You'll get your Mac news here from now on...

Help TMO Grow

Subscriber Login

Advertising Info


The Mac Observer Express Daily Newsletter


More Info

Site Navigation

Home
News
Tips
Columns & Editorials
Reviews
Reports
Archives
Search
Forums
Links
Mac Links
Software
Reports
Contact

Visit Charle Gaba's new System Shootout site for the latest shootouts!


Desktop Shootouts
New!
(U.S. $):
Apple Desktop Lineup
Laptop Shootouts
(coming soon!)
$800
Laptops
$1,100
Laptops
$1,600
"Mini-Me's"
$2,000
Laptops
$2,600
Laptops
$3,000
"LapZillas"
Apple Laptop Lineup

Thoughts Regarding Processor "Speed," Megahertz Ratings, and Benchmark Tests
by Charles Gaba - AAPLTalk System Shootout Master
IMPORTANT NOTE REGARDING DUAL PROCESSOR PERFORMANCE: "For most people, a dual processor machine will seem faster than a machine with a single processor running at twice the speed. That's because most people would be checking e-mail, playing music, etc. A bunch of things at the same time. That said, people that will be using a program that will consume all the CPU cycles of a processor (e.g. Gamers) would benefit from the faster single processor. But, as dual processors become more common, and folks write more games that are multi-threaded (e.g. John Carmack of Quake fame), then gamers will start to benefit from multiple processors."

(Thanks to Ray DeGennaro for an excellent summary of this often confusing issue!)

UPDATED 4/22/03: Intel's Centrino/Banias Laptop Processor

Intel has come out with their new "Pentium M" processor, aka Banias (branded as "Centrino", although this is actually the combination of the chipset, the Pentium M processor, and an integrated 802.11b Wi-Fi card). While this is new and there haven't been too many benchmarks run as of yet, for the moment I'm using one of the most trusted sources in hardware performance benchmarking, Tom's Hardware, as the basis for Pentium M performance equivalence ratings:

http://www6.tomshardware.com/mobile/20030205/centrino-16.html

Based on this, so far, it looks like a 1.6 GHz Pentium M system is roughly equal to a 2.2 GHz Pentium 4-M system (which, I understand, is less powerful than the desktop Pentium 4 processor, but I don't have any hard data on that, so I'm leaving well enough alone for now).

On this assumption, I'm going with a 1.4 : 1.0 ratio for Pentium M : Pentium 4 laptop performance.

As a result, here is the updated Performance Equivalence Rating Chart:

