MacSpeedZone has benchmarked the new Intel iMac, and they conclude something completely different than MacWorld did not too long ago. “We are pleased to report that our testing results show that the new Dual Core Intel iMac, which clocks in at 2X 2.0GHz is almost as fast as the current high-end Power Mac that has two Dual Core G5 processors running at 2.5GHz.” And so it seems people can’t seem to come to a consensus on anything related to the Intel iMac. Whether it be speed, or sales.
‘Intel iMac Is Almost as Fast as the Quad Core Power Mac’
80 Comments
-
2006-01-28 12:30 amrayiner
Take that mailing list posting with a grain of salt – not because Chris works for Apple, but mostly because there has been a big herd of madly skilled Linux and BSD hackers optimizing the **** out of gcc on Intel (and Xcode is mostly a fancy UI around gcc), where the PPC port of gcc hasn’t received nearly as much devotion.
That argument is misleading, since it conflates code performance and compile performance. Lot’s of people have been optimizing GCC’s code generation for x86, meaning that GCC generates better code on x86 than it does on PowerPC. GCC itself hasn’t gotten a lot of optimization attention until recently, and most of it has been directed at the C++ frontend and optimizer. These components, which currently dominate GCC’s compile-time, are fairly unoptimized generic C code, like the rest of GCC.
Of course, what this benchmark really tells us is that the G5 is a poor performer on generic, branchy integer code with lots of random memory references. This fact shouldn’t come as a surprise to anyone.
Edited 2006-01-28 00:32
-
2006-01-28 1:40 amGet a Life
Optimization was actually one of the purposes of the subtasks in the EGCS project. It’s also important to differentiate between different types of optimization. But that’s besides the point. At best the grandparent’s argument is that an Intel-built GCC is more efficient at compiling because it is built itself with a compiler more optimized for the x86. If one were to compile GCC with XLC would that satisfy the grandparent or not? If one were to explicitly disable all optimizations for both builds of GCC, would that be sufficient? What about linker considerations? Does he have any specific complaints there? Requirements of universal binaries for both?
-
2006-01-28 7:28 amrayiner
It’s interesting to look at the SPECint GCC results for the G5 and Pentium-M. The 2.26 GHz P-M gets 1994 in the gcc subtest, the 2.2 GHz 970FX gets 1130.0. Scaling for the clockspeeds used in the second benchmark I linked to (1.83 for the P-M, 2.7 for the G5), and you get the same 16% speedup reported in the posting. This implies that GCC compiled with XLC probably isn’t notably faster than GCC compiled with GCC.
The code generation issue is a factor, but its entirely hypothetical. Considerations of XLC or ICC are moot — GCC is the compiler Apple uses. The G5’s performance with any other compiler is largely irrelevent. Also, casting the blame on GCC entirely is not fair. GCC generates poor code on the G5 because the G5 is hard to generate good code for. It’s two cycle integer latency and group dispatch scheme requires a lot of smarts from the compiler.
-
2006-01-28 9:20 amstew
“GCC generates poor code on the G5 because the G5 is hard to generate good code for. It’s two cycle integer latency and group dispatch scheme requires a lot of smarts from the compiler.”
That’s in line with the RISC idea of making dumber but faster processors and letting the compiler do more. In my experience, the effect of that was also that the improvements you got from hand-optimizing code on a G5 (like placement of high-latency instructions, avoiding LSU rejects and pipeline stalls) were much bigger than on x86 CPUs.
Edited 2006-01-28 09:23
-
2006-01-28 2:05 amGet a Life
Since this is just an anecdote it’s not especially compelling. If it suggested the opposite or something Rayinder doesn’t expect to be true, he would criticize it as such. He might even say something about CINT2000 scores or what have you.
You dismiss the result because you don’t think the PPC970 backend of GCC is as optimized as the x86 backend. Even if that were the case, what would it matter? If you have a build cluster, you don’t really care if it could potentially, maybe, someday be competitive if you spent a large sum of money paying experts to improve GCC’s performance. If a build cluster of Core Duo iMacs is more efficient than a build cluster of quad G5s, then you’re just going to buy the latter. If that’s all they’re doing, and they do it cheaper while being marginally slower, that’s a no-brainer. They even take up less space! Though you’re probably just going to wait for the next XServe.
We can’t really say why this person obtains the results he does, or if they’re representative. We can’t because there’s not a lot of information. Maybe it’s because the machine he’s using doesn’t have enough RAM, maybe it’s simply demonstrative of his particular build, perhaps he’s running into some limitation of MacOS, perhaps the 4-way G5 scales about as badly as a Xeon, or perhaps the G5 really is just that bad at integer performance compared to the Core Duo.
Until someone bothers to perform real testing, no one knows.
-
2006-01-28 5:07 am
-
2006-01-28 7:45 amrayiner
Since this is just an anecdote it’s not especially compelling. If it suggested the opposite or something Rayinder doesn’t expect to be true, he would criticize it as such. He might even say something about CINT2000 scores or what have you.
