Slashdot Mirror
x86 vs PPC Linux benchmarks
Jay Carlson writes "We've all heard about how Apple's hardware is really fast compared to PCs. ("Supercomputer!" "Twice as fast as a Pentium!" "Most powerful laptop on the planet!") So, if you *aren't* going to use Photoshop and Final Cut Pro, how fast is it? I care more about integer apps like compilers, so I did some careful benchmarking of a few x86 and PPC Linux boxes. Submissions welcome."
I would have thought that, by now, almost all slashdot readers would be aware that benchmarks really ought to come with one of those warnings that you see on certain late-night commercials: "For entertainment purposes only."
A 500MHz Apple box outperforms a 1.7GHz P4 box by a good margin. The reason these benchmarks don't reflect this is that the compilers for Apple generally aren't as tight as IA-32 compilers. The Apple version of GCC doesn't even make use of half the opcodes!!
So this really proves nothing. If you want to benchmark two boxes like this what you have to do is time each operation individually. Store on an Apple box is about 3x faster and load is 6x faster. ALU operations are usually about 5x faster. It seems to me that Apple is the Mercedes of computers while IA-32 is like a Kia.
Comparing *apples* to oranges! (rimshot)
That is completely false! GCC has known problems mainly with register allocation and instructions scheduling (one of the scheduling passes is not even run on x86 for this reason) on machines with a small number of registers, like x86. Other optimizations also perform suboptimally for x86.
that, among people of equal age and education, Mac users have better English grammar and non-technical spelling skills?
Actually gcc does a poor job of optimizing on any platform - compare it to Sun's C compiler on SPARC or Intel's C compiler on x86, and it loses fairly badly. gcc's strenth really is that it's portable and free, not that it produces very good code.
Considering that was only 1 533mhz g4 processor I am kind of impressed. It held up well with/o altivec support, and a second processor. Rumormill is say'n apple is going to start going back to a fully MP pro line up again now that OS X is out. That is a smart move. Clearly a dual processor g4 could have kept up quite well with the AMD and Dell boxes. Why didn't this guy do his benchmarks with darwin instead of Linux PPC? Darwin is cross platform and MP aware. Sounds like more of a fair fight. He should of at least used Yellow dog linux 2.0. Isn't that PPC native an MP aware. Linux PPC kind of sucks compaired to what others are dishing out.
Instructions for Do-It-Yourself x86-Biased Benchmarks
... the dream CPU!)
Step 1) Disable one of the Mac's processors
Step 2) Run x86-biased benchmark suite
Step 3) Publish useless benchmarks
Step 4) Congratulate self yet again for saving money by building your own x86-compatible PC so you can use it to create useless benchmarks (Uses 5x the power or a Mac! Generates 5x the heat! May be faster under some circumstances when plugged into wall power! x86
Step 5) Reward yourself by enjoying music, movies, TV, books, artwork, advertising, and Web sites all produced and encoded on Macs by people who are too busy enjoying their work to have time to pause and make useless benchmarks
Just wait until you see my 1.33 GHz DDR (double-doodoo-rate) coleon (food)processor
Intelligent? Huh! Mr Carmack just said what others have already said, with the difference he is famous... That is all there is to his remark.
Please, be cool about him. His only a demi-god. Nothing more.
As for the moderators... Brrrrr... Post #292 even has benchmarks posted, with a link to the source, but got Score:0... Mr Carmack had vapourclaims to offer, beyond his persona.
Do we talk fat facts or do we not?
Interesting article found below as far as the "supercomputer" argument goes for scientific computing, cost/performance analysis.
r ch .html
p df
G4 fares quite well even with incomplete Altivec support in the FORTRAN libraries.
http://developer.apple.com/hardware/ve/acgresea
>>>>>>
Research from outside laboratories/developers/users
An Evaluation of PowerMac G4 Systems for FORTRAN-based Scientific Computing with Application to Computational Fluid Dynamics Simulation
by Craig A. Hunter, NASA Langley Research Center
http://ad-www.larc.nasa.gov/~cah/NASA_G4_Study.
And I've done them myself, and I'm as much of a Mac nut as anyone. And yes, the PC came out faster. But the Mac was a 500 mHz and the PC was 1 gHz, last time, and the PC came out (when I averaged everything together) a tad less than 55% faster. I.e. the 1 gHz PC was the equivalent of a 775 mHz Mac. Cycle for cycle, the Mac wins. All out, the PC wins. In Photoshop (or elsewhere where one can really take advantage of those beautifully-engineered Altivec instructions) the Mac wins, at least sometimes. Why is it so hard for a single person to swallow all three of those conclusions? It's amazing; you get people on the Mac side who just can't stand to think that their computer is slower than their best friend's. You get people on the PC side who can't stand to think that the Mac's processor really IS more efficient than the PC's. (And, given how well the P4 runs X86 code compared to the P3, this gap is widening. :) Grow up, people. Sheesh.
Now, with a well-done app on a dual-processor G4, you can see some pretty wild results. Yes, you can get the same thing with a dual Pentium, I hear. I've never seen any hard numbers on this, given different OSes (preferably X on the Mac, since 9 doesn't take advantage of dual processors). It'd be interesting to know how well a well-threaded app can do.
--Fred Fnord
Apple says that the G4 is faster because of its more advanced FPU and its far superior Velocity engine (Mot calls altivec) they make no claims that the interger calcs are twice as fast. Never have. To realise that OS X and its apps can easly be Altivec accellerated (Click chack box in compiler) would crush your silly theory. It's like saying that the PI with MMX is faster than the PI without... but if you don't use MMX apps then its not... well no shit!
The Athalon with best price/performance. Not too surprising.
...)
TO THE AUTHOR OF THE BENCSHMARK (Jay): Having the machine names in the comparison boxes is silly. We don't care what you'vr named your machines. Try CPU abbreviations next time. (e.g. P3/733-192, G3450-320,
From http://fampm201.tu-graz.ac.at/karl/timings40.html.
Numbers are relative to a G3-300MHz (higher are faster).
There are more numbers on the homepage.
Athlon 1.2 GHz, 512MB, Windows 2000 [73]: 4.78993
Athlon 1.2 GHz, 512MB, Linux [72]: 4.4734
Gateway Select 1000, AMD Athlon 1000 MHz (1GHz), 512KB L2, 192 MB, Linux [65]: 3.77305
Kryotech 1GHz AMD Athlon, 512k cache, 512MB, Linux [66]: 3.69674
Gateway Select 1000, AMD Athlon 1000 MHz (1GHz), 512KB L2, 192 MB, Linux [60]: 3.57748
Dell Dimension XPS B1000r, 512MB Ram, Win98 SE [64]: 3.38084
Dell 4100, 933MHz, 128MB, Linux [63]: 3.19988
COMPAQ AlphaStation XP1000, 2 GB RAM, 4MB L2, Digital Unix 4.0F [50]: 3.16987
AMD Athlon, 800 MHz, 512 KB L2, 256 MB, Linux [61]: 3.00154
Dual Xenon 866, 512 MB, Windows 2K [71]: 2.9618
PenguinComputing, dual 800Mhz PIII, 128Mb, Linux [67]: 2.76745
Athlon 700, asus k7m, 512 mb, win98 [43]: 2.70199
Dell 800 Mhz, Pentium III, 512 MB RDRAM, Win 98 SE [57]: 2.69821
Athlon 700 MHz, 128 MB, Linux [51]: 2.67027
Athlon 650 MHz, 256 MB, 100 MHz, Win NT 4.0 [42]: 2.60662
Compaq-Digital Alpha 8200, 625MHz, 1GB RAM, DEC-UNIX [11]: 2.48495
SONY VAIO F409 notebook, PIII 650 MHz, 128MB RAM, Red Hat LINUX 6.1 [55]: 2.21137
Athlon 650 Mhz, 32 Mb, Windows 98 [32]: 2.14558
Athlon 550MHz, 128MB, Red Hat Linux 6.1 [45]: 2.13797
Dell XPS T600r, P3 Copper wl 256kb, 128MB, Linux-2.0.36 [39]: 2.09939
Dell Precison 410, 2 PIII 550MHz, 256Mb RAM, Windows NT 4 [28]: 1.8561
Dell Precision 210, 550 MHz, 128 Mb, NT 4.0 [34]: 1.83606
Dell XPS T550, PIII-550, 128MB, RedHat-5.2, Linux [24]: 1.777
Gateway GP7-500, 500 MHz, 192 MB, WinNT 4 [6]: 1.70318
PowerMac 8500, 500 MHz MACh Carrier G3, 1 MB L2, 256 MB, MacOS 8.6 [35]: 1.68249
Instead of pitting his machines against one another, he should've assembled them into a BEOWULF CLUSTER so they could WORK TOGETHER.
give me a fucking break, you don't optimize for an architecture, you recompile with it as the target. debian probably has a horrible installer for ppc, but it has a horrible installer on x86 too, so it probably doesn't matter.
I'm a PC head but I was going to buy myself one of those nice G3/500 iBook because of the killer combo of a superb hardware feature set, Unix core and a nice GUI.
Now you're telling me that they have a SUCKY performance? I was at least expecting that the G3/500 would be able compete with a P3/750 and a G4/500 with a P3/1Ghz. Damn it!! Why must you do this to me Slashdot? Why must you cast doubt in my heart?
Or more like 3.14159 + 2010?
--
Forget Napster. Why not really break the law?
"Remember when the U.S. had a drug problem, and then we declared a War On Drugs, and now you can't buy drugs anymore?"
First of all, because my older Mac basically does what I want in a style I like. I won't give Microsoft a penny- we're not talking about that, so that's moot, we're talking about Linux platforms. I also won't give Intel a penny if I have a choice- and am seriously questioning whether it's good to give nVidia money either at this point. They'll only use it to impede progress, they are already doing it, strong-arming vendors.
Second of all, the CPUs are so different anyway that it seems crazy to try and compare them. It's like the x86 are model airplane engines screaming at 60,000 rpm and the PPC is a truck rumbling at 2000 rpm. It's a difference between top-end horsepower and bottom-end torque. PPC is register-rich and has a relatively shallow pipeline. I daresay the version of GCC used could have been better, but even if it was fully optimised, the 'torque curves' of the chips are DIFFERENT!
The x86 has been designed for years to scream for doing certain very narrow tasks- the simple repetitive processing of games, the focussed processing of well optimised OS routines. In many ways this is the most common situation (though I tell you, I've seen PCs sag and go unresponsive... admittedly running windows...). However, PPCs are a decent general purpose tool for _broader_ tasks. Anything that can use all the registers at once, slog through really big amounts of data in complicated ways... hence, the way Photoshop filters keep coming into the spotlight.
I don't know what all this proves, nor do I especially care since any of it is 'good enough'. I guess the bottom line for me is that I can't see performance metrics as being the end of the story. When you look beyond the actual performance and consider what happens as a result of your buying decisions, it becomes clear that the better immediate information people have, and the more prone they are to consider nothing whatever but raw results defined as narrowly as possible, you see the real reason why Microsoft is choking IT to death, why nVidia is currently threatening vendors to cut off the air supply of competing 3D chipmakers, why Intel destroyed other choice for so many years.