Intel Pentium 4 /
Pentium III / Celeron
(desktop processors)
Intel Pentium M /
Pentium 4-M
(laptop processors)
AMD Athlon XP /
Athlon 4 / Duron
PowerPC G4
PowerPC G3
appx. desktop P4 equiv:
appx. desktop P4 equiv:
appx. desktop P4 equiv:
appx. desktop P4 equiv:
3.06 GHz Pentium 4
3.0
.
3.0
Athlon 3000+ (2.17 GHz)
3.0
.
3.0
.
2.9
.
2.9
Athlon 2800+ (2.08 GHz)
2.9
.
2.9
2.8 GHz Pentium 4
2.8
.
2.8
Athlon 2600+ (2.13 GHz)
Athlon 2700+ (2.2 GHz?)
2.8
.
2.8
.
2.7
.
2.7
Athlon 2500+ (1.83 GHz)
2.7
Dual 1.42 GHz G4
2.7
2.66 GHz Pentium 4
2.6
.
2.6
Athlon 2400+ (2.0 GHz)
2.6
.
2.6
2.53 GHz Pentium 4
2.5
.
2.5
.
2.5
.
2.5
2.4 GHz Pentium 4
2.4
.
2.4
.
2.4
.
2.4
.
2.3
.
2.3
Athlon 2200 (1.80 GHz)
2.3
Dual 1.25 GHz G4
2.3
2.2 GHz Pentium 4
2.2
.
2.2
Athlon 2100+ (1.73 GHz)
Athlon 2000+ (1.67 GHz)
2.2
.
2.2
.
2.1
.
2.1
Athlon 1900+ (1.60 GHz)
2.1
Single 1.42 GHz G4
2.1
2.0 GHz Pentium 4
2.0
2.5 GHz Pentium 4-M
2.0
Athlon 1800+ (1.53 GHz)
2.0
Dual 1 GHz G4
2.0
1.9 GHz Pentium 4
1.9
2.4 GHz Pentium 4-M
1.9
Athlon 1700+ (1.47 GHz)
1.9
Single 1.25 GHz G4
1.9
1.8 GHz Pentium 4
1.8
2.2 GHz Pentium 4-M
1.6 GHz Pentium M
1.8
Athlon 1500+/1600+
(1.35-1.40 GHz)
1.8
.
1.8
1.7 GHz Pentium 4
1.7
1.5 GHz Pentium M
1.7
1.3 GHz Athlon
1.7
Dual 867 MHz G4
1.7
1.6 GHz Pentium 4
1.6
2.0 GHz Pentium 4-M
1.6
1.2 GHz Athlon
1.6
.
1.6
1.5 GHz Pentium 4
1.5
1.9 GHz Pentium 4-M
1.4 GHz Pentium M
1.5
1.1/1.13 GHz Athlon
1.1 - 1.2 GHz Athlon 4
1.5
Single 1 GHz G4
1.5
1.4 GHz Pentium 4
1.4
1.7-1.8 GHz Pentium 4-M
1.3 GHz Pentium M
1.4
1.0 GHz Athlon
1.4
933 MHz G4
1.4
1.9 GHz Celeron/
1.2 GHz Pentium III
1.3
1.6 GHz Pentium 4-M
1.3
.
1.3
Single 867 MHz G4
1.3
1.8 GHz Celeron
1.2
1.5 GHz Pentium 4-M
1.2
950/1000 MHz Athlon 4
1.2
800 MHz G4
1.2
1 / 1.13 GHz Pentium III
1.1
1.4 GHz Pentium 4-M
1.1
850 / 900 MHz Athlon 4
1.1
667/700 MHz G4
1.1
900 / 933 MHz Pentium III
1.0
.
1.0
1.2 - 1.3 GHz Duron
1.0
900 MHz G3
1.0
1.3 GHz Celeron
0.9
.
0.9
1.1 GHz Duron
0.9
800 MHz G3
0.9
1.2 GHz Celeron
0.8
.
0.8
1 GHz Duron
0.8
.
0.8
1.06 GHz Celeron
0.7
.
0.7
900 / 950 MHz Duron
0.7
600 MHz G3
0.7
UPDATED 4/16/03: Processor Equivalence Estimates Given More Credibility

FWIW: I'm pleased to report that my "rough estimate" formula of G4/Pentium 4 overall performance has been proven to be pretty accurate by a fourth source:

http://www.barefeats.com/pentium4.html

In a nutshell, they pitted a PowerMac G4 DP 1.42 GHz vs. a Wintel Pentium 4 SP 3.06 GHz w/hyperthreading. They also threw in a PowerMac DP 1.25 and PowerMac DP 1.0 system as well.

I know BareFeats catches a lot of flak for their methodology, but in this case they seem to have their stuff together--The PowerMacs had 1.5 GB of DDR PC2700 RAM running OS X 10.2.4; the Pentium system had 2 GB of RDRAM PC1066 running WinXP Pro; all systems had 7200 rpm hard drives (80-120 GB), so I'd say they were fairly evenly matched up aside from the processors.

They ran a battery of tests, including Cinema 4D XL, Photoshop 7 (both multiprocessor-aware and single processor-only filters), and Bryce 5. I'm not a benchmarking expert but this seems a reasonable range of things to throw at the systems.

In any event, when you tally up the results:

3.06 GHz Pentium 4 w/HT system: total time 150 seconds
1.42 GHz DP PowerMac system: total time 162 seconds
1.25 GHz DP PowerMac system: total time 170 seconds
1.0 GHz DP PowerMac system: total time 225 seconds

Now, the System Shootouts "ballpark performance equivalence" formula has stood at:

1.0 GHz G4 appx. = 1.1 GHz Athlon appx. = 1.5 GHz Pentium 4, assuming 133 MHz FSB & SDRAM on the PowerMac.

For Dual Processor G4 systems, I've been figuring:

DP 1.0 GHz G4 appx. = SP 1.5 GHz Athlon appx. = 2.0 GHz Pentium 4, assuming the same as above.

