Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Mav451 said:
I dunno, they are both roadsters...not exactly meant with "luxury" in mind. Personally, I don't care how much the Honda s2k costs, I would take it anyday over bmw. It just so happens that it is also cheaper to buy, faster, and cheaper to maintain (bimmers are notorious for high maintenance costs)

Just my nickel.

I am a Honda buyer since 1987 (when i turned old enuff to drive)

Loved the look of the S2000, but the Z4 gives me something the S doesnt. It gives me leg room. :)

So for me I HAVE to buy the Z4. Knee problems are not worth saving 10 Grand.
 
Loyalty vs. Commerce

uzombie said:
Sorry to jump into this flamewar but...hasn't anyone noticed that 1000 G5's takes away from the 1000 consumers that really NEED a G5 Xserve (or the processor for their box)? Afterall, its your loyalty and purchases that contributed to the Mac. Since Apple dropped the ball on the .edu, I guess you can wait...

If I was selling anything and had a choice of delivering 1000 units to one customer vs. accruing 1000 orders from loyal users, I wouldn't thomk twice...the bug buyer gets them. It helps the bottom line, it helps work the bugs out of the new product, and it drives down the relative research/production costs on subsequent machines.

If you want to see a company go out of business, look for one that puts large volume purchasers at the back of the line.
 
Not getting it

I've read through most of the messages here, and still don't quite understand what all the fuss is about. I'm getting a new Mac soon, and while I wouldn't mind waiting 6 months for another revision or improvement, I can afford it now. I can only justify about $2000 for the new computer, so the new Dual 1.8 looked really nice. Correct me if I'm wrong (which, I admit I may very well be), but 2x 1.8 is a lot faster than 1x 1.6, which was my main concern. I'm using a 500MHz G3 now, which still works, but is definitely showing its age, and it's not that I'm running out of memory or need a better graphics card. So maybe I'm exactly who Apple was targeting for the 1.8 model. That being said, I was going to upgrade to 512 Mb RAM and the 9600 Video Card.

So I guess my question is, what is the difference between the motherboards? And what impact would that have unless I'm really pushing the system?

By the time I need 4 Gigs of RAM, I'll have bought a new computer anyways. (Maybe in 6 years.) $50 doesn't seem like much to upgrade the video card to something better. Granted, yes, it seems like poor marketing; lure the customer in with a lower price, then charge a lot to upgrade to what it should have come with. But, is there any reason the actual hardware on the 1.8 is inferior?
 
exodius said:
I've read through most of the messages here, and still don't quite understand what all the fuss is about. I'm getting a new Mac soon, and while I wouldn't mind waiting 6 months for another revision or improvement, I can afford it now. I can only justify about $2000 for the new computer, so the new Dual 1.8 looked really nice. Correct me if I'm wrong (which, I admit I may very well be), but 2x 1.8 is a lot faster than 1x 1.6, which was my main concern. I'm using a 500MHz G3 now, which still works, but is definitely showing its age, and it's not that I'm running out of memory or need a better graphics card. So maybe I'm exactly who Apple was targeting for the 1.8 model. That being said, I was going to upgrade to 512 Mb RAM and the 9600 Video Card.

So I guess my question is, what is the difference between the motherboards? And what impact would that have unless I'm really pushing the system?

By the time I need 4 Gigs of RAM, I'll have bought a new computer anyways. (Maybe in 6 years.) $50 doesn't seem like much to upgrade the video card to something better. Granted, yes, it seems like poor marketing; lure the customer in with a lower price, then charge a lot to upgrade to what it should have come with. But, is there any reason the actual hardware on the 1.8 is inferior?

You're not missing anything. The dual 1.8 is a great computer. If you don't mind a refurb, you might look at the dual 2.0 on Apple's web site for $1999. If you want a new one, consider one of the catalog places who gives free RAM, printers, etc with the computer.
 
jragosta said:
Oh, no.

Tortoise will insist that they don't know what they're doing, either.


Au contraire, I would insist that they only intend to use it for a very narrow application space. The Earth Simulator in Japan is even more extreme in this regard, being narrowly tailored for only a tiny number of codes.

