Slashdot Mirror
Apple's G5 Speeds Challenged
An anonymous reader was the first of a seemingly infinite stream of people to submit a URL to an argument that makes the case that the G5 isn't quite what Apple wants you to think of it. The evidence? Apple's own press material. Worth a read.
...and benchmark different too!
Is this truly the only Earth I can live on?
They're giving us a desktop UNIX running on 64-bit hardware, what else can you ask for? sheesh
Sheesh, evil *and* a jerk. -- Jade
translation
i am too stupid to round up.
Are you MORE than your SPINAL COLUMN?
While the methods Apple used may not have been in the best of intentions and possibly missleading, this just underscores the greater difficulties of benchmarking across platforms, specifically processor architectures. The playing field will never really be level using SPEC. The only way to truly determine which machines are "faster" is at the application level, where real work is done.
Apple is always a little sketchy when it comes to speed measurements. I can't count how many questionable run-offs Steve Jobs has demonstrated during his keynotes.
They're always a little suspect. I love Apple as much as anyone, but their talk of the megahertz myth and the amazing clock cycle of the G4/G5 and the biased tests they use are starting to sound a little shrill. Apple needs to admit that their machines aren't as fast as the fastest Intel has to offer. They're much cleaner and much more elegant, though, and that's why they're in the market. That's what they should stress, since it actually attracts customers -- rather than THE NEED FOR SPEED.
Al Gore has definately been able to spread his political spinnery into Apple's culture pretty quickly since joining their board. :-)
OMG, you mean benchmarks are subjective? Marketing execs get a hard on the size of Georgia when they hear the term "benchmark." Let us all hope and pray AMD and Intel don't hear about this, lest we never be able to trust an ad campaign again!
The author of this little essay is a known troll in the Mac community. His previous essay made sure to bash Apple for copying the original windows GUI for the Mac(!).
This guy is a known troll. He MAY have valid points but his credibility is zero.
CDE open sourced! https://sourceforge.net/projects/cdesktopenv/
Lets hope we can look at some independent tests in the coming days and see which unit is really value for money, because if Dell's benchmarks are correct their unit is 20-30% faster and only 2/3rds the price.
Economic Left/Right: -0.62
Social Libertarian/Authoritarian: -3.69
He had me thinking he was insightful and thoughtful until the end where he replies to all of his hate mail individually. Woulda made his point better if he just left it alone.
slashdot: where everyone yells sarcastic metaphors to themselves to understand the issue
Apple is the new Sony: give people the gadgets they want, and charge a premium for it. Speed
is not the killer argument. As long as it's
fast enough to digest whatever you throw at it,
who cares?
Ceci n'est pas une signature
YOUR HOME TOWN (AP) -- You were not "the most handsome boy in school," contrary to what your mother may have said at the time, officials today announced.
... no shiat! Apple marketing spins things; Dell marketing spins things ... everyone spins. Don't take it so seriously.
"Mothers always say things like that to their gangly, awkward teenage children," one official said on condition of anonymity.
----
Point is
Mindy: "Well...desserts aren't always right." Homer: "But they're so sweet!"
hyperthreading is like having 2 processor, when you only have 1 physical processor.
So If you want compare a hyperthreading processor vs G5, you have benchmark 1 (single) hyperthreading CPU vs 2 (dual) * G5.
We have to be fair.
Consensus is good, but informed dictatorship is better
Anti-Apple Troll vs. Apple Troll
How terribly interesting.... o_O
Well, it certainly isn't the first time that a company has used a benchmark to make a product look better than it is, and it certainly won't be the last time. I think what we should all learn from this is as follows. Don't worry about Statistics, Benchmarks, or any Media Hype. Just go to the store, buy whatever kind of computer you want that floats ur boat, Be it a Mac, Linux Box, Windoze Box, or god forbid, a compaq. Set it up, get broadband internet, and read lots of Slashdot and play Starcraft.
I have no regrets, this is the only path.
My whole life has been "UNLIMITED BLADE WORKS"
imagine that...a company fudging the specs to sell product. this is absolutely unheard of!! we must stop this travesty now!!!
;)
(this is sarcasm - and should be modded as such
We're like rats, in some experiment! -- George Costanza
I mean computers are so fast that there's very little that I might want to do at a consumer level that makes a difference. Most applications are responsive on my ancient 500MHz Pentium 3.
The only things that really need speed are things like 3d rendering, video compression and compiling large appllications. 3D rendering in games is influenced by the speed of the graphics card a lot more than the speed of the CPU, so we're left with the long slow scenes. Personally, it makes very little difference to me if a rendering a scene or compressing a video takes 30 minutes rather than 40. If I can kill 30 minutes, I can kill another 10 quite easily.
In the past, I'd have been able to tell you whether I was using a 20MHz or a 25Mhz 386 just by using it. I can hardly detect the difference between a 1.5GHz machine and a 3.0GHz machine without using a benchmark.
In the end, it's just numbers.
The Big Difference may not be the out and out raw speed of the hardware but I'm betting that in conjunction with their new 64 bit OS it will scream and I don't know of any other company putting out a 64 OS to the desktop.
"Talent does what it can; genius does what it must."
Hadn't Microsoft just done the same when they compared Windows SMB performance against Red Hat. Sad to see Apple is using same crappy methods.
“Wait for Hurd if you want something real” –Linus
Ultimately, it will be interesting to see the real world performance of the G5. I own a 2100+ amd athlon, but I don't feel much of a speed increase from my old 1 Ghz. As usual, a processor is only one part of a computer's performance, and the 1Ghz bus that the G5 will use will greatly contribute to the percieved speed of the system. Also, the interaction with OS X will be important. I use a G4 powerbook running jaguar, but occassionally there are slow downs - not sure why.
"When interpreting these FP results, let us keep in mind that most people use Integer (not FP) most of the time. Therefore, integer results (SPECint) are much more important than floating-point results (SPECfp). In other words, most people should ignore floating-point results because they do not use floating-point anyway (or not much)."
Well that may be true in Windows, but from what I recall the Cocoa API and PDF display model relies on floating point exclusively for screen coordinates. IOW, floating point may not be so important on Windows but it is on the Mac.
I also recall the implementors of Mathematica complaining about the integer centric nature of the Windows API.
I watched the video. (http://stream.apple.akadns.net/ - requires QuickTime). Now, I'm sure there's many ways you could tweak the benchmarks and so forth but the Photoshop and Mathematica benchmarks rocked. The G5 was 2x faster than the Xeon.
I used to get involved doing benchmarking back in the good old days of Whetstone when I worked on supercomputers. Every manufacturer had a different nasty tweak to the compilers that were pulled out only when it was time to do benchmarks for a customer. The mantra then as now was: the best benchmark is the app you want to run (since most buyers of supercomputers write their own apps, porting them for a benchmark was a possibility).
The G5's may not be the hottest thing on the planet but they're close enough to get Apple back in the ball game. Nice systems architecture, nice case and the claim is they're quiet as well. Oh, and don't forget you can put in 8GB of RAM. Now even OS X doesn't need to swap :-)
I still havn't seen an independent organization that does performace comparisons that I would trust. If you need to REALY know, do the testing your self.
Sure it would be expensive for a private user, but for corperations it would be feasiable. I still haven't seen many companies do their own performance measurement.
Come the revolution, the Bourgeois, Capitalistic, "A PARKING STICKER HOLDERS", will be first against the wall!
And what's more, when you start running programs that use more than 2GB of data, the 64-bit machine is going to beat the pants off the 32-bit one, since the 32-bit machine (i.e. intel) is going to have to resort to slow and hacky solutions such as segments and paging. The intel may me "faster" but only as long as 32-bit are enough for you. The days of 32-bit machines are numbered, just as they were for 16-bit machines when 32-bit machines started to appear.
Stick Men
The 1GHz backplane is the real news. No processor benchmark test really takes into account the total real speed of the system when running applications.
The fast backplane will speed up IO, which is a common bottleneck. 1GHz for a PC backplane is huge. The only machine I had seen a 1GHz backplane in so far is a HP-UX server. It cost wayyy more than $2000 or even $3000.
I really believe that with this new chip alliance with IBM Apple will finally be able to put that "the OS is really cool, but PCs are always faster" stuff behind them.
Yesterday was a good day for apple.
The big new is that they have new chips, and these suckers can scale and perform well with SMP, who cares if there not the fastest in the world, If apple can't outperform they'll just stick more processors in
It seems to me that if somebody wanted to use an inferior product, the first thing they'd do is develop a thick skin and at a minimum ignore the criticism being lobbed at their platform of choice. That, or choose to adopt something that seems to work better for the majority so that they don't have to feel left out all the time; obviously when you get to the point of chewing out people who are trying to show you why your choice is flawed it's become a popularity contest for you already (competing, not computing).
Try not. Do or do not, there is no try.
-- Dr. Spock, stardate 2822-3.
I am a Mac user. I have been using Macs for years. I am writing this article on a PowerMac G4. I enjoy using Macs.
Finally, I'm sure the real world testing, once available, will be of more interest to most people than any of these silly lab tests.
SPECfp ratings don't matter! Thanks to *mumble* Apple SPECfps are faster per SPEC mark. Stop buying into the SPECfp myth!!!
--- Often in error; never in doubt!
Strange that they didn't bother to test serial ATA on the Mac vs serial ATA on the PC. Seems kinda bogus for them to use two different busses. Look up Serail vs Parallel ATA on a PC and you will see that kind of performance boost.
Last time I checked Apple is also running a somewhat "cooked" version of Bonnie for these tests.
-DU-...etc...
"Don't sweat the technique."
Steve Jobs' assistants roll out the two carts... they open Photoshop... they click on the "Gnorglize foodlefacets" effect... the progress bars start moving... the Pentium's pulls ahead... the Pentium completes... the Mac is still grinding away... ...and Steve Jobs says, "Well, as you can see, the Pentium really IS a little faster at this particular task. But, hey, the Mac has a prettier case and it's easier to use?"
I'd really like to see this someday, but somehow I don't expect to.
"How to Do Nothing," kids activities, back in print!
First nVidia, now this... how am I supposed to go bankrupt buying more computing power than I could ever hope to use?
seriously... so now we might think that in real world usage, the G5 is maybe just a little faster than the x86 competition instead of a S*** load faster. Considering the performance point of Apple's previous offerings, I'm not exactly dissapointed
Just as irrigation is the lifeblood of the Southwest, lifeblood is the soup of cannibals. -- Jack Handy
He turns hyperthreading on but only counts one processor at 2ghz for the apple. Either leave them all on or all off.
That guy is a dipshit.
Anyone who's followed the computer industry for more than a couple minutes knows that there are lies, damned lies, and benchmarks.
Go use a machine, for tasks you'd typically perform -- that's the only benchmark that matters.
But if you must assign a number to the size of your virtual phallus, by all means, benchmark away...
Perhaps the benchmarks and fanfare are only being used to push those who were interested in purchasing a Mac for the OS but were not going to spend $3000 on a G4 when a G5 was expected.
Those who are deeply entrenched are not going to switch sides because of a benchmarks. There are still people swearing that AMD chips are faster than Intel chips, and when the benchmarks show they are then the Intel fans will swear it's not true. Blah blah blah. Who cares?
People use Macs for reasons other than Benchmakrs against Intel/AMD chips. I have used PCs for a very long time, but I have also been using Mac OS X at work and as often as I could. It really is a terrifically designed system which offers significant benefits. The only thing holding me back what the overpriced G4. Now there is nothing, thank you Apple.
The List of Grievances with Slashdot.
Benchmarks aren't what sells apples and price certianly isn't the drawing point. People use macs because they like macs. Hence why the mac market doesn't increase that much, they're too pricey and don't act like a PC. Granted as a user who uses windows, linux, and Mac OS, and all the subvarients between I can tell you that there are perks to all the operating systems. But as far as hardware goes x86 wins hands down.
Why is x86 better than apple? Simple, they're more tweakable, upgradeable, provide more selections, and are used by more people. Apple makes up for the "not used by many people" by making every mac an exact clone of another. Hence why when you get a file for a mac to be installed you just drop a binary in, every mac is the same (to an extent), whereas every PC is not, but the components are the same some just perform better than others.
Apple's prices are outrageous, and let me get into it a little more. A first time computer buyer is wary of a computer. They don't want to invest a whole lot of money in something they don't know if they're going to be able to use. But for $600 they can have a pretty decent machine that plays most every x86 game out there and runs most every x86 OS out there with little or no trouble. For $600 you might be able to score an old iMac. That old iMac MIGHT be able to run Mac OS 10.2, but it's going to be hella slow and not be able to do half the things the same priced PC will be able to do.
People who buy computers are looking for the most they can get with the least amount of money. Most people's computers are still beige. Most peoples computers have all the same applications. And Most people rely on somoene other than themselves for computer help, hence more PC's more help available.
I like OS X (especially with a two button mouse). I like linux (especially when everything works right). And I like windows (especially when XP loads correctly and doesn't crash and doesn't require me to kill processes in the task manager all the time to get some of my memory back).
All of these systems have their perks and they all have a place in the market, just they all want more of a place in the market, hence the competition. If Apple wanted to procreate so much they'd come up with a bargain computer other than the eMac or iMac. Something that has the ability to be upgraded (even if the user never wants to) and has the ability to run popular programs, hence MS, hey MS if I buy a copy of Word I want to be able to install it on either my PC or my Mac, I don't want to have to buy two different copies.
Anyways, these computers will be blasted out of the water in no time when Intel and AMD roll out their 64-bit badboys. Remember the 970 is actually an older chip in comparison to the AMD and Intel varients. Granted x86 isn't exactly new ... but neither were the moto's.
Ignore the "p2p is theft" trolls, they're just uninformed
I can vouch for those unfortunate enough to have worked in the grocery industry (and have an idea of what that kind of mechandising entail) but this is hardly insightful. It happens on nearly everything that you buy.
When asking the pricing managers (which work for the chain, not an individual store) they replied that there was a study once done, indicating that there is a psychological tendancy to shy away from certain "maker" numbers as being too big. For example, the masses statistically believed that twenty dollars was too much to pay for item x, but for some reason, nineteen ninety-nine was not too much to pay for the same item. Funny thing is that with the same item, eighteen dollars would again be too much, but seventeen ninety-five wouldn't.
Even if the study is flawed or bogus, it is still being taught in the "front-line" marketing schools, (ie. grocery, drug-store, clothing, etc.) , and so I expect we will see nineteen ninety-five for many many years to come.
The real issue is that you need a dual CPU on the Mac to be comparable to the single CPU on the PC (for some applications/uses). When comparing prices, look for machines that can do equivalent amount of "useful" work (whatever is relevant in your business/pleasure). I think that you'll find for most users (business, home, academic) that the PCs prices out at 20 to 30% cheaper (after adding software, monitor, other stuff). Of course, for some applications the Macs will be cheaper and for others perhaps an XBOX or homebrew is appropriate. Remember, many other companies quote prices that include monitors (apple doesn't, except for the built-in e/iMacs.) Also, if you need it, Office is much pricier on a Mac than as a included item from a PC manufacturer. (Also, I don't think there is a boxed cheap version of home/academic Office for the Mac.) Don't forget to add a couple of bucks for an adaptor if you don't want to be a pricy (but gorgeous) Apple monitor.
Misleading Prices
Both Apple and Dell are guilty of using misleading
prices. For example, Apple gives the price of the
low-end G5 as "$1999", and the high-end G5 as
"$2999". In other words, they have subtracted
$1 from a $3000 computer to make it seem cheaper,
which is absolutely ridiculous. This
demonstrates that both Apple and Dell are willing
to mislead people when stating their prices.
Geez, that's so tricky of them, I fell for it!!
Fools, this is not something that Dell or Apple invented, this has been a standard marketing trick for at least 50 years if not more...and EVERYONE does it!
"Some things have to be believed to be seen." - Ralph Hodgson
I think we still have to pay for it?
However, Apple have the marketing power, user base, application software, user-friendliness and pretty GUI to make a commercial success of their 64-bit desktop machines (which, incidentally, run at a competitive speed).
Stick Men
I'm leaning towards a similar conclusion (``My next boxen will be PowerMac G5'') for when it's finally time to retire my NeXT Cube.
The new case design addresses most of my complaints about the old G3/4 design (funky round handles and irregular surfaces make stacking / arranging things around those problematic, noisy (but grant it is quieter than my NeXT Cube) drive panel access---I guess the SuperDrive has no buttons on the face plate beyond eject?)
and Panther finally brings most of the missing features from NeXTstep (Faxing, PostScript support, speed) and Mac OS 9 (Labels, apparently working QuickDraw/GX like font support).
I'd give my interest in Hell though for a way to change the monolithic, immovable main menu to a movable vertical menu a la NeXTstep (w/ top-level Print and Services!), esp. w/ tear-off sub-menus, and really wish that there was a language option which would give one concise NeXT-style menu shortcut descriptions....
William
Sphinx of black quartz, judge my vow.
That the Pentium III made the internet faster and the P4 makes digital media better. Both camps pump out a lot of bullshit. I actually expected Apple to go the 64bit@2Ghz == 4Ghz @ 32bit route, so this isn't so bad.
"Windows Me offers tremendous reliability and stability improvements..." -- Paul Thurott
it would be Veritest. After reading the PDF myself, they draw their own conclusions that the G5 outperforms the two Dell machines they tested it against.
So how would Apple be at fault for saying they have the fastest CPU benchmarks based on independent testing? Unless of course "a very stupid man decided to take the large stack of money."
In C++, friends can touch each others private parts.
http://www.apple.com/powermac/graphics.html
337 fps with a twin 2.0 GHz and Radeon 9800 Pro @ 1024x768, 32 bit color
I really can't understand why the author of this piece takes the SPEC numbers provided by Intel and Dell at face value, rather than investigating them in detail the way he has with Apple's; Those guys have certainly done as much twiddling to perform well on those tests as Apple has.
And I can't understand why there's a problem with using GCC on the intel over ICC. Sure, GCC doesn't produce the fastest code for the x86. But it doesn't produce the fastest code for the PPC, either; For that you'd want to use the IBM compiler.
And the repeated claim that for "most people" integer performance is what matters is somewhat stupid: For the "most people" who are mostly exercising integer performance (i.e for web browsing, emails, word processing), a top-end box like the ones being compared here is overkill. For the people who do need this sort of speed, it's much more likely that there will be a large amount of FP in the mix.
You have to have a license to drive a car, but any asshole can have a website.
Has anyone ever known a cause so perfect (much less a cause like "computer technology evangelism") that it didn't have at least a few vocal morons supporting it?
But I still fail to understand why. If you don't want a faster computer, then don't buy one. But you're completely wrong that "it's just numbers." Sure, one step up a mountain is only one little step where no one can tell the difference. But then you take another step. And another. Before you know it, you've travelled 10, 100, 1,000 feet. That 40-minute video compression might take 39 minutes on the next step up, then 37, but eventually it will only take 1 minute, or less.
So don't dismiss numbers, especially if you can't see far enough to add them up!
Moderation: Put your hand inside the puppet head!
The "I get to pick and choose what comments I respond to, and I get the last word at the end, so nyeah!" section at the bottom reveals this article for what it is â" flamebait.
I do not fully understand how the writer could claim that Apple have done something wrong when benchmarking in the "there is a faster Pentium"-aspect. The 3.2GHz was just released when the new G5s were presented, and even if Apple have had access to a preview sample, I do not believe that they would ever use such things in benchmarks. I mean, what if Intel suddenly decided not to release the 3.2GHz at that time, then Jobs would have stood at the keynote saying "well, we have benchmarked against a processor that is not available yet". It would not seem credible.
All the people commenting about it.
Next question please.
The author of the article makes the point that most programs use a single processor unless specifically written for using two, so we should downplay the dual processor results. A good point on the surface but examine it more deeply and it has two flaws:
1) This is Apple's Pro machine and many of the users are in the Graphic Arts, Audio and Film industry. The most siginificant programs in these fields do get optimized for the Mac platform.
2) I don't know about you, but it is normal for me to be doing several things at once on my computer. Listening to music, downloading email, munging video, plus about a hundred background tasks. The OS itself balances these separate tasks between the processors, so there is a very real and significant advantage to the dual processor even if the individual programs don't take advantage.
-I have no Sig yet I must scream...
I seem to recall some rediculous test of a Gateway flatpanel iMac wannabe, where the machines were compared by examining start-up times. The catch? The iMac was timed until classic desktop was completely started, as in; the iMac had booted two OS's in slightly more time than the Gateway had booted one.
The gateway is therefore 1.6 times faster!
Just as irrigation is the lifeblood of the Southwest, lifeblood is the soup of cannibals. -- Jack Handy
Just my two cents
I do believe Apple used Windows XP for the tests between the G5 and the Dell PC. They would have to disable hyperthreading because Windows XP only supports up to 2 Processors and Microsoft does count each processor w/hyperthreading as being two processors (dual Xeons would count as 4 processors). If they wanted hyperthreading enabled on both processors they would have used Windows 2003 Server or Windows 2000 Server.
I always chuckle when I see someone spend extra for a 100 Mb or 1 Gb ethernet card, then plug their PC directly into a cable modem that does 1.5 Mb. I want to tell them to go to their local PC builder and get an ethernet card from the spares pile.
I thought that a big part of the excitement over the new Mac products was the high performance bus? Raw processor speed is always only part of the story. If the Mac can move data amongst ram, disk, processor, and video card much faster than the Intel/AMD systems, the net result is a more powerful PC.
I run a 633 celeron based eMachines. I pulled the OEM hard drive and replaced it with a 7200 RPM model. My user experience performance doubled - although the bechmarks wouldn't have changed. This is reasonable, since on an average desktop machine, the user's waits at boot and when loading an application from disk. Both those processes ran much faster with the high speed hard drive.
Just a note, besides web browsing/word processing I develop in Java. Tools such as Ant do a very good job of incremental compiling so I never have to kick off a make script and wait.
And I got modded as flamebait and troll yesterday for stating the same thing.
Zealots suck.
The ratio of people to cake is too big
If I had a dollar for every time I've seen thius EXACT same comment. Seriously, do you just cut and paste this thing every time there is something related to apple on slashdot?
The multithreading in the G5 architecture is tight. If you were performing enough integer math to saturate the CPU, which is the only time these sorts of comparative speed evaluations are relevant, the G5 damn well would be taking serious advantage of the second CPU.
Everyone has an agenda, and this article's author is no exception.
Kevin Fox
SPEC still indicate integer performance because some people care!
...any publicity is good, and good publicity is even better.
Everyone is talking about the G5, and from a marketing standpoint...chingChing!
You chumps are a laugh a minute...want some cheese with that whine? Apple is laughing all the way to the Bank! Go Steve, go!
there is a bit of a bias there. He complains about Apple tweaking its benchmarks. I have no problem with that. Companies should get blasted for running bogus benchmarks. But then he compares Apple's results to Dell's and AMD's without questioning their tweaks.
Perhaps what he meant to say is: "If we are going to use bogus benchmarks, let's compare them to the bogus ones from the competition."
http://www.apple.com/powermac/graphics.html
There you go. 337 fps for the g5 vs 275 fps for a 3 ghz P4.
Benchmarks are shite...the simple fact is that Apple has some fine machines coming out that are a huge improvement over what they have now. Oh look they render in the same amount of time, or the Dell is .5 seconds faster.
It's all marketing spin. Of course Apple is gonna optimize things to make it look better. It's not like noone has done it before.
Hyperthreaded cpus show up as 2 in Lin and Win, so there should be a clear cut application for this type of 'marketecture'. Either it is or it's not and since it does show up as 2 cpus, then I don't mind if it's disabled for benchmarks.
That said tho, the benchmarks are not taking full advantage of the cpu, so yah, that's screwy.
On the top of it, I don't mind the skew because, at least, it's not some Windows dominion marketing deal. But in the honest scheme of things in the galaxy, yah, that's pretty lame of Apple to pull that shit.
Will it prevent me from buying a still phat 64-bit (tho I honestly won't use the 64-bits) OS super system? No. But remember, it's not just the cpu but also the bus/hyper transport, the faster DDR and busting the 4GB memory barrier. Regargless of the fudging, I still think this is a great machine. Linux will stay on my servers while OS X will stay on my desktops.