The anecdote is what it is: an account of a person who uses both systems can offer a performance comparison based on their own usage. It could be more scientific, but its hardly any less useful than the tests Macworld conducted.
Also, I expect certain things to be true because I have reasons to expect them to be true. I know the G5’s compile performance is poor — I have one and an Athlon64, and its obvious to me which one compiles code faster. The SPECint.gcc results for the P-M and 970FX show the same thing. I posted the anecdote because I felt it was an accurate account of the G5’s performance in a “real world” situation. I’d give SPEC scores, but the PPC fanatics don’t believe in SPEC…
Edited 2006-01-28 07:47
-
2006-01-28 8:50 amGet a Life
Do you have a 2×2 G5 stashed away somewhere that you conduct useful comparisons with, or just the 2×1 G5?
As for the MacWorld tests, though I only vaguely recall the specifics, I remember them being crappy.
-
2006-01-28 8:53 amGet a Life
Is there some manner of magic dance I must perform in order to gain the privilege of being able to edit my comments, or must my lackluster language skills remain this obvious?
-
2006-01-28 12:26 pmnull_pointer_us
Neither, actually. It’s quite simple.
To enable editing, you must go into town and tell everyone you meet that OSNews.com is the best site ever while you jump up and down wearing a t-shirt painted with a giant green foot and your nick. When word of it gets back to the staff here, they’ll enable editing for you.
My advice: wear a mask.
EDIT: See, it worked for me.
Edited 2006-01-28 12:27
I just picked up one of these for my Mom yesterday and spent most of the day setting it up/playing around…I don’t care what the benchmarks say, the thing simply feels blazing fast, not to mention it’s oh so very easy on the eyes to look at. Very impressive indeed.
-
2006-01-27 9:50 pmrockwell
//not to mention it’s oh so very easy on the eyes to look at.//
Luckily, folks that use the computers I build DON’T MIND that their computer looks like … well, a computer.
Sheesh. Praise the technical merits, but I’m sick of fanbois telling me how “pretty” the iMac is.
It’s. A. Computer. Not. A. Sofa.
-
2006-01-27 10:01 pmjayson.knight
Ok, I’ll reword it; in this case form defines function. She had a very small space to work with (and was laptop only up until her iMac purchase)…the fact that Apple managed to cram all that power into such a small footprint is what’s impressive. I don’t mind the way most computers look either (2 large towers sitting under my desk), but one thing that has always helped sell Macs (besides the OS of course) is how visually appealing they are.
-
2006-01-27 10:03 pmpiquadrat
It’s. A. Computer. Not. A. Sofa.
It’s a computer you are looking at all the time while working with it. That’s why design is important for many users, even more so because absolute performance doesn’t really matter that much anymore.
-
2006-01-28 12:48 amsomebody
It’s a computer you are looking at all the time while working with it. That’s why design is important for many users, even more so because absolute performance doesn’t really matter that much anymore.
Yeah, and based by the factor that 90% of people has computer under the table? do they live under the table too?
-
2006-01-28 5:39 amApproachingZero
>>Yeah, and based by the factor that 90% of people has computer under the table? do they live under the table too?
Um, why do you think they hide it under the table? Because it’s big, it’s loud, and it’s ugly. People would much rather have the drives/ports within easy reach, but the design of these monsters make that impractical.
So why did they buy it?
Probably because when they bought it, they didn’t know there was anything better out there, or because they thought they needed Windows in order to email pictures and instant message. The vast majority of computer buyers in this country for the last four years have probably never seen a Mac since the ’90s, and they have never heard of nor seen OS X or iLife.
Going forward though, you’re going to see the looks of a computer become more and more important to users who are shopping for a replacement for that beast breathes fire under their desks. Apple just very recently became a viable and affordable alternative for these people, and I expect to see Mac sales growth outpace the industry average for the next five years straight because of it.
Think about it, what kind of computer do you want to have next? A Dell? An HP? A Lenovo? A Sony? Yawn. Boring. They all look the same, and they all run the same shit Windows XP. The only interesting new hardware and the only interesting new software are both coming out of Apple these days.
And I say this as a person who loved Windows in the ’90s, considered it the best OS on the planet, and called Apple “crapple”. I used to believe a Mac on someone’s desk meant they were an idiot.
I don’t think I was wrong. I just think Jobs and Ives have done a tremendous job in turning the company around in the last four years or so, and have completely leapfrogged Microsoft’s OS and PC makers’ boxes within the last two years.
Apple is the one setting the bar higher right now.
-
2006-01-28 8:27 amalcibiades
Well, here are the tradeoffs.
1) Buy an all in one mac, a bit slower, no upgrades except memory, can’t even open it. Disposable – can’t reuse the screen if upgrading. Can’t supply more than a few percent of the market even if they wanted it.
2) Buy a Powermac. Huge, heavy, noisy, unique perforated metal case. Probably can’t even supply one percent of tower case design demand.