Being fast for one type of code does not make a system generally fast. DSPs are extremely fast for some codes, but we don't make computers out of them. The G5 is like this in terms of its application to supercomputing. Fast in a very narrow space, but not fast in any kind of general purpose sense.
 
G5 SPEC scores?

maverick13 said:
The Opteron looses to G5 in floating point performance.


What are the SPEC numbers for the G5?

For Opteron 2.2GHz:

spec2000fp: 1691
spec2000fp_base: 1553


I can't seem to find PPC970 numbers that beat this. Just curious...
 
tortoise said:
What are the SPEC numbers for the G5?

For Opteron 2.2GHz:

spec2000fp: 1691
spec2000fp_base: 1553


I can't seem to find PPC970 numbers that beat this. Just curious...

ROTFLMAO.

My 'I know everything about supercomputers' wants to rely on a benchmark that's probably more subject to compiler oddities than anything else and which isn't even capable of handling multiple processors.

Forget SPEC. It's a lousy benchmark-unless you happen to be using one of the SPEC applications every day.
 
LOL - true words !!

jragosta said:
Forget SPEC. It's a lousy benchmark-unless you happen to be using one of the SPEC applications every day.


for XX from set {SPEC, Photoshop, Linpack, Maya, FCP, Quake, ...}
[
write ("Forget $XX. It's a lousy benchmark-unless you");
write ("happen to be using the $XX application every day.");
]

True words, but it applies to all benchmarks - not just SPEC !!
 
-edit-
Ok, I see you mentioned the Civic in there. The civic, is not bought because it is "cheap". You can get Kia much much cheaper. But most people will not do that because they KNOW that the Kia will be ultimately unreliable, with a higher TCO in the end. People buy the civic for its high MPG, reliability, and even for the personality. I rarely hear people buying a Civic "because its cheaper". Hell, you can buy a Focus if you wanted to do that.[/QUOTE]

Yeah your right a civic is probaably the cheapest car 5 years down the road. All the kids at my high school have one because of low cost (intial) good realiability and great milage. A civic is just a great value. ;)
 
jragosta said:
My 'I know everything about supercomputers' wants to rely on a benchmark that's probably more subject to compiler oddities than anything else and which isn't even capable of handling multiple processors.


No better or worse than any other benchmark, except your own code. Interesting to see your sudden loss of interest in benchmarks. SPEC gives some kind of baseline and at least SPEC is somewhat general (unlike LINPACK or Photoshop). Since the multi-processor architecture of the Opteron is superior to that of the PPC970 (one thing that I don't really think is arguable), multi-processor benchmarks won't really help prove that the Opteron is weaker.

I haven't found SPEC fp benchmarks for the IBM compiler, so it is hard to say how fast the PPC970 can be driven. The Opteron benchmark I quoted was done with the Pathscale compiler, a new Opteron specific high-performance compiler. Use any compiler you want, any oddities will come out in the wash. Hell, the LINPACK benchmark that you are in love with has the "oddity" of measuring how fast a system can do mult-add operations, and it is also a lot narrower benchmark than SPEC. You have not provided a single bit of evidence that the G5s are terribly good at anything other than DSP benchmarks.

But all that doesn't really matter. The most damning benchmark is STREAM, which is much more architecture neutral and one of the most important benchmarks used for supercomputing purposes. The very worst compiler performance for Opteron is a full 50% better than the best compiler performance for the PPC970, including the IBM compilers.

Unless you actually plan on running LINPACK, Photoshop, and similar DSP-ish codes on your supercomputing cluster, you won't do well with a G5 cluster that is only capable of benchmarking well on those types of codes. Pick the right tool for the job. If you are running DSP codes, then you should go with a G5 cluster.

I've always acknowledged that the PPC systems are extremely good for DSP codes, which they are, but then the ignorati try to automatically translate that into meaning they are good for all other kinds of codes too. Some people are running supercomputing codes that will excel on the PPC architecture, but the majority are not.
 
jragosta and dr.Zoidberg:

It's a pleasure to read your two comments. However, scottkle does bring up an interesting point.

scottkle said:
I never agreed until now. If there was a "cheap" Mac. One that could get PC users excited. One that is at THEIR price point.