There is only one rule --- WIN.
I have worked for six different computer companies over the years. All they ever wanted to do was win ONE test. This is so the literature and marketing droids could focus on that test showing that we had the faster computer in the known universe.
At one place there was a choice. We could have a C compiler that either ran the customers work faster OR gave better spec marks. I don't have to tell you which one management picked.
The results are never that useful. Each manufacturer runs their soon-to-be-released hardware and software against the competition's already released product. It is always unfair. Everyone in the industry knows that and no one really cares.
Apple now has a machine that stands up to the best for performance. Recognize that and move on. Because next month someone else will have another machine that gives "better" numbers. The only thing any of us care about is -- is it fast enough for what I want to do?
Enjoy your job, make lots of money, work within the law. Choose any two.
I am trying to imagine a world where manufacturers hand over the idiotic marketers that think they are going to get away with stuff like this. It isn't so much a desire to severely beat a marketing exec, but perhaps a few examples would lead us to.. honesty in marketing? ::gasp!::
First it was NVidia/ATI, now Apple - at least in recent months. I am positive there have been many more dating all the way back to the 8-bit days. Some may slip through the cracks, but with the proliferation of the Internet it takes one curious individual to bring to the house of cards.
-iCoach
"Never upset a goalie, getting hit with a blocker is an unpleasent experience - facemask or not." -Me
and pretty much ignored the overall system performance. Back in the days, Byte magazine had a "WinMark" (or something like that) benchmark that ran real world tasks on various machines. For Apple pro machine users, Photoshop and Mathematica are probably pretty good real world benchmarks simply because so many Photoshop users (both PC and Mac) spend so much time sitting in front of their screens waiting for things to happen.
That said, his points about the bogus tweaks for the G5 and the bogus downgrades for the Intel systems seem legitimate. After all, it was Apple who put the SPECmarks out there for comparison.
Get one? :) Free OS + cheap AMD procs == good times!
Why not get two, like me
I'll never give up my Linux for a fruit company, let alone a company who claims that their software is "micro"!
...You are perpetuating the lie that Al Gore claimed to invent the Internet.
http://www.snopes.com/quotes/internet.htm
"640K ought to be enough for anybody." Well, I think that 4GB ought to be enough for Personal Computers for quite some time. What possible use is another 4GB for personal use? I can see it in use for corporate or scientific uses, but personal? Apple's trying to make it seem like 4GB is insufficient. They sound like a used-car salesman. (Before you bash me: I'm going to get a G5 early next year. OS X is beautiful and my 800MHz Mac will be showing its age by then.)
Totally Life!
ALL replies
OK, if he is, supply us with some facts to back up your conclusion. Your arguments will have much, much more weight.
Who CARES!? I'd stand in line for a week to get kicked in the balls and throw the best PC money can buy in that trash, just to use a Mac as fast as the new G5s....
Screw PCs. They suck. Macs rock.
When will the trolls understand that:
WE DON'T GIVE A SHIT ABOUT HOW FAST THEY'RE FUCKING PCS ARE,
HOW STUPID ALIENWARE CAN MAKE A CASE AND PAINT IT NEON GREEN,
HOW MANY FPS GAMES THERE ARE AND HOW MANY FPS THEY CAN ACHIEVE.
OFFICE SUCKS,
EXPLORER IS A DUMB IDEA
BRUTE FORCE DRM IS A DUMB IDEA
WINDOWS LOOKS STUPID
but most of all....
T-H-E-Y A-R-E N-O-T M-A-C-S.
Jesus, get it through your thick, fucking skulls.
Well, it doesn't seem like we are getting what they claim. Head over to the Apple Store, under 'Choose Your Powermac G5' it still lists the old G4 specs!! I guess they were a bit freaked out by the previous info leak.
Seriously. Apple has such a history of doctoring benchmarks to make their own proccies look faster that I didn't even parse the headline as "World's fastest PC!" Rather, I read it as, "Capable of being doctored to look like the world's fastest PC!" Which, you have to admit, is a quantum leap over the G4.
That said, I use Photoshop a lot, and given their claims I would expect it to be at least competitively fast at Photoshop. And hopefully we'll get an optimized version of POV-Ray for the G5 (more likely MacMegaPOV), which would be great for my render jobs. While compile speed will probably remain slower than the competition, Apple's new IDE should soften that with its precompiling etc.
So when it comes down to it, I'll probably get one of these systems, maybe after they drop in price (combo drive anyone?) and sport Panther. After using OS X, Linux and Windows for a few years, I'd say that the Mac operating system, unlike the processor, is decidedly superior, and when I'm writing a term paper, surfing the web, or serving a low-traffic site, that's what counts more than speed. While the price is an issue, it's worth it not to need tech support.
I hereby place the above post in the public domain.
I wish I got hatemail like he did...
The SPEC bench marks don't show REAL world performance. SHOW me how these things run Photoshop and don't turn anything off because as everyone knows, it's not going to be turned off when a user runs it and also Adobe and others will make versions optimized for the G5. If I can get Photoshop running faster (or any other app) then thats a good thing! Personally, the G5's are still cool....regardless of how fast they are, they will still do everyting I want and more.
Gorkman
"The fastest" desktop -- they mean the fastest desktop that is RUNNING Apple applications and Apple OS. AND since no one else can legally manaufacture Apple hardware, they win by default.
Imagine how embarrassed Steve would have felt if some Taiwanese close mfg. came out w/ an Apple clone that was 30% faster and cost 30% less AND came out a week before their "fastest desktop" in the world launch...they wouldn't even be able to claim that it was the fastest Mac for a week.
Apple turned off hyperthreading in the Dell precision machines, and disabled SSE2. These are modifications you're gonna notice using photoshop, so those benchmarks say nothing.
---
"The chances of a demonic possession spreading are remote -- relax."
With the single exception of Photoshop (and maybe anti-aliasing), I don't see why you couldn't do all the rest of that stuff at the same time. Maybe you just needed to learn how to use Solaris!
Moderation: Put your hand inside the puppet head!
Anybody who buys a computer based on benchmarks or hype gets what they deserve.
Buy a computer based on what it can do and how it meets your needs.
Sheeh people.. It's marketing hype. Please don't tell me this is a new concept to you. If it is, I have a few bridges and plots of swampland on the moon to sell you. I have the deeds right here. Please email me for further details.
abcdefghijklmnopqrstuvwxyz
The difference is that this guy has actual facts to back up his claims.
> A G5 is faster than the fastest Intel box with Linux. Read the benchmark whitepaper. It describes the testing methodology in precise detail. In a side-by-side, controlled test, the single-processor G5 was 10% slower on integer performance but 20% faster on floating point performance than the Pentium 4 with Linux.
Apparently they never got so far as Chapter 1 in Hennesy & Patterson, where you learn the mantra of "make the common case fast".
Sheesh, evil *and* a jerk. -- Jade
Go read Jon "Hannibal" Stokes article about the world of benchmarking., over on his site, Ars Technica.
h tm l
http://arstechnica.com/cpu/2q99/benchmarking-1.
This will give you at least a basis for understanding why benchmarking is used, and what makes or breaks any given set of results. Also, feel free to argue about anything and everything that is said about these benchmarks, since, apparently, everyone of you is in the benchmarking labs day in and day out, testing systems and looking at the results on a scientific level.
I also think benchmark scores are, quite frankly, marketing bullshit. A processor designer can tweak a program and a compiler any number of ways to increase thier scores. The true test would be to use the SPEC benchmark suite with no flags set on the compiles for either platform. That way you are testing just the base processor, with no SIMD instructions, no disabling of the software prefetch algorhythms, no "cheats" as it were. Then test those same systems with every trick in the book thrown in. Then look at the difference. This will probably give you a better picture of the performance you will see in real world activities.
If you have a machine that absulotely sucks donkey when using no "cheats" and then you see this amazing boost in performance when the "cheats" are enabled, you probably are dealing with a highly optimized and specialized instruction set, which can be either very good for specific applications, but absolutely horrible for programmers who don't have access to, or don't bother to research, the abilities of that processor.
These are the benchmarks I'm interested in most. And it'll be at least late September before we see any of that.
Also, while all this is interesting, in an intelllectual sort of way, what about the actual perfomance gains over the current crop of G4's? Why not take a look at the difference between the SPEC scores of the dual 1.42GHz G4 towers, vs. the dual 2GHz G5's? That alone will tell you more about the increase in speed and power that has been delivered. If Apple had been smart, instead of trying to impress and piss off the x86 sparkheads they should have posted those scores as well, to give a real side by side comparrison between the speed and power of the G5 vs the bottlenecked, processor starving, gimp that is the G4. But that would make too much sense, wouldn't it? And you know marketing is all about confusing your consumer into beleiving that the latest and greatest is really what they want, not some old machine from 3 months ago...
Don't Ask Questions. I don't know the answers and even if I did I wouldn't tell you.
Just wait and see if Apple releases benchmark numbers to spec.org. There, they would have to pull out all stops (not use GCC, etc.) to get as high a number as they possibly can.
Considering the IBM pSeries benchmarks already trounce the P4 and Xeon using 1.7GHz POWER4 CPUs, it would be interesting to see how the G5 does with its smaller cache but at 2GHz (don't forget the 1GHz bus, either).
I think we would find the benchmarks at Apple.com were off, but probably not by much. Another thing that is not denyable is that the G5 scaled to two CPUs much much better than the Xeon (look at the rate numbers--this is unsuprising given the POWER4 heritage).
Healthcare article at Kuro5hin
Say after me: there is more to a computer's speed than the CPU's speed.
This 2ghz G5 may only be as fast as an Athlon XP 2800+, but it has got a ONE GHZ FRONT SIDE BUS (!!!) man.
How long has it been since computers last used x2 multipliers: the likes of the DX66, DX4 100, and Pentium 133mhz.
So while the CPU may be a tad slower than the very top AMD & Intel chips (which cost insane amounts of money and nobody much buys for good reason), the significant front side bus advantage will give overall better system speed.
Of course MacOSX is slow and bloated, but so too is Windows XP.
Probably the Gensi Pegosis and Castle Iyonix are the snappiest computers for everyday computer use - yet they both use 600mhz CPUs.
A computer's usage speed is dependent on the following in this order:
*software efficiency (by far the most important)
*front side bus speed / ram bandwidth
*hard drive speed and transfer bandwidth
*CPU (least important)
but that list has been reversed in popular perception.
Hence why billions is spend on making CPUs more efficent and faster every year, while front side bus speeds increase at a crawl and software takes a full 100x more resorces to do the exact same thing it did a mere decade ago.
Our priorities and perceptions are every bit as messed up as Apple's marketing. If modern OSes like WinXP and Linux only used 4x as much resources as Windows 3.1 to run, which is not unreasonable, then we'd get better real-world performance out of a PII 500 than our P4 3000s give us.
So what if the new G5 PowerMacs are not as fast the PIV or Xeon counterparts? Are they faster than the G4 machines they replaced? Does it now appear that Apple has processor that will continue to get faster at very nice clip (3GHz in 12 months). Hell, yes!
Geez, I looked at the discrepancy between the apple results and the official SPEC results and I thought it was because GCC was having trouble with Intel for some reason. But disabling SSE2 and using the 387 instead? That alone makes the floating point results worthless.
Using a tweaked malloc? Using special hardware prefetch tweaks? Disabling hyperthreading? Gah...
The following sentence is true. The preceding sentence was false.
First off, yesterday we have the day when all the Mac fanatics go overboard. Hey, I'm one of them and I went overboard. Enthusiasm goes right over the top and reality slowly slips away inside the Reality Distortion Field of the great and mighty Jobs. Yesterday was for the Mac users
Today we get the backlash and debunking. I honestly don't know if it's completely true or not but I'm inclined to believe it. I've grown accustomed to the idea that benchmarks and anything else like them (side by side tests of any kind) can't be trusted so this seems to fit.
The only thing that really makes any difference to me personally is how much faster the G5 is than the G4 it's replacing. The rest of it I just don't care about.
I use a Mac for a lot of reasons and flat out speed isn't one of them. It has to be fast enough obviously but it doesn't have to be the fastest and never has had to be the fastest.
I use a Mac because I have found it to be very stable and a pleasure to work and game on. If the benchmarks were rigged then it's a shame. They didn't need to do it and it wasn't worth the risk of negative press IMO.
Appended to the end of comments you post. 120 chars.
In some stores, you can tell the commission a salesperson gets by the last digit in the price of the item. For example, if something is $24.55, they get 5% commission on that sale.
But as another poster mentioned, most places just do it because it does have a psychological effect on many people; $299 really does look significantly better than $300.
As for the benchmark part of this article, it's good that people investigate this sort of thing, but every company stacks benchmarks in their favor. Always. They do it because they want their stuff to look the best so that they can make more money, and because they know all the other companies are going to do it too. This is less blatant than some I've seen over the years (no, too lazy to find examples).
WMBC freeform/independent online radio.
I dont know how much this would affect the benchmarks, but Apple made a point of using standard GCC for all of the benchmarks, rather than the platform specific benchmark tweaked compilers. If this does lower the dell/intel scores, maybe Dell and Intel need more attention??
How about a dual Opteron 244? Not even close.
Hate Mail: Oh no! Apple produces kick ass machines once again and someone wants to piss on our parade.
Response: I apologize for pulling you out of your wet dream and bringing you back to reality.
Honk if you're horny.
A. Benchmarking is a black art, and benchmark results more often than not bear little or no relation to reality (i.e. the actual performance you will get, today, running your particular workload). Talk to anyone who does it for a living and they are the first to admit that.
B. Benchmarks are very rarely impartial. Whoever is footing the not inconsiderable bill for a properly-done benchmark will have a result they want to see, and the benchmarkers can do a lot to make sure they do see it.
C. "Perception is reality" is a well-known saying in marketing. It doesn't actually matter whether the perception is correct. If Joe Sixpack believes he has bought the fastest PC in the world, he will be happy. More so since he most likely has nothing on hand to compare it to.
D. The speed this industry moves at, there will be a faster one along in a month or less, so if you really want something faster, wait for it.
E. All this debating about which is faster is more like masturbating. And "Masturbation, although an inherently pointless way to pass time, is at least enjoyable. Comparing PC performance is equally pointless, but rather less fun. The conventional epithet applied to those who engage in the former to excess is equally applicable to those who persist in the latter."
Benchmarks produced by the company selling the product are tweaked to make the product look better. Excuse me if I'm not shocked.
Spec results for the P4 aren't defined by what Apple (or Dell) say - they're defined by the listing at www.spec.org. Surely anyone who is actually buying for the processor speed (ie, not many Mac users) can just go and look up the numbers themselves.
I'm sure that the actions, filters, and files that apple uses for their photoshop performance displays highlight the mac's prowess as much as possible while, at the same time, try to bash the x86 machine as much as possible.
I think apple's purpose for these claims goes beyond the fact that their trying to sell machines. They're trying to exterminate all of these myths that have been going around for the longest time about their hardware/ software. 90% of the people I know that don't like macs don't like them because the ones in their middleschool/highschool were horribly upkept and would not work or crash too often to be usable. I think that a major reason why apple went with the BSD underbelly in OSX.
Also, these fucken trolls on slashdot with that story of "my 350mhz g3 is barely usable if I'm copying a file and playing an mp3." Fuck that, I had my 132mhz 7600 (604 based machine) running fucken OS 7.6 and I could download, listen to mp3s, chat, and surf the web with minimal problems. Granted, I had 256mb of RAM in there, but it was fine. Only when the applications started getting more robust that that computer began getting unusable, and by that time, I had a 450mhz G3 which is STILL in use.
Although, apple does piss me off sometimes with their claims which, although true, ARE misleading and cause these mac fanatics to make outrageous claims based on Apple's statements/ demonstrations.
Although many mac users (fanatics?) are idiots, I think that there's a much higher percentage of windows users who are, too. And the windows users are much more likely to pick a fight about it.
Platform choice is a preference! Use what you like. Use what likes you. Use what makes you whole.
...spike
Ewwwwww, coconut...
I could swear I've seen this somewhere else today... hmm... yes that would be this article on AMDzone. Take a look at the second page for the SPEC score comparison...
we are all invisible unless we choose otherwise
He had to switch to the MIS major after finding out he couldn't hack Software Development II.
Its obvious that the new PowerMacs are aimed at early adopters doing things that people have not really caught up with. DVD burners are cheap now. Make your own movie. Play with the iMovie effects plugins. I'd like to see (later) DV/HDTV rendering performance compared on different systems: AFTER the iMovie plugins crowd has a chance to catch up.
Oh, and in case you don't understand where games are going or you never saw "The Matrix" or any other VR sci-fi, convincing virtual reality relies on MASSIVE databases of objects filtering out the things that would get obscured by other objects, and streaming them to a rendering engine/GPU. I could just say CG animated movies, but really we will be playing *in* the CG scene and not just watching it play. I want to see the NEXT game made for the PowerMac.
Also, benchmarks are putting the cart before the horse. A new architecture or platform is a challenge to programmers: Saturate THIS! Imagine as a programmer if you took turns completely exploiting a machine at a time and simply reported the results. If you do a test that is a greatest common denominator of two platforms, you ignore the value of the incompatible feature sets of each respective platform. A real benchmark illustrates the full potential of each compared system, which provides an illustration of their differences. What happens when there isn't really any software to exploit the potentials of either/both of the platforms?
PCs are cheap and fast, but not really that advanced. There isn't much unexplored potential to attract the early adopters and the fatter profit margin supplying their hardware. I understand if you want to get the most for your money, but for some people, money isn't the top criterion.
Consider as a side note that after a year, top end Mac computers only lose half of their market value. So, after a year, you can almost trade-in your old high-end Mac whereas you're stuck with the comparative PC model. What can you get with a $1500/year budget over 5 years? Can you push the envelope on a PC? That's a tough question.
--- Nothing clever here: move along now...
Patrick Doyle
I mod down every jackass who puts his moderation policy in his sig. Oh, wait a sec....
With regard to price, if you're after a high-end system, he represents that the high end of the Dell line comes in at $3680, yet rapidly returns to promoting the idea that a $2000 Dell is equivalent. In an effort to configure up an Intel system comparable to the new high-end Apple PowerMac G5, I ran the Dell configurator. It clocks in at $3939:and that's with a lesser video card and a smaller, slower IDE hard drive (add $840 for SCSI, a better comparison with Serial-ATA). I don't think I was being unfair in my selection of components. (OK, add $30 for a USB floppy on the Mac if necessary)
This guy certainly has a point about the non-optimized Intel benchmarks, but he reveals his prejudice by not offering a fair price comparison.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
Just how much x86 code out there is SSE2 optimised to the extent that Intel probably did with the SpecFP benchmarks? 5% ? Maybe 10%? Even MMX isn't always used when it could be, and that's been around much longer. Personally I think the Intel P4 spec scores are a little misleading as well.
http://www.amdzone.com/articleview.cfm?articleid=1 296
The author claims that the P4 is "Faster on integer single-processor tasks, which is what most people use most of the time."
This is patently false. Typically, users run more than one program at a time. At the very least, there's an application, and the operating system. The machine I'm typing this on has 40 processes going, totalling a few hundred threads. Single-porcessor systems may have been king back in 1995, but these days you can typically make excellent use of multiple processors.
No doubt in a couple of days time a new Amiga will be announced... Twice as fast as the G5, with integrated hardware um stuff, and all that.
Then we can really start the fan wars.
The Mac zealot replies to this guy's site were pretty funny, but Amiga worshippers are in a different league...
I know Intel has its own compiler, but the AMD 3200+(2.2 GHz) beat it out too. Is there an AMD specific compiler, or do they use GCC?
Also, Apple did add G5 specific optimizations to GCC for this benchmark.... that's probably not on the scale of the Intel compiler, but it certainly makes a difference.
-=-=-=-=-=
I'd rather be flamed than ignored.
So now Apple gets flamed for only putting one button on their SUPERDRIVE? Exactly how many buttons do you expect on a drive??
I am the hub of Jack's digital lifestyle.
As the site is alrady down, here are some email that the author (www.haxial.com) recieved.
Answering the Hate Mail
Idiot Got Single CPU Test
Need not say more. If you had actually read beyond the first table of results, you would have seen that I go on to discuss the "Rate" benchmarks, which are for multi-processor computers (according to SPEC).
PC invasion? Hey PC weenee stay on your PC boards. Longhorn won't even be out 'til 2005. Most of the cool new features M$ is touting are in OS X NOW! Plus Panther, which WILL be 64-bit aware will be out before the end of the year further extending the gap. Longhorn won't be competing with Panther it will be competing with whatever iteration Apple has out in 2005. 10.5? What's faster than a Panther? uh, a Cheetah is faster than a Panther. Cheetahs are the fastest land animals, not Panthers.
Oh no! Apple produces kick ass machines once again and someone wants to piss on our parade. I apologize for pulling you out of your wet dream and bringing you back to reality.
Apple basher... This isn't the first time this guy's posted negative anti-Apple comments on his "soapbox." Definitely a PC whiner disguising himself as a Mac user. Apple's new boxes will leave PC's in the dust and they can't stand it. Now, not only do we have a superior OS but we also have superior hardware to run it. Now all you need is a superior race of humans to use these new superior macs.
Back on top! Damn the PC trolls are coming out the woodwork aren't they. There was more than a Photoshop bake-off today. Luxology (founders of LightWave) previewed their 3D animation program and the G5 outperformed your beloved Xeons. Dual 2GHZ 64-bit processors, 1GHZ bus, up to 8GB DDR SDRAM. Eat your heart out PC trolls. The Mac is BACK! Beloved Xeons? I don't even own a Xeon. But you obviously do own your beloved Mac.
This guy is the same guy that wrote all the Haxil programs. He is a extremely large Mac bigot. (And an idiot too) That screenshot is from a windows machine (as you can tell by the style of Anti-Aliasing) http://www.haxial.com/spls-soapbox/ read any of them. He says Copy/Paste is easier than Drag & Drop (Look at the example he used) and he says the screen menu bar is the worst idea he's seen. PC weenie. One that can't make valid points either. Oh yes, the "And an idiot too" part will definitely convince adults to switch to Mac. Also, what you think is Windows-style anti-aliasing is actually NO anti-aliasing. That's right, anti-aliasing is DISABLED in that screenshot, I turned it off.
Did he even watch Steve. He said that the single proccessors are slower in some test by a small margine. The really designed the system for MP. I can't find the 3.2Ghz Xeon he's whinning about. I think he got those mixed up with the P4s. I put a Dell together and Apple was right on the money with a similar equiped machine. During the presentaion they gave real world test. Was this guy not paying attention. This guy says he is a Mac user. I hope he gets a PC. He is making the rest of look bad if he is. Whiny beotch. Xeon, P4, they're all the same, they're all just evil demons from hell, right? I never said there was a 3.2 GHz Xeon. I said there is a 3.2 GHz P4. Since you apparently have difficulty distinguishing between different types of processors, I will help you find where the 3.2 GHz P4 is mentioned on the Dell website:
http://www.dell.com/us/en/bsd/products/m odel_dimen _dimen_8300.htm
what an idiot.. the proof is in the PUDDING!!!!! specmarks are just crap anyways.. look at REAL apps and you'll obviously see the G5 beat everything easily.. where is mr. haxial (pirate) now???????????????? so why isn't PS or mathematica or luxology faster on PCs?!??!? people are forgetting specmarks are just crapmarks.. The pudding? So you're suggesting I won't be able to understand the glory of all things Mac until I eat one? Maybe I should take one to bed with me too, to help me learn to love them. As for "specmarks" being "crapmarks", Apple chose to publish SPEC benchmarks on their
Ug... /.ed so soon.
Since I was bored, I went to look at apple's site. Something struck me as odd. I think they are using the same chip for all three models. From a "hurry and get these out the door" view it makes sense to use the same chip, why take the extra effort to detune them, just run it at a lower FSB.
1.6GHz PowerPC G5
800MHz frontside bus
200*8=1600
1.8GHz PowerPC G5
900MHz frontside bus
225*8=1800
2GHz PowerPC G5
1GHz frontside bus
250*8=2000
Quad pump them and there you have your 800,900,1000 FSB.