3) Buy an SFF, like the Shuttle or Asus T series, desktop probably. Fast, reasonably upradeable, quiet, quite good looking, but there are limits to how many hard drives you can add. Swap the box and reuse your screen if you want.
4) Buy a tower and put it under your desk. A decent one will leave you with reasonable fan noise levels, unlimited upgrades, as many disk slots as you ever want, and it will be fairly inexpensive for what it is. Its what the world runs on.
In terms of the aesthetics, I find the Shuttle or Asus with a good screen more attractive than the iMac – that white plastic looks really tacky, to me. But people’s tastes differ. Class however is one thing the iMac does not have. Sex I guess is in the head of the user.
The thing I wish Mac advocates would look in the face when advocating their solution, is the practicality of it. The fact is, its completely irrational to advocate that everyone switch, which is the corollary of berating them for making a different choice, because if they all tried to, the Apple business model could not supply them. So of course they buy something else. They actually want to have a computer. Stop telling them to do something that collectively is impossible for them.
Or do you go around saying that everyone should drive a Caterham 5? Never heard of it? No.
-
2006-01-28 9:16 amApproachingZero
alcibiades, I think you forgot that the Mac mini exists and is much, much smaller than a shuttle. You didn’t mention it in your options. The Mac mini is the Mac that I see most people getting.
And there are no practical limits to how many hard drives you can add with either one, this being the era of USB 2.0 and all.
The problem as Mac advocates see it is that most of the people who have made the “choice” to buy a Dell/HP/Sony etc in the last four years really didn’t know that the Mac was an option too, so they really weren’t making a fully informed decision. I don’t tell people that they have to buy a Mac, but I do tell them that before they buy anything, that they should at least visit the Apple Store first so they can see what their options really are.
And since everyone I’ve sent to an Apple store ends up choosing the Mac after playing with OS X for ten minutes, I’m pretty sure that if Apple could get it together and advertise the Mac and OS X in a compelling way, many, many more people would also switch.
And with Apple’s choice to partner with Intel for its processors, chipsets and motherboards, the supply problems that have long dogged them should dry up once production gets ramped up. Intel makes enough chips to feed the Dell beast, after all. There’s no reason Apple couldn’t ramp up production quite a bit now that they’re not constrained by IBM & Motorola.
-
2006-01-28 12:04 pmSoulbender
“I don’t tell people that they have to buy a Mac, but I do tell them that before they buy anything, that they should at least visit the Apple Store first so they can see what their options really are.”
“And since everyone I’ve sent to an Apple store ends up choosing the Mac…”
If they went to the Apple Store first, played with Mac and decided to buy it they still didnt make an informed decision. To make an informed decision they’d also have to visit the PC store and play with XP and Linux too.
-
2006-01-28 6:35 pmApproachingZero
>>If they went to the Apple Store first, played with Mac and decided to buy it they still didnt make an informed decision. To make an informed decision they’d also have to visit the PC store and play with XP and Linux too.
These people have been “playing” with XP for five years now, I think they know what it does.
-
2006-01-28 2:13 pmhenrikmk
Um, why do you think they hide it under the table? Because it’s big, it’s loud, and it’s ugly. People would much rather have the drives/ports within easy reach, but the design of these monsters make that impractical.
Yup. But it’s not just the looks of the Macs that are important. Anyone noticed something else too about the Mac Mini and the iMac? Namely: Minimal desktop footprint.
You can hardly use less tablespace than with an iMac. It can even hang from the wall with a VESA mount if you want. I think thats pretty important for someone like me who is stuck with 2 19″ CRT monitors on my pretty big desk. Suddenly its not just design, but the sheer practicality of being able to put a lot more stuff on the table.
The Mac Mini on my table uses a little more, but it’s nothing compared to the PC tower I have under the desk, collecting dust and is hard to vacuum around.
It just a whole lot more practical and easier with a small box that happens to be nice looking than with a big box. That is definitely a good reason to buy a mac.
-
2006-01-28 7:45 amStephenBeDoper
Why are you looking at your computer while you’re using it? Personally, I tend to look at my display, but maybe I’m weird.
-
2006-01-28 11:27 ampiquadrat
Why are you looking at your computer while you’re using it? Personally, I tend to look at my display, but maybe I’m weird.
Have you ever seen an iMac G5? The display IS the computer. So, yes, I look at my computer while using it, and I’m very happy it has a beautiful, unobtrusive design (other than most cheap LCD displays you get with the usual computer from the shelf).
-
2006-01-28 8:40 amSoulbender
“It’s a computer you are looking at all the time while working with it. ”
Most people tend to look at the screen and what’s on it while working, not on their pretty case.
-
2006-01-27 10:11 pmmarcushe
That’s the mentality of PC user alright. Obviously, your customers have never seen an iMac. Class & sexy sells, as people want it.
-
2006-01-27 10:32 pmGet a Life
The iMac’s sales are not competitive with the sales of those much maligned boxes. So then does one presume that your idea of sexy doesn’t sell or that there are other considerations at play when people purchase computers?