So, I decided to see how far off the mark I am (at least with the DP systems).

Based on the above, a DP 1.42 G4 system w/a 166 FSB and DDR PC2700 should be roughly equal to a SP 2.84 GHz Pentium 4 system w/533 FSB and RDRAM PC1066, all else being equal.

150 seconds (3.06 P4) / 162 seconds (DP 1.42 G4) = about 8% faster overall. (3.06 / 2.84 equals...about 8% faster overall. So far, so good.)

The DP 1.25 G4 system should be about equal to a 2.5 GHz Pentium 4 system if I'm correct. (150 seconds / 170 seconds = about 12% faster. )

3.06 / 2.5 equals...about 22% faster. Whoops; the DP 1.25 is actually closer in real performance to about a 2.7 GHz Pentium 4 (3.06/2.7 = 13% faster). Still, not too far off.

Finally, the DP 1.0 GHz G4 should be about equal to a 2.0 GHz Pentium 4. (150 seconds / 225 seconds = about 33% faster.)

The DP 1.0 is closer to about a 2.3 Pentium 4 (3.06/2.3 = 33% faster).

My conclusion? The DP G4 systems are at *least* equal to a double-clockspeed *single* processor Pentium system, possibly more at lower speeds, but I'll leave it as is for now for the sake of simplicity. So, if you want to know ROUGHLY how your G4 stacks up:

For *single* processor G4s: Add 50% to your clockspeed to get the "Pentium 4 equivalence"

For *dual* processor G4s: Double the clockspeed of one of the processors to get the "Pentium 4 equivalence" (single processor P4, that is).

For "Athlon equivalence": add 11% to your single-G4 clockspeed or 50% to your dual-G4 clockspeed (one processor only) for the equivalent Athlon.

And yes, I know how grotesquely simplistic this is, but hey, ya gotta have SOMETHING to go by.

UPDATED 7/19/02!!!

I just received some reasonably solid benchmark support for the System Shootouts' simplified "Pentium 4 Equivalency" rating system.

As regular visitors to AAPLTalk know, I've taken Steve Jobs' RDF "twice as fast" marketing claims, added assorted hard-core benchmark testing results from Intel-vs-AMD hardware sites such as Toms' Hardware, Tech-Report, etc., thrown in a few Bare Feats scores, added a dash of common sense as well as my personal, if limited, hands-on experience, and up until now, boiled it down to an extremely simplified rule-of-thumb:

1.0 GHz G4 appx. = 1.2 GHz Athlon appx. = 1.6 GHz Pentium 4

(the above assuming the latest chipsets, similar hard drives,video cards, etc.)

Up until today, I was only about 60% confident that I had this fairly accurate. Today, thanks to a benchmark duel posted at Digital Video Editing's Web site, I can increase that confidence level considerably:

http://www.digitalvideoediting.com/2002/07_jul/features/cw_macvspc2.htm

Now, the site itself ran about a dozen common DV and graphics tasks between 3 similarly-configured systems: A 2.53 P4, a dual-processor 1.67 GHz Athlon, and a dual-processor 1.0 GHz PowerMac.

Numerous respected PC-Only benchmark sites seem, to me, to generally conclude that the Athlon-to-Pentium 4 performance ratio comes out at around 1.3:1...ie, a single 1.67 GHz Athlon approximately equals about a 2.1-2.2 GHz Pentium 4. This is even better than AMD's own branding, which calls the 1.67 Athlon "2000+" meaning that they claim it's equal to a 2.0 GHz P4. This is where I get my Athlon/Pentium 4 ratio:

1.2 GHz Athlon appx. = 1.6 GHz Pentium 4

Since this is already established elsewhere, and the Pentium 4 system used is a single processor unit, I can really only fairly match up the Athlon dualie vs. the G4 dualie.