I'll be willing to bet that someone figures out how to make a the 1.6 a 2.0 within two months. Then again, I've been way off before. The MB could be waaaaay different.
Very good point - The Opterons munch thru data real well, and hopefully with a little sales and some price drops will become a great platform. I cant wait to get me some of that 64 bit action...
Take this benchmark for instance. Apple disclosed all the information they had to. They never LIED to the public (at least with this), but by burying necessary information deep and showing only numbers they have managed to mislead anyone who is too stupid to do further research. If you can't find the little link underneath the data shown and click on it, they figure you deserve to know only what they say.
CODITO, ERGO SUM: I Code, therefore I am.
Apple used G5 specific optimizations in GCC. They also used a specialized malloc(), which they didn't use for the PC. Also, they disabled SSE2 on the PC. And hyperthreading on the Xeon. And they used specific hardware tweaks on the G5.
(besides, even if GCC isn't poorly optimized for the x86, one could argue that the NAGWare Fortran compiler, used for most of the floating point tests, is.)
The following sentence is true. The preceding sentence was false.
Interesting that they didn't benchmark against a dual processor Intel system...
"No one likes working in a hamster wheel, and your shop smells of cedar shavings from here." - TaleSpinner
I especially liked the hatemail at the end. This was my favorite:
People that go to ivory league schools that live in trailers are a very low population, lets say 1% so since there are 99% of the people living in houses then you can clearly see that people that live in trailers are stupid, when compared to they're counter parts. Or put it this way, any finite number divided by infinity results in a number so small it does not exists. So any people that live in trailers that go to ivory league school you meet in passing are just figments of your imagination. Point is if you like Macintosh use it, if you don't then well don't use it.
Heh. That sure dispels this guy's critique of Apple's benchmarks, eh?
There is much cruelty in the universe, John.
Yeah, we seem to have the tour map.
8GB is not the maximum of a 64 bit address space, it is just the maximum of this particular machine...
A 64 bit machine gives an address space of 4G*4G (A lot).
btw those things are hacky solution because you need to do even more dirty tricks to get your programs to work... Programs expect a flat address space...
Jeroen
Secure messaging: http://quickmsg.vreeken.net/
The most interesting thing about that page is the collection of hate mail that he got. Nevermind all that benchmark stuff that he yammers on about - everybody knows benchmarks are bullshit anyway.
Those comments really demonstrate the applicability of the bell curve to real-life situations, especially things like intelligence of a population.
The dangers of knowledge trigger emotional distress in human beings.
I hate to say it, but I told you so.
Ok, ok, if you use both processors on an integer task, continuing to ignore floating point and bus performance, all you have to do is use a different benchmark on the Intel box to show the Intel box being a hair faster.
No comments on using the G5 on appropriate applications or application mixes.
Why rain on Apple's parade like that? They continue to do amazing work. The G5 appears to be dramatically faster than the competition in some perfectly realistic applications and at least comparable everywhere else.
The people giving this anti-Apple rant any credence seem not to have read it very carefully. It exemplifies exactly the sort of spin-doctoring that it claims to be offended by.
mt
Does this mean that my Pentium III doesn't really make the internet any faster? It was all a facade to sell me more processing power? Oh, the humanity.
-=-=-=-=-=--=-=-=-=-=-=-
What would Yossarian do?
The POWER4 and PPC970 specs have the FSB going at 1/4 of the chip speed, DDR. So whatever the processor speed is, the FSB will be half that.
http://weblog.infoworld.com/yager/
I think it may be that there is a good reason for a discrepancy in the testing, then again their may not. A Windows user can speculate, a mac user can defend, yet at the end of the day neither will change their mind. I have a mac at home (PBG4 & PMG4), and they are much slower than my work PC, but the difference is not the speed, its the feel of it. The user experience, the crispness of a mac vs the sloppiness of the windows machine, or sorry to say linux especially. I think osx is a bit more refined in some ways, personally I like it much better than Windows 98/2000, now I think that I would like XP a bit more, but I haven't gotten to play with it much. Thing about it, I like computers - as do all of you - only I don't really care one way or another! They both have usefullness, one is just faster than the other, but also onle looks a lot better than the other. A corvette is much faster than a Rolls, but one takes you there is style and class... but if all of this is true, and Apple has deceived us with marketing!! Well thats not what I buy a Mac for, and I definately don't like being lied too! I think that when the new G5 is shipped another side by side comparison should be made - for all of us too see? Who would do that, without bias?
What can I say. I like my games -- which is the one of the primary reasons I have a Winblows box. Also, I have a rather large investment in PC/Windows based software. When I can, I just into Linux. As far as the Mac review. I seen no problems here. He makes some good points, some arguments could be refined and other points are not brought to light. What is the fastest desktop anyways? How many Teraflops can it do? It's all a bunch a numbers anyways and there doesn't seem to be a great to definatively measure the difference in performance between a Mac and PC -- fundamentally you are comparing Apples and Oranges (he, he). I don't think he was down on the Mac, just wanted to shed light on some misleading info by the Apple marketing crew. There are a lot of good things about Macs, raw speed has never been a major part of it. Really, how long did it take Apple to make it to 1Ghtz? The whole theology of Mac, is "speed isn't everything" -- which tends to be THE Intel way of thinking. Personally, I don't own a Mac and don't plan to right now. However, I can see their use -- especially in certain niche markets. For general consumer use -- it depends. If you do standard stuff and not much gaming then a Mac is good for that. Especially if you gravitate to the geeky *nix side of OS X. For heavy gamers, I just don't see it. Yeah, there are more games out for Mac than there ever was, but typically *if* there is a Mac version coming it tends to be at least 6 months after the PC version. Works for other types of software too.
SPAM solution made easy: 1 spammer, 5 cords of rope, 5 hourses, and fireworks. Be creative.
I really don't give a shit if Apple fudged their benchmarks on SPEC. I expect as much from them, AMD and Intel. I take those numbers with a grain of salt. However, it's difficult for me to believe that they could have fudged the systems to such an extent that the G5 is twice as fast in Photoshop and Mathematica. I don't care what the benchmark is, you can't have results like that without the chip actually being faster than the competition. At this point, it's just a matter of degree. Maybe the G5 is only 1.8x faster in Photoshop under ideal conditions in both tests; who knows? Do I care? No, because it'll be faster.
... but come on Apple; where's double-precision AltiVec?!). The audio tests were also very telling. While I'm a bit skeptical about the applications not being the same, I think it does say a lot about the audio capabilities of the G5 and what it can do with a scant 25% of its CPU power.
The guy from Wolfram Research made it clear that the G5 outclasses the Pentium 4 in the scientific computing arena to such an extent that it doesn't even compete with it anymore; it competes with high-end UNIX workstations (and beats them, too, apparently
Bottom line, people are starting to try and eek out the edge on Mac vs. PC performance, and that's a good thing. With the G4, that was impossible because the G4 boxes were outclassed by such a huge margin by the x86 ones. Any way you look at it, these machines are competitive. And they run Mac OS X; the Pentium 4 does not. Therefore, I'll be buying the G5 next because I'll get competitive performance with the best OS on the planet.
Apple's hardware focus is misdirected. Instead of catering to the vast majority of Mac users (like yourself), Apple has the mission of migrating (the "Switch") PC users to the Mac world. Although Apple does say "our new box is faster than our old box", it attempts to use smoke and mirrors to get their machine up to par with the newest offerings from the PC world. Then, Apple goes so far as to claim that their machine is the fastest a consumer can buy?
One thing car salesman figure out quick is to never claim the trophy - only claim rights that his car is "one of the best".
Apple would be smart to recognize that a person that has used a PC since "DOS became graphical" will more likely stick with a PC (and its ultra cheap components) than make the switch to the proprietary Mac world. Instead, APple risks its credibility to make PC consumers think they are buying into inferior hardware.
Knowing that companies taint their benchmarks is definitely common knowledge; however, that does not make it a good practice. nVidia caught major flak from the 3DMark benchmarks. AMD is currently catching hell from their XP rating system. Apple should focus on their user base and continue to offer the public high quality components, software, and services - which all leads to increased market share.
What's left? Just as you said - the G5 is indeed faster than the G4 that it is replacing. It's technologically superior and still does a hell of a job for the Mac consumers out there. There was a point in time where I figured Apple would simply advertise "We're Apple - we're not a PC" and get the point across.
Apple should drop its comparison with the PC world considering the application and core hardware base are vastly different.
Ayup
"In other words, they have subtracted $1 from a $3000 computer to make it seem cheaper, which is absolutely ridiculous."
Those lying fucking bastards. I've never seen that before in my life. Never - I repeat - never, have I seen a product priced at anything less than a perfectly round figure. I'm so glad I read spl's soapbox. I mean, I went to the Apple store, and saw that it was $1999, and I admit it, I said "I could afford this." But thanks to the philanthropy of spl, I was forced to examine it further. If you actually sit down and do the math, $1999 is not, in fact, a thousand dollars and a little more - no, innocent consumer! $1999 is nothing less than a dollar shy of $2000!
c-hack.com |
Apparently they never got so far as Chapter 1 in Hennesy & Patterson, where you learn the mantra of "make the common case fast".
Apparently YOU never got so far as Chapter 1 of the Mac OS X system architecture guide where it says "we use single-precision floating point arithmetic for everything."
When you draw something to the screen in OS X, whether it's through Quartz 2D or a higher-level API, you specify pixel locations as floats. That's right, your window is 200.0 pixels by 300.0 pixels. And you can use fractional values, too. A pixel at 200.5 gets antialiased between 200 and 201.
Mac OS X is a VERY floating-point intensive operating system. The only ints in the system foundation APIs are bools and enums. Everything else is floats, floats, floats.
(By the way... aren't you the Black Parrot who about a year ago got stone-cold-busted for making false claims about Microsoft? Something about your company being put out of business by Microsoft or something? Claims that later turned out to be false? Just asking, because I want to know whether I'm talking to a troll or not.)
So, I guess you guys all benchmark your cars before you buy them, right? I mean, you study torque curves, run them on the dyno, analyze fuel consumption, record 0-60 acceleration and skid pad figures, right?
If all those numbers are the best, it must be the best car, right?
Or do you just go test drive one? Maybe two? I'll bet most of you don't have a clue how much horsepower you make at the drive wheels.
Yes, Apple hasn't fooled you with thier bogus numbers for a second... But somehow Intel and AMD are completely honest. And Windows really is worth all the trouble just because it's mainstream and you can play MOO3 a few months earlier. And Linux is worth all the trouble because it's free.
Uh huh. Have fun.
How can you say that using a mac is using an inferior product? Sure the G4 cant keep up with the P4 no matter what apple sais. But anything lost in processor speed is, for me at least, gained back by the OS. Now that the G5 is out and around the speed of intels 3.06ghz processor, how is the new powermac an inferior product? It has faster hardware and i think most of us can agree OS X s a "better" operating system then XP, hands down. I see what you are saying and you're partially right, just dont go calling macs inferior no matter how inflated the benchmarks are.
Okay, you convinced me then -- when the faster-than-a-Dell model G5 comes out, and it's the most super-duper top speed computer ever -- and not to be surpassed by any rinky-dink peecee -- then I'll put down my shiny nickels.
You just lost yourself a sale, Jobs!
DO NOT LEAVE IT IS NOT REAL
Apple is making silly claims about their processors being faster than they really are, that nobody else has made desktops with a powerful RISC architecture, and backs it up with benchmarks based on SPEC and Adobe products? Wow, what a surprise.
Why is this news? Apple has been spewing this BS since they started using the PPC architecture. The same goes for Sun. Nobody believes it, and in the long run, people buying systems for serious performance will stick with intel, who doesn't need to make any speed claims because their CPUs speak for themselves.
If Apple's marketing department ever starts speaking the truth it will be taken as a sign of the apocalypse. Not that the truth about Apple systems (That speed should not be the big concern when you want solid hardware and a great OS.) is a bad thing, but really, I think if they stopped spewing hype, we would all die of shock.
I hope this doesn't make me sound naive, but didn't they actually do comparison tests with Photoshop, Mathematica, Logic, and a couple others in front of a live audience? Didn't the G5 cream the competition? C'mon people, yesterday was a reason for Mac fans to get boners, and today some guy who really thinks Apple and Dell are horrible companies for knocking $1 off their prices is gonna give the PC fans boners? Enough with the boners already, they're just benchmarks and they mean next to nil.
I'm still waiting for my 4 simultaneous streams of HDTV amiga promised me about 4 years ago.
So? I can write one program that has a check:
for (big loop benchmark)
{
if (systemType = "Apple")
sleep(100);
}
and compiles for every OS. Everybody is running the same thing then but I can assure you that you'll have very different results. It *is* possible for a compiler to produce fast code on one architecture and suck on others. For example, prior to version 3 of GCC, gcc did fairly well on load-store architectures like SPARC and SGI machines but was horrid on Intel x86 compared to other compilers available on those platforms (such as Intel's and Microsoft's).
Who uses machine over $1800 these days?
... but to say that even a quarter of the professional market is not willing to pay for the high end $3k machine is crazy. The fact is, neither of us have numbers to talk about, but I can safely say your $1,800 for a graphics machine statement is at best influenced heavily by your opinion and not by fact.
*ahem*
Apparently you. Maybe even a couple other people
A lot of people pay for the top of the line stuff because if you think for a second that a $3k deduction and the productivity you gain by buying the top of the line is not worth it, well, you're not considering all the factors.
Everyone runs their own business with the same "I'm right" attitude, but buying a new $3k machine every 2 years makes more sense for my business than a $1,800 machine. Now, if I was to fill out a workshop full of 20 machines for production-grade work... I think the low or mid grade G5 would be more than capable at (drum roll) $1,999. That's just me, but I can afford to pay someone cheaply to work an extra 20 seconds on something if I deal with volume.
"Politicians find new names for institutions which under old names have become odious to the people."
This guys site is jumping all over Apple about being slower...when using a single processor!
This guys site even says:
"SPECint_base2000 is a single-processor test, so in the following results, where the computer has a second processor, it is either disabled or not used." then goes on to say after the benchmarks using only single processors: "As you can see, the PowerMac G5 is NOT the world's fastest personal computer. In fact, the Dell Dimension 8300 beats the PowerMac G5"
Well a big DUH is in order. Steve Jobs even SAID it was slower. He had a graphic up that showed how the single processor G5 was slower on INT based benchmarks etc etc. It was when they used benchmarks using DUAL processors that it really shined.
Yes, after all this in the article, THEN he goes on to rate the dual processors, but not before he trashes the Mac on something that the Mac had already admited to. I mean, that's pure trolling.
Bottom line, Apple used certain results in all the tests to market the new computer...just like this guy used the same tests to filter out what HE wanted everyone to see.
Also, Apple should never use benchmarks to market anything. No one should. It's too easy for others...no matter what the system to say "well, if you configure blah blah blah with this and compile with blah blah blah you'll see the Commodore 64 is really blah blah blah.
Enough already.
"Music is everybody's possession. It's only publishers who think that people own it." - John Lennon.
Not very well...
Mostly because of reduced memory bandwith...
HI O WISE PRINCE. WHT TOOK U SO DAM LONG?
and now have egg on their face for it.
I'm a happy 17" powerbook user, I like Apple's products, but they've never been a perfect company. This incident is an example of that.
Frankly, I'm a bit surprised, Steve Jobs is a very smart guy - he knows that the Internet works this way, that this benchmark would be studied and exposed as unacceptably flawed. Of course, all cross-architecture benchmarks are controversial, but intentional crippling is a bit much. All that needs to happen now is a posting on News.Com and Apple has a PR fiasco.
The first order of business, perhaps, could be to take down the SPEC benchmarks from their marketing web pages and focus on the performance increase beyond the G4, which is what existing Mac users really care about.
Beyond that, they better work on other ways to get Intel people to switch other than processor speed - it seems they're still trailing Intel on this curve, and will be for years.
-Stu
Correct me if I am wrong, hasn't Sun been selling its own 64-bit desktop solution, Blade 1000, for several years?
The reaction to Apple's announcement shows that it hit the mark very well.
Suddenly, the Wintel world is on the defensive and they will grab at all sorts of straws to explain away the news - just like the Apple fans did when they were on the defensive until yesterday.
We will hear about flawed benchmarking, false advertising(like Intel did not do it the MgHz myth), pricing manipulation(like no one else does it), and so on. The reality is... evolution. The 32 bit replace the 16 bit and gradually the 32 bit will give way to the 64 bit.
I am sure that no Wintel supporter will change their mind over the G5 claims but they will hurry to buy the next Wintel machine sporting a 64-bit capability.
Cheers mates, thanks for making my day.
Well, since a 64 bit machine can address 8 TBytes natively using a 44 bit addressing and up to 2 Exabytes with 64 bit addressing, I don't think 8GBytes will be that big of a speedbump
Food not Bombs is a nice platitude but it breaks down when you notice that the Bombees are usually well fed
Apple used gcc, which did give the 970 an advantage - but is hardly 'tweaked' for the Mac. I'd like to see IBM's own 970 numbers, using a compiler optimized for POWER/PPC in the same way ICC is optimized for x86. I haven't checked them personally, but somebody in the G5 announcement article has, and pointed out (too lazy to link up, sorry) that if you look at both optimized sets of benchmarks, the magnitude grows, but the relationship stays pretty much the same.
Anyway, as a longtime Mac guy, it's nice to see some LEGITIMATE argument about who's fastest, and it gives me a warm feeling deep down inside to know that the *initial* release of the 970 is this fast - I look forward with much anticipation to what IBM does with this chip in the next 12-18 months (coincincidentally the practical timeframe for replacing my dual 1.25, which suddenly doesn't have nearly the appeal it used to...)
Facts do not cease to exist because they are ignored. - Aldous Huxley
Sorry, ``buttons'' was a poor choice of words. Probably should've said ``controls / widgets''.
As an example of what I mean, I've pried the faceplate off of my PowerMac G4 (Yikes! 400MHz to be exact) and that reveals:
CD-ROM
- headphone jack
- volume control
- eject button (I really hated the clunky drop-down door)
- access light
Zip drive
- eject button
- access light
All but the eject button on the CD-ROM were hidden---and I've found the Zip drive's eject button invaluable for ejecting bad Zip disks infected with the ``click of death''
William
Sphinx of black quartz, judge my vow.
You mean to tell me you guys actually thought material that starts with "As PC users know only too well" [http://www.apple.com/powermac/architecture.html] is credible? It reeks of deceit and lies.
People, people. Fruits are not faster than Pentiums. Don't believe the hype!
Must-not-watch TV!
I know people are all ga-ga over the "G5" CPU and all of it's 64-bit capabilities, but the main thing for a LOT of people that will make these new Macs a lot more attractive is the rest of the system. We finally get a decent system bus, PCI-X and Serial ATA. Imagine how much this is going to better music and video production, being able to shove data in and out of various sub-systems that much more quickly. Top that off now with the CPU, and these machines DO smoke !!
...
;-)
The other issue raised in the article is "most user" tasks depend on integer performance, rather than FP. Well, at our institution (where we do medical imaging), we like (rather NEED) that FP rating. Right now, the system of choice has been Dell PCs, that people set up on their own with RedHat or Mandrake in dual-boot. We used to buy ASL, but have had just too many component failures to keep buying them. What people do is they do their processing and analysis in Linux (FP) and reboot or VMware into Windows to do their papers and presentations. So while we are not a huge market, the reasearch community is NOT a small one, and we are less price-sensitive than home consumers or industry. We need hardware to do what we need it to do, and we don't mind forking out the extra $$ to get it.
These new machines are better for us on two fronts. We can now do our analysis on a single OS and not have to bother about 2 OSes on our systems and also, we don't have to worry about setting up Linux, and the compatibility issues with hardware, etc
So when the time comes, I'll be asking my boss for one of these
Using SIMD instructions is a huge part of the performance of modern processors. Heck, even using SSE2 for scalar arithmetic is faster than using the old 387 instructions.
Saying "disable SIMD -- it's an optimization" is almost like saying "don't use shift instructions for power-of-two multiplies -- it's an optimization". Or "don't keep loop variables in registers -- it's an optimization." If the compiler can do it without weird tweaky flags turned on, then let it be done, i say!
The following sentence is true. The preceding sentence was false.
You are correct in that modern operating system kernels attempt to take advantage of that fact that there is more than one processor in an SMP box. However, applications written for SMP machines take more advantage of more than one processor than the kernel simply selecting the processor affinity of particular thread (or process).
Apple chose processor intensive benchmarks in its suite of testing (which makes perfect since). Applications that are processor intensive and SMP-aware usually are designed to calculate things in parallel (spawning N threads on N processors to calculate an end result). Applications that are not processor aware will usually spawn 1 thread on N processors - this makes perfect sense.
It is very true that more than one processor helps take the load off the system. I/O, music, background threads, etc. all consume processor time. Modern kernels know how to distribute the threads to the processors for maximum responsiveness. But 8 processors does not make a calculator produce a result faster if it is not SMP-aware.
Apple's disabling of hyperthreading takes the edge away from Intel that it worked hard in implementing for the Pentium 4. Hyperthreading for the modern Pentium 4s (3.06GHz+) and the P4-Xeons are considered "standard equipment". Although minor issues have cropped up with hyperthreading in the Windows world, Intel's logical processor implementation is excellent and considered stable for production use.
For the scenario you provided, a single HT-enabled P4 would not receive much benefit for background threads (such as I/O, for example) since it still only has one physical processor layer - just two perfectly copied states of the physical layer. As such, no performance advantage would be necessary gained.
However, SMP-aware applications that parallelize their calculation to the number of processors in a machine would do so in a manner that the P4 could implement the instructions in a more parallel fashion on the physical processor layer, thus increasing performance (albeit marginal). But even still, this marginal performance increase makes the modern P4s what they are. Intel would not have included the technology had they not wanted it to be a part of their processor.
1 HT-enabled P4 would still be slower than 2 HT-disabled P4s.
The yester-year equivalent (roughly) would be Apple comparing a floating point benchmark with their 68040 processor with a FPU-disabled 486DX processor.
Ayup
You, sir have just been marked as my friend :)
Apple tweaks the specs to make their system look faster.
Dell tweaks the specs to make their system look faster.
Apple does not use Dell's tweaked specs but instead chooses to cripple Dell's machine for benchmarking purposes. Dell did not cripple an Apple machine for comparison.
Any questions?
Ayup
Assuming it's bang/buck, and buck = 0, then bang/buck is Undefined. (division by zero!)
Facts do not cease to exist because they are ignored. - Aldous Huxley
I find this "article" highly dubious... and am amazed slashdot let this slide... actualy im not... it seems /. is intent on fueling the platform war fires for more hits and such, because this was blatant flmae-bait trash.
While it has been discussed that the differences between GCC and ICC(??) on the two platforms makes for a supposed im-balance, the subjects are highly subjective. The main grip is that it is extremely hard to determine what IS a fair test of raw CPU power. This guy continuously claims that the number one score to rate by should be the specint_2000 test (which favors intel considerably) because most people will "never" use FP...
that is a pile of bullshit... aqua alone (doing all of your interface goodies) must use FP constantly, in fact i would dare say that FP is constant and highly important on OSX, it contributes immensly to the look and feel of it. dual-processing is NOT an exclusive to specialy coded programs. That was true to OS 9, but more recent builds split tasks far better than before (although nowhere near a perfect balence).... and so buying a dual machine isnt a pointless expenditure.
the guy claims that you can buy a dell at 3.06 for less than 2.5k with ALL of the equivilent features of the powermac..... i encourage you ALL to go to dell.com and config a dell with a DVD-r along with firewire, serial-ata (cant do it), bluetooth, 1gig ethernet, 802.11g, PCI-x (cant do it), 8gb RAM cpacity (cant do it) etc.... Now you cant add some of these features from dell... which will send some of you off to www.pricewatch.com to find the ABSOLUTE rock bottom prices on the extra parts and pieces... but you can all see the price difference for an equivilent system is tiny if not reversed.
if someone could link me to a PC with equivlient features (pci-x, agp 8x, 802.11g, 1gig ethernet, serial-ata, p4 3.06, bluetooth, etc...) i would love to see honest comparisons of the price. The 8gb limit simply can't be done, so the value of it is hard to quantify, as a music guy myself i can actualy see it's value, but many people can effectivly argue it is above and beyond current consumer demands. but current consumers arent looking for pro boxes.
my main point of contention is that there is too much LACK of knowledge in regards to spec and architecture. We don't know if these tests are balenced or not, or why they were performed however they were performed. This may have been the best way to compare the two on a bit for bit level, or maybe not. Apple has it's own intrests, how solid is veritests credibility? did money change hands? how much? etc.... GCC has been reputed to being apples baby as far as compliing.... but does that mean it was fully optimized towards PPC, and the vice versa agasint the PC?