Anyway, the iMac has an ugly white space at the bottom that doesn’t fit the look of the device. It would be much more attractive without it. It would also have much less space in which to fit its rather snugly-placed components. The iMac is otherwise an excellent form-factor for a computer if you’re short on space in the room you want to keep it in.
-
2006-01-28 11:35 amhalfmanhalfamazing
————–That’s the mentality of PC user alright. Obviously, your customers have never seen an iMac. Class & sexy sells, as people want it.————–
LOL!!!!!!
Hardly. All one needs is a http://www.lianli.com and apple’s “class” and “sexy” suddenly becomes….
Irrelevant.
Apple’s branding is surrounded by many myths, and this is one of them.
Lian Li makes much better cases than apple does. Like the PC 6070.
-
2006-01-28 6:42 pmApproachingZero
>>Lian Li makes much better cases than apple does.
Only someobody who works for that company could say something like that. Their standard cases look no better than a Dell, and their pyramid, snail shell and mini cases are extremely gaudy, and HUGE compared to the Mac mini. Lian Li also has no all in one offering that is in any way similar to the iMac.
-
2006-01-29 3:05 amhalfmanhalfamazing
I agree with you about the small ones and the snail shell.
However, their standard cases look much better than a dell. Especially a black-aluminum one.
I personally own a PC70.
And I’ve never gotten the appeal of the iMac. It’s definately “different”, but not “sexy” or “elegant”. The thing that makes a lian li sexy and elegant is not just it’s looks, but it’s function. It’s a big heatsink. 😛
(as well as being lighter, more durable, and easier to work on. no cuts on the fingers)
-
2006-01-31 12:24 amJohann Chua
An all-in-one design is a little much to expect of a case maker. Apple makes the whole widget, remember? Besides, I remember old Compaqs that basically had the iMac G5 design, only uglier. The Mac mini is very good for what it is, but PC users like having more choices when it comes to hardware, which the ultra-slim form factor limits; external peripherals kinda negate the whole point of making it so small, IMO.
-
2006-01-27 10:18 pmKelson
But then there are people who spend alot of money decorating their homes, picking nice furniture, etc.
They do not want to ruin the entire look of a room w/ a crappy looking beige box, and maybe they don’t want to have to hide their computer way out of the way because it’s ugly.
Oh…and just because some have different criteria than you doesn’t make them ‘fanbois’.
– Kelson
-
2006-01-27 10:44 pmmicroshag
“It’s. A. Computer. Not. A. Sofa.”
I’ll second that. Makes you wonder if these same people also look for elegant design in their toasters, blenders, vacuum cleaners and microwave ovens as well.
-
2006-01-27 10:49 pmrayiner
Um, yes? A toaster, blender, and a microwave sits in plain sight for everyone to see (much like a computer, hmm). Our toaster at home is a shiny chrome retro one, and our microwave has a nice stainless steel front that matches the refrigerator.
I wonder if people like you ignore elegant design in your shoes, shirts, trousers, suits, etc.
-
2006-01-27 11:01 pmGet a Life
Everyone has a different idea of what ‘elegant design’ is. I don’t personally care about clothing, and I simply dress to meet the expectations of others because it’s financially beneficial. Quite similarly I have numerous pieces in my kitchen that aren’t the same color as my refrigerator or my toaster. They are however spaced and stored in such a manner as to maximize my convenience when using them to prepare food. Preparing actual food is a task that I find far more interesting than how the instruments are colored. I consider their individual excellence in their respective tasks and the efficiency of their arrangement to be ‘elegant.’
-
2006-01-27 11:05 pmmicroshag
Dude, it’s a computer. The vast majority of the world has computers which you Mac users claim to be “ugly”. They don’t care and many of them are richer and probably have more style than you. They know it’s just a tool, and for the most part they don’t care what it looks like.
And yeah, for the record, I’m not “elegant”. I don’t need to impress with my wardrobe. That’s the yuppy way But I guess if someone either needs to make up for their personality or are not afraid to show how materialist they are, yeah…it’s all about style.
Edited 2006-01-27 23:07
-
2006-01-27 11:16 pmrayiner
Dude, it’s a computer. The vast majority of the world has computers which you Mac users claim to be “ugly”.
I certainly didn’t claim PCs were ugly. I just claimed that looks do matter to people buying computers (and toasters, etc). Judging by the fact that Dell, HP, Sony, etc, don’t sell beige PCs anymore, I don’t see why that’s such a controversial remark.
They know it’s just a tool, and for the most part they don’t care what it looks like.
If they didn’t care, Dell’s computers would still look like my old D300: http://www.andovercg.com/ebay/images/dell-dimension1.jpg
instead of like this:
http://www.ixbt.com/short/2k4-11/Dell_XPS4.jpg
And yeah, for the record, I’m not “elegant”. I don’t need to impress with my wardrobe. That’s the yuppy way But I guess if someone either needs to make up for their personality or are not afraid to show how materialist they are, yeah…it’s all about style.