All else being equal, you would expect a 1.67 GHz system to take 60% as long as a 1.0 GHz system to complete a given task:

100/1.00 = 100.0
100/1.67 = 59.8

Moving to the actual benchmark results, let's tally up all 10 or so tests and see what the total times to do all of them--a full range of DV rendering, compositing, filtering, etc. adds up to:

SP 2.53 Pentium 4 system: 20:33 minutes
DP 1.67 Athlon system: 18:54 minutes
DP 1.0 PowerMac G4 system: 30:43 minutes

Does this pretty much squash Jobs' "twice as fast" claim? Well, in all honesty, yes, but I always thought that was a stretch anyway; even most Mac users I know don't buy that one completely.

The Athlon system at DP1.67 GHz came in at 62% of the time it took the G4 DP1.0 GHz system, or slightly higher than the 59.8% noted above.

As a result, for future comparisons, I'm going to modify my formula a bit, bringing it down to:

1.0 GHz G4 appx. = 1.1 GHz Athlon appx. = 1.5 GHz Pentium 4,
assuming 133 MHz FSB & SDRAM on the PowerMac.

--Charles Gaba, AAPLTalk System Shootout Master

August 10, 2001:

It's amazing how a thing like a computers' processor clockspeed can get people so bent out of shape.

I've been doing the occasional "system shootout" for about about a year or so, and have received quite a bit of mail regarding my previous efforts in the entry-level arenas: comparisons of the new iBook and new iMac against other low-priced PC offerings.

However, it wasn't until my recent higher-end PowerMac G4 system shootouts hit the web that I felt the full impact of how much fuss there has been over the latest offerings from Apple, Intel, and of course AMD.

I have been inundated with two separate floods of e-mail on the subject--the first, upon initially publishing the chart; and the second, roughly a week later, when Scott "Damage" Wasson of The Tech Report wrote this scathing opinion piece regarding what he saw as an unmitigated piece of pro-Mac propaganda.

Now, the question of whether my charts are propagandistic or not can be easily summed up: Of course they are. I'm a Mac guy. There are disclaimers all over above and below every chart, in bold face, capital letters, and blue lettering, clearly stating--repeatedly--that these charts are my opinion only, and the my results should be taken with however large a grain of salt as the reader wishes. Some of his readers speculated, in his sites message boards, that I might possibly be on the take from Apple, or that I started this whole thing in order to beef up AAPLTalk's site hits (certainly this has happened). Nice theories, except for two things: One, I've already disclosed that I own a few shares of Apple stock (100 shares to be precise--I'm not exactly rolling in dough here); and two, you don't see any banner ads anywhere on the AAPLTalk site, do you? On the other hand, Mr. Wassons site does have banner advertising, and his diatribe against the AAPLTalk Shootouts certainly seems to have increased HIS server load considerably. Take this to mean whatever you wish. I hasten to add, however, that Mr. Wasson himself never suggested that I had any hidden financial motive, although he certainly seemed to be irritated by blatant admittance of my Mac bias. We'll come back to him later.

Let's go back for a moment, though.

Several weeks ago, during his Macworld New York keynote speech, Steve Jobs performed one of his (in)famous "Photoshop Bakeoffs" in which he pitted a new, 867 MHz PowerMac G4 system against a 1.7 GHz Pentium 4 tower. He clearly stated that both systems had the same amount of RAM (1 GB), the exact same hard drive, and the same graphics card. As usual, he ran both systems through an extensive array of heavy-duty Photoshop actions and filters, and in the end, the 867 MHz trounced the 1.7 GHz Pentium 4, finishing a whopping 58% faster. Now, if you were to go by these Photoshop test results only, one would conclude that an 867 MHz G4 is equal to a 2.7 GHz Pentium 4.

In addition to his standard Photoshop demonstration, Jobs also did another very heavy-duty system-intensive demo of Media Cleaner 5, which included completely deinterlacing and encoding an entire movie trailer via the new Sorensen 3 codec. In this comparison, the G4 wiped the floor with the 1.7 GHz P4, which wasn't even halfway done by the time the G4 finished. Again, using these two comparisons only, one would conclude that the G4 has upwards of 3-4 times the processing power of a Pentium 4 at the same clockspeed.