What we DO know is that the PPC is scalable in MP configs, it is a FP monster, and the P4 is NOT MP (hence the xeon tests, which really are not consumer class chips) and the P4 is an INT monster. I expected the P4 to smoke the G5 on int, and it didn't (smoke it, just beat it well). As well i find it laughable that the first fab of a new chip (970) running 30% slower (in mhz) performs so well against the top p4. The G5 has come out the gate on a level of extreme competitivness and has a significant space to grow in... 3ghz in under 12 months? putting the G5 at 3ghz about when the p4 will hit 4ghz. The speed tests there will probably favor the G5 FAR MORE than these.
One last point, we can say much about how these tests are scewed towards apple in that the Compiler along with the "cheats" all seem to point towards manipulation.... but at the same time two HUGE benefits of the G5 arent being quantified either.... 64 bits AND 8gb of RAM are huge advantages that if fully utilized would lend such a huge performance gap it isnt funny (look at the genome matching with 40 bit words). So in a sense maybe these tests were balanced.
Too many variables to call this one, the G5 i suspect is faster on FP, but not on int (as apple indicated)... but
--Idiots, Every single one of YOU, A flaming mass of conglomerated morons, hey wait a second, isnt that how RAID works?
He specifically acnowledges that the Dell numbers are boosted. They are boosted by means of using the Intel compiler, which produces extremely efficient code. Well, this would seem to make sense. If youa re going to test two systems against each other, you would want to use the most optimised code in both cases.
However, supposing you don't want to use Intel's compiler for some reason, you should then at least do an equal compilation. Use the same revision of GCC with the same options on both systems. If you want to use SIMD (AltiVec/SSE2) do it on BOTH system. If not, don't use it on either.
Thing about SPEC is that it is a highly compiler dependent benchmark since it is distributed as source. Taht is actually part of the idea, to be able to bench compiler efficency. So, when benching hardware you need to decide how you want to do it. The accepted standard for the stuff that gets submitted to SPEC most of the time is to use the compiler that does the best job. In the case of Intel chips it is, unsupprisingly, Intel's own compiler.
Now this isn't a cheat, Intel's compiler doesn't do anything unfair. It's not like it cuts corners so the code won't run or has bugs or anything. It simply does a better job at optimization than GCC does.
So again, if you want a fiar benchmark you either need to do best optimisation in both cases or as close to equal as possable optimisation in both cases.
If you visit the Apple store, click on the big advertisement on the center to select your PowerMac G5. In the upper-right part of the screen it states:
"Just how fast? Get the proof here.". Following this link will take you to Apple's own site where you can read details about the benchmark.
What's missing?
The comparison between G4-optimized benchmarks and the current G5-optimized benchmarks.
Ayup
The idea that an FPU is totally useless is untrue. glibc will use the large FPU registers for optimized memory copies if they are not being used for real floating point operations.
I work on an embedded PowerPC product that has no FPU so we had to build a special glibc that does not use the FPU registers.
I don't know if Apple is using a glibc with these optimizations. If they are then their customers could appreciate some use of the dual FPU cores in everyday integer types of computing.
If you open contextual menus on a .app there is a menu item called 'Show Package Contents' to step inside the .app folder and inspect it.
Okay, I did RTFA, and well, as boring and tedious as benchmark tests result analysis is to me, I generally glossed right over it. What I found to be the most compelling part of this article was the "hate mail" section at the bottom. I read each one, and came away scratching my head.
I simply do not understand how people can be so consumed with obvious hatred for another person debating COMPUTERS! Why do Mac users feel so threatened? Why do Linux users feel so threatened? Why do Microsoft users feel so threatened?
I've been using Macs (since 92), Windows (since 93) and Linux (since 96) and FreeBSD (since 96) for years and well, I have yet to find anything about these systems that demand that I stand up and scream at the top of my lungs how wonderful any of them are, and to attack with such spiteful hate those who don't just fall in line.
Having started out in the computer world as a designer, I used Macs. I like them, they are cute, and fun and make many things easy. They are also slow, crash a lot and the cause of a lot of frustration. I started using Windows (3.0) because I wanted a PC, but couldn't afford a Mac at the time. Windows was cool, it crashed a lot, and I had the hardest time trying to configure hardware with it, but I got the job done. I was introduced to Linux looking for a way to get up to speed with Unix. I had a hell of time first installing it, it was cool, seemed very powerful (I was in over my head) and never crashed. Same with FreeBSD. But I still have yet to understand the mindset required to say things like: "This guy is an idiot, and his article should be pulled and his email box should be flamed."
or:
"I can't believe the haxial web site is still up, you would think by now someone would have hacked it."
Good grief, what is WRONG with people???
A while back I chose Linux as my primary OS for my day-to-day computing, on an Intel chip. I love it, its fun, its cute (thanks KDE) and it hardly crashes, and low and behold, I get my work done. My girlfriend (she's a designer) has a few Macs. I like them, but, well, it doesn't feel right to me so I stick with Linux. Sure, we get into our little OSX vs. Linux debates, but it never gets down to where she threatens my life and I launch DoS attacks on her machine. They usually end as "we should all just go back to Amiga" or something like that.
I would love to ask someone who is so delluded in their thinking to feel real hatred for someone who simply prefers not to use the computer/os/whatever that they use, what exactly do they have to fear? Why the need to act like a savage? Is it just because they are posting in a message board, and well, its time to be macho, because its safe and anonymous, and well, the need to act the tough-guy just overwelms better judgement?
sad robot making broken music
64bit isn't just about greater than 2gigs memory either. The availability of 64bit machines, for instance, will permanently solve a lot of nigly little programming problems such as the reams of software that use time_t for time definitions.
Also, good filesystems already use 64bit addressing.
64bit will matter for graphics and video too in a big way: all your doubles will now nicely fit in a register which will make for much nicer machine code.
Do not spread "09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0" over the internet, thank you.
This simple method is how they could make "top of the line, maxxed out" G5 vs. the top Dell workstation face-off rock:
Run a CAD app (or DB or whatever) with a seven gigabyte dataset.
G5 100x faster, 200x? Would the PC even finish it?
Duh. Intel knows everything about Intel chips. They designed them.
Intel's ICC won't produce code nearly as good on AMDs, and won't produce anything on non x86.
Let's not go around talking about how gcc sucks because it doesn't -- and can't, and never will be able to, unless Intel opens up all of the specs -- compete with Intel's ICC.
GCC is designed to compile code on many different platforms, to unite development efforts as much as possible accross different CPU types.
social sciences can never use experience to verify their statemen
That is a little wierd, although not wholly surprising. It's kind of a variant of the little man syndrome, where folks who adopt a platform not in the mainstream have to justify themselves either via inflated specs in some obscure area or, failing that, vitriol.
.sig...
Yes, except in this case, the vitriol seems to be
coming from the X86 fanboys and the PPC fanboys in
equal measure. Guys like the one that wrote the page
we're all talking about now lost a lot of sleep
last night poring over the specs in an effort to
debunk Apple's claims. Much like the Apple fans
went out of their way to defend them.
Your assertion that Mac users are the only ones
looking at little man syndrome are a bit one-sided
at best, patently absurd at worst.
It seems to me that if somebody wanted to use an inferior product, the first thing they'd do is develop a thick skin and at a minimum ignore the criticism being lobbed at their platform of choice.
A thick skin only lasts but so long when you have
people with nothing better to do but take shots at
you because of your perceived "difference". The
thing about being in the minority is that the
majority automatically thinks their better than
you, (whether its' true or not) and some of them
take every opportunity to poke you with that stick.
That, or choose to adopt something that seems to work better for the majority so that they don't have to feel left out all the time; obviously when you get to the point of chewing out people who are trying to show you why your choice is flawed it's become a popularity contest for you already (competing, not computing).
Your argument right there reveals your bias quite
plainly. Just who are you to judge what the best
platform is for me? No one. You're as much a pawn
of that popularity contest as anyone else, hell,
probably moreso, becuase you've fallen in with the
other 95%.
Perhaps you should better heed the words in your
Mod me down and I will become more powerful than you can possibly imagine...
While we may debate whether using a compiler optimized for a 64 bit computer is fair when comparing it to a 32 bit computer (but hey - isn't the 64 bits part of the attraction of this computer?) there was another section of the demonstration that used application testing to demonstrate the speed of the G5.
/.'ers will like the appearance of the new machine. Now, we just need a MacLinux if we don't want to live in BSD :-)- and the new Panther will include XWindows.
The applications included:
Adobe Photoshop - rendering a poster of Looking For Nemo - out did the Double Xenon 3 (not a single xenon ) by 2X
Adobe Acrobat - individual page rendering in a long document. Outdid the Dell by about 2
Mathematica - which relies heavily on FP calculations. Fractal or pattern creation. Outdid the Dell.
Music Playing/Rendering - Score by the composer of the music to the Matrix. The Dell hiccuped and then choked totally. I was quite surprised to see this happen and don't exactly know how or why it happenned.
Studio Quality Video - codec made for Quicktime. This will be beyond what most of us need or use. Microsoft has gone far with the new media player. This is for a niche audience.
There is much that Geeky people do that requires FP calculation. There is much that arty people do that requires graphics, music, video
While one would expect application benchmarks to favor the platform - these were all very impressive. And reality is usually more important than just looking at a set of numbers that does not represent what use you will be making of the machine.
A combination of the numerical benchmarks and the application benchmarks are what is needed. There will be more on the way.
And - I think most of us
Aside: There was a funny scene in the keynote with a video of a stealthy looking black panther slinking through the jungle - then a comparison to the competitor - a Long Horn chewing its cud in a grassy field with Home on The Range playing.
Conclusion: Don't rule this computer out. The architecture is interesting. We've been waiting for a 64 bit machine. This one has a choice of impressive graphics cards. The limiting areas seem to be the memory and disk speed. Next, we need to see how it plays games.
All of the talk about benchmarks is basically noise to me. I could care less about the tests that ALL companies in this industry run. Here are some suggested benchmark tests that WOULD matter to me (an average user):
- Time it takes to order the computer.
- Time it takes for computer to arrive.
- Time it takes to unpack and set up hardware.
- Time it takes to first boot up and configure for use.
- Time it takes to install standard applications (MS Office, Netscape Communicator, or Kazaa for example).
- Time it takes to start up said applications on subsequent uses.
- Time it takes to access removable storage devices (like CDs or DVDs) for directory listings, opening files and playing media.
- Time it takes to shut down machine.
- Time it takes to "restart" machine.
- Time it takes to change users.
These are some basic tasks that most average users perform day to day. The configuration of the machine mentioned in step 4 should be basic, with no "options" that the average user wouldn't know to use. I have no idea who would win these tests. Now, I don't know if this is the market that Apple is going after, it's just me.More important that speed is what I can run. Can I run my favorite games? My favorite browser? My favorite office applications? Apple does fairly well here except in the game category (at least last I checked... which as an average user was a while ago).
Come play Moral Decay!
For what it's worth, at WWDC the Blizzard guys said that Warcraft 3 was running around twice as fast on the dual 2GHz G5 as on a dual 1.4GHz G4, and that it ran out of the box. So, for this single data point, performance scaled faster than clock rate, which is a nice change for all of us who experienced the P3 to P4 transition.
Running our app in W2K/VMWare(single threaded) on Linux on a dual PIII/800Mhz was faster than running our app on a native P4/1.3Ghz.
Joe
Joe Batt Solid Design
I can understand wanting to state your opinion on something. I can understand doing something to state that opinion. But I have a hard time reading an article when his whole point is to show how justified he is in being upset at apple "Using publicly available information, I am going to show you how Apple is attempting to deliberately mislead its loyal customers and fans when it claims that "The Power Mac G5 is the world's fastest personal computer"". I would like to read a truly objective article on this subject. But this article is hard for me to take with any seriousness. What you do speaks so loudly that I cannot hear what you say. â" Ralph Waldo Emerson. 1803-1882
A lot of people seem to be missing his point here, saying that "Well, AMD and Intel probably manipulated their results too, but he accepts those." His point isn't that Apple optimized the benchmark so that their system would perform well, his point is that they crippled the competition, turning off important new features. There is no doubt that AMD and Intel had every optimization turned on when they did their tests, and that's fine. The problem that he raises is that Apple disabled the competition in their own tests.
I went from an Atari 800 with a Atari 410 Cassette drive to a Compaq Plus with a Floppy Drive & a 10MB Hard Disk Drive.
I refused to use the hard drive for like six months because floppy disk was more than enough for me.
Of course, I was like eight at the time too and had no real idea what a system running from a hard disk would perform like...
As a rock-in-roll Physicist once said, No matter where you go, there you are.
The reality of benchmarks is that they are almost completely useless.
Even if you could make all variables identical on each computer (RAM, HD, compiler speed etc), it's going to whip out some arbitrary number that is larger or smaller than your competitor. But that number is based on the set of operations defined in the benchmarking tool
In daily use, your mix of operations will vary. It may be similar to the benchmark's set, it may not be.
If you do a ton of database and 3D rendering, your FP speed is going to be much more important than your integer scores, as opposed to a person who does a lot of programming.
Why do you think Apple always does Photoshop tests in their introduction presentations?
and
These are the same reasons they show movie compressions, audio programs etc.
For many Mac users, these are more important than arbitrary benchmarks.
"Reality is merely an illusion, albeit a very persistent one " -Albert Einstein
How much does it cost to add an extra Intel 3GHz CPU to your personal desktop? Certainly not enough to make up the difference between the Dell's and Apple's 2CPU computer.
In terms of price/performance, x86's are still the best. You just need to add and extra CPU for ?300? bucks (I haven't kept up to date on CPU prcies). So, why would anyone who wants CPU-performance waste the money on a Mac?
All of this obsession with CPU-performance is pretty lame, in my opinion. My 1.1GHz is still plenty fast. More important than a fast CPU is a fast hard drive. Most wait-time is waiting for programs to load, since most ordinary uses of a computer aren't CPU-intensive. And of course RAM.
Spend your money getting faster hard-drives (e.g., 10,000rpm ATA-166 hard-drives) and faster RAM (e.g., DDR RAM).
If your a gamer, don't be fooled by the CPU-obsession. GPU's is where gaming performance is at. Getting a CPU twice as fast might increase your fps by 2 frames per ssecond -- for another $100 bucks. If you're a multi-media person, again, the graphics card (GPU) is where it's at. If you're a casual or amatuer, you can just get the gamer-line GPUs. If you need perfect quality, you'll probably want the QUADRO GeForces.
The only people who really *need* CPUs faster than 1GHz are people who do a lot of number-crunching. Usually scientists. And maybe people who compile their own software, if you want it to compile faster (though the whole point of compiling yourself is to get better performance without having to upgrade your CPU). A better thing to do if you want to compile your own software (e.g., if you use Debian, *BSD, or are a developer), would be to find a high-end *nix computer that you can use to compile it on, with options for your computer.
Don't buy the GHz hype. More GHz will not make your programs load faster, and will most certainly not make your computer much more responsive.
social sciences can never use experience to verify their statemen
-- sarcasm on --
Guess what everyone! I just found out that sales, marketing, and advertising departments are PAID to present corporate products in the BEST POSSIBLE LIGHT, even if that means not telling the reader all the facts! Who could have possibly guessed?
-- sarcasm off --
OK, for those of you who don't already know this: There are a lot of caveats and unlevel playing fields in the world of marketing, advertising, and sales. Companies pay these people to sell product, even if it isn't the best or fastest or whatever by whatever standard someone likes to choose. Every company does this -- whether they're selling computers or trucks or detergent. This is not news.
So Apple presents the benchmarks in the best possible light and puts a lot of fine print in to cover themselves. Who cares? The benchmarks are only vaguely interesting in an academic way. The only real way to know how the system will perform is to do your own homework.
You don't like being lied too by corporations?
let's see.
do you:
A) not drink beer because gigantic parties with tons of super attractive women don't show up scantily clad happy to dry hump dance all over you?
b) not drive a car, because you don't get that 'feeling' that Toyota claims or you can't go winding down a canyon road to cure your hiccups without getting arrested or flying off the road abecause the suspension on the car can't do what it did in the commercial?
c) not live in a house because you can't get a mortgage rate the rates advertised because your credit rating isn't in the right phase of the moon to get the right score to get the super low rate that made you call them in the first place?
Find me a corporation that doesn't mislead you, and we'll find one that goes out of business.
Lying is a strong word, misleading would be a better one, or even misdirection. That's what marketing is. Creating a non-existant need for you to want their product.
As a rock-in-roll Physicist once said, No matter where you go, there you are.
The reason people care is when you go to the Apple site there is a big headline that the Apple G5 is the world's fastest desktop computer, when in fact any way you cut it, it is not. If any other company pulled this kind of shit it would be ridiculed in a minute, but Apple abuses the loyalty of its users. As someone who uses macs occasionaly at work and home, I like the product but hate the BS tactics of the company and stupidity of a small but vocal portion of its user base.
Well, WebSphere Studio reccomends 768MB. I also need to run test browsers (Moz & Konq), Konsole, gimp, gaim (inter-office messaging), etc. In order to avoid paging like crazy, it would be nice to have enough ram to keep everything in physical memory.
Back before System 7, I used to load entire applications into ram (I had 4mb in a Mac Plus) to improve system speed. In portable enviroments (PB 165c) I used to do the same (under System 7) to conserve battery life by not spinning the HD at all.
Additionally, tasks like DVD authoring would benefit from enough ram to hold most of the DVD in memory at once.
RAM is good
t'nera semordnilap
A lot of people argue about whether PC's have more bang for your buck than a mac. I happen to agree with them, but I still own two macs (a 7 year desktop and a 1 year old laptop) mostly because I find the mac operating system more open to finding and fixing problems, and I happen to love how beautiful OS 10's quartz extreme works for displaying 2D graphics.
I have felt for years that a lot of people's problems often stem from the fact that money enters into decisions that they make. Consequently, whenever possible I try to leave the cost out of the decision. Once a decision is made, I this do whatever it takes to make that decision happen.
Now, I'm fortunate in that I make enough money to be able to afford a new mac every couple of years or so (not so for many, I realize), so I have been lucky in that I could decide on a computer and not have price be that much of an issue. Assuming that price did not come into your decision, would people prefer a PC or a mac? Let's assume that for some reason you had a rich uncle or aunt that offered to spend a couple of thousand dollars on you to buy your dream machine. What would it be? I'm not that experienced with PC's, but I'm thinking that the following might factor into your decision:
I'm curious to hear what people think.
-Dan
That's true, but for decades Apple has had the perception in some circles as a company that does things "differently." Hell, that's their whole reason for existence, right? So I think there are people who will be genuinely surprised to discover that Apple has so THOROUGHLY cheated. I mean, this makes nVidia look like saints.
No, I am not a Mac zealot who thinks that Intel or Gates, or whoever it is that day is the devil.
I think that's obvious from the fact you aren't threatening to kill the guy who wrote the article. Evidently unlike most of the people writing him email. Seriously, what's with these people coming so unglued?
-Looking for a job as a materials chemist or multivariat
That won't be a problem to achieve.
Power of Linux desktop will be mainly in corporation use at first, and Apple is just too expensive to be used for that purpose.
But bussiness desktop is still a desktop, so... Desktop market counts home and company desktop computers.
Home desktop computer figures would be different probably, but that second opinion raises question, if it has some sense to use different OS at home and at office. Second favorable look in that direction is Joe User. Joe User likes system that he's used to. And if that's Win then Win, Linux then Linux, OSX then OSX.
Signature Pro version 1.13.2-3 release 83.5 beta3try7 after-breakfast edition
Which is exactly why I wish mac zealots would get less offended when it's pointed out that Macs consistently benchmark slower.
I use all three major platforms, myself. I like the Mac's software support and desktop, I use Win2k for gaming, and Linux is where my development work happens because of the plethora of free tools.
Sure, in raw clocks the Linux/Win2k box outperforms the Mac. But Mac has other things going for it.
So stop trying to claim you're faster in raw speed. You're wrong, and more importantly, your fans don't really care.
"America has done some terrible things. But I know that Americans don't cheer when innocents die." -Dave Barry
Atleast if they 'cheat' by making GCC much better for their G5 chipts (they haven't), we can not only see exactly what they did, but those changes also become part of GCC. It's a double win.
I really think that using GCC is the way to do. If the video card companies had to do something like this maybe we wouldn't be wondering which video card is actually faster now...
So yesterday I was modded as Flaimbait for suggesting the new Mac wasn't actually the fastest desktop in the world.
So here's a big Told-You-So to the modders.
And why is this benchmark cheating not a surprise? They've been fudging facts for years.
Now what gets me is why companies do this? Don't they know they'll be discovered? Unbelievable.
.sigs are for post^Hers.
...right up until he wrote about how Apple and Dell are "misleading" customers by pricing their their boxes at $x-1 instead of $x. As if their sales really are dependent on nitwits saying "hot damn, I can get that machine for under $x! I never would've bought it for that much!" In short, his whining about a "marketing ploy" is a complaint about a pricing policy so common that, well, almost everyone, from computer companies to car companies to the corner grocer selling apples for 99 cents a pound are "misleading the public". And if people can be so mislead by a one cent/dollar reduction in price, who would be intelligent enough to read his review?
He also casually mentions that "most people use Integer (not FP) most of the time. Therefore, integer results (SPECint) are much more important than floating-point results (SPECfp)." What qualifies as "most people" using integer "most of the time"? Doesn't he have any data on FP usage?
And was I the only person that noticed almost all the results he posted had a caveat that such a benchmark "is a single-processor test, so in the following results, where the computer has a second processor, it is either disabled or not used."? Aren't these dual processor Apples we're comparing with single processor Oranges? (sorry, couldn't resist.) It might make sense if Apple actually ships machines with useless second processors where architecture and OS make them essentially uniprocessor machines. But if Apple does indeed sell multiprocessor machines, and I understand it, they are, shouldn't that be taken into account? What I read was not that Apple claimed that the PPC 970 is the fastest chip, but rather the dual processor G5 is the fastest desktop computer.
About all he convinced me of is that Apple perhaps twisting benchmarks for their own ends. But his review is hardly a clear and unambiguous refutation of Apple's statements.
B
"I'm payin' taxes, but what am I buyin'?" -- James Brown
What amazes me is that people don't know that the SPEC database of results exists for a reason - they are the only OFFICIAL results SPEC recognize. To get your figures posted, you must follow the so-called "reportable run" guidelines which dictate what you can and cannot do with compilers, etc.
Until SPECint/fp figures for these new macs appear on http://www.spec.org, take whatever you hear from ANY vendor (G4, G5, P4, Xeon or whatever) with a BIG apple-sized grain of salt.
in the most favorable light (for Apple). What is the world coming to?
;-)
Better go tell your nearest open source or free software advocate that they had better change their tactics. After all, using the words (in all their senses) open and/or free is a form of marketing...
Seriously, whether you agree with the benchmarks or not, hate or love Apple, at least Apple has included some transparency in the process. That's more than most companies are willing do to these days. We don't want more cases of Benchmarks In, Gospel Out.
"All the darkness in the world can not quench the light of one small candle."
would be to have each vendor pick their machine, optimize it to their hearts content software and hardware wise. Then load 3-5 common apps like Photoshop, MS Office, Illustrator, Premier, etc. Working with the same graphics or test files on both platforms, run them through their paces. Just to spice things up, have the persons doing the testing be an "expert" for each platform - a power user that prefers a PC and a power user that prefers a Mac.
Then see how long it takes each to accomplish 1) particular tasks like color rendering a photo, and 2) the total overall time it takes to complete all the tasks.