Scott, there are two problems with this argument:
  1. It assumes the average PC user wants to switch platforms.
  2. It ignores the fact that the system purchase is only the start of the cost, regardless of price point. The potential switcher also will have to buy new versions of his/her software, and quite possibly replacement hardware, too.

One that makes it seem like they are getting a deal, Apple would increase that marketshare 2x. It's only a matter of mindset in the PC world. No one hates Apple. They just don't see why they should spend so much money on something they can get for so much less.

Now you're getting hot... You see, the problem is that people in the PC world have been practically given hardware for so long that when they actually have to pay for it, they don't know how to react. Even if the end users don't think of it as such, companies in the PC world have practically been prostituting themselves just to get their hardware noticed.

Also, a lot of people try to compare (pardon the unintended pun) apples and oranges. On the one hand, they can buy lots of parts and build their own system fairly cheaply, and on the other they're looking at a whole, brand-named system.

Typically, there are two reactions amongst the PC user world. One is just the perception that Macs cost more than PCs. Again, they find ways of getting a "cheap cost" on the PC side and deliberately compare it to a whole, manufactured system. The other is from those who don't mind building their own systems and -- naturally -- cough up a fur ball over the fact that you cannot do that with a Mac. You can dismiss them as being "tinkerers", but they exist nonetheless.


I have been using Macs since 1983. I guess that when you haven't been, you don't understand what all the fuss is about. My friend asked me why he'd pay $2,799 for a 17-inch laptop with 1.5 ghz as the speed. When I explained to him what else was in those specs, it all went past him like a train. Didn't get it at all. It was 1.5. That's all he was able to take from it. I believe all the years of hearing the megahertz myth has really brainwashed him and 92 percent of the PC market. That's what they believe.

Ding ding ding! Got it in one, Scott.


The switcher campaign on the surface was a great idea. But, what Apple should do, in addition to coming up with this "cheap" Mac, is educate the masses en masse. In some way that could get through to people like my PC buddies.

The Switcher campaign was actually one of Apple's more successful efforts, both in confirmable numbers and in anecdotal stories of changing people's attitudes. It's amazing how many people I have spoken with who connected at least on some level with those ads. Even hard-core PC techs, albeit not of the switcher variety, got a good laugh at how on-target those commercials were in their criticisms of Microsoft's OSs, past and present.

However, there is something to be said for the ignorant masses out there (some might refer to them as young and old skulls full of mush) who truly believe that "cheaper is better". I have already given a long dissertation on this elsewhere, so I won't repeat what I've said before, but the essence is that there are people you'll never convince because you're not offering them anything they have the knowledge, understanding, comprehension or intelligence to want at the price point you're offering it at, regardless of any other justification for purhcase. The only thing you can do is forget them and move on. If, however, that group comprises the whole of a given market, then you're basically screwed. Apple is, fundamentally, banking on the computer-using market not being entirely comprised of idiots like that. So far, they are being proven correct in that assertion.

A brief example I could use is this: Paramount has decided that they want to show the origins of the Star Trek mythos by going back to the beginning of things -- hence Enterprise. So, we now have a show who's events take place starting 10 years before the founding of the Federation (show starts 2151, Federation founded 2161), and approximately 100 years before Kirk and Co.

It's a world where people are afraid of transporters because it is a new technology and the thought of having your body broken up into sub-elementary particles and reassembled is scarry because that kind of transport is not yet commonplace. People take shuttles everywhere. Ships dock to exchange personnel. Tons of alien planets and races are question marks or out-and-out unknowns. Enterprise "polarizes it's hull plating" instead of raising shields -- because they don't have shields -- because they haven't been invented and/or discovered by Starfleet yet. "Photonic" torpedos are just coming into use, and only the inventiveness of Lt. Reed has increased the power of their "phase cannons" to a useful level.

In other words, this is a "where we came from" type show.

Now, there are a lot of viewers out there who are dissatisfied with the show because it is "primitive". Comments I've heard are "Why are they all so stupid?", etc. The point is a large section of the (potential) viewing audience lacks the mentality or capacity to "get it". I respect Paramount for staying the course with this show instead of just giving up, but the truth is that unless they can generate enough of a viewing audience, this kind of public mentality will kill the show in time.