Oh don’t lecture me about materialism. People buy nice looking things because they like to be around things that look nice. Most people are this way, to one degree or another. There is no point denigrating them for that.
Edited 2006-01-27 23:18
-
2006-01-28 4:21 pmyorch
“It’s. A. Computer. Not. A. Sofa.”
yes please, explain that to my girlfriend
she will pay 50% of the thing.
-
2006-01-28 10:17 ampphahnl
I think most people buy not so good looking computers, because they’re cheap – it’s the same with other things like sofas, toasters, stereo equipment, etc.
But there are always people who want better looking (“design”) stuff and who are ready to pay more for it. That’s the market for Apple – and in this regard, it’s for computers, what Bang&Olufsen or Bose are for Hi-Fi-Equipment.
Regardless, what we all *really* want to see is XP running on a Mac and a side-by-side comparison of Photoshop etc. This will either proove, or dispel 25 years of Myth and Lore from both sides.
And of course it’s faster anyway; the G5 is really really old. The MegaHertz Myth has, and always has been pretty much true; but the G5 went out of date a long time ago and Intel Processors surpassed the G5, MHz Myth and all. (So all you people proclaiming the MHz Myth bunk, wake up and realise that whilst the G5 stood still, Intel chips progressed)
-
2006-01-27 10:38 pmGet a Life
The “MHz Myth” is the matter of higher clockspeed=better. Were it true you would be clamoring for a Pentium D in your iMac, rather than a Yonah. Or for that matter, the Opteron would be at the mercy of the Xeon.
The moral of your story is that progress is better than its absence.
-
2006-01-28 12:01 pmhackenbacker
The G5 is NOT really, really, old.
The overall PowerPC architecture has been around some time, but the G5 was only introduced a couple of years ago. It hasn’t stood still either – progressing from 130nm->90nm production, and incorporating dual-core. As a 64-bit chip, and with HyperTransport bus (that is over twice the speed of Intel’s FSB) it holds two technological advantages over the Core Duo.
Intel processors have not surpassed it – they have more or less caught up. The advantage that Intel does hold over IBM/G5 is in the manufacturing process.
-
2006-01-28 5:25 pmrayiner
Intel processors have not surpassed it – they have more or less caught up. The advantage that Intel does hold over IBM/G5 is in the manufacturing process.
Intel processors surpassed the G5 (in desktop stuff) shortly after it came out. Moreover, its silly to say that Intel holds the advantage over the G5 in manufacturign process, since the Core Duo runs at slower clockspeeds than the G5. In reality, the Core Duo is dramatically faster than the G5, in integer code anyway, simply because its a much better design.
-
2006-01-28 8:24 pmGet a Life
It isn’t the age of the G5, it’s the design space of the processor. Clock-for-clock Yonah is better at integer code. It’s weaker in vector processing. This isn’t simply because of the differences in the AltiVec and SSE2/3 instruction sets. It has sophisticated power management, rather than fitting into its thermal profile simply because of its manufacturing process. I’m intentionally ignoring the compeitiveness of the Pentium 4 and earlier Pentium M releases.
The thing Intel has over IBM really is an interest in designing general-purpose desktop processors. You’ll notice that IBM happily sells Opterons to segments that don’t want POWER. And that they happily design and sell embedded and specialized processors in the PPC family.
Let me laugh a bit: 2 x 2.0 Ghz faster than 4 x 2.5 Ghz? 4 Ghz faster than 10?
You don’t even need to take account into details (ej: cache size, etc); dude, Core Duo IS fast, bust is not THAT fast.
It’s easy to get 4x comparisons comparing macbook pro against a powerbook: Powerbook is a single G4 processor (not even G5, not even a low-power G5) and core duo is a modern processor with *two* cores (ie: SMP, etc)
It’s easy to get 2x comparisons with the G5 imac since G5 only has 1 cpu, not 2 like core duo. Apple (not suprisingly) didn’t update the imac with a dual core processor like it did with the powermac
Oh but benchmarks like this will make second-hand powermac quads cheaper….BRB
-
2006-01-28 12:06 pmhackenbacker
Two points you missed:
– the PowerBook isn’t even a dual-core G4, which are specced to run at the same clock speeds and with the same FSB as a Core Duo (and with a similar power signature too)
– the G5 in the iMac always had it’s FSB vastly underclocked compared to what it could be (667Mhz, vs. 900Mhz+ in the PowerMac). The fact that the FSB is the same as the new Intel iMacs contributes to the benchmarks.
…Apple is the new Ubuntu.
Why so many articles? I really don’t care.
What’s with the name “core duo” anyway? What was wrong with “dual core”? It’s like they decided that they needed something more “touchy feely and hip” for the Mac audience.
-
2006-01-27 11:35 pmAndrew Youll
can’t trademark common phrases apparantly…. *looks at MS… and shrugs sholders* Dual Core is a common phrase.