However, thanks to TechTV's botched coverage of the keynote (Jim Louderback repeatedly interrupted the speech to announce that Jobs "never gave out" the amount of RAM or graphics cards used in the systems, implying that the demos were fixed), a lot of people didn't hear Mr. Jobs give these figures. Of course, if he had been listening instead of giving color commentary, Jobs clearly stated the equality of the systems used (hear for yourself; the entire keynote can be played back via QuickTime HERE (scroll forward to the benchmark tests at around 1 hour & 20 minutes in).

Now, don't get me wrong--I normally have a lot of respect for Mr. Louderback, and I'm a huge TechTV fan, and I do appreciate their coverage of the keynote. I was severely disappointed at his snide comments in this situation, and hoped that he would make up for it by doing the right thing--simply stating that he was point-blank wrong, had made a mistake, and apologize. I wouldn't even have minded an additional comment regarding the complexities involved in attempting to compare processor performance, which is why he kept making his comments in the first place.

Instead, however, TechTV and Mr. Louderback compounded their error, attempting to save face by conducting an absurd in-house series of "tests" which they apparently thought would "prove" that the Macworld demos were false and that the G4 was not faster than the Pentium 4. They ran a half-dozen very simple Photoshop tests, and concluded that each system won three--even-steven.

Unfortunately, this in-house test was even more biased than the Macworld test was. TechTV used a last-generation, 733 MHz PowerMac G4 (instead of the 867 MHz system)...which favors the Pentium. They used the 1.8 GHz Pentium 4, instead of the 1.7 GHz...which favors the Pentium. They did give both systems equal RAM (512 MB), but, ironically, never stated what graphics cards or hard drives either system had--which is ironic considering that this was the same charge they had falsely made against Jobs! In an additional insult, they claimed that "With 512MB of memory in each computer, there was little else we could do to make these truly comparable systems" which, of course, is nonsense.

And yet, even with all of these things favoring the Pentium 4, the G4 system--running at only 733 MHz--still equalled the 1.8 GHz Pentium 4. Further, if you tally up the total numbers of all six tests, it's worth noting that the G4 actually finished the entire series of tasks in 81.9 seconds, 17% faster than the Pentium 4s 98.1 second total.

Now, am I saying that the G4 is 2.5 times faster than a Pentium 4? Hardly.

...Although that's certainly what Apple and Mr. Jobs would have you believe. However, this is stretching the truth, which is only fair considering that Intel is the one who started this whole MHz-is-everything nonsense in the first place. Both companies--as well as AMD--are trying to get people to buy their product, so naturally will show their wares in the best possible light.

The problem, of course, is that--as MANY, MANY PEOPLE HAVE REMINDED ME--there are other tests in which the Pentium 4 will most likely out-perform a G4. Photoshop 6 has been optimized for the G4s "Altivec" technology (Apple uses the more public-friendly term "velocity engine" which I personally think sounds kind of silly). Other applications may be optimized for the Pentium, while others may not be available for one platform or the other. There are a thousand other things to consider as well--there's the hard drive size, speed, throughput, etc.; the graphics card type & amount of video memory; the amount of RAM (all three of these were considered by Apple; one of them was by TechTV). There are other factors such as the TYPE of RAM (SDRAM? DDR-SDRAM? RDRAM?), the system bus, the memory bus, and the frontside and backside caches (levels one, two, and three). ANY of these things can affect the performance of a given system. Some can be equalized (such as the amount of RAM) while others simply cannot (such as the type of RAM--is RDRAM "better" than SDRAM? Beats the heck out of me, though many people have debated this topic as well). The point here that Apple was trying to make is that although Intel would have the world believe that clockspeed = everything, the raw MHz number of a given processor is only ONE of MANY factors to consider in judging the overall performance of a system. They even had Jon Rubinstein, their Senior VP of Hardware, come out and give an interesting but oddly-timed lecture about what Apple calls the "Megahertz Myth."