It's a bit like having a show featuring, say, the British Navy of the 1700s and 1800s and having viewers complain that they aren't using GPS systems on the ships.

Of course, the difference here is that Paramount only has investment costs involved with the show, and if it goes south, too bad, but ultimately no big deal. Apple has their entire business tied up in the Mac platform.

Besides, as I've said above, there is a lot of truth to be had from the fact that Apple's development costs are higher than other PC manufacturers', and in any event, it isn't like this stuff is free for the PC companies, either. Good PCs (and I use that term loosely) are expensive, too. Consider Sony's VAIO line of highly multimedia-oriented PCs, or Alienware's systems. They're not cheap, and they're not ever going to be cheap.

To an extent, maybe Apple is wise to occupy "niche player" status because it means that the masses of dead-heads who are a part of the WinTel world are idiots that Apple doesn't have to take care of.
 
Travis Novak said:
Are you out of your mind?!?!?!?!?
If price always won would there be any BMWs in their "fancy cases"? Lets compare stats...

bmw z4 3.0i: $45,000
leather interior
power everything
automatic softop
5 speed auto transmission
0-60: 5.9

honda s2000: $32,000
leather interior
power everything
automatic softop
6 speed manual transmision
0-60: 5.4

Now we have a compareably equipped honda for tons less than the bmw. It even outpreforms the bmw. But when you drive both you know the honda's not a bmw. There's certain small things that make the bmw what it is. Some intangable, some probably mental, but that alone in some people's minds is enough to make the bmw worth it for them.

Furthermore Apple provides the most complete package you can get. Unlike windows it includes a great set of bundled software and good realiable hardware. If price won all the time we would see everyone in a Honda civic, in a one bedroom apartment, eating ham sandwiches three times a day. Some people need a little bit better... so they buy a mac!


FIRST OFF,, it's one thng to compare PC's to Mac , but don't even go there with me when it come's to cars that's my specialty.

only a complete IDIOT would pick a an over priced Z4 (looks like an over priced Mazda Miata), over a complete Kick ass car like the S2000 (one of my personal fave's). Honda's are much cheaper to maintain (someone already mentioned that). S2000 is an incredible car by any comparison , the damn thing is a technical marvel, these to cars are not even in the same leauge.

there is more to a car then just 0-60 times , the S2K has world class handling , and a high speed ride that's just makes u feel like ur about to explode.

anyone who buy's a S2K dosen't see it as a Honda like you do , they See ABSOLUTE VALUE a kick ass Sports car at an incredible price, just like with the 350Z, Lancer Evo, WRX STI and SRT-4.

basically 1 car is to rich Daddy's Boy's who want to impress thier gold digging Girl friends with a BMW logo, the other is for real car enthusiasts who want to tear down the road.

i would easily buy a MR-2 Spyder for 25K or a usedZ3 (they looked better even 007 drove 1) before that crappy Z4.
 
agreed

MikeTheC said:
...To an extent, maybe Apple is wise to occupy "niche player" status because it means that the masses of dead-heads who are a part of the WinTel world are idiots that Apple doesn't have to take care of.

By dead-heads you dont mean those of us that enjoy Jerry Garcia, right?

Also, i agree with both your conclusion and example...The funny thing is, people dont even realize how Apple pushes the industy...i gave a presentation in my communication course last semester on why you should buy a mac...i compared my Tibook with a new Dell....they have a similar style, but of course mine is from January 2001...they just arent aware...i could go on and on with them, but its like talking to a brick wall sometimes...
We could give out free cool-aid in every city and display Macs at the same time, maybe that would work!
 
tortoise said:
What are the SPEC numbers for the G5?

For Opteron 2.2GHz:

spec2000fp: 1691
spec2000fp_base: 1553


I can't seem to find PPC970 numbers that beat this. Just curious...

http://www.apple.com/xserve/
dual specs in gigaflops/sec

Xserve Dual 2GHz PowerPC G5 9.0 Gigaflops

IBM eServer x325 Dual 2GHz Opteron 5.91 Gigaflops

Pretty much clear.