“We are pleased to report that our testing results show that the new Dual Core Intel iMac, which clocks in at 2X 2.0GHz is almost as fast as the current high-end Power Mac that has two Dual Core G5 processors running at 2.5GHz.”
I don’t think so, if they properly utilize the full capacities of both the Quad will stomp the iMactel processor wise. In fact most DP PowerMacs will too.
I’ve been watching this pretty closely and I have concluded there is a lot of FUD and counter FUD going on.
One of the reasons for the FUD (even from Apple) is that the new Core Duo’s are not all that much of a performance leap actually if compared to equally matching G5 processors.
However they do run cooler, which makes them attractive for other reasons.
Take a look at the X-Bench scores, which test all parts of the machine independantly. The overall score is not much better than a Single iMac G5 and less than the baseline of a DP 2 Ghz PowerMac.
The FUD and counter FUD is designed to mask this reality to avoid a serious loss of sales if the facts were told.
How a iMactel Core Duo beats a Quad is beyond me.
http://spiny.com/forum/viewtopic.php?t=342
http://db.xbench.com/compareindex.xhtml
http://media.99mac.se/g5_dualcore/
http://www.barefeats.com/quad02.html
Edited 2006-01-27 21:56
-
2006-01-27 9:59 pmrayiner
One of the reasons for the FUD (even from Apple) is that the new Core Duo’s are not all that much of a performance leap actually if compared to equally matching G5 processors.
Define “equally matched” G5? There *are* no dual-core G5s that would fit into an iMac! The dual-core 2.3 in my Mac runs hot, and its got a giant heatsink and four fans! Putting one into an iMac would melt the damn thing. A single G5, or a high-clockrate G4 is the equal match of the Core Duo, in terms of things like die size and power dissipation. The 970MP is a 170mm^2 chip that dissipates ~100W. The Core Duo is a 90mm^2 chip that dissipates around ~30W. In something like an iMac, where you’ve got tight cost and power dissipation constraints, a single G5 is barely a fair competitor to the Core Duo, much less a dual-core one.
As for XBench, its bunk. XBench says a 1.42GHz Mac Mini is as fast in the CPU subtest as the 3.8 GHz DTK machines. It says the Intel Macs have very low UI performance, but everybody who uses one reports that they are very snappy.
Edited 2006-01-27 22:01
-
2006-01-27 10:19 pmJustAnotherMacUser
It’s obvious I’m discussing performance.
Sure a Core Duo can do more work than a Single G5. But it’s not nearly as fast as a Quad (or most DP PowerMac’s) like the article mentions.
Performance per watt is irrelevant, case design is irrelevant.
So is your opinion of X-Bench.
The user interface “feel” is much faster, but “feel” doesn’t neccesarily translate into performance under load.
One can make a PowerMac G5 “feel” snappier by simply using a RAID O pair or a single 10,000 RPM for a boot drive and upgrade their video card from the slow stock one.
Slow boot drives is what makes Mac OS X “feel” slow
-
2006-01-27 10:42 pmrayiner
It’s obvious I’m discussing performance.
That’s not how things work in the real-world. If we were discussing just “performance”, we’d be comparing the 970MP to an Opteron (a comparison it would lose). The iMac is a small-form-factor machine. It has a thermal envelope that cannot be exceeded. Arguing about solutions that exceed that thermal envelope is meaningless — its completely hypothetical.
Sure a Core Duo can do more work than a Single G5. But it’s not nearly as fast as a Quad (or most DP PowerMac’s) like the article mentions.
Did you read the article? That’s not what the article says!
Performance per watt is irrelevant, case design is irrelevant.
Sure its relevant. Consider the P4? With its long pipeline, its thermally limited, and can get up to nearly 7 GHz with appropriate cooling. But nobody considers 7 GHz P4s because the power and cooling requirements are totally unrealistic. We’re talking about iMacs here. A dual-core G5 is about as relevant to the discussion as a 7 GHz P4.
The user interface “feel” is much faster, but “feel” doesn’t neccesarily translate into performance under load.
You’re grasping at straws. Is it really easier for you to believe that XBench is correct, even though it tells you something that is contradicted by experience? XBench says a 1.5 GHz G4 is faster than a 3.8 GHz P4. Are you seriously suggesting that this is an accurate assessment? Isn’t it easier to just admit that XBench may just be broken on the Intel Macs? Historically, its been poor comparing Macs that spanned different architectures. The G5s looked much less impressive relative to the G4s based on their XBench scores.
Slow boot drives is what makes Mac OS X “feel” slow
I’ve had a Raptor as my boot drive in my DC G5 (until I took it out because the noise was unbearable). It doesn’t make the machine feel appreciably faster.
-
2006-01-28 2:21 amPowerMacX
“Performance per watt is irrelevant, case design is irrelevant.
So is your opinion of X-Bench.
The user interface “feel” is much faster, but “feel” doesn’t neccesarily translate into performance under load.“
X-Bench is NOT a trustworthy benchmark. Period.