The end result of "X>Y, therefore all your base are belong to X" is cartels, stagnation, and the choking off of true progress. This is the case even when X is indeed >Y. You can't have a market if you're only allowed to buy one thing- and if you're not free to do whatever the hell you want, for any or no reason, you're being restricted by your own anal-retentive perfectionism and playing right into their hands. Two years from now these CPUs will ALL look like crap, but if you can successfully and publically make the case that there is only one greatest choice and nothing else will do- why, you are part of a market force handing complete dominance to that choice (like with Microsoft) and trusting it to keep on deserving your support AFTER it has no competition left and can do what it wants: and when have we EVER seen ANY company deserve our trust after it has controlled its market? It's not in their nature.
Which is to say- I'd still get a Mac. And people can do as they please, but I really can't have much respect for those who'd denigrate me for my choices- I have enough time and patience to run an older 300Mhz G3 machine in relative comfort, so that counts as 'enough', plus I cannot forget the larger situation, all these companies busily trying as hard as they can to do away with all capitalism and become the single source for whatever it is they do. That disgusts me- it's not what I call a suitable model for society. So rather than whine or write long dissertations about it, I ACT in accordance with my beliefs.
If any of you guys REALLY believe that PPC should go away- you should be running Windows, not Linux. People have all kinds of motivations for what they do, and in the larger scheme of things, 50% or 80% or even 300% processing speed disparity is pretty insignificant. Have some historical perspective.
End long, rambling, crotchety ol 'rant ;)
The sole reason Photoshop is faster on Macs than PCs (recently), and the reason Steve always brings it out to show off, is altivec. We need benchmarks of vectorized vs non-vectorized versions of the same task, running on the same G4 or against an x86 box. Unfortunately I can count the programs that use altivec in the real world on one hand; hopefully this will change with OS X arriving (altivec in system libraries).
You're overlooking the obvious: solaris is a very slow operating system. Especially its filesystem and disk-access layers are very slow, and the entire OS has been optimized for the MANY-processor cases. Running solaris on a box with less than 8 CPUs and less then 4GB of memory is just silly. And if you can afford a box that big, you can afford hardware RAID and VxFS and so on to cover up Solaris's atrocious performance.
XFS on solaris would indeed be nice; UFS is slooooooow. I'd even settle for ext2 - no journaling but it's a helluva lot faster than UFS. Actually it'd be nice if XFS/linux worked on bigendian systems too.
I guess you've never tried Linux on sparc64. I consider this the best-supported of all architectures. I've used m68k, ppc, mips, mips64, sparc, sparc64, and i386 and have used linux for over 7 years in total, including a fair bit of mips and sparc hacking, so I consider myself in a good position to judge. When I install Linux on a sparc64 system it Just Works; the last time (last year) I installed on x86 I had no end of problems - different problems on each of the 40 or so systems I was maintaining at the time. I'll never go back to x86; draw your own conclusions.
As I said, Solaris is optimized for very large machines running many processes simultaneously. If you can afford to drop a million bucks or two for 64 CPUs and 64 gigs of memory and lots of striped disks on FC controllers, Solaris performs fairly well. Not as well as IRIX, but fairly well nonetheless. OTOH Solaris is all but useless for the 1 and 2 CPU systems commonly used for workstations and smallish servers. The hardware is great, however, and I highly recommend running Linux instead of Solaris. You don't seem to lose anything in stability but the performance increase is very nice indeed.
Unfortunately unless you only use gcc, it isn't a very good benchmark for CPUs. Unless every CPU ran in the same otherwise identical system, other differences would greatly affect the outcome. The gcc test is in fact a fairly good exercise for the system as a whole - it covers disk i/o, memory bandwidth, cpu power, and the abilities of the OS. Unfortunately the CPU is not usually the bottleneck in this scenario. If you have little memory, it's the disk subsystem. If you have lots of memory, it's either the ability of the OS to use it, or the memory bandwidth itself. I'm actually a believer in the gcc benchmark - but not for CPUs.
It's not that the processor sucks. It's just that Apple has put a heck of a markup on them, so that they make a hell of a lot more off of them than the manufacturer does. This is the opposite of the situation in the PC world, where the PC builder makes much less profit than Intel... so what should be a superior processor architecture gets marginalized.
Thank you, Apple.
(currently testing something about signatures here)
I know what you're saying, but if the Athlon really were like the PPC, it would run with 10 watts. It doesn't.
I know it's probably too late for the PPC, it probably became that way when Apple axed the clones. But we've lost a lot of potential because of that. I wish I could get a desktop with a Transmeta processor. Nice, energy efficient, no fan in the living room I can hear from the bedroom. Oh well.
(currently testing something about signatures here)
Although I do have commodity memory in my iBook, apparently Apple's latest firmware fixes change things to that a lot of commodity memory doesn't work. Of course, maybe you'd better let them gouge you, just to be safe...
I wish PowerCAD would come out for Linux. I wouldn't be chained to the Mac any more.
(currently testing something about signatures here)
And the author said he didn't use p3/p4 specific benchmarks in the test. Just generic x86 stuff, which is what you get from Debian.
(currently testing something about signatures here)
Or, in perspective, weekly Seti units per MHz:
;)
Athlon: 0.021
Intel: 0.024
Motorola: 0.025
This indicates that the G4 is slightly better for SETI@Home than an equivalent x86... and consider further that the x86 version is probably more heavily optimised. Ultimately, it's irrelevant as you can't get 1200MHz Macs, so the less efficient CPU wins because it still get's more done.
BTW, command line SETI client on a dual P2-450 under Win2K: at least 65 WU / month (I think I'm averaging 11hrs / WU)... so you could say my aging P2-450 is better for S@H than a G4-400
A dual machine does give a small benefit: the SETI process gets a bit more CPU time over the period of a day as there is another CPU to share the load with everything else I run. I mentioned that I have a dual CPU machine as it gives a more accurate picture of the CPU dedication to S@H for the period of a month... I wasn't being ignorant and thinking that S@H was taking advantage of it.
I did try running the command line version and screen saver on the machine thinking that 3 threads on 2 CPUs would be better than running on 2 computers... for some reason, it didn't work out as well as I expected.
In general, though, scientists don't care about integer results-- floating point is more important. I have heard that some Crays were notoriously slow at integer arithmatic.
My apologies for the confusion if it seemed so.
Er, that is to say, I got 1,261 search results each representing a patent. I don't have 1,261 patents myself. :-)
Thanks for clearing that up... with the USPTO's track record, I wouldn't have been much surprised to learn that you actually obtained 1,261 patents in the course of looking up some patents. ;)
But it seems to me that GCC builds much slower code on powermacs than whatever it is Apple builds their binaries with. (If Apple's using a modified GCC, which wouldn't much surprise me, I sure wish they'd throw us their patches.)
Now, if I'm right, and GCC produces relatively unoptimized binaries, shouldn't you compile GCC itself (and maybe even the rest of the system) with another compiler before pitting machines against each other in a compiling race? It seems to me that you're using a slow, badly-compiled GCC binary, otherwise.
Granted, this seems an excellent real-world test, since nobody does that. But I can't help but feel that we're (by "we", I mean the Community[TM] )currently incapable of exploiting the PowerPC, and it seems unfair to blame the chip.
Maybe I'm wrong. I would love to see some discussion from the GCC team. I just thought since nobody else seemed to have brought this up...
The drawback is that the larger number of instructions required to complete a given task is likely to take up a larger amount of space in RAM and on disk.
On the other side of the coin, most RISC chips also have more general purpose registers, and so can dispense with much of the register/memory shuffling that x86 typically has to do.
Just getting rid of some memory accesses, as well as removing redundant instructions, also speeds up execution as register/resgister operations are MUCH faster than even cached memeory accesses.
The results look nothing like the compiling benchmark, and have convinced me to start a web hosting company using OSX Server on Macintosh hardware.
Maybe I'm a hopeless aesthete, but I wouldn't do that yet until Apple releases proper rackmount servers with redundant hotswap fans/PSUs/HDDs. They know how to do it (They even used AIX), it just hasn't happened with a G4 yet.
Also, I'd hope that they'd release a server with an improved memory interface (multichannel PC2100) because web servers love memory bandwidth..
Your Working Boy,
- Otis (GAIM: OtisWild)
Welcome to the wonderful world of vendor-supplied memory. All the big Wintel vendors rape you equally when you buy their RAM, except for the occasional promotion for "FREE 128MB RAM with system purchase, for a limited time!!!" that you see in the mags and in all the mailorder catalogs.
It's not just Apple.
However, to give you a bit of good news, all Macs sold in the last several years have user-installable RAM, and with the exception of the original (Rev. A through D) iMac, it's very easy to do - easier than in many PC's.
The new G4's use standard PC133 SDRAM, all other model desktops use PC100 (the Cube and iMac), and the TiBook uses PC100 SO-DIMMs, the iBook (old and new models) uses PC66 SO-DIMMs, though PC100 works fine, too.
PC133 SO-DIMMS seem a tad flaky so far - I just got a Gateway 9500 laptop at work and still haven't gotten a 3rd party 256MB SO-DIMM that will work with it (we've tried Samsung, Micron, and Hitachi, with Infineon on the way). Apple will probably start using them with a TiBook revision at some point, after the interoperability issues vanish.
- -Josh Turiel
-- Josh Turiel
"2. Do not eat iPod Shuffle."
The AltiVec is primarily intended for graphics and audio processing. It's analogous to the Motorola 56k in the original NeXT machines.
There's absolutely nothing sexier than the titanium powerbook however. (no notebook computer anyhow ;). That's the *real* reason everyone wants one, even if they won't admit it.
According to an article in the linux journal
the Altivec unit handles only single precision
floats
-> only useful for special tasks or for calculations were you can take care for the
reduced precision,
-> not for general scientific numerical calculations
Unless you think paying $400 for something that you can get for $26 is somehow fair, I think it's you who needs to stop 'beeing a troll'. Looser. What about the fact that you can use that $26 256 meg chip in the Apple machine? Why should we be surprised that Apple tries to make money on RAM? What PC builder doesn't?
Know what I like about atheists? I've yet to meet one that believes God is on their side.
IMHO, I think a list of hardware, bios settings, and hardware and software tuning parameters are required in benchmarking results.
Otherworld - damn it, replay this series! http://epguides.com/Otherworld/
As a long-time PC user, I just ordered my first Mac last week. I've used them before at many jobs, but never owned one of my very own. I chose to get the new iBook.
Why? Because Apple actually has the x86 world beat in the notebook category. The cost of my iBook (I work in education) was $1545 + $237 for the AppleCare warranty. Try and configure a PC laptop the same way for $1545. My iBook has an XGA display (12" yeah, but still XGA) AirPort card, built in ethernet and 56K modem, as well as DVD. I swear by Dell systems, but I couldn't come close to touching the iBook for the same price, and I would have had to tolerate an external wireless antenna, as you can't have ethernet and 802.11b in a laptop yet...