This is the kind of "benchmarking" that exists in the real world and I want to see a test of that. No whining about "Altivec" or "SSE/SSE2" - it means nothing to me and won't mean anything to me unless the sofware I'm using happens to take advantage of it.
To me, this eliminates the "But they left out BLAH optimization and disabled BLAH functionality that REALLY speeds things up!" If the vendor can't figure out how to optimize their own hardware, then that's something you have to consider when deciding your purchase - along with the final cost of all the "mods" and "extras".
THOSE are the results that would impress me more than some vague reference to floating point vs. integer performance.
blue
Once again, math can be used to show that 2ghz 3ghz.
Nope, I'm saying the previous AC poster knows nothing about compilers or much about benchmarking.
I'm not a zealot, whether or not the Mac "kicks ass computationally as well as in all the other ways" doesn't give me wood.
However, I am convinced by the evidence that Apple's benchmarks are flawed and, as such, are pure marketing, typical of what we've seen in the past.
...using a G5 that fell off the back of a truck and the latest computer from Dell. Borrowing Apple's technique of "tinkering" with the systems, I optimized the Dell system to it's highest level of performance. I made only a single modification to the Apple system: removing its power cord.
Interesting enough, the Dell system matched the numbers found on the SPEC website, but the G5 was unable to complete the benchmark.
I think that this test, which can be easily duplicated, shows conclusively that Apple's G5 marketing is a complete lie.
144l. ph34r my 133t l3g4l 5k1lz!
I've seen no good explanation to explain how if the Xeon is really so much faster, the app demos were so much better on the G5. Obviously there are a lot of things like internal bandwidth that help out the apps - but isn't that what we should be really concerned with? Instead of comparing compiler writers to see who can wring the most out of a CPU alone.
The reality is that we need to wait until the machines are shipping and then evaluate them, to get a more objective picture. The Dell benchmarks are as full of exaggeration as the Apple ones, and until you compare real world apps you can't really say which is faster or by how much. Until then the application demos alone are compelling enough to suggest the G5 really is a fast machine.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
What I find funny is that most of us that read /. have BOTH systems available to us (maybe at a friend's place, maybe in an adjascent office), and can recognize that a machine's OS and hardware play second fiddle to the EPERTISE of the person using it. I can guarantee I could give a top-end Windows, Linux, OR Mac machine to my father and any one of them would be on fire within days.
-Lucas
So that is why there was no Athlon benchmarks in the comparison -- because the athlon does not care what compiler you use.
"Fighting terrorists with millitary might is like killing a mosquitor on your Dad's forehead with a rifle."
The emagic comparison seems a little fishy though
Not sure if it's on the site, or just on the tech specs pdf (where I'm pretty sure I saw it), but the audio comparison is really bugged out. First of all, the title bar says "Logic vs. Cubase," but we're not comparing apps, we're comparing procs. Does that mean they're running Logic on the Macs and Cubase on the PCs? If so, the comparison is absolutely meaningless.
Then they say "with 5 plug-ins." Which ones? Which brands? Waves? Different plug-ins take up different amounts of proc usage. I can only do one high-quality reverb at a time, but lots of eq's. And were there 5 plug-in on each track, or 5 plugins that each track was routed through?
Then their results: 52 track, 115 tracks, etc. That's a lot of tracks. But again, meaningless.
A good benchmark would be:
Digidesign's Pro Tools Free, no audio hardware except what came with the computer. Record 8 tracks of 24-bit/96khz audio (or whatever). Then pile the plugins on, and use the exact same ones in the exact same places on each machine. Let us know when the interface gets sluggish, and then when it craps out. Also let us know how it craps out. Crash? Doesn't respond? But for crying out loud, don't give us worthless numbers.
c-hack.com |
I agree in general that this guy's reputation does not make his point invalid, but I can't resist bringing up this soapbox about Infinity and the Universe. It's totally hysterical. Among other things, it refutes Einstein's theory of relativity and the law of conservation of energy. It's got absolutely nothing whatsoever to do with the topic in hand, but if you have a basic knowledge of physics or mathematics, it's a great laugh.
All I want is a secure system where it's easy to do anything I want. Is that too much to ask ~~ Randall Munroe
More to the point, Intel drive their performance forward on two fronts, architecture design and clock speed. While the G5 is a strong contender from an architecture stand point, that still isn't enough on it's own over throw Intel's crown.
Fortunately, if Steve J's prediction of 3GHz in a year are on the money, the new G5 systems will be considerably more competitive on clock speed than their predecessor, complementing the G5's design strengths even further.
Given that Intel aren't planning any more speed bumps to the Pentium 4 for another year (http://www.theinquirer.net/?article=10121) until the Pentium 5 emerges, they could find themselves playing second fiddle to the G5 on anyone's benchmarks within 12 months from now.
Should be interesting.
Macka
[n/t]
Black holes are where the Matrix raised SIGFPE
"Only in their dreams can men truly be free 'twas always thus, and always thus will be."
--Tom Schulman
He's not going to post every single mail he gets publically.
He's just going to post the stuff that really stands out.
retrorocket.o not found, launch anyway?
Umm... maybe read this whole thread and the posted article? It pretty much states why the Apple benchmarks are flawed.
But... since Mac zealots have problems reading information that casts their beloved machines into non-favorable light or speaks disparagingly of their their beloved Temple Apple, here are but a few (summarizing the original link and pretty much these ~1000 forum posts):
- explicitly disabling SSE2 optimizations on FPU intensive benchmarks for the x86
- using gcc instead of the better Intel compilers for the x86 while using an optimized G5 gcc compiler
- using a speed optimized malloc libraries for the G5 tests but using the standard malloc libraries for the Intel tests
At work, we just bechmarked the Dell systems a month ago and got very similar results to Apple for the "base" rate. The article seems to be quoting the "peak" rate for the Dells. It's not valid to compare peak rates yet because gcc 3.3 and os 10.3 aren't really fully optimized yet.
The article also complains that using the NAGWare compilers is not a valid test since they're too slow. But I think the NAGWare compiler is a more vallid comparison than intel's compiler because most real-world computing is done with NAGWare because it fully implements the F95 spec and is more portable. In addition NAGWare is well tested for accuracy and it also very much cheaper.
The Dell benchmark numbers are pure fantasy. They never occur in real-world use.
There are 10 types of people in this world, those who can count in binary and those who can't.
I don't. However, his analysis is easily replicated and interpreted. Instead of discounting what he says because he frequently debunks Mac propaganda, why don't you attack him on the points that he made? I'd be interested in seeing that.
Additionally, being a mac user, why would it make sense for him to have it out for the mac?
-Looking for a job as a materials chemist or multivariat
The limit as x approaches 0 of n/x is inifinity.
No, it isn't. Only one side of that limit is infinity; to see the other side (assuming n is positive) try evaluating at x = -.1, -.01, -.001, etc. Then the limit is negative infinity.
If you try to put those both together into a normal limit expression, you can form series with no upper bound, series with no lower bound, and series with neither. It doesn't get much more undefined than that.
Is this guy serious when he says that the number of Floating Point Operations is irreleveant.
Hello genious, super computers are measored in FLOPS (Floating Point Operations).
Not all conservatives are stupid,
but it is true that most stupid people are conservative.
- Hume
I just read the veritest document myself. This guy did not show the complete picture at all. First, in favour of Apple, these tests were run on Mac OS X 10.2.7. I don't think this is the 64-bit kernel that can really take advantage of the G5. I'm not a mac head, but I'm guessing this is Smeagol?
So right from the start the G5 is seriously crippled in these tests. Especially if they don't even take advantage of 64-bit as seems to be the case. Now, on the the other points that the spl dude makes.
Sure the special malloc library seems a bit unfair. But then again, do these tests really focus on memory allocation? I'd think they are limited by CPU power more than memory allocation. In any case, it'd be nice if we saw results without this library.
Now for the comfusing part. The 8300 only has a single cpu. For the base tests, they use hyperthreading and an SMP kernel. They do the exact same test for the 650's base test, hyperthreading with a single processor and an SMP kernel. The G5 system is run with a single processor as well.
So it seems this test is well balanced and fair. The confusing part is the rate tests.
For the 8300 they have no hyperthreading and a uniprocessor kernel. For the 650 they have no hyperthreading and an SMP kernel with two processors. The G5 system is run with two processors. It's unclear why they chose not to use hyperthreading on the rate test. It could be that hyperthreading actually reduced the scores of these tests. I'm no expert on the SPEC tests and hyperthreading, but what I do know is that hyperthreading is an intelligent technology. It can't always increase speed, it depends on what kind of code it's running. In the rate test it's possible that hyperthreading is unable to yeild any improvements, in which case the overhead of enabling hyperthreading may make the scores worse than without hyperthreading.
At anyrate, the tests were a LOT more fair than the dpl guy makes them out. And considering that the G5 could be seriously crippled by not running 64-bit and who knows what other optimisatoins, I'd say that the numbers are still impressive.
..because most people have apparently not read through the whole thing.
"[...] In other words, most people should ignore floating-point results because they do not use floating-point anyway (or not much)."
This is utter bullshit. Floating point is extremely important for many productivity applications--anything graphics, 3D, modelling, scientific, CAD, etc. Ignore floating point?! What the hell crack is he smoking?
The whole article is filled with this kind of fart-biting. The data are far more interesting without his stupid inane conclusions muddying the waters.
When I see something that costs $1.99, mind thinks "One hundred and ninety-nine". When I see something that costs $2.00, my mind thinks "Two". I can't imagine that I'm the only person with this reaction, either. I'd like to know if there have been any sort of rigorous studies done to show which sorts of prices people perceive as being higher.
I'm not too happy about the fact that Apple may have tweaked the benchmarks in their favor--but everyone does that now, so they're just playing "fair."
What's really important about these G5 systems is the newness of the architecture--each processor runs full speed on the system bus, which runs at 1GHz. That's very difficult to benchmark when you're just doing math, but this foundation seems much beefier & smarter to me than Intel's current offerings--8GB/sec memory rate _per processor_. Since Apple is just at the beginning of the G5 changeover, I expect we'll see huge things in this area in a very short time.
Interestingly...if you go to go to mac's store right now, they put up the G4 specs under the G5 area! This is the exact opposite of a couple days ago. If it's not there anymore, here's proof: http://us.share.geocities.com/richardsonke/G5.bmp. It's on geocities, cause i'm at work and i don't have access to my normal ftp server. Can someone mirror it? Thanks. And NO, this has NOT been photoshopped.
"Men lie."
"Yeah, about sleeping with other women, but never about bioluminescent plankton."
-Dan Brown
is it's all crap. I don;t care if your Intel, AMD, or Apple. They all lie about how fast there cpu's are and benchmarks can always be easily scewed - whether its by turning off SSE2 (as in this article) or if they change the bios settings from the default on one machine (as toms hardware said AMD recommended for their 3200+ benchmarking) - its all made up.
I'm never going to ALWAYS have the latest and greatest system (My pc is a P4 1.6 and the mac i'm using to type this is a Dual G4 500) but what makes my choice for what System to use comes down to one simple thing....
WHAT THE SOFTWARE I AM USING RUNS BETTER ON.
When I want to use telnet, ssh, ftp (cuteftp, dreamweaver), or any internet related app - I find that for my setup they seem to work better on my pc for some reason, when I want to run Photoshop (although when working with files over 80megs it seems to open faster and run filters faster on my PC) or Illustrator or even Quark they seem more responsive on my Apple.
If I had to make a choice and choose only 1 system (glad I don't have too) I would probably choose a really expensive PC (like a Dual P4 3.2ghz with HT or something) ONLY because it would be cheeper for me to build (read not purchase a Dell) than the cost of a single mac.
But in real life I don't have to make a choice or rather i've made the choce to use both - and hay I can still play Diablo on both too - although if I want to play more games like Arcanum or Final Fantasy or Neverwinter Nights i'm kinda stuck only using my PC.
Ave Molech Setting
Gee, a manufacturer (in any field) messes with the numbers to make its product look better... imagine that!
But did anyone notice that the author plugs his own business while stating the obvious?
-A
i think the most telling "benchmark" on the whole site are the mac users comments.
I realize that every community has it's crackpots, but as a long time linux user looking to go to a G5 from intel chips, it kinda makes me want to distance myself from the Mac crowd as much as possible.
I think they convey the exact type of blind fanatacism that has dissuaded mainstream computer users from linux for years.
Might want to turn the mirror on that one there - you'll find that most people using Intel processors (like me) really don't give a shit. Kind of hard to be a zealot when you don't give a shit. It's you, a mac-head with the permanant chip on your shoulder, who is using exclamation points, and referring to the ever-powerful "get a life" argument. This coming from a guy that spends probably 5x more money on computers than I do. I mean, here you are screaming because someone had the audacity to say your favorite computer isn't the fastest on the planet, and you're telling *me* to get a life? Uh-huh.
I'm still waiting for one of you mac-heads to actually analyze the evidence as opposed to going nuts, saying it's all a conspiracy by the intel "zealots." Hasn't happened yet.
For what it's worth, I actually like macs, and would own one if I could afford a decent one. But I don't like being blatantly lied to. So if you have a rational argument for why Apple's benchmarking techniques are legitimate, I'm listening. If not, well, sounds like it's time for your meds anyway.
-Looking for a job as a materials chemist or multivariat
It's not surprising to see Apple try to make their products look so wonderful. If I had a company I would want to make my products look wonderful too, especially compared to the competitor's. Now, the ethics of doing such is questionable of course, but not really shocking. Shouldn't we all have learned to take marketing with a grain of salt so to say?
You are an idiot. void* is not even an integral value.
Just tell me how fast Lightwave renders. That is where time actually equals money for me.
"Derp de derp."
I checked out VeriTest's config file for the PCs. Conclusion: they cheat! Someone needs to get this PC hardware, rerun the tests using some better opts, and debunk those fools. They make the Intel PCs use some GCC switch called -fbranch-probabilities that causes disk accesses and bogs down the program w/ debugging type crap, because it keeps a histogram/count of the result of every conditional jmp and branch and writes it to a file. The Mac conf file doesn't do that! They also use slower, smaller hard-drives (no brand given), but they use a fast sATA drive on the mac. See pp. 23,24. Talk about invalid, unfair, and biased benchmarks.
The biggest trick the devil pulled was letting lawyers become politicians so they can write the laws.
Yeah, here's a question: Where exactly did Apple tweak their specs to make their system look faster?
They didn't use the IBM compiler. They didn't use the Altivec optimization.
They used the worst case scenario for both machines and theirs came ahead.
Well, when we talk about "all optimizations on" we're really talking about real world application performance, since software vendors determine what "all optimizations on" means in practice for the software we actually use most of the time.
In this case, that means an independent comparison of Photoshop, 3D rendering, etc.
Apple's Photoshop tests suggests that the two platforms are at least about on par now, which is all that matters. This simply means that one can now choose a platform based on OS, not on performance.
That being the case, I would have no hesitation to buy a high end G5, whether or not it is slightly faster, or slightly slower than the current top of the line intel box, because I'd much rather be using Mac OS X than Windows (Linux is not an issue since it runs on both platforms).
Here are some SPEC results I googled for, commisioned from SUN on their Xeon based Fire V65x, running a single 3.06 GHz Xeon. You'll notice that they, too, disabled Hyperthreading. Obviously, Sun would have wanted these benchies to be as fast as possible. So, probably, the single thread used for SPEC scores is best suited by TURNING HYPERTHREADING OFF.
s 20 03q2/cpu2000-20030520-02193.pdf
Meaning, if Apple's results are reliable (which I think they are...levelling both machines by optimizing them for neutral operations and having them run neutral code), they tuned the Dell FOR SPEC. They didn't decrease its performance -- they probably increased it a bit.
http://www.specbench.org/osg/cpu2000/results/re
Just because you put the words "Fast" or "Hyper" in front of a chip's feature doesn't automatically make it faster, as any BIOS hacker knows.
Hey freaks: now you're ju
...or maybe not -- We'd have to check with Alanis on this one.
Anyway, thought it was funny/interesting that this comment is right under the banner ad which happened to be for Intel's compiler.
Targeting this? Compile with this!
There is much cruelty in the universe, John.
Yeah, we seem to have the tour map.
who cares how fast it is? it's a MAC !
...scores, against Opteron, Athlon, Itanium, Power4, and real SPEC scores for the P4 and Xeon. What I found was a huge gap in peformance. I find it difficult to believe Jobs, and I found that Opteron, a the true first available 64 bit workstation CPU, which is what the Power Mac G5 is, far outperforms it. Of course it has twice the cache, and the same HyperTransport. Many have raised the GCC compiler issue. I say when you benchmark something you use the best available for that platform. Apple did not, and they also did not have the guts to reveal the hardware specs of the P4 and Xeon system they tested. Jobs is smearing the PowerPC970 and Apple because of this. I find that sad, and I feel Apple is better of without him, and needs the clones back.
ignorance is bliss. googlefiberatx.com
Even if you do not have any single application that needs lots of RAM, it doesn't take too many big apps running at the same time to start a system swapping. The more RAM you have, the more apps can run without tedious system delays.
On the other hand,if you're happy running XP on 256M, then you have no reason to change. But I doubt that you run Word and Outlook and Excel and Powerpoint and IE at at the same time you're gaming.
Information is not Knowledge
"The other feature he's complaining about is the disabling of hyperthreading. From other benchmarks I've seen before, hyperthreading in SMP systems usually results in equal or slower performance, or at most a 10% addition in certain benchmarks. It was probably better to leave it off."
= 94 75
g /H yperthreading.html
Which benchmarks would those be? Please reference.
I found this one:
http://www.theinquirer.net/default.aspx?article
which is from AMD. No numbers are published in that article.
Now here is an article:
http://home.insightbb.com/~george/Hyperthreadin
that shows 0 to 60%+ improvement with hyperthreading.
Other pages seem to show an improvement of 0 to 20% with hyperthreading.
You dismiss a 10% improvement. Wbat do you think hyperthreading is?
The principle is that a processor execution core can be tied up waiting for a cache line. But, there may be live data belonging to another thread
in the cache. Instead of just waiting, the processor can be internally switched to the other thread, and can make progress in the (otherwise) dead time. A damn good idea. An interesting architecture. Explain again why this was disabled for the SMP tests.
Ratboy.
Just another "Cubible(sic) Joe" 2 17 3061
Are G5 faster than the G4 machines they replaced?
Yes, they are. The real argument is just about *how much* faster? (That and "did Apple cheat?", which is simple to answer: yes, they all do - all the time).
While the processor may not be suitable for the title of "fastest desktop" out there, you still have to give Apple credit for redesigning the entire architecture, which is more important than raw CPU power for Mac users. Because, you see, Mac OS X with all its little gizmos is really hungry for hardware acceleration of any kind. So graphics and memory throughput are important. While the G5 may not "compute" as fast as the newest Pentium, it will finally resolve the sluggishness problems of Apple's operating system. It will feel really fast. Who's going to notice raw CPU speed anyway? *cough* OK, but that may not have been the most pressing issue with Macs.
It's really not all that shocking that Apple padded the numbers... but there is something about this article that doesn't sit right. I am not sure that the author knows exactly what he is talking about... I get the feeling he is cutting and pasting together an article without understanding how the pieces fit together. I think the jury is still out on this one.
There's always a way to contrive some benchmark that will make system x seems faster than system y -- try hard enough, and you'll find a way.
But what matters is whether or not the benchmark is in some way relevant to the work you're doing. I can tell you, for the stuff that I'm doing, this benchmark has relevance. The article's biggest complaint is that gcc3 is being used on all the processors -- well welcome to the real world buddy!
I use a an analysis package called Root all day long. Go looking through the makefile for Root and you'll see that when it's compiled on macosx it uses gcc3 and when it's compiled on linux running on intel processors, it uses gcc3. So these benchmarks reflect the kind of performance I should see -- hence it's relevant to me and thousands of other people... that makes it a pretty good benchmark IMO.
DFDie Menschen verhoehnen was sie nicht verstehen. -- Goethe.
So, I guess what really should happen is Apple should take a new G5, load it up with Darwin (not the full OSX) - then take Intel boxes and load them up with Darwin as well (since it runs on x86, if I recall correctly). Compile KDE/GNOME/etc. for both, then compare performance of the two side-by-side.
;)
... take it or leave it. This is from a guy who builds his servers by hand, with an AMD system, a dual P3 system, and a P75 firewall - but does all his work on a 400MHz G3 iMac. Its all about what you get for what you pay - and how much that's worth to YOU.
Problem with this is that it would satisfy the geeks, not the designers and home users that Apple would like to market to as well (despite this being announced at WWDC). For them, something that compares Photoshop, After Effects, Final Cut Pro, etc. and then maybe the performance of Safari v. IE 6, Office X v. Office XP, etc. - this is what other folks want to see. Never mind the gamers that'll want to see comparisons of Unreal Tournament play at insane resolutions
Just my $0.02
Worst.
Analogy.
Ever.
- A.P.
"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?"
Say I've got a 32-bit Penitum 4 which has 32-bit address registers. The machine has a 36-bit address bus, so it can access 16*4GB of physical memory. That's 64GB of RAM. My operating system takes up 1 gigabyte of the processes adress space, leaving 3 gigabytes to the process. My process wants to allocate 8 gigabytes of RAM. How do you address 8 gigabytes of RAM with a 32-bit pointer? How do you map 8 gigabytes into a 4 gigabyte address range?
Now, I have a 64-bit CPU that has 40 of its address pins wired up. It has 64 gigabytes of memory. The kernel takes up 1 gigabyte of process address space. My application wants 8 gigabytes of RAM. It has 64-bit pointers.
So, wiseguy, how does your 32-bit machine do it? Does it by any chance allocate four 2 gigabyte "chunks" (segments?) of memory, and select each as needed? Isn't this how the 80286 with it's 16-bit address registers got around the 64k limit? Have we come full circle?
Stick Men
First, I'd like to point out that you are absolutely insane.
Oh, please. Methinks the Intel zealot doth protest too much. If you *really* didn't give a shit, you wouldn't be posting about the issue here on Slashdot, would you?
I talk about a lot of shit on slashdot. If that made me a zealot about every issue I discuss, I'd be in the padded cell with you. Note that your very argument concedes that you are a zealot, and that you very much give a shit. You assume that other people feel the way you do about your computer, but you're dead wrong. To me, my computer is a toy, something to work with, and occasionally an interesting problem (getting things to work in linux), but it's not a part of me the way yours clearly is you. And if I were you, I'd want to think about how healthy that is, when it drives you to start flaming people who rationally discuss your computer.
Be real. You obviously care. You have much time, money, and expertise bound to the intel compatible platform.
My computer is a $500 piece of shit. Additionally, I use macs at work. I actually like them. But that doesn't obligate me to swallow every line of shit passed down from on high at Apple corp.
Like everyone else with a brain, you know that Mac OS X is a superior OS to either Windows or the various other OSs that run on intel (some of which also run on the Mac anyway).
I assume you're trying to troll there - however, I completely agree! Yes, MacOS is the best OS I've ever seen, being a good marriage of unix with a useable window manager. So you're barking up the wrong tree there. Consider that - it actually is possible to criticize Apple and like macs simultaneously!
So it is a bit frightening to contemplate the possibility that you have spent all those many hours backing the wrong horse, i.e., the one that can run the dog OS, Windows, but can't run the sweet OS, Mac OS X.
Again, you've got some problems. However, I'm not backing any horse. If anything, I'm backing them all - at work, I use or have used Linux, FreeBSD, Windows 95,98,2000 and XP, Solaris, Irix, and Mac OS 9 and X. And as said, with some issues, Mac OS is probably my favorite overall. Note that this completely takes out your argument of my having so much time invested in intel, or any particular OS. I like learning new OS's. I like learning new hardware. I just won't pay $3000 for it.
So first, I will again beg for someone to actually debate the issues rather than shooting any messengers with bad news. Seriously, think mac heads (those who have responded with rants) - if you can't even rationally discuss any issue related to your beloved mac, doesn't that bode ill for your mental stability?