Ironically, this is very similar to point that TechTV was apparently trying to make about Apples Photoshop tests--that the specific test used is only one of MANY possible ways of measuring performance (the Media Cleaner video encoding bakeoff, oddly enough, seems to have gotten lost in the ensuing shuffle). Instead of specifically trying to trash the G4, I now believe that Jim Louderback was trying to make a valid point--that you can twist any given benchmark test to your benefit if you try hard enough--by doing the same thing themselves. Judging from much of the e-mail that I've received lately, this point seems to have gotten confused as well.

Now, let's get back to Scott "Damage" Wasson and the Tech Report.

Just as I never expected a professional technology/processor reviewer to devote an entire column to tearing apart my little charts, it turns out that HE never expected HIS opinion piece to be linked to over on the SlashDot (read: /.) message boards...but it was. Someone submitted his article, which in turn linked to my site, and all hell broke loose. Next thing I knew, the /. boards were buzzing, the Tech Report message board servers were overwhelmed, and I was deluged with hundreds of e-mails. I wrote Mr. Wasson a personal note in which I took him to task for not having the courtesy to at least send me a quick e-mail notifying me of his column.

To his credit, he apologized for not doing so, and we have since exchanged a few notes on the issue, in which he sympathized for my sudden e-mail flood; it seems that the main problem is the slash-dotters, who are quite a vicious bunch (even worse than we Mac users, it appears!)

Mr. Wasson has put forth a call, as have many others, for some way of conducting a bias-free, no-holds-barred, across-the-board comparison of the PowerPC, Pentium, and Athlon processors to definitively clear up this confusion...not an easy task. Having read up on some of Mr. Wassons other work, I've concluded that he does seem to have a lot of respect for the G4s engineering, and though he has little respect for Apples marketing claims, he also seems to have quite a bit of disdain for Intels as well. His general overall opinion seems to boil down to this: The G4 has excellent engineering, and does perform considerably faster than the clock speed alone would indicate, but for Apple to claim 2-2.5x overall performance based on Photoshop alone is just as insulting as Intel claiming that the Pentium 4 is the fastest processor around. In other words, yes, the G4 is faster than people think--but HOW MUCH faster is difficult to ascertain. Apple claims twice as fast; Intel claims not at all; the truth is probably somewhere in between.

And this, of course, leads to the AMD Athlon.

With all the e-mail that I have received since this whole mess started, the number one thing that people griped about the most was the lack of any AMD Athlon systems at all. Countless people slammed it into my head that the 1.4 GHz Athlon Thunderbird kicked that behinds of ANY processor from either Apple/Motorola or Intel. In a further irony, if true, this serves to add further evidence to Apples "Megahertz Myth" argument; in fact, AMD themselves has started their own marketing campaign to get the word out.

I've received links to numerous benchmark studies involving Photoshop, Quake 3, SiSoft Sandra, C Linpack, Cinema 4D, Lightwave 6, Mathmatica, SETI@home, etc. from sources including Tech Report itself, JC News, Tom's Hardware Guide, and several others. Most only cover the Pentium and Athlon chips, although the JC News link has some good info on the PowerPC/G4 as well. Unfortunately, it only covers the 500 MHz G4, so some speculation is still necessary. In any event, after slogging through all of them--on top of the pile of e-mails sent to me, and the hundreds of message board postings I've read--I've boiled it down to the following:

1. In my oversimplified opinion, the G4 is somewhere in the vicinity of around 1.6x faster than a Pentium 4 at the same clockspeed.
2. In my oversimplified opinion, the Athlon is somewhere in the vicinity of around 1.3x faster than a Pentium 4 at the same clockspeed.
3. Therefore, in my oversimplified opinion, the 867 MHz G4 appx. = 1.4 GHz Pentium 4 appx. = 1.1 GHz Athlon.

