Apple GHz vs. PC GHz

hey daniel you came at a good time
i think thats why apple switched to intel to get rid of the megahertz myth now i dont know much about what other pc's you have been looking at but as far as i know the core duo is supposedly the lowest wattage highest performance portable chip available as it is being implemented in other laptops such as dell acer and hp

but one thing i want you to remember is that people people buy from apple not for the hardware but for the awesome operating system that runs on it OSX
but the switch to intel has made it much sweeter

No simple answer. Every processor line is different. For instance, on the PC side...

An AMD Athlon 3700+ runs at 2.2GHz and is VERY ROUGHLY equivalent to a Intel Pentium 4 running at 3.4GHz. Then you throw in the Intel Celeron, AMD Semperon, Intel Pentium M, etc, and everything becomes a little less clear.

Let's assume that this is a hypothetical question, considering that the MacBook Pro is not available with a 1.5 Ghz Duo Core processor (for arguments sake).

With the G4/G5 processors, Apple ran slower in Ghz, but how the processor processed the information is what you need to take into consideration.

Let me use this as an analogy. We have two pizza parlors. For lack of better terms, let's name one PC Pizza and Apple Pizza (kinda like a desert pizza, eh?)

You order a pizza from PC Pizza, and you stand there and watch them make this pizza and you notice that it takes a total of 21 steps to make this pizza. And you notice that the time to take this pizza was 60 minutes.

Next time, you order Pizza from Apple Pizza. You wait and watch, as you did with the PC pizza place, only Apple's pizza place makes this pizza in 7 steps, and you only wait 20 minutes for it to be made.

The point I am trying to make is that the PPC was more efficient in sending the data through the pipleline as is the intel's way of doing things.

But, the playing field is level now. Now the question is who's operating system will make better use of the processor? Just think of Windows registry, Antivirus, Adware, and bloated OS on top of things, and compare the MacBook Pro 2.16Ghz vs. Beige box 3.4Ghz, I think they would be about equal. They would both do what you need them to do, but in my opinion, Apple will cause less grief with the OS as compared to Windows (pick your flavor).

So after all this blathering I have just done, you can check out this web page for more information. http://www.jmusheneaux.com/6000j.htm

I'm sure others will chime in on this, but here are a few things to keep in mind.

First and foremost, the CPU speed is only one factor in the perceived speed of a machine (Mac or PC). RAM, front-side bus speed, disk speed and video processing (GPU and VRAM) all have an impact on perceived performance (sometimes more so than chip speed).

If you're going to compare CPUs across product lines, make sure they are equivalent. For Macs, this got a lot easier with the switch to Intel by Apple. You can now compare their dual-core chipsets to those used by Dell, Acer, IMB/Lenovo, etc in their laptops and desktops.

I wouldn't, however, compare the dual-core Intels in the iMac and MacBook to single-core Pentiums or Athlons, etc. They are different architectures - 2x2GHz means two processors running at 2GHz. Comparing that to a single processor running at 3GHz (with a different chipset supporting it) is sketchy at best. You might expect the dual core to be generally faster (2x2GHz = 4, which is greater than 3, right?), but it's not that simple, as mentioned above. Some single-CPU, processing-intensive activities might be faster on the single 3GHz chip.

If you try to compare different CPU architectures altogether (like PowerPC vs. Pentium), things get much murkier in terms of hard numbers. The closest thing to quantitative measures you can make are the SPEC benchmarks (SPECint, etc.), and/or wall clock measurements on real-world activities (like "open, save and close these 100 reference Word files"). But even these measures can be subjective - the Word performance measurement doesn't mean as much to someone that uses OpenOffice, for example, or to someone that predominantly does image processing in Photoshop. And SPEC benchmarks don't tell you how the mix of operations will play out for your activities (e.g., faster floating-point ops doesn't mean much if your bottleneck is loading stuff from disk).

The best and most accurate test is to try out a machine, doing the kinds of activities you expect to do with it. If you get a chance to do this for a decent interval, you'll get a good feel for the speed and responsiveness of a machine.

Dale -

That explains the core operation - but not the core limitations.

PC Pizza only has 64 registries (or employees).
-Or rather, the x86 based CPU chips run in CISC. Complicated Instruction Set Computer. The instruction is bigger, with less registries to handle.

Apple Pizza has 128 registries (or employees).
-Or rather, the PowerPC based CPU chips run in RISC. Reduced Instruction Set Computer. The instruction is reduced, with more registries to handle.


So, to simplify this to make use of your analogy - think of the Superbowl. Typically the biggest pizza day of the year.

Pc Pizza has it's 64 employees - And because PC's instructions are so complicated - some of those employees have to walk around checking to make sure the size of the pizza crust is handled correctly, some have to run around making sure the boxes are being assembled properly, some have to check to make sure the delivery drivers are being assigned properly, etc. So not only do you have less staff, those you do have are being used to do tasks that just be automatically handled.