The iMac Core Duo has an ATI Radeon X1600 GPU, the iMac G5 has an ATI Radeon X600 – and the 20″ Core Duo has 128MB GDDR3 (with an upgrade option to 256MB) vs. 128 DDR for the 20″ G5.
Remember Quartz Extreme? (silly name)
-
2006-01-27 10:04 pmJustAnotherMacUser
Here’s some actual user results from a Quad
http://www.tow.com/2005/11/25/quad-g5/
-
2006-01-27 10:11 pmKelson
If you had actually read the article, you would never have written this post. They state that according to the way Macworld did their benchmarking, a Quad-G5 and iMac Duo are similar in performance. They then go on to completely debunk that and make clear they are not really saying a Quad-G5 and iMac Duo are on the same level performance wise, but that Macworld had flawed testing methodology.
You are tilting at windmills.
– Kelson
I know, I know, many people are so hyped and happy to see some really good looking shiny things appear to market (And I want to point out – I am happy about that too, so…). Apple, for good or worse, has created their momentum with their products and, want to believe it or not, they deserve their share of fame.
Personally I just happy that anyone talks about Apple which releases REAL products and crunches OS X releases one after another, improving their server solutions, etc. It is just feels strange that company which are actually are NOT in software business have just rockin user friendly operational system. So, more power to Apple. More power to Ubuntu. More power to Fedora project. And ecetera. Because they have REAL stuff for users coming, not just plain marketing agenda.
BUT have said all this, I would like to ask one thing for Apple fanboys and haters – please, stop beating this dead horse. It is over. Apple has momentum and whatever you want to spin it or not, nothing will change it. Like Google, Apple have worked hard to achieve this, so…
Just shut up. If you have something to say about OS X, or iPod design analysis, or something else..then ok. But please stop “Intel will be death for Apple/Apple will own the world/We will get OS X86 on common PCs” discussions.
Intel won’t be death for Apple. Maybe they can miscalculate something, but I really doubt that.
Apple won’t own the world. They can own home entertiment center market, however.
And no, we won’t get official support on common PCs for OS X86. Just won’t.
Did anyone discussing the headline on OSNews see how I used quotation marks?
Ok, thanks for noting.
-
2006-01-27 10:37 pm
-
2006-01-27 10:38 pmThom Holwerda
What is this punctuation and grammer of which you speak?
Be careful with those multi-syllable words here .
-
2006-01-27 10:47 pm
-
2006-01-27 11:08 pmThom Holwerda
Actually, the word is ‘multisyllabic’.
Thanks for pointing that out. I was in doubt (English isn’t my native tongue, you see).
-
2006-01-28 12:14 amnull_pointer_us
>> Actually, the word is ‘multisyllabic’.
Thanks for pointing that out. I was in doubt (English isn’t my native tongue, you see).
Actually, I think you’re both ignoring proper punctuation here.
When you name an English token (or a single character, for that matter) such as the word multisyllabic in a sentence, you should use italics. Quotation marks are to be used for words or phrases with nonstandard meanings and for citations (of course).
Additionally, parentheses are to be used for content which fits grammatically into the sentence while dashes are to be used for things which are interjected – this independent clause is one such example – into an otherwise-correct sentence.
Someone will probably reply and correct me, too. Don’t you just love English? 😉
-
2006-01-28 1:34 amDittoBox
We don’t need to get this technical, this isn’t slashdot.
Thanks however for the pointers, and I believe you are correct in (most of) your assessment.
There’s some interesting commentary on the use of quotation marks in the computer world here:
http://www.catb.org/~esr/jargon/html/writing-style.html
Though it’s not entirely the problem I think consistent use of quotation marks rather than italics stems from the obvious reason that most communication mediums of the day -and indeed many today- do not/did not have the ability to communicate in anything but plain text.
Also might I add that you used the dash incorrectly (at least for American English). The dash you’re using (again, we’re facing a technological problem, no em dash on the keyboard) is an en dash (I think). The proper usage is to provide a space between the initial sentence and the interjection, then an em dash, then immediately following that you place your interjection text, no spaces.
There, I’m done.
-
2006-01-28 3:15 pmbiteydog
I was always taught that the word was ‘polysyllabic’. But then I am a gigayear old and half-Scottish, so I’m maybe out of date. (I was also taught never to start a sentence with ‘but’ as well – so that’s put me in my place.)
This author draws attention to slowness (a problem) in the software of apple’s encoder not a slowness in the hardware. Notice what he said when he was using G5 quad system and then the %CPU time was just ~50; this is a problem of the lazy encoder which was not able to really benefit from the other half of that system, leading to effective similar performance like that for Dual core systems wheather its IBM or Intel.
As I always noticed in the computer industry, software in general is way behind hardware in optimizations and excellence.
this is getting rather lame. We know for one that it’s faster than the old G5 iMac, because it has a 2-core chip. Second, who cares about anything else, besides getting some pro-apps(get off ur butt ADOBE…u should of had a cocoa app out before).
Personally, I think this processor will give any 970 a run for it’s money!