I think that the G4 desktops are still overpriced, but the iBook line is very reasonable. The iMacs are okay, but the 15" CRT is dead, Apple. If they came out with a 17" version, they'd see a renewed interest in them...
---
When in danger or in doubt, run in circles, scream and shout. --Robert A. Heinlein
Well, if codes are optimized for Altivec, will it be fast as Apple says? (700-something Mhz PPC is
....
as fast as 1.5GHz Pentium IV? )
Well, although I don't try it, but I'm very pessimistic on that. Because the Altivec, SIMD, is not for increasing general performance.
Only if there are lots of array calculations, etc, Apple H/W will be fast as much as
Uh.. slower than what Apple says..
Macs have never run NT or OS/2 for PPC. Those OSs run on the Motorola PPC workstations, which are a fairly different kind of beast.
The benchmarks were not adjusted for MHz. Apple's claims about speed are valid; they're testing their software, which is as optimised for PowerPC as much as gcc on GNU/Linux is optimised for x86.
Comparisons between platforms should be done with similar memory and processor speeds; a while back, I benchmarked my G3/450 (100 MHz memory bus) running NetBSD against my girlfriend's IBM pentium III/450 (100 MHz memory bus). That tells a tale: the G3 was, on average, 1.2 times the speed of the Pentium III.
Otherwise, these results should be adjusted for MHz.
I'm probably a lot like your accounting prof as far as Macs go. I run LinuxPPC as well as MacOS 9 and X, but only because the PPC box was available. I'd have no problems with running it on an x86, but I've never owned an x86. I've owned three Macs, from a 68040 to a G3, and an Apple IIGS. I stick with Apple stuff basically because it treats me better. There's no one thing about it, but I have way less trouble with any of my equipment than anybody I know who does anything comparable on PCs. My next computer, coming up shortly, will be either an iBook or a Titanium PowerBook for those very reasons.
:-)
PCs win on a lot of points, but Macs are better for me in a, well, holistic sense.
I'm just trying to say that people like me who stick with them aren't mindless zealots, but there are actual reasons why we do what we do.
Mod down posts with a "Free Mac Mini/iPod" sig, they're spam!
Why oh why oh why does everybody bring up Apple's memory prices? Nobody in his right mind buys memory from Apple! Believe it or not, regular PC66/100/133/whatever memory works fine in Macs.
Mod down posts with a "Free Mac Mini/iPod" sig, they're spam!
The fastest Mac CPU hasn't been a G4/450 for quite some time now. It's now a G4/733, which is still not as fast as the top of the line P4 (again, depending on what you're doing), but it's better.
Mod down posts with a "Free Mac Mini/iPod" sig, they're spam!
I can, but it doesn't apply here. Do you know what "holistic" means? It's not synonymous with "touchy-feely". It means taking the whole picture, looking at the whole device. I feel that Macs are better for me when you take the entire machine into account. That's all.
Mod down posts with a "Free Mac Mini/iPod" sig, they're spam!
I don't know where you get your information. I know for a fact that Apple machines have worked with commodity memory since the days of SIMMS. Certain iMacs require portable-style memory because of the form factor, but it's still the standard stuff.
Mod down posts with a "Free Mac Mini/iPod" sig, they're spam!
A dual P2-450 running SETI doesn't get you crap because SETI doesn't launch a number cruncher for each processor like d.net does. Unless you're running two different SETI clients in different directories or something your second processor is running idle. I average 10 hours/WU on my dual P3-500 which only half of the processing power gets used. With d.net on the other hand I get 100 utilization.
I'm a loner Dottie, a Rebel.
Yeah, there are lots of ways you can "benchmark" a particular processor or operating system. The problem is not all OSes or processors are made the same. I'm not defending Apple or Intel or AMD either. If this was an AMD/Intel benchmark the fact a 733 was headed against a 1200 would get tons of people's panties in an uproar. Secondly, you know marketing is bullshit. How many PC makers tout their boxes as the fastest PC you can buy or some shit. Even between individual chips, Intel claims they're the fastest and so does AMD. Benchmarks are not magic bullets, they're merely magic bullet theories.
This benchmark in particular is bullshit. What should be looked at is peak work done per clock and total price per clock if you want a price comparrison test. Or work per watt or maybe even cool facter per work/clock. PPC chips have an advantage of not needed an instruction decoder and a larger register space. You can do a bunch of LOAD operations on the same clock so more operations are running on the next clock than with x86 based systems. On the otherhand x86 processors have the advantage of a higher clock speed which begins to negate the lower register space. In general, anything you compile with gcc is going to suck ass. Do you REALLY think Apple uses gcc? Goddamn they use Motorola's compiler! They give you gcc because they don't want to pay the licensing fees on the fucking thing for every copy of OS X they sell. Motorola's compiler is oodles better than gcc ever will be, ever. In the benchmark's preamble he says "well I don't use Photoshop" well goddammit the Apple statement just became invalid as did his entire premise. Since he wasn't benchmarking Photoshop 6 he ought to have been using a 450MHz P3 and Athlon rather than shit that is OBVIOUSLY going to be faster in mere clock speed. If you can't tell from the tests the compilations have alot to do with moving data between the processor and memory. The G3 in the iMac and G4 are already at a disadvantage due to their bus speeds while the Athlon gets to chug along on its EV6 derivitive bus. Testing the processor ought to involve shit that can be held entirely in the chip's cache (which can be considered part of the chip because it afterall it's memory cache). Then the chip's FSB speed and register space size become important. You can then get a good measure of how much the processor is doing on every clock and how many clocks it needs to get a particular job done. At this point you've got a real test. Work done per clock or per second or watts used per work cycle or something. Here is where you say X processor is more better than Y processor. This is why SPEC tests cost so much money. They have LOTS of different tests that test things in different ways and they still don't show a true measure of real world performance.
In the real work you might be running Final Cut Pro or Photoshop 6 in which case you damn well better have a system they are optimized on (yes I'm aware FCP is only available on Macs). If your real world workload involves compiling go with the system that works best for your particular task at hand. This dude did a compiling benchmark test, he has benchmarked COMPILERS compiling binaries for whatever. Everyone here's been arguing "well such and such is faster than such and such" fuck that. This is not a x86 vs. PPC test or something. This is somebody feeling righteous because they keep getting reeled in by marketing quotes and feel like they've been lied to. Use what works not what a marketing department says works.
I'm a loner Dottie, a Rebel.
Okay I see your point then. It's a shame SETI hasn't taken a que from d.net and done a better job handling multi-processor boxes. A majority of the clients they have running are Win32 (which usually means Win9x/Me) but if they launched a number cruncher for every processor or ran two crunchers on the same WU the bigger boxes some people run S@H on would benefit alot.
I'm a loner Dottie, a Rebel.
If you bothered to actually read the article you'd have noticed that the author actually did a bang-for-buck table for the tests he run.
Oh yeah, nothing's worse than those damn noisy Intel chips.
--
--
Mod up a post Rob doesn't like and you'll never mod again
Except that many people use and enjoy their high resolution laptops, including myself. It was the only thing I really shopped for when buying a laptop. I would not consider buying a laptop with less than 1400x1050.
LCD screens are very crisp, clear and sharp. You shouldn't have to squint. If you find yourself doing so, you should see an eye doctor.
robert
Well, *586* optimisations would be pretty stupid.
Pentium optimisations generally reduce
performance on P6 or later generation cores.
Dunno how much difference 686/ K7 optimisations
make.
Rob
No, gcc isn't a good x86 compiler. GCC is a multi-target compiler, designed for RISC processors, like the PowerPC. In x86 hardware, gcc produces about 10-15% slower code than Intel's own compiler in integer code, and about 30% slower code than Intel. If he had used gcc-3.0 targeted to pentiumpro (which he didn't, he used 386 target) the x86 wold be better stil.
The only thing that is fast in PowerPC is "altivec" instructions (faster than MMX and SSE), but nobody really uses it (except in hand-optimized code).
Seems that you didn't read ther article, because You are plain WRONG!
He is comparing two CROSS-COMPILERS to MIPS code, so all that you say don't apply. The cross-compiling is the same task in each case.
Please, go and read the article.
Thus any scores over .61 and .72 respectively, indicate that the PowerPC is doing more per clock cycle than then PIII. If Motorola can ever get their act together (and that is not a certain), normal code on the PowerPC will run every bit as fast and faster than the x86 processor
This illustrates a common fallacy - if chip A does more work per cycle than chip B, then chip A is "better" and as soon as those "idiots" who make it get the clockspeeds up there it will perform better. This neglects the fact that B may do less work per cycle precisely because it is designed for extreme clock speeds, and in fact there are plenty of instance where the "speed demon" cpus (high clock speeds, simple instructions) outdo the brainiac chips (lower clock, beefier instructions). The reason Pentium 4 trounces any PowerPC is that it is designed to scale to 2GHz. Of course it does less work per clock, but overall it does more per second, and that is the more important metric.
No, the Apple Store price I quoted in that table is for a single-processor G4/533. Go price it yourself.
(I'd give a URL but the Apple Store is too sessional.)
All this means is that compiling on a RISC architecture is bound to be a great deal slower.
I'm aware of that, and I talk about it on that page. See the "Choice of workloads" section.
The install-egcs test does measure native compilation performance. This is relevant for people who use the box for development.
The install-glibc and cross-gcc tests are both compiling to a single RISC architecture, little-endian MIPS. The amount of effort required to optimize for PPC or x86 doesn't factor into this.
If you don't care about development compile times, just look at the cross-only numbers.
An important difference you neglected to mention is the x86-emulation front-end built into all current x86-class microprocessors.
Yes, they are still superior (at higher clock speeds) to PPC chips for integer performance... but they also dissipate 10 or more times as much current... which generates an incredible amount of heat, and, arguably, wastes a lot of energy.
How fast do you need to do most of the things you do? For desktop use, the machine will usually spend 90 percent of its time idle. It's waiting for you, not the other way around.
"Bang for your buck" is not a simple function of hardware throughput. What the OS does for you, what your apps do for you, whether or not you need two or three fans to keep your CPU from turning into plasma, these are all very relevant factors which influence the choice of a platform. Specific combinations of hardware and software are inevitably going to be more productive for certain tasks. For the arts and humanities, Macs offer some well-integrated technologies and some unique advantages in desktop applications that remain unsurpassed. If you're Jay and you want to hang out and build your OS over and over again, you need an Athlon running Linux or BSD.
No one is forcing you to buy memory from Apple. It's the same stuff that you use in a PC. Buy it somewhere else.
It's much easier to troll when one is not completely ignorant of one's subject.
I'd have to agree based on my subjective, biased observations and experience. FWIW, compiling the BeOS Tracker on PPC with mwcc (the Metrowerks compiler) is a hideous saga. Compiling the same code on x86 with various versions of gcc is not much worse than your average *nix kernel. Yes, I realize we're comparing apples and oranges.