PS: TRIPS is the best processor :eek:
 
MikeTheC said:
jragosta and dr.Zoidberg:



To an extent, maybe Apple is wise to occupy "niche player" status because it means that the masses of -heads who are a part of the WinTel world are idiots that Apple doesn't have to take care of.

Reality check here. Keep in mind that the 'masses of -heads who are a part of the WinTel world' use their computers to do basic stuff like word processing, number crunching, email, web surfing, etc. Their computer is like an appliance. Your toaster makes toast, your computer allows you to shoot off an email. That's all it is; it's not a way of life. For most people, computers are a value not a nuance-play. ;)
 
completely ridiculous specs - from Apple, of course

maverick13 said:
http://www.apple.com/xserve/
dual specs in gigaflops/sec

Xserve Dual 2GHz PowerPC G5 9.0 Gigaflops

IBM eServer x325 Dual 2GHz Opteron 5.91 Gigaflops

Pretty much clear.


Did you know that these "benchmarks" don't even require connecting the computer to the wall outlet? You don't even have to put the CPU in the computer. You can leave the cooling pump and the Rube Goldberg radiator turned off.

The "Apple GFLOPs" are done with a calculater, reading the CPU spec sheet.


SPEC and the other benchmarks discussed here actually involve booting the computer, and seeing how fast real programs run. They exercise different parts of the CPU, memory system, I/O system, disks, graphics - in short, the whole system.

SPECfp, in particular, not only tests the floating point arithmetic system - it also is a good test of deliverable memory bandwidth. (And contrary to an earlier comment - the SPECrate numbers do test SMP systems.)

One can argue about whether these benchmarks are representative of one's particular workload - but they are real and are representative of some important workloads. (Did you know that SPECint tests "gzip" file compression and "gcc" program compiles among other tasks?)

Apple's numbers are theory....
 
i agree...

kangaroo said:
Reality check here. Keep in mind that the 'masses of -heads who are a part of the WinTel world' use their computers to do basic stuff like word processing, number crunching, email, web surfing, etc. Their computer is like an appliance. Your toaster makes toast, your computer allows you to shoot off an email. That's all it is; it's not a way of life. For most people, computers are a value not a nuance-play. ;)

i agree, there are many people like that, and they wouldnt even know what to do with a 3Ghz machine....and no one really needs a G5 to shoot off an email....and if most people that buy iMacs only use it like a toaster, they sure wont need a G5 in it, unless they really were trying to make toast.....i wonder, has anyone tried cooking with their machine yet?
 
When you do the research...

Travis Novak said:
-edit-
Ok, I see you mentioned the Civic in there. The civic, is not bought because it is "cheap". You can get Kia much much cheaper. But most people will not do that because they KNOW that the Kia will be ultimately unreliable, with a higher TCO in the end. People buy the civic for its high MPG, reliability, and even for the personality. I rarely hear people buying a Civic "because its cheaper". Hell, you can buy a Focus if you wanted to do that.

Yeah your right a civic is probaably the cheapest car 5 years down the road. All the kids at my high school have one because of low cost (intial) good realiability and great milage. A civic is just a great value. ;)[/QUOTE]

First, if J.D. Power has any credibility, KIA ties with Toyota in customer satisfaction. According to Edmunds, Consumers rank Kia Rio at 8.1 and Hyundai Elantra (Corporate Cousin) at 9.0, compared to 9.1 for the Civic.

Now, as for TCO, I have had Fords, Chevys, Honda Civics, VWs, a Nissan Sentra, Oldsmobiles, Buicks, Plymouths, Dodges, and a GMC truck over the past 45 years. Hands down, considering all exenses and residual value, the 2000 GMC Sierra Pickup is the cheapest vehicle to operate over the first 50,000 miles, the Sentra was next. The 1983 Civic S and 1992 Civic DL finish near the 79 Chevette, the 72 Olds Cutlass (which was literally falling apart rolling down the road when I traded it), and the 73 VW 412 as the most expensive cars per mile.

Most fun? 1965 Ford GT Convertible (352 4V), 1968 Mustang Convertible (289), and 69 Mercury Cougar XR7 (Cleveland 351).

