Need Apple vs. PC resources

[Intaglio]

Hi -

A person in my math class was talking about how Macintosh CPUs are far more efficient than PC CPUs today and I was wondering if anyone knew of some online articles or other resources I could look at the check the validity of this. He said that macintosh CPUs do more cycles per minute than IBM compatible CPUs and other stuff.

Any help would be appreciated.

-Eric

 

[gsaldivar]

Tell him to wake you up when Macs start placing at the top of SETI benchmarks.

You can also view RC5 benchmarks at this site.

Fastest Mac G4 1066Mhz, Speed = 9,494,325
Fastest AMD Duron 1400Mhz, Speed = 4,960,256
Fastest AMD K7 TBird 2000Mhz, Speed = 12,008,943

 

[damocles]

Heh

I'm sure that's a 'real' Northwood on that list

 

[gsaldivar]

Can you suggest a better site?

 

[damocles]

Can you suggest a better site?

Nope, but that site really doesn't say much. You can't just pick a particular benchmark that a processor does well in and use that to justify anything (that's an Apple trick )

See if this article helps a little- G4 vs K7

 

[Phatty106]

<< Tell him to wake you up when Macs start placing at the top of SETI benchmarks.

You can also view RC5 benchmarks at this site.

Fastest Mac G4 1066Mhz, Speed = 9,494,325
Fastest AMD Duron 1400Mhz, Speed = 4,960,256
Fastest AMD K7 TBird 2000Mhz, Speed = 12,008,943 >>



These benchmarks only prove the original theory, which is that Mac prodessors are more efficient than PC processors. Theoretically, the TBird 2000MHz should be turning out twice the work units as the G4.

ph

 

[Zenmervolt]

Actually, your friend does have the basic idea right, although his terminology is completely wrong. MHz is nothing more than a measurement of cycles per second, so IBM/Motorola Power PC microchips (the CPUs used in Macs) do not run at more cycles per second than an x86 cpu. However, a Power PC chip can actually perform more operations per clock cycle than an x86 cpu, which means that at equal MHz, Power PC cpus do have a slight advantage over x86 cpus. However, this advantage is not more than a maximum of 100 MHz (ie a Power PC 866MHz chip would be roughly equal to a PIII 966MHz chip). Unfortunately, 866MHz is as fast as Power PC cpus get, and for the price of an 866MHz Power PC equipped Mac ($2,499) one can easily buy a 1.4 GHz Athlon system from almost anywhere (~$1,400) or a 2.0GHz P4 IBM ($1,958) and add several options before you come to the price of the base 866MHz Mac. So although the Power PC does have a slight advantage clock for clock, the advantage is negated several times over by the tremendously low maximum speed and the high price of a Mac. Also, the SpecPref benchmarks (a cross-platform unoptimised benchmark suite that is very good at measuring the CPU alone) shows that the Power PC cpu does not perform significantly better than an equally clocked x86 CPU.

Zenmervolt

EDIT: That 1066MHz G4 (IBM PPC 7xx) is a severe overclock and the fastest officially available PPC chip is the 866MHz part to the best of my knowledge. To get a PPC 7xx chip to run that fast requires some serious cooling, dedication, and luck. Also, just for fun, inform your friend that the IBM has least a half stake in the technology behind the Power PC chip, Mac people hate to be reminded that their supposedly wonderful computer owes at least part of its existance to Big Blue.

 

[gsaldivar]

I hear ya damocles.

Unfortunately, the reality is that most Mac users simply won't make any further effort to understand the finer points of each processor. Mac users are people who just point to whatever "benchmark" Apple has on their website as evidence that the Mac can beat all other computers hands-down.

While RC5 and SETI aren't a complete picture of the power of the processors, that doesn't mean that the benchmarks aren't useful. All I was trying to do was find some of the more "unbiased" ways of measuring performance between the platforms.

"...These benchmarks only prove the original theory, which is that Mac prodessors are more efficient than PC processors..."

Are the PowerPC processors more efficient than the P4 & K7? Yes.

Does this mean that they beat the P4 & K7 hands-down? No.

It seems that the recent Macworld has simply made many of these guys much more annoying than they ever were. Always it's the Photoshop filters vs. a crippled Pentium and guess who "wins"...

 

[zephyrprime]

You can also view RC5 benchmarks at this site.

I'm looking at the site and the numbers are all over the range and seem very unreliable. Check out the numbers:

5753 Intel Pentium III 667 Windows 98 2.8010 RC5
7204 Intel Pentium III 667 Windows 2000 2.8010 RC5
4853 Intel Pentium III 667 Windows 2000 2.8010 RC5

The deviation is too great! Hard to make any good conclusions across platforms.

 

[gsaldivar]

Um... the number in the *left* column is the Record number, not the Benchmark speed

try looking again at the *rightmost column* and I think you'll see that the benchmarks fall into a pretty predictable curve.

yes, there are a few false values, and here is the link to get those removed from the list.

it's hard to cheat at this type of ranking if the false speed you submit sticks out and isnt comparable to those with the same configuration.