One thing that used to amuse me is that mwcc could take up to -O7 for an optimization option. Also wryly amusing is that if you google for "mwcc", the first result that has anything to do with compilers is called "abiword-dev Archive for June, 99: Re: Stupid mwcc compilers". Heh.
I dunno which one actually compiled faster, though.
If the commodity ram you put in a machine with updated firmware fails to work, it's because it's out of spec. I paid $190 for 2 256 PC100 SODIMMs which are working just fine in my firmware-updated TiBook. The reason a lot of commodity ram fails to work is becauswe it's out of spec. PC100 means more than "probably runs on a machine with a 100MHz memory bus"
oops - i replied before I even read the article! my comments still stand, but it seems he was doing cross compile not native compilation, so it was a fair benchmark.
The point is that gcc is going to be going a different amount of work compiling the same program for different targets due to the different work done in the code generator and optimizer for the different targets; therefore, if you want to compage gcc running on different platforms, you should compare them compiling to the same target (i.e doing a cross compile - which is what he did) rather than compiling to the native target.
I disagree. It's easier for a code generator/optimizer to generate code for a simple/orthogonal (RISC) instruction set than for a CISC one. I agree though that compiling native code isn't a good benchmark since it's doing different things on different machines - better just to compile for, say, x86 on all platforms if you want to use it as a benchmark.
Compiler optimizations patented?! :-(
Do you have any examples?
Surely they can't have patented emitting certain code though - just the algorithms they use to produce it?
Most of my noise is from my 10krpm drives... the CPU fans are nearly inaudible. The drives and their cooling are the much greater part.
--
"It's tough to be bilingual when you get hit in the head."
By no means was I try to imply that mac-addicts are mindless zeolots :-) There are plenty of things in my life that i've bought, which while I could've gotten the same basic thing for a lesser price, I went for the more expensive item "just because I liked it more".
S.t.e.v.e.
It seems that the PC comes out on top when compared to apple. However did anyone really expect anything else? It always suprises me when people bring up apple's quotes (super computer, faster than a p3 etc.) What do you expect them to say? "Well our G4 is slower than the other new chips out there but costs more! Please place orders here!" I mean come on. The one thing that apple has going for them is their incredibly loyal userbase. My accounting professor uses Macs. I have gone over the benefits of PC's many many times with him. And on a lot of points he agrees that the PC is better.
He just bought a titanium powerbook.
'nuff said.
Although that cinema display is super slick (as it should be for its price tag...).
S.t.e.v.e.
-------
CAIMLAS
~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
Though it might more accurately be described as a GCC/Linux PPC vs. x86 benchmark, it's still interesting to those of us who are committed to GCC and Linux. Those two items, and their spawn, suck up almost all the (used) cycles on all of my machines.
Geeky modern art T-shirts
That's why he compiled for MIPS on both platforms.
He covered that in the article, as can be learned by reading it.
I hear this song and dance from folks that used
to point out how slow Alpha is on linux comapared to XYZ. Until a few years ago I couldnt refute their claims. That is until Compaq ported their hand tunned UNIX C/C++/Fortran compilers to Linux. Behold,The Alpha spanked everything in sight. The same deal goes here with PPC.
GCC was not optimized for speed. It was optimized
to be as portable as possible first. Where does GCC get worked on the most? X86 of course. So it's a big surprise it's the fastest supported
platform?
A FAIR test would involve using a compiler from an
independant vendor like KAI or the Portland group.
Then use SPEC XYZ , STREAMS, DRYSTONE, etc etc. That's a fair test. Alas I know KAI doesnt support PPC except on AIX and I don't know about Portland. Can Linux/PPC run AIX binaries? Linux/Alpha can run DUNIX binaries.
Another reason why a compile test of CISC vs RISC
is unfair. RISC being a reduced instruction set (hope I didnt surprise anyone with that) requires the assembler to put together more instructions
to build said binary. This takes more time and also results in a larger binary. This is where a RISC implementation needs to be FAST. Alpha has that advantage (the huge/fast cache and memory throughput helps ALOT), Apple/PPC doesnt.
Well how do you solve that problem then? You put
people on GCC or you contract Cygnus to do the work for you. API Networks has done the later and the changes have been folded in to GCC already and will be there in 3.0 .
For Apple to do such a thing their marketing/sales
department would have to show that spending money
there would increase hardware sales. That or the press alone is worth the cost. We're not talking millions of dollars here but it's not small change either.
The people who I would bring to their attention is IBM. With all the work going into S/390 of course they're going to want those compilers up to snuff. Get them to work on glibc too if you can.
So ladies and gentlemen. There you have it.
Peter
--
www.alphalinux.org
www.alphalinux.org
Hi. I'm a GCC maintainer. I don't work on this part of the compiler, but I can speak to this point:
Nearly all of the optimizations in GCC are not machine-specific. Those kinds of optimizations, ones which are specific to the processor, are called peephole optimizations, and while every little bit helps, they don't make that much of a difference. The big ones are done at an intermediate level, before the compiler "knows" what processor it's using and starts to chunk out the opcodes.
More specifically, unlike the Linux kernel, glibc, and other major projects, GCC is not designed for and targeted primarily at Intel chips. The x86 is just one more back-end like any other; sometimes it falls behind and sometimes it pulls ahead, development-wise.
Some have, some have but won't be in the upcoming 3.0 release in a few weeks, and some are yet to come.
The biggest problem is that many of the really cool optimizations -- the ones that make a big difference and aren't CPU-specific -- have been patented by IBM and other major players.
You cannot apply a technological solution to a sociological problem. (Edwards' Law)
I can give you over a thousand examples, with the help of our f[r]iendly patent office (my tax dollars at work). Just go to http://www.uspto.gov/, look under the green Patent Grants area, follow the Advanced Search link, and search on "compiler and optimization". Doing this today, I got 1,261 patents, but some of them don't apply here.
Er, that is to say, I got 1,261 search results each representing a patent. I don't have 1,261 patents myself. :-)
You cannot apply a technological solution to a sociological problem. (Edwards' Law)
That's got to be wrong. My 750 PIII runs Seti@Home at a SLOWER speed than my G3 (note the 3) 400.
Yes, my PIII is hobbled by Win 2K.
Seriously, this is the same as I've seen for every Pentium I've installed SETI on...a PII 300, PIII ~600, and a PIII 750. The PIII 750 is the first one to actually get in the ballpark of a G3 400. Now, the G4 500 was still blowing them out of the water... 6 hours or so.
Count yourself lucky you don't have to use any other UNIX vendors compilers. Sun's compiler is about another factor of 2 faster than HPs aCC or SGIs compiler (obviously running on different, though nominally comprable hardware).
I made the experience that a PowerPC at 400Mhz can easily outrun a P-III-550 at compiling an application with GNU C++. I think it was almost twice as fast actually. The problem I currently see with PowerPCs is their lack of proper XFree X-Servers. Some 2D-accelerated X-Server would pretty much make me happy but all I got working was that Framebuffer X-Server which makes that box unusuable for all that requires the desktop. I'd really love to be able to run Linux properly on one of those systems, I'd gladly switch my home pc from x86 to PPC.
Lies,
Damn Lies, and
Benchmarks
I read through these comments and get the impression that slowly a conclusion is being reached: PPC-type hardware is good for some things, x86 hardware is good for others. Nothing really new there, is there? For running Linux, it seems from this little (and far from in-depth) benchmarking session that PCs are a bit better, especially given costs. You can probably get a 1.2GHz Athlon box for the cost of a 533MHz G4, and it'll be better for Linux, so if you run Linux, why not?
MacOS X, stuff like Maya, Final Cut Pro, etc. etc. quite obviously runs better on PPCs, barring some strange circumstances. I imagine that with enough "brute force" (RAM, dual processors, etc.) one could get a PC to run this stuff faster than a Mac.. but what's the point? You might as well just keep it simple and buy a Mac that'll run it pretty well outta the box.
I agree though, that cost is an important consideration. With 760MP around the corner, if it ever does surface in quantities making it available, dual Athlons might give dual G4s a bit of a whippin', especially considering AMDs prices as of late. In general I find you can buy a PC with a much faster proc, more RAM, etc. for the same cost as a Mac from the Apple store.
Still, even a 1.2GHz Athlon would probably choke on OS X, and the G4 will at most hiccup...
"Caffeine is not an option. Caffeine is a way of life."
check out dell's site. they charge almost as much as apple.
upgrading from 128 to 256 is $130
from 128 to 512 is $370
(this is for pc133)
it's one way for the companies to make up for a low profit margin.
> When will this obsession with speed end ?
When we developers don't have to wait a full hour for a full rebuild when we're compiling today's games. (If you think that's bad, the Windows 2k guys had to wait *12* hours for a full build.)
Hardware is SLOW SLOW SLOW.
Since the others who corrected this statement did so in AC mode, I really thought that I should correct this in a way that will get read.
Most instructions in either a RISC or CISC chip take more than 1 clock cycle to complete. This is why pipelining works. Tyipcally the RISC processor (like MIPS or PPC) will take more machine instructions to accomplish a given task. The advantage here is that the chip can be better optimised for a smaller set of possible instructions and is likely to finish each instruction in a shorter amount of time. The drawback is that the larger number of instructions required to complete a given task is likely to take up a larger amount of space in RAM and on disk.
When dealing with a CISC processor like that x86 family, you have a larger number of instructions to choose from and can thus accomplish a given task in a fewer number of instructions. Because the length of each instruction can vary, the most used instructions are usually the shortest and can therefore save RAM and disk space. The drawback is that the processor is less optimised for each particular task that it knows how to do and may therefore require more clock cycles to execute each instruction.
Please note that my comments above are generalizations. There are a large number of tradeoffs involved in creating a processor. Consideration must be given to cost, power consumption, supported instructions, compatability, clock rate, pipeline length, and so forth. This is not an easy task to achieve because there are so many different variables. This also means that when you are evaluating the finished product you must consider more than just the number of instructions (CISC vs. RISC) or the clock rate or the number of instructions executed per second (MIPS).
Furthermore, the processor in only one part of the whole computing solution. In addition to the processor, you need the supporting motherboard and RAM systems which introduce a whole slew of bottlenecks to the system. You also need a good compiler which optimises appropriately for your priorities (program size vs. execution speed, etc.). Also consider that a system may be well optimised for a problem that you are not trying to solve. Some chips are better at solving floating point problems while others are better at solving mostly interger problems.
________________________
I don't want free as in beer. I just want free beer.
Should be able to charge faster due to 2 AC adapters :)
Seriously, I don't know about the harddrives, but if its twice as dense rather than a second platter it should increase the transfer speed somewhat as more bits will pass under the read head in a given timeframe.
Doesn't help seeks at all though.
Rod Taylor
Maybe not.....as he stated at the top, he wasnt interested in showing how fast a G4 was if you tweaked and optimized it to hell.....he just wanted to see how it compared doing the things he actually did......
I don't believe he was trying to prove any point, or show either platform in any particular light.....he was just trying to see what results he would get with what he has...