-Looking for a job as a materials chemist or multivariat
Very very interesting... I'm reading through the PDF now.
you are moving from what most would call a "Personal Computer" into the workstation market. At that point, the high end G5 is pretty low cost at $3k, but you would also have to see how well it would perform such applications against an SGI or other workstation architecture.
IANAL... But I play one on
Umm... maybe read this whole thread and the posted article? It pretty much states why the Apple benchmarks are flawed.
But half of the posts on this thread explain the reasons why the benchmarks are valid and the article is wrong! Not the least of which is that the person who wrote the "article" (if you can call it that) is a known anti-Apple zealot who will say ANYTHING to make the Mac or Mac users look bad.
And the other half say they aren't valid.
- explicitly disabling SSE2 optimizations on FPU intensive benchmarks for the x86
And explicitly disabling Velocity Engine on the G5. Level playing field. Also, note that they left SSE on. Next!
Actually, this is not entirely true. Intel standard practices are that the compiler uses the SSE2 register set for FPU operations if present, even for scalar FPU operations. The reason being that the x87 stack is, and always has been, nasty. So, for all intents and purposes and according to Intel standard practices *and* the way compilers work, even if you aren't using SIMD, you *are* using the SSE2 registers simply as a set of general purpose FPU registers. The fact that the x87 FPU stack is nasty is why Apple chose to benchmark this way, even though compilers compiling for P4+ architectures will use SSE2 regardless of any SIMD. Apple knows the 970's FPU is better than the x87 stack therefore, go against common (and recommended) practice to make the numbers look better. In a nutshell, code compiled with relatively new compilers will use SSE2 by default on P4+ architectures. Forcing x87 is actually going against the norm.
- using gcc instead of the better Intel compilers for the x86 while using an optimized G5 gcc compiler
There's no such thing as "an optimized G5 GCC compiler." They used GCC 3.3, just like you can download from anywhere. And also, it makes ZERO sense to test one compiler on one machine against another compiler on another machine. That test tells you nothing about either the compiler or the machine. You have to isolate all variables except the one you're testing for, which in this case was hardware performance running the benchmark.
Are you benchmarking the compiler or the hardware? If you are benchmarking the compiler, then use the gcc on both. If you are benchmarking the hardware, use the best available compiler for each machine. It just so happens that gcc is probably the only working compiler available for the G5 at the moment. It also doesn't hurt that gcc on Intel x86 isn't that good. Last time I checked, the Intel compiler was freely downloadable. The only reason to not use the Intel compiler is marketing. Again, are you benchmarking the compiler or the hardware? I would rather see what the hardware is capable of, not what gcc is capable of.
- using a speed optimized malloc libraries for the G5 tests but using the standard malloc libraries for the Intel tests
You misunderstood THAT one too. Apple's standard Jaguar malloc() includes code for cache-coherency for dual-processor systems. Jaguar does not include a uniprocessor malloc(). So for the SPEC base tests (which run on only one processor) they replaced the dual-processor malloc() in 10.2.7 with a single-processor malloc(). Again, this puts both systems on a level playing field, because Red Hat 9 does not put a multiprocessor malloc() on a system unless it actually detects an SMP motherboard.
Not entirely. Not only was the malloc code simply a non-cache coherent malloc library, it was one optimized for certain types of memory allocations, probably ones that work well with the SPEC benchmarks. So, you have two optimizations - uni-processor and speed/specialist optimized - for the G5 tests while you simply have the one optimization for the Intel - uni-processor. I have used many malloc libraries in my career - optimized for large allocations, for small allocations, for allocation, for freeing, for garbage collection, single-threaded
Did someone really have to go through all that
trouble to prove to everyone that benchmarks were
rigged? People just need to learn to use the
hardware/software they like and not get up in a
rage when other people don't.
I think it's just fine if apple uses a specialized malloc() library. It was built to take advantage of the G5 processor, and they are candid about that. What is really going on is that they are trying to even the playing field a bit, because the compilers used on the intel-type machines have been in development on those processors for a long time. They already have a lot of optimization built-in. The G5s are supported, but haven't been for long. Apple is trying to simulate the malloc() library that might be used later, once they can makemore efficient. Oh, and not to mention, this 'argument' he puts together depends on you not reading the same PDF he claims to have read. Go download it and make your own informed conclusions. From the perspective of a long time programmer and x86 user...this doesn't appear too misleading. Anymore than what intel clims HT can do (another flaw in this guy's argument).
====
Crudely Drawn Games
Additonally, note that the cash register was first patented in 1883, so my explanation still makes sense in regards to n.99 and n.95 prices showing up in advertising in the 1880s.
c as h_register.htm
http://inventors.about.com/library/inventors/bl
My journal has hot
If you were to get a new Xeon system with your choice of linux/freebsd and compile a benchmark yourself, that's probably the results you would see compared to a G5 unless you shelled out the extra cash for the Intel compiler. Who's to say that Intel isn't cheating by tuning its compiler to take shortcuts on the SPEC benchmarks? It's not like compiler writers haven't been doing that since people started using standardized benchmark programs.
The real test is always to pick the app that needs the most performance, try it out on both platforms and pick the one that you can afford that does the best.
the good ground has been paved over by suicidal maniacs
As I watch the keynote yesterday, I was dismayed by a couple of the claims Steve made. I use Macs, Wintel, and several "proprietary" Unix workstations heavily and am quite familiar with the advantages and disadvantages of each. That said, I am undeniably fond of my macs running OS X, and use my iBook 600 more than any of my other computers.
But when Steve introduced the new PowerMac G5 as the "worlds first 64 bit desktop personal computer" that tweaked me a bit. I've used 64-bit DEC, SGI, and Sun desktop systems for more than a decade. Don't flame me with the "PC vs Workstation" argument. Most of those Unix workstations were smaller than the G5. And yesterday's demos show Apple is undeniably targeting the same high end multimedia, graphics, software development and scientific markets.
But the SPEC benchmark claims set my BS senses tingling. I too checked out the Veritest results yesterday after Apple's claimed Intel SPEC results didn't jibe with the official published numbers for the same Dell 650. I was annoyed to read that the "independent" tester didn't attempt to maximize the results for all contestants. Granted Apple [probably] paid for the testing, but they should be outsource the evaluation for objectivity, not to have someone lie on their behalf.
It has been known for years that SPECmarks are an indication of CPU performance, but a poor predictor of overall system performance. There are several application benchmarks that are better indicators of performance for certain classes of applications (database, web serving, desktop applications, etc). Apple doesn't seem to publicize these, (other than the perennial Photoshop demo). If "honest" benchmarks don't support your marketing case, I believe it is better to remain silent than to deceive.
I do believe that the PowerMac G5 really will be a very strong contender in the high end desktop market. I do believe that the new PowerMac G5s are probably performance and price comparable to the high end 1st tier Intel boxes. I don't believe the old "macs cost %50 more" or the new "the G5 is $1000 less" arguments. I know from experience that when you kit out these things with the hardware and software needed to get real work done, the prices are comparable. I did say 1st tier manufacturers - not some OC'd LAN party generic white box that's been riced out with mercury cooling and neon.
However, for more than %80 of the work I do, my 600Mhz G3 iBook is more than sufficient. And it's easy to carry around. The other %20, however, pegs my PowerMac G4. It also pegs my Athlon 2200 box. I will probably replace the G4 within the year. The only question is: Dual 2Ghz G5 this fall, or Dual 3Ghz G5 next year?
I often read that the littney that MS stole the graphical interface from apple, but that is ok since apple took it from Xerox. For the record, apple paid paid Xerox for the idea. Last I heard, paying is not stealing or "borrowing." Just for the record,
-Iowa
"He who laughs last, didn't get the joke."-Cap
Well if I'm not a mac user, what is a TiBook doing in my computer room.. I just received an ibook too.
...and *surprise* the Mac always blows away the windows machine. Then a few weeks later when the machines ship and real people start testing them, turns out they can't keep up with Pentiums.
I'm not saying the G5s aren't impressive, but Apple has burned a LOT of karma points over the years by fudging on the speed issue. I'll believe they're fast when someone who's not on Apple's payroll gets their hands on them.
Never let a lack of data get in the way of a good rant.
With Apple's price point at $3K, they're priced up there with the entry level high-end workstations. HP's Itanium 2 workstation sells for around $3.3K. Sun's base 64-bit workstation is a little under $2K. So Apple's 64-bit offerings have to be compared with the expensive boxes, not what's selling at WalMart.
Apple is probably ahead on price/performance and usability in 64-bit desktops, but they're not first.
It's interesting that many people are critical of the benchmark results Apple trumpets. Certainly there are many that can address that topic. Depreciating the numbers, however, doesn't address the more interesting topic, which is:
How do you like the rest of the motherboard?
1 Gig frontside bus X2, PCI-X X3, Serial ATA X2 with separate channels for each drive, USB 2.0/FW 800 & 400, AGP 8X, DMA for every I/O function without bandwidth contention, etc.
Seems to me this may be the most advanced motherboard ever put in any desktop, affordable computer.
And even if you don't believe the SPEC stuff, how about the software demos? Just lies, I suppose.
If y'all had taken the system diagram Apple is showing and substituted 2 Pentium 4's or 2 XEONs for the 2 970's, it would be touted as an Alienware-buster and proof that Wintel is King.
Better check your bubbles for bursting.
I want to address the definition issue.
If your definition of "fastest personal computer" means "best performance of single-CPU/memory sub-system in integer operations using the most optimized compiler for each platform", then everyone's tests show this to be a P4, and the use of gcc on both platforms puts the P4 at an unfair disadvantage, and disabling the second G5 processor makes sense. This seems to be the point of the author of the original article. It seems to be a definition closer to "who has the faster CPU"?
You can grab IBM's numbers for the PPC970 @ 1.8 GHz from their web site:
Compare these to a P4 3.06 GHz as reported at specbench.org:
One would imagine that a 2GHz PPC970 would narrow the gap a smidge, and we have no idea what compilers/optimizations were being used with the above numbers. It's probably fair to say that the P4 is the faster individual processor, but not by a lot. Did anyone else notice that Intel announced the 3.2 GHz P4 yesterday?
You could also try to define "fastest personal computer" as "best performance of the single/multiple-CPU/memory subsystem in a weighted comparison of integer and FP operations using the same compiler with similar optimizations", then using gcc makes sense, and it is only fair to compare the dual G4 to a single P4. It appears that Apple did not use "similar optimizations", but otherwise the comparison seems reasonable. They also compared against the dual Xeon platform, which also seems fair. This definition seems to be closer to "who has the fastest hardware".
It is also worth noting that commercial software is probably not optimally compiled, since it wants to run on a broad spectrum of similar processors, so they compile to a lowest-common-denominator.
The rest of us probably care more about application-level benchmarks, taking off-the-shelf software and running it in potentially useful ways. Apple found some tests that make it look good. They did not appear to be terribly contrived examples, certainly benchmarking Photoshop is useful for Mac users. Hopefully the various hardware review sites will give us some of these.
Finally, you can define "fastest personal computer" as "the benchmarks that make us look best", which is what you do if you are trying to market a new product.
Guess which definition Apple used?
I think the upshot, as far as we can tell without impartial reviewers (if there are any) getting their hands on the hardware, is that Apple is performance competitive. Are they the fastest? Their claims are no more tenuous than anyone else's, and maybe they can legitimately be called the fastest for a month or three.
There are two kinds of societies: sustainable and doomed.
Some situations, sarcasm is expected, and applauded.
/hd/crap/chicken/wing/feather/hda1
Take the Linux vs Windows crowds. On any thread involving discussion of one or the other ( both are not necessary ), you expect to see posts like:
"Ahahah, I paid Bill Gates $299, and all I got was this stupid Blue Screen of Death"
"I'm a first time Linux user, and it was so easy to install. After only 10 hours to build my slackware install, it was all ready to go:
Now, I simply went to the console and typed:
>creoni tarball gz==d | sdfnh!!@ fek fsk-f++++
>grep | -lba+12 cron | (dsfd)sdfsa0/hda/dev0
>ld -abcdefg$$$556%0 - |R|G|T|Y
>mnt
>die ?|TY
It was that easy!"
This is accepted, and expected.
But apparently, you are not aware that PC/Windows vs Mac/OSX is a VERY tender subject on Slashdot. Be careful where you step, there are lots of irrational folks with a fast shot in these discussions.
Man is the animal that laughs.
And occasionally whores for Karma.
...of that Haxial article is Adam Hinkley, the guy who made Hotline. Just in case anyone cares.
The author claims the test is biased mostly because:
1. On x86 hyperthreading was disabled
2. on x86 SSE was disabled
3. on PPC a custom malloc was used
4. on PPC a different set of optimizations were used
1. I admit is seems odd that this was disabled. I think it's effect would be little, but it should be turned on
2. So was the PPC's AltiVec. I recall that SPEC wants FP and INT performance from the ALU sections, not SIMD
3. And I'm sure that there are many "tweaks" for x86 that are transparent within the GCC 3.3 code generators
4. Again, each CPU has different optimizations, either allow them all or disable them all - on both platforms, command line switched or embedded
What I think would be interesting for Apple to do to help settle all this (You know, spread around some of that $4B+ they have lying around):
Purchase two of the fastest model of 1st tier systems they can get that run on x86.
Using four different testing labs, send one machine to each lab (2 x86, 2 G5). Instruct each lab to perform any software/configuration optimizations they feel necessary to get the most performance out of the machine. Then they run a standardized set of benchmarks. They each fully document the changes they've made and the results.
Apple (or perhaps a 5th lab) colates the data and produces a final result.
Or some open source minded person with some extra bandwidth(ha) could create a web site where PCers and Macers could post their own results from the benchmarks. With sufficient results posted, the "noise" would get filtered out and the results would become statistically useful.
Article X: The powers not delegated... by the Constitution...are reserved...to the people
...people in general need more of this kind of distance and insight.
:-/ Too bad...
Oh wait. Damn, you posted as AC.
I don't think I've ever seen a numerical benchmark that has been really satisfying. Cook-offs are really the way to go IMO, and Apple blew the Dell away with Photoshop, PDF viewing, and Mathematica. What should matter is how well your applications perform, not what arbitrary benchmark number you've managed to come up with.
You can get real-time insights on OS X performance using MenuMeters (free, it's on VersionTracker). It's all stuff you can get from top, but more accessible. This puts little meters in your menubar, and after a week of watching RAM and CPU usage in near real time I see the CPU is the real performance limiter on my system. With luck you will find your machine could benefit from more RAM - which is pretty cheap nowadays.
In a 64-bit address space, all your pointers take up twice the space.
This IS SIGNIFICANT in many applications, and they must be run in the 32-bit mode.
The extra address space helps you only if you're willing to spend another several hundred dollars to get over 2 gigs of RAM, or are willing to put up with a huge swap.
I hereby place the above post in the public domain.
It's amazing how much flaming the author received for his analysis. People were calling him all sorts of names simply for pointing out that Apple's benchmarks were not fair. I think it's important to keep companies honest.
But as is often said, the CPU processing speed isn't the main selling point of a Mac. They've been behind for quite some time now, but people are still buying them. This is a great advancement, bringing Macs up to speeds relatively comparable to that of the rest of the market. The 970 is a new chip, and IBM needs time to ramp up the clock speed. P4's didn't get to 3.2ghz in one day.
*stands up and starts clapping* I agree with what your saying. Why just settle for one. You really limit yourself by doing so. If you love technology try to get your hands on all of it. I use a pc most of the time, but I use a mac for protools. Mostly because digidesign won't update versions for the pc.
Your analogy is flawed, since the house you claim to have purchased is not new. If this were a used computer, yes I agree that your analogy is true, and anyone claiming that the OS is a "tax" is off-base. If you bought a *new* house, and each such house had any of those conditions you state (mandated by some large construction company, not "defacto" fees), then you sure as hell would call it a tax.
I guarantee you that a Dell works flawlessly with windows, whereas a noname machine, even if it has "better" parts, might have trouble
YOU guranatee? I really doubt that. Tell that to my parents whose Dell HD broke down months after their 1 year support contract expired. My home-built PC, with no such support contract has had minimal problems the 2 years it's been operating. It's a known fact that PC's from large manufacturers use shoddy components unless they're high-end (and thus high-margin) models.
Make sure everyone's vote counts: Verified Voting
I have no tag line
Apple does not use Dell's tweaked specs but instead chooses to cripple Dell's machine for benchmarking purposes. Dell did not cripple an Apple machine for comparison.
I think your missing something here- Apple didn't cripple the Dell, it came with Windows!
Creationists are a lot like zombies. Slow, but powerful and numerous. And they all want to eat our brains.
As the author of a "hate mail" (starting "Flamebait- My question is: Why is MacNN giving any attention to this?"), I can verify that my post was taken from MacNN and not mailed to the author. It was meant as a critique of the article, not an expression of hate to the author as it is presented.
/. (and that's saying a lot) is picking apart these numbers, and that a detailed analysis is poking through the posts. But I'm saddened that they are buried amongst a landfill of posts from passionate, persuasive people spending their neural energy on "mac fanatics" and "deceptive marketing."
So let me take a moment to reiterate the original point.
I am, like any Mac user, deeply concerned when somebody claims that Apple is using false numbers. The author is welcome to his or her opinion, but I found the claim - that the Veritest numbers are false - was never substantiated: they're as true as any other benchmark. So, no biggie.
But it it the tone of the article that got to me. Claims like, "Apple is attempting to deliberately mislead," and "Apple cheated" and "a significant percentage of [Mac users] are crazy fanatics" have no place in a technical discussion of benchmarks, and undermine the author's believability. All authors have a point of view, but bias is another animal altogether. Authors need to be open-minded to be believable, and this author's use of hyperbole and emotional phrases betrays a certain zeal. Despite what may have been the author's best efforts, the article is not a level-headed, rational discussion about benchmarks. It is a fanatical rant.
And, hey, I'm all for fanatical rants. Not only do I enjoy them, but I am the source of many. My objection, though, was to the editors of the Mac News Network (MacNN) for posting this article, unqualified, as news. It is not news. News informs, and a fanatical rant actually does the opposite: it polarizes. People take a side and stick with it, regardless of facts. The speed of the new G5s is a very very important issue, and this article is a step backwards in understanding these highly complex comparisons.
So, needless to say, I find it *highly* telling that my editorial objection was co-opted by the author as "hate mail." And the response to it just further underscores my point that this is not a rational investigation, but a crusade.
I'd also like to note that, for whatever reason, MacNN has since removed the news article from their site.
I'm glad that the comparative level-headedness of
would like to send my $6000 to buy a top end G5 as well as a top end PC, I would be glad to perform independant, unbiased benchmarks on productions systems :) Anyway, I think that all this crap about arguing over Mac and PC is really dumb. Use what appeals to you. I think that OS X is probably a great os, even though I have not used it very often. I use Windows XP every day and I can say it is also great. The one thing I can say is that I think that Windows has a much better way of managing individual windows, even with the new Expose system mac has devised. To me the way OS X switches windows seems like something akin to what I find on my pocket pc (at least the old way, I'm guessing Expose is a huge improvement, but still not as good as having the task bar.) Anyway, this is just my opinion, I think that everyone should just save their energy when trying to argue that Mac is better than PC or that PC is better than Mac. I use an Alienware box, and I must say that I absolutely love it and I'm sure that Mac users feel the same about their purchases.
SIGFAULT
Oh for god's sake just drop it already. We all know you CANNOT compare PPC and x86 processors! With clock speed, it's like Miles and Kilometers, and every system setup is different, conditions are different, hell even temperature can affect performance. Comparing the Apple results to the Intel results... its more likely that Intel has ALSO manipulated their tests!
Who cares about SPEC benchmarks. Get real-world tests, which the G5 clearly spanked the Xeon system on.
It's like saying I get higher 3dMark scores than you do, but I get less FPS in games. Which computer would you rather have? You want real benchmarks, not artificial ones.
Orange
If someone releases bench marks, of course the results are "optimized" - as if you do statistics with a goal in your mind or whatever biased undertaking. You'd optimize compilers, maybe even the integer unit designs. Point is, however, what's the big deal ? is there any objective benmark out there anyway ? do you cheat 5% ? Is it 10% or 25% ? Well, guess it's more like 5-10% at the most. So, who cares ? Do you notice the speed difference that was presented maybe inaccurately ? h.
The hyperthreading is truely concurrent and it looks like 2 CPUs, but it does not give you MORE than a single CPU's maximum throughput. It does NOT make task switches perform faster or any such tripe.
Sometimes it works to your advantage (in I/O bound situations), but sometimes the extra overhead makes it slower than if it was disabled (as in the case of some intense computational benchmarks)
So if they disabled it in SPEC_int (single thread), that was the RIGHT thing to do.
Black holes are where the Matrix raised SIGFPE
Blinking the cursor in Microsoft Word is almost certainly an integer operation. I think it's vitally important for the user to understand whether he would get better performance in this, the most common of end-user computer tasks, on a 3 GHz Dell Xeon or a 3 GHz 64-bit G5.
Request your free CD of my piano music.
Since the site was ranting about hyperthreading off .. heres some more benchmark confusion for you guys:
http://www.amd.com/us-en/Processors/ProductInforma tion/0,,30_118_3734_3750,00.html
You kiss your mother with that mouth?
Here is an idea, lets look it up:
2. A task exacted from one who is under control; a contribution or service, the rendering of which is imposed upon a subject.
3. A disagreeable or burdensome duty or charge; as, a heavy tax on time or health.
4. Charge; censure. [Obs.] --Clarendon.
seems like its a tax to me, asswipe.
The Kruger Dunning explains most post on
Does anyone have benchmark information on the IO abilities of the G5? Raw CPU power isn't the only important factor in a modern computer's performance.
Actually, the opposite is true. Most Mac users adopt their platform because of informed decisions, knowing they're going against the grain.
Most PC users adopt the PC for one of three reasons: 1) they're technically ignorant and get what everyone else has or what they have at work, and/or 2) they want to get the cheapest thing possible, and/or 3) designing, building, and maintaining a computer is their hobby.
All three of these classes of users are threatened by Apple and the possibility that the alternative to PC might be superior. Class 1 because it reveals their ignorance, class 2 because maybe a low price of entry isn't the only measure of value, and class 3 because it's possible that a mass-market design is as good as their custom-crafted perfection.
Mac users catch crap from outspoken (and usually ignorant) PC users on a daily basis. Over time, this builds up and when a fool, like the "debunker" in question, makes a huge public spectacle of crapping in the middle of a parade, people vent on them.
why could that same person put some nickles in there pocket?
The Kruger Dunning explains most post on
XEROX; The original copied company.
The Kruger Dunning explains most post on
The more people use your platform, the more valuable your platform becomes. Never before has this principle been more important than right now in the computer industry.
Success of a platform brings hardware support. It brings software support. It brings applications. It raises volumes, lowers prices, and expands the universe of clients that you can communicate with. Almost as importantly, the expectation of future success is nowadays (rightly) an important consideration in any investment decisions one may make concerning future technology commitments.
You have only to look at OS/2 to see how a technically superior platform can lose out and die from lack of users.
Macintosh users are deathly afraid that their platform will be marginalized into irrelevance, not from technical shortcomings, but from lack of a userbase. As much as Apple continues to thump the theme of open source, the fact is that the Mac platform needs healthy corporate support to survive, and the only way companies can make money is if they have users. Of course the exact same thing is also true of the Windows platform, but Microsoft has a lot more money and a lot more users than Apple.
Where do I fit in to all of this? I am a devoted Linux user who used to passionately advocate for Linux. But that was before I realized all of the information that I conveniently gave to you above. Now I don't advocate Linux anymore, because I realize that Linux doesn't need advocacy to thrive.
Why?
Because Linux is not dependent on corporations and does not need revenue to survive. Linux does not need a large userbase to sustain its development. With Linux, even a small userbase can sustain very productive development because such a large percentage of the users are also developers.
I am far more worried now about laws like the SSSCA that would have made Linux illegal. I am worried about companies like SCO that are claiming ownership not just on IBM's infringing code but on all versions of Linux, past present and future. But I am not worried about Linux dying from lack of users, because free software is different and all our old notions about software platforms no longer apply.
You don't think losing by 10% with a clock speed handicap of 30% is fast?
