CPU Speed Mac<--->Pc

[Aperture]

Hey guys. I was wondering what the CPU speed of my 1.5Ghz Core Solo Mini would compare to in Windows. I know they used to compare the PPC Chips to a PC CPU so I am just curious about the intel chip. Would it be compared to a Windows CPU with the same specs it has on a Mac because it is intel? What I am asking is what is the Core Solo comparable to? I want to know a speed estimate and then a chipset it might be comparible to.
Ex.

Intel Pentium 4
Intel M
Etc.

Just curious.

Thanks, Kevin

 

[count chocula]

i dont knw about what chips it is comparable to, but i would think it would be basically the same as any pc with a 1.5 ghz cpu. i'm no expert, but that would be my guess.

 

[bbrosemer]

Its still a 1.5 Core Solo in windows hardware doesnt change nor does the speed of the processor the only thing that changes is how well the code is written that is going through the CPU and if you were wondering a Clean version of windows is slightly slower... barely not much there was some bench somewhere ... dont remember....

 

[brbubba]

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2648

These benchmarks should give you a good idea. They use a core duo in the tests but its unknown which tests will utilize the second core.

 

[gekko513]

It's a bit faster than a Pentium M and they used to say a 1.7GHz Pentium M was comparable to a 3.0GHz Pentium 4.

 

[Mord]

and a 1.5GHz core solo is about the same as a 1.7 dothan

 

[Timepass]

Umm CPU speed reading like how you want does not matter what the OS it is for comparing.

Now CPU is hardware and you can only compare your hardware. Since intel finally stop pushing the mHz myth it really comes down to how the chip is set up.

A solo core intel chip of 1.5 ghz is the exact same speed on a PC or a Mac because it is the same Chip.

Problem is bettween intel pushing the mhz myth for so long and apple showing how it was a myth for so long people have come to believe that the cpu speed on an apple computer = a much high speed number on a PC. It doesnt. It comes down to comparing the CPU to another CPU and if they are the same CPU archtecher then they are clock speed is a great comparison since there is no diffences in how the chip is set up.

But if the cores are diffenent there are a ton of other factors that matter and clock speed is a poor way of comparing them.

An example to look at from AMD64 line is there is a 2.2 ghz 3200+ clawhammer corre out there and there is a 2ghz amd64 3200+ newcastle core. The claw hammer clock speed is faster but the over all cpu speed is exactly the same because the core are different core designs.

 

[Eraserhead]

One of AMD's best ideas is the 2200+ thing, it enables you to compare chips easily, Apple/Intel should adopt it too... (especially Apple) as it would stop the "slow" myth some more...

 

[Timepass]

One of AMD's best ideas is the 2200+ thing, it enables you to compare chips easily, Apple/Intel should adopt it too... (especially Apple) as it would stop the "slow" myth some more...

problem is the AMD naming system was so they could compare with intel chips and how intel oringallly named there chips.

AMD naming system is going to have to change some time as well since it no longer cover what they oringally had designed it for.

3200+ was about the same as a 3.2ghz intel chip. So you really can do that. Plus it intel name system was based on clock speed. So now you get the idea where the problems start.

 

[Mord]

they should just number them by their SPEC integer and FP scores added together

 

[wonga1127]

What Windows chip? That could be anything, even a 486 if you wanted. Apples to Apples (no pun intended), it wouldn't make a difference what OS you are using, just what program. Apples to Oranges, only benchmarks can ring true.

You know what I just figured out? About a year ago, my old Dell Dimension 1000 from 1999 crapped out after a power surge fried the motherboard. I replaced it with a Dell Dimension 3000, 2005. I finally realized why, after customizing the Dell to my liking, it felt slower than my Pentium III 1000 compared to the 3000's Pentium 4. First off, the Pentium 4 is the Prescott, one of the slowest, hottest P4's to date, and the Pentium 4 could actually be theoretically slower than the PIII, due to pipeline stage numbers, etc. Bottom line, I miss that old Dell 1000.

 

[Mord]

Windows XP and especially vista are slower than OS X on the same hardware, they consume more resources and don't handle resources so well benchmarks have shown a 2-5% speed increase in non graphics card bound tests (graphics card drivers tend to be better under windows)

 

[Sun Baked]

A solo core intel chip of 1.5 ghz is the exact same speed on a PC or a Mac because it is the same Chip.

Yes, it is sort of like asking how much faster the Red Honda is than the same Honda in Grey with real bucket seats.

But some people may "think" OS X feels slower because it doesn't always react instantly to a mouse click.

 

[Mord]

if the driver (OS) is more comfortable then he's more likely to be able to driver faster without distraction, also bucket seats are lighter so the car will be slightly faster (os being less resource intensive)

 

[w_parietti22]

this is no longer an issue now that both PCs and Macs use the same chips.

 

[JFreak]

CPU does not care which program runs what code, it just executes operations after another. If company A uses the same CPU than company B, then their systems have same CPU performance. If there are benchmark differences, those come from the code. Logical.

Now the interesting question is: why the benchmarks differ? Is the quality of software code so important? It is. Here's an imaginary example to make things clear:

If there are X amount of cpu cycles available and the operating system of company A utilizes 10 of those, then there are (X-10)/X=90% cycles available for applications. On the other hand, if the operating system of company B utilizes 20 of those cycles, then there are (X-20)/X=80% cycles available for applications. Now if the operating system of company A is so efficient that adding another application into execution only makes the OS to consume 5 more cycles, but the operating system of company B consumes 10 more, then there are (X-15)/X=85% and (X-30)/X=70% cycles available for applications in systems A and B, respectively. Does begin to look bad, right? Yep. But let's assume there will be two more applications running, which would make the numbers (X-25)/X=75% and (X-50)/X=50% -- this would mean that the more efficient system A would use one quarter of cpu time for operating system but system B would need to give use half of the cpu for the operating system; half of which is basically "nothing" as demonstrated by system A of this example.

In real world these numbers are a lot closer to each others, but just for example's sake I made it a big issue. Nevertheless, the more efficient system can do more real work with the same hardware than the less efficient system. THAT's what this performance talk is all about today: the software.

But some people may "think" OS X feels slower because it doesn't always react instantly to a mouse click.

That's exactly why some other people may "think" OS X feels faster (in doing real work) because it doesn't always consume cpu cycles for faster-than-fast-enough GUI refreshing and such.

they should just number them by their SPEC integer and FP scores added together

Why bother; just sell a model "2006/Q2 Duo Core" or "2007/Q1 Quad Core" chip. We all like car analogies, but not many of us talk about cars by production numbers; at least I recognize cars better by car type and model year.