If you change the oil and rotate the tires, about any car you buy today will run reliably for 100K miles. Buy the KIA and use the change to get anew Mac!
 
AidenShaw said:
SPEC and the other benchmarks discussed here actually involve booting the computer, and seeing how fast real programs run. They exercise different parts of the CPU, memory system, I/O system, disks, graphics - in short, the whole system.

SPECfp, in particular, not only tests the floating point arithmetic system - it also is a good test of deliverable memory bandwidth. (And contrary to an earlier comment - the SPECrate numbers do test SMP systems.)

One can argue about whether these benchmarks are representative of one's particular workload - but they are real and are representative of some important workloads. (Did you know that SPECint tests "gzip" file compression and "gcc" program compiles among other tasks?)

Apple's numbers are theory....

Only Apple's Gflops are theoretical - and they don't try to hide that.

Go to Apple's web site. You'll see many true application benchmarks. I'm sure that the vast majority of Mac purchasers are more likely to use Photoshop or any of the other apps listed there than SPEC.

It's amazing that you're still touting a benchmark that doesn't even use the second CPU on Macs. That by itself makes it largely meaningless. Not to mention that SPEC is really a compiler benchmark more than anything else. It's far more sensitive to compiler selection than any benchmark I've seen.

If you run SPEC apps all day, it's probably useful. But for 99% of the world, real, commercial application benchmarks are far better.
 
Most fun? 1965 Ford GT Convertible (352 4V), 1968 Mustang Convertible (289), and 69 Mercury Cougar XR7 (Cleveland 351).

Cool, I have a Mustang 641/2 convertible 289.
 
jragosta said:
If you run SPEC apps all day, it's probably useful. But for 99% of the world, real, commercial application benchmarks are far better.

Note that I said "SPEC and other benchmarks" - and in other comments I've agreed that you should benchmarks the apps that you use. If you don't use Photoshop, don't pay much attention to Photoshop benchmarks.

On the other hand, SPECcpu (SPECint and SPECfp) are far from being as useless as you imply. In particular, to get a good SPECfp score you must have a system with good deliverable (vs. specsheet MHZ) memory bandwidth. Even if you don't need FP codes, a good FP score is worth noting.

As far as the dual CPU thing goes, "meaningless" is far from the case - for two main reasons. First, good single CPU performance is pretty much a requirement for good dual CPU performance. Second, many important applications scale poorly on dual CPUs, so single CPU numbers are better predictors for those. Third (that's 3 reasons), dual CPU Intel and AMD systems exist as well....

And, finally (make that 4 reasons) - look at the SPECrate numbers (http://www.specbench.org/cpu2000/results/) - these are for SMP systems - with number for systems with up to 256 CPUs listed!

While you're at the SPEC website, (http://www.specbench.org/benchmarks.html) look at the *other* SPEC benchmarks. There are SPEC benchmarks for Maya, 3ds max, PRO/ENGINEER, Solid Edge, MPI, Java client/server, mail servers, file servers, web servers. There's so much more than just the SPECint that you seem to distrust.
________________________________

In the end, however, looking at lots of different benchmarks will give you a better idea of how balanced a system's performance is.

Unless you are only interested in a single app, and unless there's an open, reproducible standard for benchmarking that app, you'll want to look at several. (Even if you use Photoshop, should you trust that Apple's private Photoshop tests are fair and balanced - or would you assume that the scripts "just happen" to show off AltiVec?)

I don't think we really disagree that much - but I think that the way you simply kiss off SPEC is wrong. Take SPECcpu with a grain of salt, but don't ignore it.
 
AidenShaw said:
Note that I said "SPEC and other benchmarks" - and in other comments I've agreed that you should benchmarks the apps that you use. If you don't use Photoshop, don't pay much attention to Photoshop benchmarks.

On the other hand, SPECcpu (SPECint and SPECfp) are far from being as useless as you imply. In particular, to get a good SPECfp score you must have a system with good deliverable (vs. specsheet MHZ) memory bandwidth. Even if you don't need FP codes, a good FP score is worth noting.