I went and looked in my bios, and there is no hyperthreading option. ;)
I wonder if I can get an bios upgrade for this 386?
The Kruger Dunning explains most post on
Right. They also claim to be the first 64bit desktop. Sun, Compaq and a variety of UNIX vendors have been shipping 64bit desktops for years. Frankly the only major vendor that's left to do this, as far as I know, is Microsoft.
Based on upvotes, Ageism is the only "-ism" Slashdotters care about and think isn't SJW
he clearly states that the Dell benchmarks are higher from using an optimized compiler and even makes a comparison to prove the point between the low apple score and the dell one which is double that... and he goes on about hyperthreading etc
if you're refering to the part where he's taking quotes of the veritest results pdf and showing where apple used G5 optimizations for their benchmarks etc well, does Dell have that kind of information even available about the desktop model he is talking about? my guess would be no since it's not there and this guy went out of his way to be thorough in every other aspect
Thank you, Apple, for a fine lesson in how to lie with statistics.
Well, Eight out of ten people know that 78.3% of all statistics are made up.
Maybe I'll be accused of trolling (again), but I have to ask, what's wrong with just basing evaluations like this on unoptimized vanilla out-of-the-box systems using only software available to all systems?
I do understand though that when it comes to GCC, the PPC support is less mature than x86 support, resulting in a possible disadvantage.
Say it isn't so.
that back in the 19th century newspapers used to be one cent so newspaper companies would talk store owners into changing their prices so that people would end up with pennies and thus be tempted to drop them quickly on newspapers
;)
i've also heard the psychological effect theory whereby your brain automatically looks at whole rounded numbers as more expensive or better quality or whatever.. (so sometimes you will see higher end things that end in rounded prices if they're for rich buyers) and that somehow 9.99 looks cheaper than $10
of course who knows if any of it is true
Intel standard practices are that the compiler uses the SSE2 register set for FPU operations if present, even for scalar FPU operations.
You can't use SSE2 registers for double-precision floating point arithmetic. An 80-bit double won't fit in a 64-bit SSE2 register. So enabling SSE2 (1) would not have helped in a double precision benchmark and (2) would have given a misleading result in a single-precision benchmark.
Last time I checked, IEEE Double precision floating point was 64 bits. 80 bits is extended mode and non-standard.
Are you benchmarking the compiler or the hardware? If you are benchmarking the compiler, then use the gcc on both. If you are benchmarking the hardware, use the best available compiler for each machine.
You got that perfectly backwards. If you're trying to test the hardware, you must isolate all other variables. Using different compilers on different machines (when identical compilers are available) yields results that are perfectly meaningless.
Gotta call BS on this. Verion X of the compiler produces great code on architecture X but ultra-crappy code on architecture Y. That is a fair comparison? I don't think so. If you are looking for SPEC numbers, which are the theoretical peak performances for the hardware, you use the best compilers and compile options you can. That is the precise reason why you must clearly document exactly what compiler, version, and compile options you use so that they can be reproduced. In addition, most of the hardware that posts SPEC numbers do not even share the same compiler vendor, much less version, and they are still ranked. So, you are saying that SPEC (as well as www.top500.org and others) are all invalid because they don't use gcc.
I would rather see what the hardware is capable of, not what gcc is capable of.
And I would rather see what the hardware is capable of, not what Intel's compiler is capable of.
Ummm.... the compiler produces code for the CPU to run. Crappy code runs bad. Good code runs good. Good compilers produce good code. Crappy compilers produce crappy code. Again, I could write a version of a benchmark that specifically targets certain architectures and slows down by a factor of 10X. Since it is the same version of my benchmark, by your logic, it is a valid test of the two architectures.
Hell, if you want to do an end-to-end system test, the right way is to build the benchmark with Microsoft Visual Studio on the Dell and Project Builder on the Mac. These are the build environments that people actually use. But since they weren't using Windows on the Dell, they were using Linux, they went with the Linux compiler instead: GCC.
Call up Microsoft and ask 'em what compiler they use to compile Office. I guarantee you it's not Intel's.
Seems to be missing a point.
Not only was the malloc code simply a non-cache coherent malloc library, it was one optimized for certain types of memory allocations, probably ones that work well with the SPEC benchmarks.
And you know this... why? Oh, right, because you have no problem just making shit up to try to bolster your argument.
Ummm... because I read the post. Apple documented the fact as per this excerpt:
"Installed a high performance, single threaded malloc library. This library implementation is geared for speed rather than memory efficiency and is single-threaded which makes it unsuitable for many uses. Special provisions are made for very small allocations (less than 4 bytes)."
(Page 5, also see Appendix E, Page 26, Veritest PDF)
This tells me two things... one, it's single-threaded. Two, it was optimized for speed rather than efficiency. Hence my previous statements.
IF you know anything about malloc libraries, you'd know that malloc libraries are expensive bits of code. Not only is there typically a lot of code there but they have nasty cache properties so that
The G5 is an out of order processor and the P4 and Xeon are in order processors. So the G5 has in built parallelism even in a single threaded case while the P4/Xeon has to be explicity told what is parallel (for threads).
This is speculation: Most everyday apps are still single-threaded or a single thread does 90% of the work. Apps are not inherently parallel. So comparisons with hyperthreading would have been justified.
However.. all this depends on what apps you plan to run on the machines. If you have highly parallel apps then the benchmarks with hyperthreading on would have been a help.
Is it just me or has anyone else noticed that as Steve Jobs gets older he seems more and more like this sort of evil Mr. Rogers? "Won't you be my user, boys and girls?" It even comes across in the pro-Mac rhetoric, "Don't let that Mr. Gates touch you, he's bad." "Won't you please, please won't you be my, Mac-user." [-)
What would Richard Feynman do, if he were here right now? He'd do some math and he'd follow through!
Yea, Apple probably tilted in their favor. Or maybe the testing company did because they didn't want to give Apple bad SPEC results for thier big summer show.
Then again, don't you think Intel does this to AMD? Vice versa? Wern't there just a whole large string of stories about graphics card manufacturers using special optimization benchmarks for their GPU comparison tests? This is nothing new.
SPEC tests are only as unbiased as the people who run them.
CAn'T CompreHend SARcaSm?
"512 Gigs of RAM ought to be enough for anybody." - Bill Gates, Nov 13, 2003.
Glonoinha the MebiByte Slayer
Maybe read this article too, while you are at it. Pay particular note to the SPEC performance charts on page 2.
d =1 296
http://www.amdzone.com/articleview.cfm?articlei
When I looked at the benchmarks as posted by VeriTest, and then read this article I thought, wow this guy is really worked up today. Apple produces a new machine using a very fast processor from IBM, they hire a independant firm called VeriTest to benchmark it, and they publish the full results, as well as their take on them in their marketing information. Whats the problem? We all know it is impossible to make a fair benchmark, and that the only real test is to use it for your application and see how it performs. I read the benchmark and thought it was less biased than about 80% of the ones I've seen. Could it have been better? Probably, but the nitpicking this guy goes into is a little extreme. You can tell he's really reaching when he starts to complain about the $1999 pricing as being deceptive, I mean come on already, this is done for just about every product sold in this wacko country. The photoshop test was all it took to sell me, since it is an app I actually use on a regular basis, and the tasks performed were real world, rather than theoretical. The only real question for me (and many others I know) is, when can I get one in a powerbook?
You are the moron.
Go read ANSI C standard. Pointer is not an integral value.
I don't know why I waste my time on someone who obviously doesn't even know assembly... Here's a clue - mov eax, [ebx] - contents of ebx is treated as unsigned value.
Actually, I believe this particular comparison was made more as an attempt to pump up sales of Logic software than anything else.
The fact is, for years now, Apple has been slipping in the MIDI and music composition market. Pro Tools is probably their only "saving grace" in the recent past, but it's really a hard disk recording/virtual studio product -- not a MIDI sequencer/music composition package.
Apple bought out EMagic because they realized they needed control of at least *one* respected MIDI/music product. (Otherwise, there was a very real possibility that EMAgic would dump Mac support and only produce Logic for the PC.) They're already suffering because of quality products like ACID Pro, and Sonar (formerly Cakewalk Pro Audio) that only exist in PC/Windows versions - and convinced many a Mac die-hard to install a PC in their studio.
I've worked with CuBase a little bit myself, and what I've found is this: The product has recently undergone MAJOR revision. (Perhaps, a total rewrite, even?) The "CuBase SX" product has a totally revised interface from CuBase 5.0 and earlier versions. Not only this, but they're working on both a Mac OS X and a PC version simultaneously. The last "bug fix" they put out for SX on the PC broke as much as it fixed for some people. Right now, it's probably an opportune time to pick on CuBase SX by finding complex musical scores that crash it.
After CuBase SX "matures" a little more, I'm not so sure Apple could still pull off a comparison like they did with Logic on the Mac.
Windows 2003 Server (Datacenter edition on 32bit cpu) will address and use 64G, I think.
... a sixth of a million dollars, or the price of a house.
:: can be played on any field up to the size of Utah.' Unfathomable.
I remember a few years back reading through the Netware manual and it claimed to be able to use up to a maximum of 2G of RAM. At the time my Spectrum class MPE (HP mini-computer) had 32M of RAM and most of the desktops had 4M or maybe 8M. Two gigabytes of RAM - I couldn't fathom that much memory, given that at the time memory cost somewhere in the range of $80 a meg
At the time it was pretty much the same as saying 'Football
Now that RAM is ten cents a meg (I just got a 1G stick for my server for $100, on sale) we have home machines with an address space the size of 'Utah' and all of a sudden that isn't big enough. For the price of a night out with your woman you can stick 2G of memory in your machine - ahhh something I think we are going to appreciate about the future.
Glonoinha the MebiByte Slayer
>Turning off hyperthreading on a dual-processor system is FLAT CHEATING.
Actually on some benchmark turning off hyperthreading has IMPROVED the score.
I doubt that it is the case here, but I just wanted to tell you that hyperthreading is not necessarily a magic thing which always improves performance.
Sure, marketing's often dirty. But you're confusing Apples and oranges.
This is a case where the need is not perceived but real - and felt quite keenly both by the manufacturer and its market. The present deception has sprung from Apple's desperation over its declining Powermac sales as well as its users' reluctance to buy what has been widely-perceived (fairly or not) as outmoded, inferior hardware. Apple knew it needed a hit to revive sales. Apple consumers weren't going to buy until Apple had that hit.
Why did Apple lie? Not to create perceived need, but to appear capable of meeting very real needs everyone alreadly agreed upon. Apple has merely been caught lying about size; apparently, it has a case of benchmark envy. ;-)
(read other fucking articles) :)
As other people have pointed out, Altivec was also disabled, so the SSE2 argument is a red herring. This also makes the rest of the article suspect: the guy looks at what was disabled on the PC, but ignores what was disabled on the Mac.
What I would like to see, is someone from SPEC to comment on what flags they used/didn't use.
They need advocacy so that hardware vendors are more receptive to writing drivers or releasing info (to sell it). ^_^
Black holes are where the Matrix raised SIGFPE
Did you come up with that one yourself? I've never heard it before! (Thanks stewie for the inspiration!)
Let me continue:
Worst.
Reference.
Ever.
C'mon, Simpsons isn't as funny as futurama or family guy anymore, and the CBG is as funny as "In soviet russia..." gags. Get over it, and stop referencing things that aren't funny.
Thanks! I was starting to wonder when someone here was going to bring up this point!
The big deal I see about the G5 is that it's at the very beginning of its life - and seems to be designed with considerable "ramp up" potential in mind. (IBM pretty much promised it would be up to at least 3Ghz in one year.)
If it performs relatively "on par" with the fastest Intel P4 Xeon offerings at its initial launch, and is conducive to rapid speed increases over the next year or two - it's going to be a strong contender in the long haul.
Instead of bickering over the supposed validity of particular benchmarks - the only strong negative I see to buying a PowerMac G5 is the question of future CPU upgradability. There seems to be no mention of whether or not someone buying a $3000 G5 tower today will be able to remove the 2Ghz CPUs and upgrade them to, say, a 2.4Ghz pair of CPUs after they become available.
On the Intel side, it's generally always been possible to swap CPUs without requiring a whole new motherboard and new RAM (within reason).
Sometimes, Apple made this possible too. Other times not. (There was a lot of room to upgrade "logic boards" in the old PowerMac 7500/7600 series, for example. Even the beige G3 took several different speeds of CPU by moving jumpers on the motherboard. My dual G4 1.42Ghz tower seems to have no upgrade path whatsoever, though.)
That may be true, but I don't think the actual relative performances of the machines is relevant as much as the intent on Apple's part to deceive. Is the new mac faster than the fastest PC's? I don't know, and not being in the market for an expensive computer, don't care. Basically, I don't companies that would try to deceive consumers. I know that all companies spin their products, but it seems that what apple did was a bit excessive. Considering the totality of the memory tricks they did with the Mac and disabling hyperthreading and more efficient float processing...that's not cool.
However, speeds of processors asside, if you want the Mac, buy the Mac, if you want a Windows machine, buy a windows machine
That will certainly always be true. Mac people buy macs, windows people buy windows, etc. But there are subtler considerations - like a mac owner who's expecting the G5 to be a significant upgrade over his current machine, and maybe it's not as much as he would have expected. Also, there are people who are considering either windows or mac (there must be, if the switch ads are successful) - and these people might not be getting a clear picture.
Again, there are enough flames going around already, so I'm not comparing macs to pc's - I just think Apple could have done better for their customers. I'm sure the G5's are going to be great machines, which is why it's sad they resorted to this.
-Looking for a job as a materials chemist or multivariat
You don't have to be Carl Jung to see that the taunting of Mac owners by PC fans is a patterned ritual, akin to the jocks giving the artists a hard time in school; or that, in the reverse, Mac owners berating the vanilla plainness of the PC world is a prototypical underdog's gripe about the herd.
The really funny thing, and what makes it so amusing for me to play this game (I own a PC and a Mac - so there, prlrlrrlrlrlb!), is that this stuff is obviously deeply personal to so many. It's close to the psychological metal, you migth say. The hardware and the OS are just handy symbols for having at it - for, as you nicely put it, "competing, not computing." It's a terribly amusing hobby, and I'm grateful to everyone here who blows their top in pissing matches. ;-)
I agree: but that's exactly the problem. It's well know that HP is slower for single threaded aps, and faster for aps that take advantage of many things at once. These benchmarks had HT turned ON for single threaded aps, and OFF for mutlithreaded aps: exactly the opposite of what it does best.
Assuming it's bang/buck, and buck = 0, then bang/buck is Undefined. (division by zero!)
The original poster is right, it's +INF. Here's a nice introduction to hyperreal numbers.
"Nope. If it's not threaded properly, it will not take advantage of multiple processors. The application will run on one processor."
Yes, except for one thing: you generally have more than one application running, and each of them can have more time because you have two processors. Thus, one process (such as one application) can receive a huge speed improvement simply because it doesn't get switched off of the processor as often and doesn't have to make as much room for other things to run.
If you don't believe me do the simulation yourself. Draw out 10 processes on paper, don't forget the OS kicking in to decide who goes next, and assume a Shortest Job First, Preemptive (you can ignore all of the other fancy things, this is just a simulation).
Integrate Keynote and LaTeX
It almost goes without saying that Apple fudges their benchmarks a bit. There is a bit to the linked criticism. Apple's choice of compilers, tweaks, and options, indeed makes G5 seem better than it should. Still, the moral really is that Apple now is in the same ballpark with the fastest x86 machines, rather than lagging behind as yesterday's G4/G3 models do.
:-)). The only users who have any real reason to care about the high-end specs are those who perform complex scientific modelling, graphic processing, or a few other CPU-intensive niche areas. In other words: the users who rely on FP and SMP.
But the critiques claim that "most users" use only integer apps, not FP, and that most apps aren't multiprocessor enabled, completely misses the point. NO ONE needs the latest Xeon, Opteron, or G5 to run typical home/business apps. A G4 or P3 is plenty fast enough to run Mozilla or OpenOffice (I won't even mention proprietary apps of similar purpose
A G5 is as fast as anyone can need for average applications, and probably still a little slower than the best x86 machines for high-end FP/SMP apps. But not so much slower that the other Mac benefits can't sell machines.
Buy Text Processing in Python
I know, they could take several applications that people on the mac use, like, oh, I don't know, maybe photoshop, mathematica, and maybe even a game like Quake, and run them through some stuff and compare times. Better yet, they could do it side by side with a Dell machine and we can watch it first hand. Wouldn't that be awesome? What? Wait? You mean they did that yesterday already and would know that if you watched the video stream?
Oh, right, it is better to be lazy, and make wild assertions about which benchmark is better and have conspiracy theories about fixing the results. Otherwise, this could be a serious conversation and not slashdot.
No, I am not trolling. I am just annoyed. What else could Apple do? They released benchmarks. They even said the Pentium is even faster in some cases with single processor. Then, they ran comparisons using famous, commonly used programs that are processor intensive. That is fair. God did not say that Intel was always the fastest. Get over it.
-Iowa
"He who laughs last, didn't get the joke."-Cap
Well, yes, ok, but I'm not getting the same conclusions as you are, and I assume I just don't see the data the same way that you do. I'm hoping you could be a little more specific as to how you arrived at your conclusions.
Anyone else find it ironic that an artical about the G5 has a G4 picture next to it?
--
Sitcking feathers in one's ass does not make one a chicken.
Seriously, I have a Sparc 2 lying arround in my room that I turn on from time to time, I certainly would not call myself a SUN user though.
T Money
World Domination with a plastic spoon since 1984
This guy just wants to cause a stir.
That's Dell's job, they are the ones whose computer are be compared to, and made to look inferior.
People have been knocking Macs for years. Most have never used one for any length of time. This has put Mac users on the defensive, and that's why this guy received all the hate mail.
I hope Apple sue his pants off, as he has made statements which are more misleading than any of the benchmarks.
Slashdot readers finally discover "marketing" is an euphemism for "lying"; introduction to the real world sad and pitifull for geeks of all stripes. Film at 11...
Insanity is the last line of defence for the master diplomat. But you have to lay the groundwork early.
So your saying, beyond a doubt, that this guy and his analysis are correct and accurate, while Apple AND Dell are BOTH wrong.
He posts an article with more distortions than Apple, and you take his side immediately.
I have no idea what country you live in, but in the U.S.A. and most of the world there are laws AGAINST false advertising.
More FUD
Er, I thought the Pentium 4 isn't SMP capable ..?
have you ever used a $658 dell? i'm not sure what they did, but i think the memory, video and hard drive are all coming through a layer of cheesecloth in the north bridger. *electric* cheesecloth, mind you.
combine the shitty architecture with the omnipresent windows XP, and you have a really really slow and frustrating computer to use. when you start an app the whole UI will lock for 3-4 seconds at a time, no repaints, no mouse movments, no nothing.
the three or four times you ever need to do something processor-intensive, sure it might finish fast (if the dataset is small enough to fit in L2 cache, of course). other than that you're screwed.
the new G5 systems have incredible bandwidth. incredible. combine that with the BSD scheduler and cache management, and i bet even Panther will seem fast.
like you said, if it seems fast, it is fast. and dell's sure ain't.
Tetris rules.
Where have I seen something like this, just before the guy goes on to abuse the product?
I am a Mac user. I have been using Macs for years. I am writing this article on a PowerMac G4. I enjoy using Macs.
Oh yeah, the start of the page this discussion is about.
More FUD
It's not cricket to turn off SSE2 because SSE2 is faster at scalar floating-point than the X87 unit. Intel decided to sacrifice a bit on the X87 and put the FP muscle into SSE2 for the P4. Vector or scalar, if you're compiling floating point code on the P4, you should use SSE2.
The closest Dell Dimension 8300 I could configure was $1777. At $1777, the Dell Dimension 8300 has a smaller, slower disk drive and a lesser video card than the $1999 G5. When you take the aforementioned differences into account, they're about equal values.
Incidentally, I tried similar comparisons with Dell's "Ultimate Gamer" machine, and it was also about the same price as a comparably equipped G5 before taking the slower hard drive into account. Looks like Apple did their homework in pricing the new G5s in order to minimize price arguments.
Superdrive doesn't have any buttons period... The eject button is part of the apple keyboard on the right side... so its either that or drag the cd/dvd to the trashbin (anyone other than me find this rather unlogic for a burner?)
In the case of Hyperthreading, the Xeon may have been optimized for these tests...
If your definition of user is someone who uses a platform exclusively, then no I am not a "Mac User." If your definition of a user is someone who uses a platform on a regular basis, ie weekly then yes I am a mac user.
And Chipzilla is also re-assuring its customers that its next generation chip, the so-called Prescott, is on track for a Q3 introduction, despite earlier wobbles that suggested it might retreat into Q4. In fact, general availability of the Prescott at speeds of 3.40GHz will fall into the fourth quarter
Don't think of it as a flame---it's more like an argument that does 3d6 fire damage
I wouldn't be too sure about the Photoshop benchmarks. I have seen PC's performing better than Macs and vice versa in tests. It would seem that the wide variety of tools and filters utilize different CPU's in different ways. Wanting to make the Mac look better, a person could have a set of the faster-on-Mac filters run multiple times, increasing the apparent speed difference between Macs and PC's. The same could be said about portraying PC's as faster.
As for disabling Hyper Threading, this may have been an optimization in the VeriTest SPEC benchmarking.
The only two relevant metrics here are:
a) performance alone
b) price/performance
The Apple loses on both of these. It's not only slower, but also more expensive. I don't see how there's a "clock speed handicap" -- nobody forced Apple to make them so slow.
Hmmm... I seem to have been posting this a lot today. Disabling Hyper Threading can make the SPEC scores look better.
Unlikely, but I have just as much evidence as you.
HT is intelligent, because it doesn't always work? Yes, obviously it depends on what code is running-- the code must be multi-threaded, like anything intended for multiple CPU's. Or, a single threaded application could be running on top of a multi-threaded OS, in which the OS could divvy up OS/application processing resources for best performance.
Why?
And considering that the Xeon/P4 could be seriously crippled by not running HT and who knows what other optimizations, I'd say that the numbers are still impressive.
The machine (Mac and PC) specifications are in the VeriTest document.
Although it is tempting to argue, I will agree with you. I must have missed that while reading.
CODITO, ERGO SUM: I Code, therefore I am.
Apple calls it "the world's fastest personal computer". Mac users say that it is a "workstation", so you have to compare it to Xeon processor PC's. Which one is it? Or is it both, making G5-P4 comparisons perfectly acceptable (price v. performance wise)?
I strongly disagree with his statement:
Faster on integer single-processor tasks, which is what most people use most of the time.
It is true that people need integer operations most of the time, but they don't need this kind of processing power most of the time, reading html or doing spreadsheets. They need the extra cpu power to provide solutions to the hard computational problems needed by modern graphics and sound applications.
In the not so distant past integer algorithms were all we had to do the hard stuff. As machines have gotten more powerful we have moved away from integer algorithms to perform most compute bound work. Floating point provides higher quality results for most any difficult computational problem.
We were limited to integer solutions in the past, but modern image or sound processing problems use floating point solutions today. Thats where the benefits of this kind of processing power can be seen by most people.
Although I'm sure that most benchmarks for hire are going to be a little tilted, I think I'd probably trust apple's numbers slightly more than Dell's. Having worked in a campus computer store, and having to sell the damned things, I've NEVER gotten a Dell to come out to anything near the advertised price. Apples can be pricey, but at least they are upfront about it; if it says x dollars in the apple store, then that's what you pay. Just my 299900 cents.
`which fortune`
That's right. In the REAL WORLD, companies use Intel-optimized compilers for PC applications and Mac-optimized compilers for Mac applications. That is what they should have done here.
A level ground may be useful for comparing theoretical performance of the CPU's; unfortunately though, we do not live in a theoretical software world.
Last time I checked, IEEE Double precision floating point was 64 bits. 80 bits is extended mode and non-standard.
/dev/null.
Check again. You're about 12 years out of date.
Verion X of the compiler produces great code on architecture X but ultra-crappy code on architecture Y. That is a fair comparison? I don't think so.
Of course it's a fair comparison! It's a comparison with exactly ONE variable. It tells you that code compiled with compiler X runs well on X but poorly on Y. It tells you something FACTUAL, not something INFERENTIAL.