Of course, this is a gross oversimplification, but in order to wrestle this issue down to my arbitrary "point" system, it's the best I can do. Having said that, I'm assigning:

Intel Pentium III / Pentium 4
AMD Athlon / Athlon 4
PowerPC G4
P4 Equivalence (GHz):
P4 Equivalence (GHz):
P4 Equivalence (GHz):
.
.
(Dual) 1 GHz G4
2.8
2.53 GHz Pentium 4
2.5
.
.
2.4 GHz Pentium 4
2.4
.
.
.
Athlon 2200+ (1.80 GHz)
2.3
.
.
Athlon 2100+ (1.73 GHz)
Athlon 2000+ (1.67 GHz)
2.2
.
2.2 GHz Pentium 4
2.2
Athlon 1900+ (1.60 GHz)
2.1
.
.
Athlon 1800+ (1.53 GHz)
2.0
.
2.0 GHz Pentium 4
2.0
Athlon 1700+ (1.47 GHz)
1.9
.
1.9 GHz Pentium 4
1.9
Athlon 1500+/1600+
(1.35-1.40 GHz)
1.8
.
1.8 GHz Pentium 4
1.8
.
.
1.7 GHz Pentium 4
1.7
1.3 GHz Athlon
1.7
.
1.6 GHz Pentium
1.6
1.2 GHz Athlon
1.6
(Single) 1 GHz G4
1.6
1.5 GHz Pentium 4
1.5
.
933 MHz G4
1.5
1.4 GHz Pentium 4
1.4
1.1/1.13 GHz Athlon
1.1 - 1.2 GHz Athlon 4
1.4
867 MHz G4
1.4
1.2 GHz Pentium III
1.3
1.0 GHz Athlon
1.3
800 MHz G4
1.3
1.8 GHz Celeron
1.2
950/1000 MHz Athlon 4
1.2
.
1 / 1.13 GHz Pentium III
1.1
850 / 900 MHz Athlon 4
1.1
667/700 MHz G4
1.1
900 / 933 MHz Pentium III
1.0
.
.
1.3 GHz Celeron
0.9
1.1 GHz Duron
0.9
550 MHz G4
0.9
1.2 GHz Celeron
0.8
1 GHz Duron
0.8
700 MHz G3
0.8
1.06 GHz Celeron
0.7
900 / 950 MHz Duron
0.7
600 MHz G3
0.7
.
800 / 850 MHz Duron
0.6
500 MHz G3
0.6
Intel Celeron
AMD Duron / Duron Mobile
PowerPC G3
Note regarding Processor Performance: See the processor notes page of the AAPLTalk Shootout section for a full explanation of the rationale behind the above point assignments.

...And that's what I'm sticking with until someone comes up with something better.

More thoughts on processor benchmarking from Eric H. Yang

The information above is as accurate as I could make it based on a reasonable time studying the respective companies' Web sites. There is no guarantee that these configurations, prices, and features are the best that one could come up with; assorted bundles, free internet access, free extra memory, and so on can most likely be found for any of the above systems. The point of this chart is to demonstrate that Apple's lineup holds its own against any brand-name computer, even in areas in which it has traditionally had resistance. In short, do not base your purchasing decisions solely on this chart or information. Neither I, nor The Mac Observer will be held responsible or liable for any purchase decisions you make based on this presentation.

Charles Gaba started the AAPLTalk System Shootouts at AAPLTalk.com, and has brought them to TMO. Charles is the creative talent behind Brainwrap Web Design.

AAPLTalk System Shootouts Archives

Back to The Mac Observer For More Mac News!



Today's Mac Headlines

[Podcast]Podcast - Apple Weekly Report #135: Apple Lawsuits, Banned iPhone Ad, Green MacBook Ad

We also offer Today's News On One Page!

Yesterday's News

 

[Podcast]Podcast - Mac Geek Gab #178: Batch Permission Changes, Encrypting Follow-up, Re-Enabling AirPort, and GigE speeds

We also offer Yesterday's News On One Page!

Mac Products Guide
New Arrivals
New and updated products added to the Guide.

Hot Deals
Great prices on hot selling Mac products from your favorite Macintosh resellers.

Special Offers
Promotions and offers direct from Macintosh developers and magazines.

Software
Browse the software section for over 17,000 Macintosh applications and software titles.

Hardware
Over 4,000 peripherals and accessories such as cameras, printers, scanners, keyboards, mice and more.

© All information presented on this site is copyrighted by The Mac Observer except where otherwise noted. No portion of this site may be copied without express written consent. Other sites are invited to link to any aspect of this site provided that all content is presented in its original form and is not placed within another .