Apple Pizza has it's 128 employees - First of all, there is a "pipeline" which handles all the tasks that should be 'automatic', so fewer, if any employees have to run around doing tedious tasks. So you have more employees, and more of them can contribute to doing the actual productive work it takes.

So on gameday, if there are 5,000 orders that PC Pizza has to deliver - they are bottlenecked right away because they have half the employees - and some of those employees give no productivity because they are busy handling such simple tasks. Even if you have all 64 employees running around at a speed of 2,000 steps per second, they can only get so much done.

If Apple Pizza gets 5,000 orders - they have twice the employees to process all these orders, and each of their employees are doing very simple and streamlined tasks. There is little to no bottleneck - and the pipeline is always there handling tasks automatically so the is little if any wasted productivity. So even if Apple is only running around at 1,000 steps a second - they are accomplishing twice of what PC can do at 1,000 steps a second.

First, I'm on a diet and can't eat pizza so I wish that we were talking about another food!

Second, the anologies are so good! I will use these myself. The simple answer I would have said is that in the PC world there is so much "junk" in the code that you need a huge processor just to separate all the essential from non-essential crud that goes into making it work!

Spot on! I learn so much here...

Bob

Analogies may be good enough for non-computer people, but they're usually not very good.

Note to peacheasy: computers don't have registries, they have registers. Registers are very high-speed memory locations inside the processor. While a large-number of general purpose registers is a common trait of RISC processors, registers alone are not what distinguishes RISC from CISC. Anyway, the whole RISC-vs-CISC debate has faded into the background these days, for several reasons. Here are a couple: when you get down to it, the tasks we do involve the same steps regardless of whether you're running a CISC CPU or a RISC CPU. RISC CPUs are known for performing some very simple operations very quickly. However, it takes more of these operations to get the job done. Another thing to keep in mind is that most CISC CPUs these days are actually a microprogrammed CISC implementation running on a more RISC-like core.

Bottom line: Deriving any conclusions regarding actual performance from a comparison of clock rates is only valid when both processors have very similar internal architecture.

Quad G5 2.5Ghz 4.5GB 2x250G, PB 15" 1.5Ghz,80G,1.5G Mac OS X (10.4.5)

This is deadon - the wild variances in processor efficiencies, pipeline widths, bus speeds, CISC vs. RISC (which really are no different anymore), Northbridge performance, memory latencies make comparing processors outside of the same line (eg., a 2.0GHz Core Duo vs. a 2.0GHz Core Duo) a crapshoot.

The only thing you can rely on is personal feel and benchmarks in the applications you use.

Did I type registries? Oops! haha no more drinking and typing.

As for current chips - as far as I am aware, only the highest end chips that are P6 or PentiumPro or newer have been designed as RiSC chips which actually emulate a CiSC architecture environment.

I have a friend who is a system administrator for a PC network. The topic of Macs using Intel chips came up (he doesn't know much about what Apple is doing lately).

He asked me how fast the fastest Intel processors were that Apple is using...I told him around 2.1 or 2.2 ghz.

All he could do is smile, and say, "why are PC chips still faster?" (he's thinking about the 3 ghz + chips)

I mentioned that Apple uses dual core chips. He just kept smiling and said, "yeah, but why are they still slower?"

What do I tell this guy? (other than the obvious things I would LIKE to tell him!)

Powerbook G4 17 1.5 ghz Mac OS X (10.4.7)

Because Intel doesn't produce a 3.2GHz Core Duo yet, and the 3.2GHz Pentium 4 (which he's likely talking about) are unsuitable for laptop use -- too hot, too much power consumption.

Don't even get me started on the Pentium D (64-bit Pentium 4 equivalent with dual cores)'s power consumption.

There are two issues:

1) The MBP uses a set of chips designed for notebooks rather than desktops. Your friend is comparing against a processor for desktops.
2) The MBP uses a new generation of chip, which gets more work done per GHz. This isn't a Mac vs PC issue. Dell's notebooks are using the same chips.

When Apple comes out with their pro desktop line, I assume they'll use the desktop version. It will run a bit faster in GHz. However it will still be the new generation, which will be more efficient both in power and in compute / GHz. So it may still be slightly below 3 GHz. However that chip has been doing well in tests compared to older chips running at speeds over 3 GHz. Intel's newer chips are designed so they don't have to run as fast as the older ones to get the same results. Again, PC vendors will move to the same chips when they are available. You're seeing the difference between old and new Intel chips, not a difference between Mac and PC.

Hallo,

the answer is quite simple:

Desktop PC are using a Pentium 4, Notebooks a Pentium M processor.

The Pentium M is somehow the successor of a Pentium 3, which delivers more performance at slower processor speeds (= GHz ).

You may roughly compare the performance of a 1GHz Pentium M to a 1,6 GHz Pentium 4.

So a 2,16 GHz Pentium M will be as fast as a 3,4 GHz Pentium 4.

But MBPs do have the new dual core processors (two Pentium M cores one case), so it will be as fast as two 3,4 GHz Pentium 4 machines. (Yes that's true)

Regards

Marc