Advanced users are users too!
i have a question: why do we care so much about this? why does speed even matter? i think this is a symptom of long-time computer geeks still obsessed with who has the "fastest" computer. i realized a couple of years ago that computers these days are really fucking fast. i do a lot of photoshop work, compiling and other processor-intensive tasks but i've very quickly realized that:
as a result i stopped looking at "megahertz" a long time ago. what cares to me is, can i get my work done, and is it enjoyable? as a result i use a Windows laptop at work, a Mac desktop at home and a Linux box for my server. i "know" what speed they are but that was of absolutely no consequence to me. instead i thought to myself "what is the most productive and enjoyable way of getting done what i need to do?" then i bought the computer and installed the operating system that fit.
i suggest you take a break from the benchmarks and come back to reality for a moment. if a Macintosh doesn't do what you want it to do at the right price point then fine, don't buy one, but that doesn't mean they're any less useful just because they're "slower." even Macs have their place in the world, and the slashdot geeks would significantly lower their stress levels if they stopped thinking so much about it. get the computer that works for you at the price you're comfortable with, end of story. desktop computers are more than fast enough to do just about anything you'd want to do anyhow.
- j
My first instinct says that this is just not true. If someone can explain why, I'd be very interested in knowing.
B1ood
Note to self: pasty-skinned programmers ought not stand in the Mojave desert for multiple hours. -- John Carmack
My Athlon really is faster than a PII or the equivalent PPC processor. Now if only my motherboard didn't suck . . .
http://www.dhpc.adelaide.edu.au/reports/065/abs-06 5.html
DHPC Technical Report DHPC-065
Cluster Computing with iMacs and Power Macintoshes
D.A. Grove, P.D. Coddington, K.A. Hawick and F.A. Vaughan
Archived: 8 March 1999
Published in Proc. of Parallel and Distributed Computing Systems (PDCS'99), Fort Lauderdale, August 1999.
Abstract
We investigate the use of a cluster of networked Apple iMac or Power Macintosh personal computers as a compute server for sequential and parallel programs, and describe our experiences with a cluster of iMacs used primarily for an undergraduate teaching laboratory. We explore the support for cluster computing offered by the various operating systems available for the
Macintosh: MacOS 8; the upcoming MacOS X which offers Unix functionality; and the recently available LinuxPPC for the Macintosh. We present sequential and parallel benchmark performance results for an iMac cluster, a Beowulf-style Intel Pentium PC cluster, and a network of DEC Alpha workstations. We compare the performance and price/performance of these systems, and discuss some of the less quantifiable issues such as software installation, ease of use, and system maintenance.
PDF version and PostScript version available...
To be sure, but yet Intel has been the one (along with AMD) that has been pushing the "MegaHurts Wars" full speed.
Only on desktop. Itanium runs at a mere 800 MHz (and yet has the fastest FP performance in the world).
Itanium is actually a pretty good argument against the sheer cluelessness of people who insist on doing performance-per-clock comparisons. It is manufactured in P858, the same process as the Pentium 4, yet runs at less than half the clock speed. It uses much more power, has much lower integer performance, has higher FP performance, has lower memory performance, and is more scalable for multiprocessing. What sort of generalization are you going to derive from THAT?
For what may be hoped to be introduced in the near future, a 1 GHz PPC chip should outperform a 1.1 Intel and have some other potential redeeming characteristics.
This is meaningless. For starters, Intel has a 1.7 GHz processor out now, so it doesn't make sense to compare it to a 1.1 GHz processor (or various vapor PPC products). Second, by the time a 1 GHz PPC is finally shipping, Intel will have something a lot faster. Third, you are assuming that performance scales linearly with clock speed, which is a horrible assumption (clue: the memory performance is not affected by clock speed changes).
Having the long pipeline so you can scale past 2 GHz is not all that its cracked up to be in the real-world. Mis-predicts cause too many pipeline flushes with other bad potential side effects. For some stuff its fine, for many things it ain't. The PPC runs with a very short pipe.
It doesn't matter. If you have two computers, one with double the pipeline length, and the other with half the cycle time, the misprediction penalty will be identical. One will have to recover double the number of stages, but since each stage takes half as long, it's the same.
Major Premise: Gcc does less useful optimizations on the sparc platform (than on the x86 platform).
Fact: Sparc is non-X86.
Fact: PPC is non-X86.
Minor Premise: Not well working on one non-X86 platform implies also not working well on another.
Therefore: Gcc does less useful optimizations on the PPC platform.
Is that sufficiently twisted and bizarre?
(I was going to write in prolog format, but it's been a bit too long for me to be sure, and I'm not about to waste any more time on this when I have much funner things to wasted the last few hours of this weekend on.)
As I recall the PPC and x86 architecture have a different endianness. Wouldn't there be an unfair advantage for the processor that was cross-compiling for an architecture with the same endianness? Endianness-conversion costs time, and when you have a compiler writing large amounts of files with a different endianness than it's own, that might take large amounts of time!
Of course the tester wasn't trying to isolate CPU - he wanted to test overall hardware performance. And I wonder if there's a fifth variable - how well the compiler was compiled for the host platform. Given a platform X, and a compiler xcc which really optimizes for X, if you compile gcc with xcc it will perform better on the test compilation than if you compile it with gcc. (I'm assuming the test compilation is always with gcc).
Given that your SPARC compilation was slower than your Intel compilation, there are at least three possible explanations:
- SPARC architecture is a harder target to compile for, regardless of the compilation platform.
- GCC is less efficient on SPARC
- The Sun/Solaris machine is less efficient than the Intel/Linux machine.
Use of a constant target architecture would have eliminated #1.One needs to consider the intel chip and the fan together, as a system. Neither one, alone, is going to get much processing done for very long.
Of course, it is possible to move away from the atx form factor motherboard for x86 chips, towards a design which is more ammenable to keeping the die cool without a fan, but, until we do: yes, the chip (system) generates too much noise.
--
When your computer is louder than your refrigerator, you know you are in trouble.
I wind up doing my own internal PPC vs X86 benchmarks almost every year.
I'll set up whatever current game I am working on to run with the graphics stubbed out so it is strictly a CPU load. We just did this recently while putting the DOOM demo together for MacWorld Tokyo.
I'll port long-run time off line utilities.
I'll sometimes write some synthetic benchmarks.
Now understand that I LIKE Apple hardware from a systems standpoint (every time I have to open up a stupid PC case, I think about the Apple G3/G4 cases) , and I generally support Apple, but every test I have ever done has had x86 hardware outperforming PPC hardware.
Not necessarily by huge margins, but pretty conclusively.
Yes, I have used the Mr. C compiler and tried all the optimization options.
Altivec is nice and easy to program for, but in most cases it is going to be held up because the memory subsystems on PPC systems aren't as good as on the PC.
Some operations in Premier or Photoshop are definitely a lot faster on macs, and I would be very curious to see the respective implementations on PPC and X86. They damn sure won't just be the same C code compiled on both platforms, and it may just be a case of lots of hand optimized code competing against poorer implementations. I would like to see a Michael Abrash or Terje Mathison take the x86 SSE implementation head to head with the AltiVec implementation. That would make a great magazine article.
I'll be right there trumpeting it when I get a Mac that runs my tests faster than any x86 hardware, but it hasn't happened yet. This is about measurements, not tribal identity, but some people always wind up being deeply offended by it...
John Carmack
Ah, I see.. I admit to have mis-read that pert of the article, thinking that you were testing the time taken to build a cross compiler, not to build an application using that compiler.
That is, indeed, a far better way of testing the systems. Please take my appology on this point.
I'm sure you'll agree, however, that even given that, it's not necessarily true to say that this is a definitive statement on the relative potential speeds of the two types of machine, as it depends upon the efficiency of the compilers involved. It takes a huge amount of work to get the best from a RISC processor and GCC isn't all that good in that area (relative to the manufacturer's own optimised compilers).
Agrajag: "Oh no, not again!"
Agrajag: "Oh no, not again!"
I'm sure this person THINKS he's testing the same thing on each machine. Unfortunately, he's not.
Firstly, the action of compiling on different architectures is very different, even without considering optimisation strategies. To compile code into the CISC code of the x86 architecture is very different from that of a RISC chip such as the PowerPC. For a start, instruction ordering etc. for a RISC chip, even for not really optimised code can take far more processing time. Then, if you add optimisation, which in a RISC architecture is a FAR more complex task.
All this means is that compiling on a RISC architecture is bound to be a great deal slower.
Basically, this "benchmark" is measuring not only the intrinsic speed differences of the architectures and chips but also degree of optimisation the native compilers used can cope with and the extra processing power needed to generate the code during the comile stages.
Basically, using compilation as a benchmark is not at all useful, other than to test the difference in speed of two similar peices of equipment using identical software (ie. compilers & OS) or the difference between two versions of the same OS or two versions of the same compiler.. Basically, you can only change one variable to deliver a meaningful benchmark if using the method chosen in this "study."
The only way to get a half-way meaningful benchmark for the two systems used here would be to write a program which did lots on disk I/O and integer manipulation, worrying about whether it's being biased for or against certain types of architecture or use (eg. loops sitting in cache etc.). This would give you an idea of the real-world speed differences between the two systems. However, you won't be just measuring the intrinsic speed of the machine but also the different ways the kernels have to do things on the two architectures, the degree of optimisation the compilers building the kernel and the program could generate and the speed of the hard disk built into the machine.
As you can see, it's a tricky thing comparing two types of machine.
Agrajag: "Oh no, not again!"
Doing a world build of anything using gcc is a benchmark of your hard drive, NOT your CPU.
You mentioned in the article that the test isn't disk intensive, but every stage of GCC feeds the next via a file. Tons of RAM or not, you're still benchmarking your disks.
And BTW- your hunch that gcc produces shitty PPC code is correct. Run the bytemark tests if you want a more interesting benchmark of CPU performance. Make sure to test using different compilers on the same CPUs to show how much a compiler can affect efficient CPU utilization in the software it's building.
> Actually, the line was:
:p
;)
> "This thing can flash-fry a whole cow in 40 seconds."
> "Aw... But I want it *NOW*!"
Actually, it goes like this (emphasis mine) :
Moe: "Oh, boy! The deep fryer's here. Heh heh, I got it used from the navy. You can flash-fry a buffalo in forty seconds."
Homer: "Forty seconds? But I want it now!"
Buffalo, not cow.
> And, yes, I'm a nitpicker
I guess that makes me a meta-nitpicker...
--
PaxTech
All movements for social change begin as missions, evolve into businesses, and end up as rackets.
----
Althea verified to run quickly on Mac and PC hardware.
Yeah, check out this at the Apple Store. It's the pricing breakdown of the powerbook G4.