What a great article. So a Intel iMac could beat a quad PPC G5; if the software was rewritten to remove the “bottleneck” and take full advantage of the Intel chip. Impressive. I imagine this new insight will have quad G5 owners running for the closest Apple Store.
The i-mac intel is nicely prized.Saves you a lot of money compared to the 2.5GHz quad mac.
Okay. Can we PLEASE stop now? Mod me down as much as you want, but i’m getting tired of tons of imac benchmark articles “it’s slower”, “no, it’s 100,00001% faster!”, “no, it’s 300% faster”.
Please… No more…
I couldn’t agree more. This and “We have to get XP running on a Mac!!” is starting to get old. I think people get that people for whatever reason, want to get XP/Vista running on a Mac. So please…
… it’s all just starting to get rather old.
agreed
Personally, I believe that any performance differences can be directly attributed to the sorry state of compiler optimizations(in gcc) for any architecture other than x86.
(AFAIK Apple used gcc for powerpc applications, and I’d assume for x86 development that Apple would/will be/is using Intel’s own compiler optimized for their x86 variant. Even if Apple is using gcc still for OSX(they may, as they may have relied on gccisms and may not have been able to add Objective C support to Intel’s compiler yet) they should still see a significant performance increase for x86 architecture. Not to mention that in many applications there exist hand coded/optimized x86 assembly, where no such code exists for other platforms. All of this doesn’t even drag in how Apple handled endian problems, or how vendors supported them, i.e. either stupid/easy or smart/slightly difficult. I’d bank on the former considering Apple’s market share.)
The good thing about these particular “benchmarks” is that they, apparently, relied on application type benchmarking, which SHOULD show true performance numbers, although it IS modified by the original premise of poorly optimizing ppc compilers. (i.e. They should’ve used IBM’s own compiler suite probably…)
(i.e. They should’ve used IBM’s own compiler suite probably…)
IBM’s own compiler suite is irrelevent. Version 6.x isn’t substantially faster than GCC (on the G5), and while later versions are substantially faster, they do not run on MacOS.
Later in the article:
“This is where the Macworld “First Lab Tests” article falls a little flat … obscuring the processor capacity vs processor usage problem inherent with mutiprocessor machines (or multi-core … same difference). Using Macworld’s logic we could argue, given the data above, the Quad G5 Power Mac is only 14% faster when running some of Apple’s own applications. We think that this is misleading, as we pointed out.”
The “new Intel iMac is as fast as a quad” statement is a parody of the conclusion you can reach if you use Macworld’s procedure for evaluation its performance. Clearly, they thing that such a conclusion is unfounded. On the other hand, the new Intel Mac *can* be as fast as a quad:
http://lists.apple.com/archives/xcode-users/2006/Jan/msg00680.html
That message isn’t particuarly ‘scientific’ even by benchmark standards.
Read the rest of the thread. He points out a large C++ build is only ‘seconds’ faster on the Quad. More concrete numbers are here:
http://www.gusmueller.com/blog/archives/2006/1/17.html
A Core Duo 1.83 beats a dual 2.7 G5 by a significant margin in the compile test.
I’ve already read the thread from before, because you’ve posted it before. It’s just anecdotal. Take that second link where there’s a 13.9% difference between the dual G5 and the Core Duo. If you were to expect the quad G5 to only be a ‘tick or two’ faster than the Core Duo, we’re suggesting maybe a 15% difference between the quad G5 and the dual G5.
What would be useful is the building of the same projects (more than one, and preferably several) on all three machines several times.
You have to be careful with the numbers. The link compares a dual 2.0 to a quad 2.5 G5, the second compares a dual 1.83 to a dual 2.7 G5.
The dual 2.7 gets 316 seconds, the dual 1.83 core duo gets 272 seconds. Assuming a linear scaling of gcc speed with clockspeed (iffy, but adequate — GCC is a CPU hog), we could estimate that a dual 2.5 would get 341 seconds and a dual 2.0 core duo would get 249 seconds. So if the quad is a hair faster than the dual 2.0 core duo (let’s say 5%), that suggests it might come in at around 237 seconds, or 44% faster than the dual 2.5. That’s entirely realistic SMP speedup.
Yes, the above results are extremely approximate (GCC is in reality spends about 70% of the time on the CPU, 30% doing I/O, depending on the project), but it suggests that the speedups each person encountered pass a basic sanity-check.
I was assuming the Core Duo in both cases were equally clocked (perhaps I should have looked at the link again after all). Your estimation is somewhat rough, though from my testing with an Opteron at various clockspeeds assuming linear scaling predicts results with only a 4% error with one of my test cases (building q3a). The other results aren’t as good, but that’s not especially important. The scaling from the dual G5 to the quad G5 with your estimates is rather poor. The quad G5 should be given more RAM and then compared with a dual G5 and a Core Duo iMac. Then compared on different code bases, preferably all freely available so that testing can be reproduced.
On the other hand, the new Intel Mac *can* be as fast as a quad: