and experience comes from bad judgement.
- W. Wriston, former Citibank CEO
make -j 32
Mmmm, alpha....
A year ago I had the thought in my head that alpha would be the linux platform of choice. Kind of like how sparc is to solaris (as apposed to intel), this would help compaq develop a relationship much like sparc/solaris and SGI/Irix but I don't think that the alpha hardware has come down to near the price that it needs for this to happen. I would love to see Alpha become Linux's premier platform....
-Aaron
----------------- Who is Jesus?
Actually, the K6-2/500 I have runs ~1000 BogoMIPS... they seem to do nothing faster than equivalent Intel chips... of course, that doesn't mean much 8^)
--
"It's tough to be bilingual when you get hit in the head."
I am also running 2.4.0-test7 and I keep getting this message, but never got it on 2.2.12, 2.2.14, 2.3.99-x and maybe some earlier 2.4.0-testx even though I was using more-or-less the same config file. Any suggestions how to fix it? Thanks in advance.
Awww, shuddap! I've got a lower userid than you, you pissant...
Um, dude, Microsoft doesn't run on alpha any more.
Oh yeah, just another way that Microsoft products suck.
Fire and Meat. Yummy.
Hey Mods... how's about changing the logo to the more appropriate COMPAQ logo. Digital doesn't exist any more, and this is a COMPAQ product that we're talking about.
-S
--- What parts of "shall make no law", "shall not be infringed", and "shall not be violated" don't you understand?
On first glance, one might guess that 1489 was the total # of BogoMIPS, but that just isn't right...
--
"It's tough to be bilingual when you get hit in the head."
Powerful, yes, but not as powerful as THIS
--
Chief Frog Inspector
A feeling of having made the same mistake before: Deja Foobar
So impressive...
Can't even handle the
This
Yes, but if the OS doesn't support those CPUs and RAM effectively and efficiently then it's kind of stupid... ;>
--
Garett
Why is it interesting now and not 3 weeks ago?
What's my Karma Mr. Burns? "Excellent"
I would find it unremarkable if there was some Intel chip that had an outrageously higher BogoMIPS rating that would disappear from the "running" as soon as you tossed POVRay onto it...
If you're not part of the solution, you're part of the precipitate.
Is this a breakthrough technology? No. Is this an earth-shaking legal or political development? No. Is it something that geeks the world over will have wet dreams about tonight? You bet your ass.
All the posts about "I need a towel" and "they should sell tickets just to touch it" are gonna look funny when this is just another slag heap of unusable parts.
"Why, when I was yer age, miboy, we had to put up with using a computer. That's a complicated physically connected brick of processing components. We thought a mere 2^5 processors was worth drooling over. Yes, miboy, I know your cochlear implant has more than that. You're missing the point. This thing was tremendous! It took up a whole rack: four times the size of a grown man! And all of its memory circuits were in the same cabinet, requiring massive cooling apparatus, unlike the distributed memory crystals that people embed in their jewelry."
Anyone still drooling over 2^5 address space on ferrite core memory? Anyone still drooling over 2^5 address lines? Or data lines?
[
Try the LISP programming language, its bracket heaven.
Arrgh! I read that, I was commenting about the symmetry of the whole thing - it was a subtle joke (not a good one, to be sure, but a joke nonetheless) - how can you have *S*MP, if you don't have a 'symmetric' number of procs... It's like a 3-processor SMP box... there's something just inherently wrong with that - take another one out, or replace the bad one (powers of two are much nicer than simple even numbers).
/. readers (and overestimate mine).
I should have included more text, but hey, I tend to underestimate the humor coeffecient of
--
"It's tough to be bilingual when you get hit in the head."
You started off with a great topic, namely, abusing the +1 bonus. However, by the end of your post I was seriously starting to ask myself if you weren't trolling. I mean, "glorious new tomorrow" for Slashdot? This is a website, not a country of its own. Yes, there are social phenomena and quirks that are a part of any group. That example you gave of drendite is hardly bowing down and worshipping. The user says they feel like they've seen Bigfoot, that is to say, it's weird seeing someone with an ID that low. It's a freak of nature, not an example of a "social ill". Jeeze, social ills are things like drugs, crime, and child abuse. A freak post on slashdot is not a social ill, and to compare the two is rather immature, IMHO.