The "Faster" model:
500MHz PowerPC G4
1MB L2 cache
256MB SDRAM memory
20GB Ultra ATA drive
DVD-ROM w/DVD-Video
ATI Rage Mobility 128
10/100BASE-T Ethernet
56K internal modem
Two USB ports
One FireWire port
The "Fastet" model:
500MHz PowerPC G4
1MB L2 cache
256MB SDRAM memory
30GB Ultra ATA drive
DVD-ROM w/DVD-Video
ATI Rage Mobility 128
10/100BASE-T Ethernet
56K internal modem
Two USB ports
One FireWire port
Extra AC adapter
Extra battery
So here's my question: Why is the "Fastest" G4 any faster than the "Faster" G4?
Because the hard drive is 10 gigs Larger?! (they're all 4200RPM).
Or is is the extra ac adaptor that somehow makes it "fastest"?
Friggin Apple. Buncha liars.
However there is one problem however. My website is unable to be reached! Why did Slashbot rape the webserver of Mr Ezgo? Is most rude. Perhaps because you are the hackers you have are hacked the webserver of Mr Ezgo?!! :-)))) Now now childs that is not nice! ;-)!!!!@!
So thank you for the link!! To CITY DESK COMPUTING SUPERSMALL!!! Teh WAY OF THE FUTURE!!!
Mr Ezgo
City Desk GmbH
--
--
I like to watch.
I've found some used Mac 5200's for $75 and I am going to get one to run Debian. It will be interesting to run these benchmarks on it to see the $/bogohurts ratio on that. Then do a comparison with the Compaq 5120+OpenLinux2.3 for the fun of it.
Hmmmm. The battle of the dumpster computers! I'll think of benchmarks with Perl & MySQL to give a realistic evaluation of how I use computers.
But I do physically-based animation.
The Dual G4 533 is only $300 more than a single G4 533. What I don't get is that it appears these test were run on with only one of two CPU's working, and little if any optimization for the G4's AltiVec unit. ? That isn't exactly a fair test of the state of the Mac processors, is it?
Interesting, the CISC is 3X the speed of the RISC, and the score for that machine is only 2.5X higher... Has anyone ever actually thought of comparing same clock cycles to each other, to actually try to compare Apples TM to apples?
One Token Ring to Rule them All, One Search Engine to Find Them, One WAN to bring them in, and TCP/IP Bind them...
Does anyone know what would happen if source code that violated those patents was produced in Europe (where there are no software patents), and then the object code was distributed to the Land of the Free (over the 'net)?
Also, has anyone tried contacting IBM etc. about using those patents? I know nothing here, but it might be that some of those patents are purely defensive--i.e. just to prevent others patenting them. So IBM might let GNU use them.
I doubt this guy's benchmarks are very accurate. Here's one of the Pentium II vs. PowerPC from 1997 (kind of old, but still very relevant) which proves to be a much closer battle.
If you celebrate Xmas, befriend me (538
So would this mean that EU hackers (where there are no software patents) could distribute patches to give GCC these optimisations (like the crypto people do?)
Athletic Scholarships to universities make as much sense as academic scholarships to sports teams.
are you saying a four year old (when CPU speed quadruples every 3 years, the Chips tested had about 200mhz clockrates) benchmark testing obsolete, last generation CPUs, is some how relevant to today's market?
"and dear god does this website suck now." -- CmdrTaco
Memory prices on apple.com: (pc100 SDRAM)
64 megs: $100
128megs: $200
256megs: $400
Prices for PC133 memory on pricewatch.com
64 megs: $9
128megs:: $15 256megs: $26
Unless you think paying $400 for something that you can get for $26 is somehow fair, I think it's you who needs to stop 'beeing a troll'. Looser.
"and dear god does this website suck now." -- CmdrTaco
I recently purchased a G4 after considereing that for half the cost I could have a computer that could do the same, possibly even cheaper. One thing that people fail to see is that certain industries such as graphic design are completely sided with Macs, if you want to really get in, you have to know macs inside out, so for those who must take things such as that into consideration, a cost increase might be worth it, especially if it runs those programs better. But maybe thats just me
I'm sure that Apple is really happy that this benchmark that shows that the G3 and G4 aren't really as fast as advertised is being hosted on a Mac.com site.
http://homepage.mac.com/nopea1/benchmark/
I guess this would only be useful for holy war purposes, though. It would probably be both subjective and inapplicable to most real world apps, but why do anyone of us benchmark quake/photoshop/browsing speed? anyways?
For the coolness factor, I guess. . . .
Always clueless, WhiteWolf
WhiteWolf666 an exBush supporter. All you new-school,compassionate,save the children Republicans can rot in hell
After actually reading the article, I noticed the bogohertz(or whatever) scale. But still. . . . As much as I am a huge PC advocate, I don't think these tests were particularly good general performance indicators.
WhiteWolf666 an exBush supporter. All you new-school,compassionate,save the children Republicans can rot in hell
I know--that was my next comment
WhiteWolf666 an exBush supporter. All you new-school,compassionate,save the children Republicans can rot in hell
Its opptimized for the x86. Sure OSX ships with it, but apple did do heavy optimization for it. Those chages may not have been rolled back in to the tree yet, so I would not trust these bench marks. Your running an os that is optimized for intel, and a compiler that is optimized for intel, and comparing them to a ported copy.
Most people know that gcc is slower on suns then the sun compiler. That is all about optimization. So why wouldn't it be the same for ppc?
Try doing this benchmark on darwin, and im sure the macs will do better against the intel boxes running darwin, then running linux. Im not saying that it will be faster, im just saying your comparing apples and oranges.
The P3 that the Seti numbers refer to is running Win 2000. It's also an average, so perhaps your P3 is hobbled by something running in the background.
The comparable data are:
So you voted for the guy that supported raising fuel prices to $5 and other energy prices to "protect the environment?"
Oh, and you probably voted for Davis and the Legislature which could've solved the problem before it started, but didn't cause it was an election year.....
Way to blame the problem on the wrong people.
Remember, energy was on it's way up LONG before January. Don't look to the Federal Government to fix a state power problem, look at the state that's supposed to increase demand.
Never overestimate the intelligence of the individual, and never underestimate the stupidity of the masses.
I had a Toshiba Portege laptop at work. It needed more RAM. We cracked it open and popped a laptop-format memory module in.
When we turned it on, the display said, "Please remove the non-compatible memory module and replace it with an authentic Toshiba part" or something like that.
It was the ultimate insult.... Tosh forced you to buy their super expensive memory. At least when Apple was doing that crap they'd make weird arbitrary changes to the SHAPE of the memory board, so you wouldn't feel teased!
Since someone has to point it out, I guess I'll have a go and say that those tests were/would be apples-and-apples comparisons... IF GCC and the Linux kernel are equally optimized for Apple hardware as they are for Intel! He even mentions this in his benchmarks.
Trust me, if you've ever played with the PPC ports... they're definately behind the times, and I can only imagine how far they are from taking actual advantage of the hardware (a lot of stuff is barely supported, let alone clean and optimized); Intel is Linux's home platform so OF COURSE it performs better there.
Another, more interesting, benchmark would be to do some tests from a *BSD on intel and Darwin-only on Apple hardware. Those should be a bit more "at home" and put these tests on more level ground.
Of course, gcc (egcs) is optimized for x86, and not powerpc.
Let alone Linux itself... the PPC ports of all this stuff are a bit behind the times. I'd be interested to see someone write/port some tests between *BSD on Intel and Darwin on Mac; that would be a more level comparison. Too bad I don't have much free time at the moment...
--
< )
( \
X
Friends don't let friends use multiple inheritance.
in order for RISC to make up for the lack of instructions. it has to execute more, becuase the processor know's less. as for x86, there are ton's of pre set instruction's, enabling it to process more important information. and thus performing faster.
don't forget about bus speed/clock speed ratio, and other hardware like hard disks, memory bandwidth ....
Welcome to the realy world people, your trying to compare apples to monitors. your also trying to compare performance of different chip based on the output of the software and NOT what is actually happening inside.
:) P4@2+Ghz, Athlon4, Hammer, itanium. :)
Photoshop6 on PC vs Mac are two completely different programs. open them up in a hex editor and you be hard pressed to find ANYTHING similar about them. sure they come from the same(or similar) source code, but then they go throught the copiler.
also, dont compare clock speed, clock speed does not matter*, its one small part of what makes a CPU fast. Long ago i played with digital Alpha systems and learned that an Alpha 500mhz kicked the crap out of a 500mhz p2/k62.
The PowerPC is VERY fast at some things, and average or slow at others, one thing i have noticed though is that the x86 arcitecture is more well rounded. It has good(not excelent) integer and floating point performance, and it doesnt really have a weak point as far as performance.
x86 has huge advantages as far as support, chipsets, standards implementation, OS support, availability, etc. i guess you could argue that some of these could be weaknesses like, too many chipsets and standards to be stable or easily supported. apples proprietary hardware has made it much easier for them to support.
in my opinion, this arguement is getting fairly close to useless, in just a few months their will be brand new chips and architectures to argue over about performance
A chevette with a 300 hp engine is faster than a corvette with the same. This is true of any car with more power/weight than another car. But I digress.
On the issue of mac vs pc architecture, what does compiler efficiency serve for but an excuse? We see that macs fall behind on SETI@home, on linux compiles and on everything else shared between the two platforms except Photoshop filters. O course, this could be easily explained by the mhz differences, which brings up another point.
Its interesting to note that regardless of clock speed, the motorola, intel and AMD chips all do roughly the same amount of work per clock cycle (with progressively smaller returns per increase in clock -due to latency, cache misses, etc). This seems to suggest to me that given the relatively static architecture of these three brands, that they are probably all roughly equivalent, assuming identical clock speeds. Unfortunately, its been about 4 years since intel or amd was producing 400 mhz pieces.
Even more unfortunately, everyone I know who is an avid mac user in the past year has switched over to athlon or pentium based systems due to cost and features. The only people that are carrying on the fight are trolls and morons or both (my unproven personal observation).
Sort of like comparing Photoshop on Mac to Photoshop on Windows...This is just giving Apple a taste of its own medicine.
Cedric Balthazar Rotherwood
Sun Certified Programmer for the Java Platform +
System Admin. for Solaris
www.blacklablinux.com / www.yellowdoglinux.com Go there. Be happy.
We apologise for the fault in this post. Those responsible have been sacked. -- Signed RICHARD M. NIXON
http://www.blitzbasic.com/
http://www.blitzbasic.com/
Graphics3D 640, 480
Agreed! I'd really like to see how things would have come out in more SMP friendly conditions. The dual G4 seemed to be an effort by Apple to overcome the MHz war, and was priced accordingly. OTOH, the classic OS has had only limited support for SMP, so the uniprocessor numbers aren't totally misleading. Apple released a machine a few years ago - I think it was 604 based - that supported multiple processors, but because it was the only machine available, developers never got into it.
Now, they do list what content is deemed unacceptable, but IMHO, this content does not fit any of those descriptions.
So I don't think Apple really cares.
"Empathise with stupidity, and you're halfway to thinking like an idiot." - Iain M. Banks
Users only care about one thing, value for money... and its easy to hold a showdown to show which has the best, x86 is best value and will be for the foreseeable future (for most people's purposes AMD's x86 processors BTW, not Intel's).