Of course, you may not like the results. Fine. Then get off your ass and improve compiler X. Or pick another compiler THAT EXISTS FOR BOTH PLATFORMS and run the test with it instead. Don't go throwing in random variables to muddy the results.
Heh... my app using Intel's compilers will run 2X as fast as yours compiled with gcc, that's all that matters. That's why you compile with better compilers and throw crappy compilers into
If you are looking for SPEC numbers, which are the theoretical peak performances for the hardware
Ah. I see the problem here.
NO THEY ARE NOT.
If SPEC numbers were theoretical peaks, then we would test microprocessors on multimillion dollar test harnesses instead of in actual systems. SPEC numbers are not meant to be "theoretical peak" anything. They are meant to be a BENCHMARK (perhaps you've heard that term before) for relative comparison of two or more SYSTEMS. Complete systems. Not theoretical components.
Here's what you do. You run the test on one system. You change ONE VARIABLE and run the test again. Did the result go up or down or stay the same? See? You learned something there.
If you go changing EVERYTHING, then you have learned nothing.
You don't even understand what we're talking about.
OK, I'll agree with you on the theoretical part. I misworded that one. SPEC consists of kernels of "real world" problems to give an idea of how things may perform on your system. As with anything, you have to take it with a grain of some salt. It gives you an idea of what your system will do on similar tasks. I chose the wrong word so I'll concede that argument to you and restate.
They indeed do test complete systems. Choice of compiler is a part of that complete system just as choice of operating system and speed of microprocessor are. That is why they are still valid even if using a different brand/version of compiler. You can have everything the same about two systems except the compiler and it will still be two different systems and one is allowed to be better performant than the other and still be a valid test. You would chose the slow system over the faster system simply because all the comparisons used gcc. Meanwhile, I would use the faster system and turn out more work over time and my boss would be happier.
That is the precise reason why you must clearly document exactly what compiler, version, and compile options you use so that they can be reproduced.
Right. But you can't reproduce SPEC on a G5 with the Intel compiler. There is no Intel compiler for the G5. So poof. SPEC with the Intel compiler is not a valid way of comparing a G5 to an Intel system. It's only a valid way of comparing two Intel systems.
Um, no... it's so you can exactly duplicate the hardware and software used to run the benchmark at another site to verify that the results are reproducable.
So, you are saying that SPEC (as well as www.top500.org and others) are all invalid because they don't use gcc.
Yes. That's precisely what I'm saying. If you mean to do a SCIENTIFIC, PRECISE comparison of two systems, you have to do it diligently. Otherwise the results are not useful. We're talking about differences in SPEC scores of 10% here; that's so close as to be virtually meaningless as it is, even under the most rigorous conditions. If you change architectures, chips,
I can see why this might be an issue for PC users/possible switchers, but for me as a Mac user, the fact that Apple now has fast enough hardware to even attempt to make this claim is more important to me.
-You may license this sig for only $6.99.
Oh... forgot about the IEEE 754-1985 standard... (been the standard since 1985)
/ co ding/ieeefloat.html
e e/ ieee.html
http://research.microsoft.com/~hollasch/cgindex
Last updated on Jan 02, 2003.
In case you can't cut-n-paste, here's an excerpt:
The following figure shows the layout for single (32-bit) and double (64-bit) precision floating-point values. The number of bits for each field are shown (bit ranges are in square brackets):
If you prefer a non-Microsoft site:
http://www.psc.edu/general/software/packages/ie
from which:
Double Precision
The IEEE double precision floating point standard representation requires a 64 bit word, which may be represented as numbered from 0 to 63, left to right. The first bit is the sign bit, S, the next eleven bits are the exponent bits, 'E', and the final 52 bits are the fraction 'F':
[...]
Reference:
ANSI/IEEE Standard 754-1985,
Standard for Binary Floating Point Arithmetic
I'm done. May you continue to live in academia.
Even with Dell's more P4-friendly numbers, I just noticed that *if* the SPEC scores go up linearly in regards to clock rate on the G5, then the 3.0 GHz G5 that will eventually come out will actually be a hair faster than the existing 3.06 GHz P4.
:)
This is meaningless, however, as in 12 months, we may see a 3.5, or even a 4.0 P4 on the market. Provided the P4 motherboard designs are updated to properly feed the upcoming P4's, they'll still be competitive. If they don't, the G5 will smoke the P4. Of course, just as the 3.06 P4 wasn't built in a day, neither will the 3.0 dual G5. It's very hard to judge performance based on the very first system from any company, as the performance issues which are bound to crop up won't be fixed for at least one more generation.
I'm fully aware at the moment that Apple will spin facts. So will Intel, and so will Dell. Hell, the only reason I'll get a G5 is because I'll eventually need a new desktop, and I've grown to prefer working in OS X. Simply put, the PPC970 is a sufficiently powerful chip for anything that I'll need to do for a while, so I see no reason not to buy one when the time comes.
However, to make everyone feel better, here's what I propose: Take a dual G5 and a top of the line dual P4/Xeon system. Hand the G5 to a highly talented Mac expert, and have him run the SPEC benchmarks, as well as some general application response testing. Hand the P4/Xeon to an Intel expert, and do the same with him.
Then swap. Make sure the Mac guy knows dick about Windows/Intel, and that the Intel guy knows just as little about Mac and the G5 chip. Average these (most certainly lower) scores with the heavily optimized ones, and use the average.
That's your performance level. Let them both cheat as much as they can, but let them suffer for it.
Raptor
"Procrastination is great. It gives me a lot more time to do things that I'm never going to do."
Quote "The price-to-speed ratio of Macs is not one of those great things, but just because it is poor on that particular point, doesn't mean," blah blah blah.
I just went to Dell to price out the Dual 650 3.0.6 that edged out the top G5 (according to this guy) to a comparable Mac config. and came up with $4300. Why do people keep protesting Macs are too expensive? Yes, they cost more than the usual PC, but you GET more too. And if you look around, you seem to get more for less. People just see $3000, and say "DAMN boooey!!", because they are looking at the PC package at the same time, monitor included, on the web for the low low price of $600. What they aren't understanding, is that you actually are getting more bang for your buck with the Mac hardware. Does everyone need that "extra" speed, SuperDrive, RAM, Dual Procs, 64-Bit, 1 GHz front side, USB 2, FW 800, Digital audio in/out, AGP 8X Pro, PCI X 133, etc... no. But paleeez.
I can't find other comparable models that can beat Apples prices. If someone knows of cheaper PC's with the same config/performance, please post them.
Thanks!
"The greatest obstacle to discovery is not ignorance - it is the illusion of knowledge." - Daniel Boorstin
The vast majority of The Majority makes their choice based on ignorance (i.e. "it's what we have at work"), on what would most easily alloow pirating of software, or on lowest up-front price. Not on what "works best".
int main()
}{
a.c(4): error: expression must have arithmetic type
Gee, I wonder why... Oh, I know - you have no clue!
There are things that can eat processor time without doing massive amounts of matrix multiplactions (3d), fourier transforms (compression) and general gruntwork (compiling). Take the GUI as an example. Small features you might not notice can heavily tax the CPU. Take resizing a window. In older versions of windows, a grey bar would be drawn indicating the proposed new boundary for a window. In XP you resize the window as you drag the mouse, and on this Athlon 2000+ its pretty doggish. Same with moving windows around. Your GUI is slowly improving, but the beauty of a good GUI is that you don't notice the flaws.
Of course, my 500 Mhz k6-2 has similar problems under metacity. Its a bit slower, I suppose, but they're really quite apples and oranges. The one place where one will kick the other's ass is transparency. They say that X11 wasn't modelled to handle inter-window styles like transparencies but even just a console transparent to the background isn't very smooth. OSX handles this well, as does the few applications using it under XP.
Since people are using higher resolutions, higher bit-depths and faster refresh speeds, a feature like transparency will push hardware pretty well if abundant. You can't use normal blitting instructions since they're designed to move data from one place to another. Transparencies are at the simplest, addative; most use more complex routines than that to give the illusion of depth and to avoid colour saturation.
And these are only things we see right now. In the future you might see desktops and programs that dynamically size themselves relative to the other programs running. On one hand that takes some power away from the user, to operate like I do with a single window at a time and alt-tab as desired. On the other, the idea is both neat, simple to the user, and could be made optional.
I Browse at +4 Flamebait
Open Source Sysadmin
Even by Slashdot standard, you are all making an unnecessary huge fuss over something hardly relevant.
The G5 Power Mac has a much greater performance lead (200 - 700%) over the Dell in running real-world apps (such as PhotoShop, Logic, BLAST, HMMer, Quak 3) than in SPEC.
OK, Apple may have used optimized GCC for G5 or disabled Hyper Threading and SSE2, but that's more like leveling the play field rather than cheating, when considering that
(1) GCC is much more heavily optimized for the long established x86 than any other CPU such as PowerPC or SPARC, and the G5 is still a prototype;
(2) HT doesn't always help the performance, and in any case its impact is normally quite small for benchmarks.
(3) SPECfp is purely for measuring the FPU performance which excludes SIMD operations using SSE2 on IA32 or Altivec on G5.
(4) The dual G5 Power Mac is a really great system (dual 1 GHz FSB, dual channel 400 MHz 128-bit DDR memory, 8x AGP, dual channel serial ATA), and Apple has far more leverage than resorting to cheats.
WHAT IS YOUR FUCKING PROBLEM? WHY ARE YOU THREATENED BY THE FACT THAT A MAC IS FASTER THAN THE FASTEST PC?
I'm not at all. Why are you so threatened that people say that the Mac isn't? I could care less if we were talking about two completely different processors as long as the tests were fair, which they weren't, in a bunch of folk's opinion. Is it that finally Apple presents something to you and you are so desperate to believe it that you will do anything to cling to it?
If it is that important to you, I'll even say that the Mac is faster given that test. I think the test was a waste of time and showed nothing and proof that is conclusive to me of this is given at http://www.spec.org
In fact, if you wish, we can abandon the whole issue of Macs entirely and I'll still argue the same point. Let's concentrate only on gcc vs. the Intel compiler only on the x86 box used. GCC produces an executable that runs given code at speed X seconds on PlatformMoo when using the best set of optimizations it can. The Intel compiler produces an executable that runs the same code at speed (X * 0.70) seconds on PlatformMoo when using the best set of optimizations it can. Which compiler should I use on my computational codes if I want them to run as fast as possible? I could give a rat's ass about what is more beautiful, what licensing I have to have, or who is running the tests or who wrote the compilers. FACT: the Intel compiler generates faster code for the benchmark on the given system as clearly and undesputably shown on the http://www.spec.org web pages. Answer these: which is faster? the gcc executable or the Intel compiler executable? Which reflects more accurately how the system is capable of performing on the given problem?
Heck, let's use car engines as an analogy. Take a car, time it around a track. Put restrictor plates on the carborator. Time it around the track again. Which timing more accurately represents the performance the car is capable of achieving?
I'm guessing both CPUs should also run the same binaries for it to be truly a real test, that way they have the exact same optimizations and algorithms.
if I can complete my task using CompilerB in 2 days when CompilerA takes 4 days, which is my boss going to want me to use?
But you're not talking about COMPILERS, FUCKWIT. You are talking about a combination of COMPILER, OPERATING SYSTEM, AND RUNTIME ENVIRONMENT. You know NOTHING about the compilers you tested. Got it? NOTHING.
I'm not out for an academic exercise. I'm out to see how fast my code can run. Period. If I have to spend a few days to figure out and try a number of different compiler options to get fast code and if it takes a different OS and compiler to get it done for us, we don't have a problem with that. Time is money. We turn around accurate simulations faster for more clients, we succeed. Obviously, our best solution would be hand tuned assembly running on MSDOS, but we do need to have some portability to rapidly port to the next fastest thing and we'd prefer not to have a machine with MSDOS on it.
I just want you to ignore the Mac for a moment and state which executable was faster and which more accurrately reflects the performance characteristics of the system - gcc or Intel - on the SPEC benchmarks. Remember... ignore Mac (although I know that it will be damn near impossible for a Mac zealot to do).
Well... maybe this... what if CompanyX made a compiler tomorrow that only ran on the new G5 and was very good (generating numbers even better than the Intel compiler on the x86 box). Would you still use GCC only to compile everything that will ever run on the G5? What a joke.
Why not put your opinion where your reputation is and come out from behind the AC flag? I think it actually must accurately represent your achievable performance. I'm willing to put my karma on the line. What are you afraid of?
(Personally, I would like to see the G5 be faster than the PC and I
"...suck it you Intel Weenie!"
absolutely classic.
Oh well....I am going to go "suck it" now.
(Working on a device driver for my Radeon 9000 128MB video card....)
-Hack
Got Geometrodynamics? Awe, too hard to figure out? Too bad.
The part of gcc that actually generates the code - the part that is important to the compiled app's performance - is, of course, completely different for each CPU. The rest of gcc is irrelevant; it's only interface & syntax parsing stuff, it doesn't affect the compiled code's performance. gcc/x86 and gcc/G5 might as well be completely different programs. It wouldn't matter dick if Intel did provide a G5 compiler. You simply can't "standardize" the compiler for two incompatible CPUs.
The only meaningful comparison is to use the compilers that app developers on each platform are likely to use - and for performance-oriented apps, that's the Intel compiler, since it is so much faster.
Why would anyone engrave "Elbereth"?
I think that you would find peoples' complaints disappearing if operating systems costed $10 apiece. But the fact is that Windows does not cost $10 apiece. Windows costs more like $50 apiece even for the best OEMs, despite the fact that its per-unit cost is even less than that of a modem.
You give many good economy-of-scale arguments for hardware, but many of them don't apply to software and even the ones that do apply would not be so costly as to exceed the (inflated) per-unit cost of the OS.
Actually, in the past, a lot of people did complain mightily about integrated hardware and the concomitant reduction in configuration choices. Nowadays those complaints have mostly disappeared because the quality of integrated components has risen to a high level while the cost has declined to a negligible level. The same cannot be said for Windows.
Lastly, unlike OEM Windows, your modem most likely did not come with a EULA that contains explicit permissions to return the product for a refund in the event that you refuse the EULA.
fitten, you have the patience of Jobe and I have the utmost respect for you to deal with that knob in the manner you did. Polite, elegant, factual, and you never stooped to his level.
/dev/null" and those goofy ads with the people who can barely tie their shoes slamming the Windows sytem for being so hard to use and the mac "just works". It's sad, and it reminds me of the election campaigns where senators never tell you how they're going to improve the economy, only how badly their opponent is going to screw it up. That's just tacky, and I have no respect for those individuals.
You are correct and I will back you on the fact that this is not a comparison between the same compiler on different platforms. If you are comparing CPUs, you want the best code available to accomplish the task and that would mean assembly. But since SPEC wants the code portable, it's in a higher level language. This means you want a compiler which will produce code that runs best for the particular processor. For x86 platforms, that is usually the Intel compiler. GCC produces slower code in 98% of situations. That tells you nothing about the CPU, only the optimization abilities of different compilers.
Throwing that "scientific method" argument at this approach is completely nonsensical UNLESS you're comparing the code generated by the same compiler on two different platforms, which isn't the point here.
The whole point of comparing CPUs is to say "Hey, CPU, perform this task and tell me how long you took. I don't care how you do it, as long as the end results are accurate. Take any shortcuts and speed advantages you can." By deliberately giving the CPU bad code that makes it take longer than necessary is unfair. Apple should have used every advantage they had at their disposal (Altivec, G5 optimizations, etc) but also give the P4 the same chance. Allow it SSE2, give it the tweaked malloc function. THEN see who comes out on top. Anything less is a sham and is designed to deceive potentional buyers.
If your product is as good as you say it is, it will stand on it's own merit and not require trickery. That is what bothers me about Apple. Almost every ad I ever see for them is a mudslinging campaign against the x86 platform. They have a good system, and they should point out what they have and how it is advantageous to the buyer, such as "DVD-R to let you burn movies or up to 4.7GB of data!", "Up to 8GB of RAM supported which allows much faster image processing than before!". Instead you see crap like "sending other Unixes to
> > Apparently they never got so far as Chapter 1 in Hennesy & Patterson, where you learn the mantra of "make the common case fast".
> Apparently YOU never got so far as Chapter 1 of the Mac OS X system architecture guide
Of course not. While it's reasonable to suppose that a chip designer is familiar with the contents of Hennessey and Patterson, there's no reason to suppose that a Linux dweeb would have read any of the OSX documentation.
> where it says "we use single-precision floating point arithmetic for everything."
> When you draw something to the screen in OS X, whether it's through Quartz 2D or a higher-level API, you specify pixel locations as floats. That's right, your window is 200.0 pixels by 300.0 pixels. And you can use fractional values, too. A pixel at 200.5 gets antialiased between 200 and 201.
> Mac OS X is a VERY floating-point intensive operating system. The only ints in the system foundation APIs are bools and enums. Everything else is floats, floats, floats.
That's very interesting, and it surely indicates that OSX is more fp-intensive than other platforms. But I'm not sure it means that fp operations are the common case. Surely a typical application will do more than make calls to that API? Surely all the counters, indices, and addresses that do so much of the grunt work in any program are still processed by the integer ALU?
And lots of integer operations don't even show up explicitly in a program's source code. You can't even print a string without iteratively incrementing an index. And a simple expression like A[i] hides two integer calculations required for accessing the value, regardless of the data type of the array. So unless the G5 has a special ALU in the data path for performing these rudimentary operations, almost any program is going to be soaked in integer calculations at the machine-code level which would have to be handled by the same ALU tested by the integer benchmark.
[Of course, that's not an argument that integer operations are the common case. I don't know how you could tell without instrumenting your code to count integer and fp operations.]
Also, notice that even if we suppose that fp operations are in fact the common case, that would tell us something very interesting about these benchmarks: it would say that the purported 20% fp speedup vis-a-vis Intel is no longer an interesting comparison for programs that use the OSX API; the interesting comparison would be whether OSX can do these things faster in fp than an Intel OS can do the same things with integers. The relative performances on the SPEC benchmarks are no longer useful for the analysis.
Don't get me wrong; I think the x86 architecture is a kludgey piece of crap. I'd certainly rather have a G5, if all else were equal. But KPoC or not, we should still think critically about the claims of benchmarketers and fans.
> By the way... aren't you the Black Parrot who about a year ago got stone-cold-busted for making false claims about Microsoft? Something about your company being put out of business by Microsoft or something? Claims that later turned out to be false?
Heh, I remember that thread. Some lamer was arguing incessantly that Microsoft had never crushed any of the competition, and since he couldn't actually win that argument he decided instead to crow when I didn't give him some meatspace biographical information he wanted, as if he had actually scored a point in the debate.
But I'm curious why you say 'busted'. Ignoring the slim, slim possibility that some overzealous idiot mistook a wisecrack for a statement of fact, wouldn't 'busting' someone require some actual knowledge of the fact? I.e. in this case knowing the complete employment history of the person or persons posting as Black Parrot, and going down that list to ensure that none of those companies had ever been crushed by Microsoft?
> Just asking,
Sheesh, evil *and* a jerk. -- Jade
Well, the PowerPC backend of GCC prior to version 3 has always been rather poor, and usually considered much poorer than that for x86 CPUs.
With GCC 3 things changed: the backend for x86 is much better, but that for PowerPC still needs some working.
For example, PowerPC cpus have multiple condition registers, and no open-source compiler makes use of this, which can be a trememndous bonus on unrolled loops -- also you can decide whether some operations affect condition bits or not. I hope APPLE and IBM will add support for this in GCC, you can easily get much more performance across the whole system, in all types of applications, see for example:
http://www6.cse.cuhk.edu.hk/~csc2420/lect3/power.h tm
Apple assigned a few engineers to optimize GCC about one year ago. Their work showed the first results a few months ago, and I bet we shall see further improvements pretty soon.
First AmdZone say:
Apple claims they have the fastest 64 bit desktop on the planet. We claim they have a pretty fast 64 bit workstation, but Opteron workstations available now seem faster.
To my knowledge there is no Opteron computer or motherboard with an AGP slot, as that will replace my Dual Athlon MP!
From ridiculous to:
Update: My fears about these benchmarks have apparently been valid. This site has plenty of details about how Apple twisted results and crippled the P4 and Xeon scores. You can get the details of the testing here to back it up. If you want to see where I got the SPEC CPU2000 scores then click here.
I'm absolutely dumbfounded that people who should know better are using this tripe as "proof" for their own arguments.
PC...that stands for Personal Computer
Not workstation or server
That's how people are able to comment.
Enough information has been given to reproduce the benchmarks.
And why should they put the Opteron in?
Now Apple is less shy to show SPEC results, even though their approach for "making competition fair" sounds suspiciously like tweaking a bit. This is an important change.
Furthermore, that was a developer's conference, so the developers now now one thing: Apple is working on the compiler, the tool they use most. You can expect more performance in the future on the same machines, and Apple is deadly serious about this.
Last, if you do Word, Tex, some compiles and Spreadsheets, 1, 2 or 3 Ghz do not make much of a difference.
But if you do INTEGER number crunching (cryptography or number theory research) the G4 and G5 processors can emulate with Altivec a fast 128 bit arithmetic unit. And if you do serious FLOATING point operations, you can do them in single precision with Altivec and then refine the results in double precision. I do both, and I know which CPU is now fastest.
SSE2 *cannot* do this, and SPEC does not reflect this difference. SPEC does not tell the whole story...
Here's the only thing that I don't quite aggree with:
7 69/ref=br_bx_c_2_0/102-3761900-9564136
"Misleading Prices
Both Apple and Dell are guilty of using misleading prices. For example, Apple gives the price of the low-end G5 as "$1999", and the high-end G5 as "$2999". In other words, they have subtracted $1 from a $3000 computer to make it seem cheaper, which is absolutely ridiculous. This demonstrates that both Apple and Dell are willing to mislead people when stating their prices. "
Just about any time you buy something it has a dollar or a penny missing.
Here's an example from toys r us: http://www.amazon.com/exec/obidos/tg/browse/-/172
and one from thinkgeek: http://www.thinkgeek.com/computing/handhelds/
It's just usual marketing.
Uh, in Mac OS X when you grab the disk icon of a removable media drive the trash bin changes into a universal ``eject'' symbol....
William
Sphinx of black quartz, judge my vow.
SPEC doesn't test Altivec performance.
:-D
It is all semantics, the word 'Marketing' comes from the Latin words 'marke' which means 'to create false impressions of / on / for', and 'ting' which means 'thing'. To create false impressions of this thing. Simple.
For three thousand dollars I will buy four stock P4/2.4GHz machines with half a Gig of RAM and a fast hard drive in each one, network them with a good 100Mb switch and hook them up to a KVM. I am sure a G5 can render faster than a single Xeon box, maybe not, but if you can distribute your tasks across multiple machines there is no way a single box is going to give you that kind of horsepower / money ratio.
Heck if you had four machines with better than average video and sound, you could host your own LAN party.
If you had friends.
Don't laugh. I have four computers. Sure wish I had friends.
Glonoinha the MebiByte Slayer
Everyone knows Macs suck and that Apples gay!(please note sarcasm)
It was a joke with origins in the earlier joke about Coworkers - some funny guy split it up into 'Cow' and 'orker'. Having done consulting in Insurance buildings I understand the 'Cow' reference, but what the hell is an 'orker'?
I guess you had to be there.
Glonoinha the MebiByte Slayer
Mac probably tilted the scores towards them, but who cares. The important numbers for me came about a third of the way down.
Dell Dimension 8300 $800 - $2500
Dell Precision 650 $1248 - $3680
Apple PowerMac G5 $2000 - $3000
For $500-1000 less, you get a machine that is either just a little bit faster or just a little bit slower than the Mac. There will always be a faster processor within half a year, but Intel gives you more for your dollar right now.
Aren't any of you suspicious of the 1GB FSB claim?
Apple said it's 1GHz DDR, so 500Mhz actual clock. Doesn't that seem uber-inflated?
The latest P4 get a 800Mhz QDR FSB. That's 200Mhz actual clock.
How can Apple (or anyone else for that matter) produce a CPU with an external clock that high?