As far as the dual CPU thing goes, "meaningless" is far from the case - for two main reasons. First, good single CPU performance is pretty much a requirement for good dual CPU performance. Second, many important applications scale poorly on dual CPUs, so single CPU numbers are better predictors for those. Third (that's 3 reasons), dual CPU Intel and AMD systems exist as well....

And, finally (make that 4 reasons) - look at the SPECrate numbers (http://www.specbench.org/cpu2000/results/) - these are for SMP systems - with number for systems with up to 256 CPUs listed!

While you're at the SPEC website, (http://www.specbench.org/benchmarks.html) look at the *other* SPEC benchmarks. There are SPEC benchmarks for Maya, 3ds max, PRO/ENGINEER, Solid Edge, MPI, Java client/server, mail servers, file servers, web servers. There's so much more than just the SPECint that you seem to distrust.
________________________________

In the end, however, looking at lots of different benchmarks will give you a better idea of how balanced a system's performance is.

Unless you are only interested in a single app, and unless there's an open, reproducible standard for benchmarking that app, you'll want to look at several. (Even if you use Photoshop, should you trust that Apple's private Photoshop tests are fair and balanced - or would you assume that the scripts "just happen" to show off AltiVec?)

I don't think we really disagree that much - but I think that the way you simply kiss off SPEC is wrong. Take SPECcpu with a grain of salt, but don't ignore it.

OK. I'll buy most of that. But for most people, you need to take SPEC with a HUGE grain of salt.

Download Apple's white paper athttp://images.apple.com/powermac/pdf/PowerMacG5_TO_06092004.pdf.

They list about a half dozen shrink-wrapped commercial apps. The vast majority of people who need the kind of power that these machines provide will be using these apps or similar ones. In every case, the G5 is equal to or faster than the fastest PC you can buy-sometimes faster by a wide margin.

Granted, Microsoft Office isn't in the list and it's quite likely that it's faster on a PC - although I can't imagine that anyone would attribute that to hardware issues. Regardless, even a 3 year old iMac is fast enough to run that kind of application.

Add in the Linpack performance where the G5 kills comparable x86 systems. Add in things like the gene matching program (I can't remember its name). It's pretty clear that SPEC is the outlier - probably due to the compiler issues that I've already mentioned.
 
kangaroo said:
Reality check here. Keep in mind that the 'masses of -heads who are a part of the WinTel world' use their computers to do basic stuff like word processing, number crunching, email, web surfing, etc. Their computer is like an appliance. Your toaster makes toast, your computer allows you to shoot off an email. That's all it is; it's not a way of life. For most people, computers are a value not a nuance-play. ;)

I agree however increasing numbers of "average" computer users are using their computers for home movies, digital pictures, music, etc. These needs much faster computers than simply word processing and email, etc. The increase of digital content on the internet means the need for quicker processors in addition to broadband connections.

Not to mention, at the rate MS Office is going you may soon need 3GHz and 1GB of RAM just to type a letter. ;)
 
pjkelnhofer said:
I agree however increasing numbers of "average" computer users are using their computers for home movies, digital pictures, music, etc. These needs much faster computers than simply word processing and email, etc. The increase of digital content on the internet means the need for quicker processors in addition to broadband connections.


Agreed.

And those are just the types of applications that the G5 excels at - as well as applications that really shine under OS X.
 
jragosta said:
Agreed.

And those are just the types of applications that the G5 excels at - as well as applications that really shine under OS X.

That is why I hope that not only is the iMac G5 introduced at WWDC, but that Apple begins an aggressive marketing campaign to convince people that with iLife included in the price the Mac has everything you need and it is compatible with your old files.

Secondly (and I know this would really piss off some Mac zealots), if Apple is not going to introduce it's own office suite, I would like to see MS Office added to the BTO option for Macs and the compatibility with MS Office for Windows played up. I think a collection of Word files and failure to realize that iPhoto will handle all their pictures is a major hurdle when trying to attract switchers.

One more thing that would be nice would be for Apple to introduce some sort of program that allows you to network to your old PC and quickly, easily, and an in an organized manner, bring all of you compatible Word, jpeg, etc. files over.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.