I have to say I find single instruction stream processors (including Gx all the x86's, Alpha etc etc) quite distastefull because they waste huge amounts of transistors for small increases in speed... Id rather have more explicit on chip parallelism. But as far as these go I have to say Intel's P4 is at the moment by far the most innovative of the bunch. Lets for a moment consider what they did, not only did they produce a processor with record breaking clockrates (yes the double clockrate internally still is advantageous, the huge pipeline can be a problem... but thats a compromise which has been decided in longer pipeline's favour ever since the first superscalars), but they also made the first mass market trace processor and will soon make it the first mass market MT processor with Jackson tech. Its lousy value for money of course, but hell as long as we are talking about technical superiority the P beats the G IMO. Lets see that Power4 first, and then well talk again.
Everybody knows Linux is so fast it can execute an endless loop in 5 seconds flat.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
http://www.stepwise.com/Articles/Technical/StringL ength.html
What these benchmarks really show is that the gcc compilers for 80x86 are vastly superior to those for PowerPC. It also shows how the majority of Linux (kernel and user space apps) is un-optimized for PowerPC. It doesn't mean that PowerPC is slow - just that it's full potential isn't being taken advantage of. A more usefull comparison would be to compare binaries compiled with CodeWarrior. I'm sure a 1.2GHz Athlon would still kill any PowerPC but not to the same extent these benchmarks show.
Willy
Politics at the end of the day is about economics.
Can anyone imagine what the state of play would have been, if Apple had of achieved the market penetration that the PC achieved, BTW PC means peoples computer. I've never owned an Apple Mac but it would not surprise me to find a sticker saying, 'opening this up' voids the warranty. I'm probably wrong about this, but the mentality that drives a lot of buisness, is the idea of taking over. Of building a commercial dictatorship.
I have been trying on and off for about eighteen months with different distro's, to get on the internet with Linux, finally two days ago I achieved liftoff. I'm fifty two years old, the older you get, the more difficult it gets to learn new stuff.
The relative speeds between the Apple Mac and the PC are not nearly as important, as the fact that Linux is the peoples operating system, written by the people, for the people. The whole idea that people can capitalize on ideas and not sweat, is deeply flawed.
The culture of science works because of the sharing of ideas, not so with Microsoft, you are not supposed to know how it works.
I used to code lots years ago, 6502 and Z80, as a hobby more than anything else, then because of circumstances I gave up my hobby for around ten years, then took it up again, starting with a 286, yeah when I started to learn about extended memory, expanded memory and protected mode I thought what a mess and gave up. Assembler is hard enough as it is, without added layers of complexity, then I learned about Linux.
Linux is what Dos should have been.
Yeah an Apple Mac is probably more stable than a PC running Microsoft (Not all Microsofts fault(there are some crap coders out there)) But I ask you, one company supplying the hardware and the software, might be ok for folks that use computers in a purely utensil fashion, but for the losers that want to understand how the hell a box of switches, could exhibit quite human characteristics, well I rest my case.
Peter.
It's called an elephant's trunk whereas it is in fact, an elephant's nose, a nose by any other name would smell as sweet
While I've been typing this, I bet at least five people have already pointed out the same thing... ;-)
Unsettling MOTD at my ISP.
comparing apples to oranges
http://www.barefeats.com/pentium.html.
I used to have some hope that the AIM alliance could put out some better chips, but quite frankly, after seeing the story, I feel that they are too far behind at this point to compete. Mac users deserve something better. I'm not sure anymore if OS X makes up for this much difference in CPU power.
I think that I am totally ready to have Apple move over to x86. How much farther ahead does Intel have to get before something is changed?
Actually, I'm kinda depressed about the whole thing. If anyone has anything that would cheer me up, I'd sure like to hear it...
I honestly don't care about whose processor or other piece of hardware is faster. Without software, that hardware isn't going anywhere. I don't buy computers based on what _may_ happen or theoretical performance gains that can be realized in the future, 'cause those don't mean a thing.
I buy state-of-the-art computers to run jobs that need to be done _now_. Therefore, performance of the system must include the performance of current compilers, even if they don't play nice with AltiVec, 3dNow Pro, Jazelle, or MIPS-3d.
I do fall for marketing propaganda for my personal computers, but emacs and LaTeX is all I really use.
You can find an old PowerMac 7200 for next to nothing these days. Load MKLinux on it and you've got a ready-configured dual-boot multimedia RISC Linux box for a bunch cheaper than any of the ones in this guy's test suite.
For Apple to do such a thing their marketing/sales department would have to show that spending money there would increase hardware sales. That or the press alone is worth the cost. We're not talking millions of dollars here but it's not small change either.
Got an e-mail from a friend today who has been playing around with optimization levels on the C compiler on MacOS X. He was quite impressed: apparently that compiler can generate very tight code if you ask it nicely.
Apple's C compiler is based on gcc 2.95.2. NeXT forked gcc years ago and made all kinds of changes which were never put into the Free version. Now Apple has a guy (ex Cygnus) working on the merge.
Sources: personal notes from the MacWorld Expo 2001 SF Stepwise BOF and last month's WWDC.
What Would the Fab Five Do?
You are right....
I have never tried Linux on the sparc64; at least not in the development sense, jut as a user. The thing is that these benchmarks were comparing PPC vs. x86. That was the basis of my comparison, and although my work is mostly related to the Power3 not the G[3-4], they are sufficiently similar to draw my conclusions. I've use Linux on the alpha, ppc, s390 and x86 and have a total of 6 years of experience with Linux. In my opinion, x86 and alphas are one of the more stable platforms supported under Linux, but I admit that I haven't tried it out on mips or sparc. Distribution wise, x86 is the most widely supported of all platforms. Personally, I think I've never did anything interesting with Linux\Sparc64 because at the time Red Hat was the only thing available. I started my trip into Linux with Slackware and have been a loyal fan ever since, even thought I've tried other countless distributions in the pass. Now that slackware is available on sparc, I might try it out if I could only find a machine to start it on. But for the every day normal home user, there are basically two options x86 or Apple's PPC machines, and I think that these are the basis of these benchmark results. For home usage, I do prefer the x86 because they are dirt-cheap; not exactly the best reason or the best choice, but I do think you get more for the money.
Another important thing to consider is that Linux is mostly optimize for the x86. The Power PC port works, but there is a lot of improvement that needs to be done to before comparing it with the x86 code. MM is a mayor area of improvement in the PPC code and possible the reason why the benchmark spend most of the time on kernel space.
It is also important to consider that this was not a very good comparison and I don't consider the benchmarks to be precise.
Any my Duron 650 (now like a 30$ cpu) does 846.50 mplops according to your little benchmark :).
Er mflops. Mplops is funny though :).
Ironically, Macs can use that memory you found on PriceWatch.
-Henry
"Useless organic meatbag" -HK-47
Having the long pipeline so you can scale past 2 GHz is not all that its cracked up to be in the real-world. Mis-predicts cause too many pipeline flushes with other bad potential side effects. For some stuff its fine, for many things it ain't. The PPC runs with a very short pipe.
Sex is heriditary, if your parents didn't have it chances are good you won't either.
I don't thnk that you will ever get an agument from anyone over how very, very, very badly Motorola has messed up with being unable to deliver faster versions of the PowerPC. However, you should look at a couple of issues with these benchmarks: 1) 450/733 = .61 and 533/733 = .72. Thus any scores over .61 and .72 respectively, indicate that the PowerPC is doing more per clock cycle than then PIII. If Motorola can ever get their act together (and that is not a certain), normal code on the PowerPC will run every bit as fast and faster than the x86 processor. Combined with the fact that the PowerPC has a nice quiet and fairly energy efficient air-cooled chip you might have some nice machines. Unfortunately, all benchmarks can have some rather un-intentional bias. My 1.2 Athlon would do 105 Scimarks in Windoze 98, 113 Sciemanrks in Wine under Redhat, and 119 under Windoze 2000 for Tim Wilkin's Science Mark benchmark.
Same machine, same memory, same disks, etc. the only difference was the os. Even given the same os, the tweaks that it goes through are also a function of the author's machine. Please pass the salt, I need a grain.
Sex is heriditary, if your parents didn't have it chances are good you won't either.
No it's not. The standard Apple warrenty does not apply to memory installation, though it does to PCI cards. The extended warrenty pays to have an apple service center install the memory+anything else you want put in.
My Karma is so good, I'm the Dalai Lama...or something.
Sorry, maybe that wasn't clear enough for you. You may do whatever you want ram-wise, and apple encourages+tells you how. If you get the applecare warranty, they will pay to have RAM and PCI cards (from any source) installed for you. All I had to do when I found a VAIO somebody had tossed in the trash was pull out its PC100, open the door of my G4, push the ram into the slots, and close the door. no warrenty broken, nothing un-kosher about. The apple warrenty, in my opinion, is a really friendly one.
My Karma is so good, I'm the Dalai Lama...or something.
Yes, I agree
--jeff
ipv6 is my vpn
For what I do with massive digital audio processing, single precision floating point is exactly perfect. 24 bit A/D-D/A converter maps nicely to a 32 bit float.
--jeff
ipv6 is my vpn
I wrote that benchmark in order to qualify using the G4 in an embedded system specifically made for handling massive amounts (>400) of real time audio channels at 48 KHz.
The benchmark shows that the number of usable single precision megaflops PER WATT OF POWER DISSIPATION and PER DOLLAR for the CPU is much much higher on the G4.
The algorithm in the benchmark is the EXACT algorithm that I used in various DSP's in real products. It is not a synthetic test. The G4 beats the DSPs and all other CPU's that may have been an option for the embedded system. That is what is important to me.
--jeff
ipv6 is my vpn
Hey, it is GPL'd, go for it! Add it yourself! --jeff
ipv6 is my vpn
Well, it is meaningless to compare altivec g4 versus non-altivec PIII. But it is not meaningless to compare non-altivec-non-mmx g4 versus non-altivec-non-mmx PIII.
It would be meaningful to compare altivec g4 versus MMX PIII. I would be interested to see the results. Anyone know MMX?
--jeff
ipv6 is my vpn
Feel free to email me! I'm interested.
ipv6 is my vpn
http://www.jdkoftinoff.com/eqtest.tar.gz
(gpl'd with source)
450 mhz g4:
1.7 gigaflops with altivec
410 megaflops without altivec
500 mhz pentium:
220 megaflops
--jeff
ipv6 is my vpn
You just sit and look at synthetic benchmark results all day? Seriously though what should matter to you is how well your system performs on the programs you personally use. You can toss around all the numbers you like but ultimately, whatever gets the job done the cheapest and fastest is the best for you to get (factoring in of course your personal prefrence). If you just point at a sheet with lots of large numbers from synthetic tasks, you're kidding yourself if you think you have a true picture of what's going on.
My only argument is that Mac's are overpriced, and that Apple as a company has long strived to make competition impossible. When Apple saw their profits go down after allowing clones, they took back exclusive rights. They are the IP holders of their architecture, and have every right. That just means that people who prefer to buy cheaper, yet nearly equal-power PC's will go elsewhere.
I even went so far as to give credence to the systems being comparable according to a major news source!
Just because you don't agree with something doesn't make it trolling. I'm merely espousing distaste for the managment of Apple. The facts here are true, and as someone else on this post already posted, Apple is never going to say that their architecture is slower than anything else, and hence, their website means nothing in reliable information.
In the sense that I don't have the time or care to read their benchmarks and then learn "valuable" benchmarking, their site means nothing. To the consumer, I can usually doubt most things that come from a manufacturer. I'd rather trust what comes from a independant 3rd party. Besides, this is all besides the point. I'm not buying Apple because it's too damn expensive.
Unsuprisingly, the Macatista wrote that meggy-hurtz don't matter, and besides, Mac's are 3 times more powerful than a Pentium 3 anyway! Pop.Sci. wrote back, saying breaking the Ghz IS a milestone thats important to note, and that tests have shown that yes, in some areas, Mac's are 3 times more powerful than a Pentium 3. However, these same tests show that in some areas, x86 platforms are 3 times more powerful than the Mac.
This argument has long bored me. The arcitechural differences between x86 and PPC have been vast until the last year or so. According to an article at Ars Technica, the Intel Pentium 3 chip is somewhat like the PPC, but the AMD Athlon is even more similar to the RISC found in the Mac. Even if that's the case, have you ever seen the price difference between the two platforms? Plus add in your options (and the price thereof) when buying a Mac. You take what Apple will give you. Apple's prices on memory are so laughable as to be a great stand-up routine.
Now don't get me wrong, I have nothing against the hardware, and the only problem I have with the OS as an intermidiate user is the file organisation system. I just think that Apple's managment sucks, and that I get more bang for my buck going out to a local computer store, in which I can support the mom&pop's of America. A one or two year parts, and three years labor warranty is good enough for me.
Debian 2.2 uses a 2.2.19 kernel (If you apply the security upgrades). I'm willing to bet the linux kernel PPC support has been improved since 2.2.x. I think the benchmarks would have been a little more fare if they had been run on a 2.4.x kernel.
Jay has been quite thorough, and been pretty open about how he has gone about things. His findings are broadly consistient with a study done at Adelaide University in 1998/1999 looking at iMacs as a potential beowulf cluster nodes. http://www.dhpc.adelaide.edu.au/reports/065/abs-06 5.html
Their basic conclusion was that the compiler performance was holding the performance back from being competitive ( gcc on PPC being the issue ).
Last time I checked , Apple's changes to gcc made for MacOS X and Altivec hadn't made it back into the main gcc source tree yet, although it can be downloaded from the Darwin site.
By my experience , PPC is typically about 25-30% faster clock for clock at integer, and 50%-100% at floating point. G4 will do better again if you can vectorise your code. That assumes a bunch of things like having enough memory bandwidth , not thrashing caches etc.
It could be that Intel and AMD have incorporated more of the technologies that allowed the PPC to perform well into their chips so that they now take better advantage of their higher clock speeds.
Matrathon
American ProImage
-t
I'm not going to prerelease the results, but they will be available soon. Poorly done benchmarks are worse than none at all. A benchmark based on compiling code for different processor architectures is an example of a very poor benchmark.
My benchmarks are real world, web application suite numbers. Nothing special, nothing rigged, I have a 20M website that's a combination of static pages, static images, PHP/mySQL dynamic pages, Perl forms-driven pages that write and update flat files and PHP/mySQL pages that update the database.
The benchmark itself is script driven and simulates users on the site. There are 10 different user scripts, and they run 500 times each in 10 different fixed orders - currently as 10 simultaneous users. I'm looking at adding clients to increase the number of user tests, as I've been unable to max out the OSX box with this test suite.
The simulation results are mined from the Apache log file and show the activity that you would expect from this near-real world example. CPU time is not captured, only successful page requests. Total elapsed time is interesting.
The only thing that is "rigged" in any way, is that the pages are all set no-cache, so that all images and pages are delivered each time they are requested. As far as I can tell based on status returns, there is no caching being done.
The website and the client scripts will be available to download from the benchmark page so you can run them yourselves if you wish. If you do run them, running the analysis script against the log file will allow you to upload your results to the benchmark server - should be an interesting set of data for different server configurations.
I will say that the dual 500mhz OSX Server currently out performs the Dual 850mhz Intel Linuz box by a significant margin.
I know that some will still not believe, but that's OK. You can run these tests yourself and post your findings on the benchmark page.
I'll publish the URL soon.
-t
I've nearly completed a web serving benchmark with multiple PPC configurations running OSX Server, and Intel and AMD hardware running Linux.
The results look nothing like the compiling benchmark, and have convinced me to start a web hosting company using OSX Server on Macintosh hardware.
The benchmark utilities will be downloadable and you can run them on your own favorite hardware. Benchmark requires PHP and mySQL database support, but will run on more than just Apache. I'll also set up a site where you can upload your results - configuration and resulting data.
-t
I would have thought that question about Motorola processors had been resolved long time ago. In fact, Motorola had dropped from all list of high-performing processors for quite a while now. Interestingly, the fastest MAC (733 MHz) is quite a bit slower than the slowest PIII [not mentioning Athlons]. Below are SPEC benchmarks taken from the SPEC and Motorola sites [I 'm going with SPEC benchmarks because these are the ONLY benchmarks made and accepted for cross-platform testing. They are developed by SPEC with participation from ALL major chip makers, inclusive Intel and Motorola]. Now, given poor scalability of the Motorola part, one can expect for it to catch up with the 733 MHz Intel processor only at 850 MHz in INT, and at 1200 MHz in FPU. Although there are some reasonable doubts that the current G4 will scale even that far. It looks like the average consumer does not believe Jobs' mantras about his "supercomputers" either, which is perfectly reflected in Mac's declininig share of the market: since glorious Steve returned to Apple, the share of Macs in the total dropped from 6% (end of 1996) to 2.3% (IQ 2001). SPEC Benchmarks from SPEC site Pentium III, 733 MHz INT - 35.7 FPU - 31.0 http://www.spec.org/osg/cpu95/results/res99q4/cpu9 5-19991108-03912.html
http://www.spec.org/osg/cpu95/results/res99q4/cpu9 5-19991108-03910.html
from Motorola site
Motorola G4, 733 MHz
INT - 32.1
FPU - 23.9
http://e-www.motorola.com/collateral/MPC7450FSR0.p df
I am surprised that otherwise smart people bring up Altivec (or SSE or whatever) as if it were general instructions. Come on guys, these are not instructions you can do Mathematica with!! These are special-case very imprecise operations. Intel, Motorola and others decided to use a small part of die to speed up some multimedia computations, because graphics cards were so expensive. But now when graphics cards are vastly more powerful, that argument is gone. G4 or PIII can do billions of these instructions ("gigaflops"). Well, Geforce3 card does 100 of these "gigaflops", so what. Real Gigaflops are measures with programs like Linpak.
Come on guys, give G4 a break. x86 does not compete with Mac. Mac's share now is down to 2%, and it become a truely niche product. x86 is competing against Alpha, Sun and POWER on the high end [actually, successfully, given SPECfp2000 score of 711 for 800MHz Itanium, and SPECint2000 of 586 for 1.7GHz P4 and 539 for 1.33GHz Athlon-> the highest marks in the industry], and against themselves in the mid-range. Nowhere in these comparisons one can find a G4 - for about two years now PowerPC does not enter lists of the highest performing chips. So you can say that Motorola is at war with x86, but x86 is not at war with Motorola. references: Total number of PCs sold in IQ01 - 32.5 mln.: http://www.techweb.com/wire/story/TWB20010420S0006
Number of Macs sold in IQ01 -751,000:
http://www.apple.com/pr/library/2001/apr/18q2resul ts.html
Thus, MAC's share is 2.3%
I am surprised that otherwise smart people bring up Altivec (or SSE or whatever) as if it were general instructions. Come on guys, these are not instructions you can do Mathematica with!! These are special-case very imprecise operations. Intel, Motorola and others decided to use a small part of die to speed up some multimedia computations, because graphics cards were so expensive. But now when graphics cards are vastly more powerful, that argument is gone. G4 or PIII can do 5 or 6 billions of these "operations" ("gigaflops"). Well, Geforce3 card does 100 of these "gigaflops", so what.
I doubt it will speed up Mathematica in any significant way. http://developer.intel.com/design/Xeon/devtools/ and read "Approximate Math (AM) library" entry where Intel says how wonderfully they speed up their computations up to 17 (!) times. You think they are magicians? Damn, these are not precise computations! For G4 it gets even worse -> these computations are not even IEEE-standard.
Or up to as much as $79 + shipping. For current prices on 3rd party Mac memory, go to:
http://www.ramseeker.com/
This site shows different prices from a variety of vendors and is updated daily. The only downside is that they don't show shipping costs, so you have to check a bunch of the sites to find out which is REALLY the cheapest.
This argument about RAM prices is incredibly silly but that isn't surprising since Windows/Intel users who bash the Mac usually don't have a clue about Macintoshes.
--- What?
n/t
--- What?
Actually, Mathematica could use AltiVec for many operations. AltiVec does NOT replace graphics cards -- it's additional instructions comparable to MMX/SSE/SSE2 (sans the double precision). Anywhere you've got some parallelism in the calculations (and don't need double precision), AltiVec is probably going to be useful. Chris
Improving performance from app architecture to microcode.
Huh? What does Intel's "approximate" math library have to do with Mathematica? Sure, Intel got some good speedups by using less precise math. I doubt that Mathematica could use that because of the loss of precision. But that doesn't mean that Mathematica cannot be sped up.
Improving performance from app architecture to microcode.
I read Mr. Carmacks article and thought, humm, here is an interesting tidbit. He called an iMac DV he had form over function. I think this misses the point. The form is the function of the iMac. We'll leave the G4s out of this for now.
What is most important on and iMac:
small - monitors increase in size geometricly, a 17 inch monitior is a lot bigger than a 15. I don't think you can make one with a 17 incher.
quiet - no fan. This may have a large impact on what hardware can be put inside. A g4 may just run to hot.
ease of use - look at the slot loading cd. Tbis is an important feature, it does make it easier to use. The MacOS. Real easy and cool way to get to the RAM and slot for Airport card. USB and Firewire easy to get at and use.
good looking - this is something that you live with. Haveing something pleasent in your enviroment is important.
Fast enough - just here you get to speed. If you want a web server don't use and iMac.
For the pro G4 series you have a better argument if you complain about lack of speed, but not a great one. Here again raw speed is just one factor out of many. How easy the case is to open and access is something that is considered as important and its speed. The difference isn't huge and can change at any time. What should be remembered is that you use the whole computer - not just the CPU - so you have to make the whole computer good.
On something of a side note - I read that if you took a snapshot of all the cpu cycles at some specific time you would find that the thing they are doing most is drawing screen savers. In writting this little thing I spent a fair amount to time just thinking about what I was doing. My 300mz G3 did nothing of any use during that time.