PS.Just re-read your post...the sig is not a way to prevent identity theft. People copy that all the time. The user ID is the only thing that can't be copied. Since Taco added that to the posts, there never has to be confusion as to who is the real Bruce Perens
--------
-------
"Every artist is a cannibal, every poet is a thief."
If the instruction code were compatible with the x86 set, how fast do you guys think this would run Windows 2000? 2 second boot time?
- I don't care if they globalize against free speech. All my best free thoughts are done in my head.
If you own a car that truly does 200 MPH, speed limits really don't apply to you.
The man who trades freedom for security does not deserve nor will he ever receive either. - Benjamin Franklin
Hey coward!
Read some news! IBM, Compaq, Dell and the other guys are WORKING on making 16 processors and 32 processors based server - it's still vaporware!
Also, do you mind showing me a X86 machine which supports 256GB RAM? huh?
Thought so!
Hetz (Heunique)
...'till you learn, or until I grow weary(probably the latter).
The only reason that most operating systems exist is to provide utilities to applications. Part of that utility is resource management. There is no scheduling and no management on a single task, single cpu system and indeed on systems like that, there is often little or no os. All there is is initialization code and app. Once setup, it runs. The os per se is irrelevant. The same applies to multi-tasking single cpu systems under light(aka single task) load. The os is there to provide utility, not efficiency. Why else is there such a thing as WinXX?
This can be extrapolated without much difficulty to a 32 task, 32 cpu system. Once setup, it runs. The OS is there to provide utilities like file systems, network drivers, queues and such. It does not constantly schedule the running processes around from cpu to cpu as this would be extremely inefficient, wasteful of cycles and cpu cache having to constantly flush and refill. It is much better to set the app up and let it go uninterrupted. If the os or app needs interrupts, let one cpu handle them, the others continue running. If there are many different types of interrupts, spread them across the cpus so that each gets one interrupt type, and appropriately one task. The likelyhood of having both app and interrupt code in cache is greatly increased. Once again, once set up, the os is out of the way and irrelevant, or should be. Hell, it's likely most of the OS will be swapped out anyway. It can't make much of a difference if it never runs. Believe me, a box such as this will never see your normal ps ax output of Quake, sendmail, 5 of apache, ftpd, sshd, X, java, identd, oracle8i, netscape, mutt, 5 xterms, 6 copies of bash, and your favmp3encoder, blablabla all at the same time. It might see something equivalent to 32 copies of seti@home, the distributed.net client, or one copy of oracle all of which you'll note run with very high efficiency on *any* smp capable os.
Only in semi-random high load multi-tasking systems where resources are constantly being allocated, released, and re-allocated and multiple tasks are queued up and waiting on those resources does the actual resource management, including cpu scheduling efficiency, of the os play a part. Unfortunately for folks that want to get the most out of their systems, this is just the type of environment where most low end smp boxes live and thus, the efficiency (or lack) of the os and the hardware limits such as bus bandwidth are very important and generally quite apparent.
Last but not least do the arithmetic. Even if the the app is a dynamic, high load system as above, if the cpu spends one thousandth of its time running the os code, how much difference does it make to the app if the os suddenly becomes one half as efficient? How much difference does it make if the OS becomes twice as efficient?
Good judgement comes from experience, and experience comes from bad judgement.
- W. Wriston, former Citibank CEO
If you've got 256GBytes of memory, do you REALLY need that 136MByte swap file?
"Freedom means freedom for everybody" -- Dick Cheney
You'd think it could handle the load, what with 31 processors and all.
Now, imagine a beowulf cluster of these things...
I'm pretty sure that any real thread library would need more support from the kernel than just clone(2) to work right. How're you supposed to implement an atomic test-and-set if the kernel can pre-empt you in the middle of it? With SMP, that becomes an even more difficult problem. Now I'm sure that the kernel must have its own mechanisms for mutual exclusion in order to make SMP work at all. Do you mean to suggest that there is no way for a user space program to do that kind of operation on Linux?
If so, then that's definitely a deficiency.
Like I said though, I don't know, so maybe somebody who does could speak up (hint hint).
CVS is teh suck. Use Vesta instead.
I see what happened now - the filter ate my comment in the angle brackets... forgot to set that back... which of course was the part that separated the joke from what has now been marked as "flamebait"... I hate the fact that extrans stopped working months ago and still hasn't been fixed...
--
"It's tough to be bilingual when you get hit in the head."
Anyone noticed the level of humor required to get moderated (+1 Funny) these days. At first it seems like a waste of perfectly good moderator points. But I think everyone has figured out the quickest way to the karma top is to moderate UP not down.
Maybe we need a new type (+1; Moderator needs karma)
it must be a slow news day for /. if we're posting our server logs..
- Entertaining Bits from the Ancient Kernel Tree
...on the application. In many massive computationally intense applications, the OS is effectively irrelevant since 99.99999% of the CPU time is spent in the app and not doing system calls. Wanna do a fusion bomb simulation? How about weather? Just what kind of system resources does an app like that need? They need CPU and memory. Does this box have a lot of both? Semaphores? No. Pipes? Yes, but they're static and setup at the start. File I/O only happens at start and finish. After all that they've still got multiple NICs and SCSI cards.
Good judgement comes from experience, and experience comes from bad judgement.
- W. Wriston, former Citibank CEO
Damn, I was all excited too... my 2nd PPro 200 processor for my Compaq 5500 came in just before lunch.
People seem to be complaining that "this isn't news".. no, it's not the kind of thing that you'd see on your local TV news, but it's "News for Nerds". Which is, as I'm sure you're aware, what this is all about. Why should I care that you don't like this story? I didn't like the last one, do you care?
-
Meep meep
> Can anyone comment on the SMP performance linearity of the current Linux kernel on more than 4 CPUs?
It depends upon the task being performed of course. Some tasks can never be parallelized at all, regardless of how many CPUs you have and what operating system you are running.
> Did they every sort out the issues that prevented kernel socket (or was it I/O?) APIs
> being called concurrently by processes on multiple CPUs?
Yes, Linux TCP/IP is fully threaded and will run concurrently on all CPUs, assuming that there is work to be done on all of them. (a single socket will not run on 100 CPUs at the same time, for instance)
Why the hell was this moderated as flamebait? Seems like the moderation system is at it again...
God was my copilot, but then we crashed on the top of a mountain and i had to eat him...
This means nothing. Being scalable means that performance with 30 processors is about the double of performance with 15 processors, and so forth. If this 31 processors machine, running Linux, can deliver more juice than a machine with 8 processors, I'd be surprised.
(8-DCS)
Please look at this comment for clarification...
--
"It's tough to be bilingual when you get hit in the head."
Processors 1 to 15 would work on distributed.net...
:)
Numbers 16 to 20 would do Seti.
Numbers 21-29 would run Quake3, Civ:CTP, and XWS.
Numbers 30 and 31 would run the realtime disk encryption/decryption series
I used to be someone else. Now I'm someone better.
Real life is underrated.
It's never going to amount to anything other than solving a few math problems quickly. People need to shutup about the beowulfs.
Is fascism really all that bad? People need motivation.
Ok fine, but how well does it run Quake III Arena ?
:-)
Alpha chips were a dead technology? Keep saying that, it'll drive the price down! OK, an Ask /. question here: What MPI tools are available for Linux, like LAM? IOW, how does this clever guy do useful things with his Moonbase Alpha?
I am quite civilized, and I should be brought a beer immediately. -- Bruce Sterling
that *still* depends on the workload. if it's CPU-bound, it just works. if it's I/O bound, it depends on global system recourses (bus speed and bandwith), and on how well smp-threaded that particular subsystem is on the OS. one of the big improvements of the linux 2.4 kernel is precisely a major overhaul of the tcp/ip stack to smp-thread it completely. OTOH, heavy filesystem work (lots of rename(), unlink(), open(O_CREAT), etc) is not likely to scale super-well, because of filesystem locking constraints, and that's under all OSes too, which is why you generally avoid writing your big programs in such a way that they have to be messing around with lots of little files all the time (e.g that's why you use a database backend, or some kind of db format, rather than flat files, if you want a news server that scales). beyond that, in the real high end (32 procs and more), the SMP model itself starts showing its limits, which is why NUMA (non-uniform memory access, i.e. not all memory is equally fast from each processor) was invented. Linux 2.4 has some preliminary support for NUMA, but it's still in the beginning stages.
So can I tell you to shuttup? :)
It's not enough to bash in heads, you've got to bash in minds. - Captain Hammer
How do I get ROOT?
I am become Troll, destroyer of threads
that's one side of the story. the other side is that clone() doesn't support the exact POSIX threads API, and the linuxthreads library has to bend over backward (and be slow, in the process) to add this support. the situation is slowly fixing itself, but there are some human communication problems in the middle, and it takes a lot of time for a new version of linuxthreads to make it all the way to the distributions.
If you scroll down and pause at certain intervals you can pretend its your computer booting
So is this the Jugs magazine centerfold for kernel hackers? "Look at the processors on THAT!"
You can tell from the bootlog that it's not an SMP. It's a NUMA system. Linux has virtually no numa tuning, so it will perform much worse then digital unix.
(you can tell it's numa because the first 4 cpus have higher bogomips due to their improved memory locality)
I was all happy that i finally have a Sun E-420 with 4CPU's and 2 Gb's of mem to play with today! Well actually i have to install oracle and patch it up, but still.... would have been fun... BUT NOOOOOO, you have to come and destroy it for all of us with that boot log. :)
AND just cause you did that... what the hell is up with all those warnings, freaking do make install ( i think.. i use slackware so i am not sure) and fix the System.map problems.... ohh and in case it is possible, can i get a shell account
42K bogomips.. make my 900 seem so _low!
Non-Deterministic Finite Automata
I'm not one to bitch and moan about how much cooler things used to be. But sometimes I miss the days of being able to read every single post on slashdot every day and still have plenty of time to get all my work done.
This is my last attempt to contact anyone on it, I have sent in close to 30 messages so fasr and no response ....
until (succeed) try { again(); }
until (succeed) try { again(); }
I mean you have to have a *lot* of drugs sold to do that.
Respond to s
Computing speed is relative. You can buy a faster computer, or make your mind go slower. Bring out the beer!
... why 31? Why not 32?
--
"It's tough to be bilingual when you get hit in the head."
The TCP/IP code is fully SMP threaded. It runs on multiple processors in parallel. The same is true for the VFS layer. Linux already holds the world-record SPECweb99 score on a 4-way box, which would not have been possible without good SMP scalability.
"Feh" is right. Who needs an alpha when I can get 0.0033 Bogomips on my c64. Read it and weep!
Imagine a beowulf cluster of C64s! Think of the power!
NEWS FLASH! GEEK BOOTS LINUX ON ALPHA!
Way to go!</sarcasm>
The swap looks a bit undersized for 256GB...
Here is what might have just happened : /. and decided to focus our attentions by confusing us with that headline while he's kidnaping CmdrTaco and Hemos.
1) This is one heck of a boring day with nothing happening at all.
2) HeUnique just received the rights to post and wanted to post something, anything quickly before those rights faded away.
3) Someone evil took controle of
4) I was really bored at work and decided to reply to that headline even if it didn't hold much interests.
"When I was a little kid my mother told me not to stare into the sun...
"If liberty means anything at all, it means the right to tell people what they do not want to hear"
Quoted from article:
Oh, and it compiles the kernel VERY fastGood. Maybe it'll be fast enough to start up Windows 2000 in a reasonable time frame.
<grin> Think of how much *ass* you could kick running SETI@home on that!
Fire and Meat. Yummy.
imo, what it can do well is less important than what it can't. To paraphrase Limp Bizkit, "Let's break some sh**."
I don't want knowledge. I want certainty. - Law, David Bowie
Hey, let's watch newbies drool and argue about BogoMIPS, huh wanna? k?
its porn for geeks! if you will excuse me... i think i need a towel now.
I can't resist saying this.... it'd be cool to have a beowolf cluster of these babies!!
Can anyone comment on the SMP performance linearity of the current Linux kernel on more than 4 CPUs? Did they every sort out the issues that prevented kernel socket (or was it I/O?) APIs being called concurrently by processes on multiple CPUs?
I suppose you are referring to the fact that there are 30 procs listed,
well, if you look closer, you can see the first sentence '...secondaries'
neat.
sorry i need a towel. fast. *nnnggg*- aaah... nevermind, too late...
Looking for people to chat about multicopters, coding, music. skype: gtsiros
Indeed, but other (more competent :-) companies are already
there, even with Intel CPUs. The Data General
AV25000, for example,
supports up to 64 PIII Xeon CPUs, and runs either DG/UX or Windows (or both). Of course, the
if you go the Windows route, you'll have to run multiple copies simultaneously, because Windows can only
scale to 4 CPUs on that machine. If you go the more sensible DG/UX route, of course,
it can use all 64 CPUs from a single system image :-)
"The invisible and the non-existent look very much alike." -- Delos B. McKown
Someone has to give this machine more disk quickly!
)9TSS
I did a search for "Bogo" to find the BogoMIPS and found this:
Calibrating delay loop... 1488.98 BogoMIPS
Feh, I thought, either Alpha's REALLY suck or (more likely) there's a bug there. Then I took a closer look:
SMP starting up secondaries.
Calibrating delay loop... 1493.17 BogoMIPS
Calibrating delay loop... 1493.17 BogoMIPS
Calibrating delay loop... 1493.17 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
Calibrating delay loop... 1488.98 BogoMIPS
SMP: Total of 31 processors activated (46170.90 BogoMIPS).
Oh.
--
Non-meta-modded "Overrated" mods are killing Slashdot
(Hey Ryan! Here's your proof!)
Lesee, 1488.98 *31 = 46158.38 - not 46170.9.
I would imagine the 31 processors there would be some overhead for SMP so it should actually be less than that. Where did 46170.9 come from?
There's a gorilla from Manilla whose a fella that stinks of vanilla and has salmonella.
Comment removed based on user account deletion
I'm glad you almost guarantee it, because I'd almost like a different opinion. I mean, if it's a "good" salary, I damn well better be able to afford 32 CPUs in a single machine!
/. is that machines like that make geeks drool. The fact that it compiled and booted is exciting in the same way that people who live in an area get excited when the local ball teams win games. It's not really rational, it's tribal bonding. Reports like this confirm the Slashdot tribal belief that while Linux may never be Grandma's desktop OS, it will continue to make inroads and eventually dominate the real computers of the world. This makes the Linux enthusiasts very happy, especially since financial rewards are secondary to the more important free-software geek reward of enhanced reputation (so saith ESR at least), and this is like reputation enhancement by mere association. "Sure I only run Red Hat on my aging P/75, but if I had 32 CPUs and a computer the size of an elephant, I could still run Linux. Let's see you do that with Windows 2000. Nyeah!"
For the record, I don't think the linked page is a brag. I think it's notes to interested parties who are working on porting the Linux kernel to large multi-processor machines. The reason it got on
I do not have a signature
Yeah. I haven't been coming here that long (maybe since Jan '99), but I remember /. having more stuff along these lines, and it was a much nicer place to be.
I can sort of get that kind of feeling by ignoring Your Rights Online, but not quite. We need Quickies *every day* dammit!
Is this post not nifty? Sluggy Freelance. Worshi
Look at the "active CPU mask" in the log. CPU 7 is offline.
Okay...silly subject aside, I'd love to see a pic of this motherboard (or boards) without it's case on...just too see what the bugger looks like.
See this sub-reference taken from this document
For the hyperlink impaired (and a smaller more concise list as well, although mistakes made are all mine, and there are a lot of funny numbers in the main list...):
386DX/40...........7
486DX2/66.........33
P90...............36
P133..............53
P200..............80
PII/400..........400
AMD K6/233.......466
PIII/600.........600
Athlon 600.......600
Duron 700.......1400
Alpha 21264/730.1500
31 CPU Alpha...41000
So that makes it 100 times as powerful as my home computer. Let's see. My home computer does 133,000 times as many bogoMips as my home computer in 1985, and it does 10 times as many bogoMips as the desktop I used for my masters degree 5 years ago.
So... you only have to wait another 5-10 years, and yes, you too will have a home PC that does 50,000 bogoMips.
This article is basically just someone bragging about his extremely expensive machine that I almost guarantee nobody on slashdot can even afford on a good salary.
Respond to s
You should ask Compaq why they want to run Linux on it.
They're the ones who did it and they are the ones writing kernel support for this machine.
I browse with my threshold at 2 so I can't read my own comments :-)
You used the equivalent of a megaphone on slashdot (the +1 bonus) to tell us that you like it and think it's a good thing, in only slightly more words? Did you think this was actually insightful or worthy of everyone's attention, or did you just think "I want everyone to know what I think, and though I'm not more sophisticated about my thoughts than everyone else without a +1 bonus, I'm going to go ahead and say it, because I can"?
.sig is not generated by each user when he posts a comment, where he's free to forge it. It's appended by slashdot's servers. It's an effective deterrent to impersonation, and it must again receive its prominence within the social jurisprudencial realm of slashdot.
There is an aristocracy on slashdot; make no mistake about it. When people like drendite (userid=#3) speak, people bow down in worship, simply because of his low userid. It doesn't matter whether one actually makes a true contribution to society; what matters is the aristocratic entitlement conferred by longstanding existence (not participation).
Take the British Parliament, for example. Though Britain still hasn't come close to providing universal healthcare or proper dentistry they sorely need, they have finally seen the folly of maintaining a ruling aristocracy, and have eliminated the hereditary seats in the House of Lords. Slashdot should follow their lead.
The solution is not to take away the voice of people with low userids. Nor is the solution to eliminate the +1 bonus, because it serves a legitimate purpose and is democratically attainable by all, from the oldest poster to the neophyte with a five-figure userid. The solution is to eliminate the tagging of comments with the userid of their posters.
The userid tag does nothing to help the community, and does much to harm it, by encouraging wishywashy moderators to inflate the karma of oldtimers and penalize the newguys who express controversial opinions. (When moderators waver between slamming a post or modding it up, they usually defer to the userid in addressing its seriousness and authority. This is unacceptable.)
If you're worried about fraud and impersonation, then you already have an effective means of distinguishing between posts: the signature. The
Effective policing (moderating) can only go so far. We must correct these social ills by striking at their sources -- their causes -- not merely at their symptoms. Join with me in tearing down the illegitimate reign of the slashdot aristocracy and their petty notions of insight and imformativeness, and lift up a glorious new tomorrow, where everyone, democratically, no matter what the tld of his email address or the number of his userid, shares in the same promise of opportunity for reasoned argument and receptive audience.
Thank you for your time.
Froid
But humor me here. How long has Linux been out and "developing" now?
Let's say I'm a big mover-and-shaker in the dot-com world and need a scalable, reliable, bullet proof solution NOW?
I thought so.
by the way... Alphas don't do much in the way of distributed.net keyrates. I ran d.net on an 8-way GS-140. It wasn't impressive at all. The quad p2-400 Xeon blew it away.
wolf31o2 Developer, Gentoo Linux Games Team
"I'd love to see a pic of this motherboard (or boards) without its case on"
Seen any horny motherboards lately ?
Go on admit it, you're Bender really aren't you
And I'm quite aware that BogoMIPS are an even more Meaningless Indicator of Processing Speed than the indicators that people try to take seriously... It is of practical value, but only in predicting the performance of timing loops...
If you're not part of the solution, you're part of the precipitate.
Porn for Bender...
Phooorr look at the processors on that one
its given me a full on robot chubby
So when do we get to see how well Linux handles processes across those chips?
I like to wear big-boy pants.
I work at DEC^H^H^HCompaq. We have one of these bad-boys in a lab, and I have to walk by it a lot. It's this fairly massive thing, the size of 3 or 4 fridges all in a row. They're bluish, and the heart of the beast is identifiable by the LED display that sits at about eye height, saying things like:
AlphaServer GS320
16 processors configured
And so on. I adore it. And every once in a while, when nobody's looking..
I give it a hug.
Once, I was talking to a co-worker about it.
"You know the Wildfire in the lab?" I asked.
"The what now?" he replied. So I told him about the Wildfire. Later that day, we were walking through the lab, past it.
He gave it a hug.
If anyone else wants to send the wildfire a hug, let me know.
SMP: Total of 31 processors activated (46170.90 BogoMIPS).
Who says Linux can't scale? Look at those bogomips!!!
Honk if you've killfiled JonKatz!
Same thing with Sun and their boxen...
Sheesh. "Hey look, Linux learned a new trick today everyone!"
I think we should hold a fund-raiser in which people would pay $20 to be able to see this thing in the flesh. $100 if you want to actually touch it.
:)
I'd be first in line