I was sitting last night and my roommate asked me what kind of processor my mac had. I told him 2.16 Ghz, he then said that’s kinda slow, I then told told him that it’s actually 4.32 Ghz since it is dual core. He then went off trying to convince me that I really don’t have two cores, My chip is named Intel Core 2 Duo for a reason! I told him that just because I have two cores does not necceraly mean it’s twice as fast. He then tried to relate it to RAM… go figure. You know, that 1 GB is not 1000 MB, it’s really 1024 MB, and that I don’t actually get 2x 2.16= 4.32 Ghz it’s more like 3 Ghz in total. I tried to show him, prove to him, explain to him, no dice. Tried to show him wiki, intel, apple, computing sites. nope he doesn’t get it. If I am wrong, PLEASE tell me so I can apologize to him. If there is an app any shred of evidence that I am correct tell me show I can tell him. Any computer experts here?
Does it matter who’s wrong and who’s right?
No matter what you’re discussing, there will always be this kind of situations where someone is wrong and doesn’t want to admit it or where someone is wrong and can’t or won’t see it.
Keep this in mind: in a relationship (of any kind, not just romantic) it doesn’t matter if you’re right or wrong, all that matters is how you and the other person feel.
Wrong or right, your friend might feel angry or wronged (supposing the discussion was rather intense) so I’d just tell him you’re sorry about getting upset over trivial things and bury the discussion under the sand.
Maybe he’s thinking you have a dual processor machine as opposed to a dual core machine. The Core2 Duo has two um, cores on the one chip as opposed to a dual processor machine which has two seperate cpu’s. You can run multiple threads but, as you said, it’s not twice as fast. Maybe your roommate has processor envy 
2.16 GHz are 2.16 GHz. No magical frequency manifold from multiple cores. Two cores simply means you can execute two “program streams” (threads) simultaneous, the same applies to two single-core CPUs. If the software can divide its calculations in separate threads and scales well, it does indeed reach close to twice the throughput on a dual-core CPU.
I deliberately said throughput. As analogy, imagine using two trucks to deliver stuff instead of one. Your stuff doesn’t travel faster or anything, but you can deliver twice as much in the same time. But that doesn’t help you if you only have one truck of charge and want shorter delivery times instead…
Apart from that, frequency says little about performance. The magical term here is “IPC” (instructions per cycle). A CPU with high IPC (like Intel’s Core 2 architecture) needs less MHz than a CPU with crappy IPC (like Intel’s Pentium 4) for same performance, but back then, Intel figured more MHz sounds faster, and geared their marketing to that.
Just look at some reviews, and see how much more GHz a Pentium-D (also dual-core) needs to come close to a Core 2 Duo…
holy crap i smell misinformation!
here is the truth; if anyone can contradict what I say they can have a cookie:
The processor has two cores. They share a level of cache on the processor die. I am not quite sure what the level is, but it is likely L3 or L2. These are very fast temporary storage areas for the code to be executed next.
The processor in your mac very rarely runs at 2.16 ghz. When surfing teh intarwebz your processor is likely running at about 1.0 or 1.5 ghz. It does change speed almost instantly, so no worries there. When you start a render it works at its max speed, 2.16ghz.
THE PROCESSOR DOES NOT RUN ANY FASTER THAN 2.16Ghz. no matter what.
It has a different amount of processing power than two independent processors or one single core proc. Two independent procs don’t share that on-die memory. This makes them faster for some things, slower for others.
Its all about parallelism in the code. If the code can be split into two tasks that work at the same time, using the same code to do it, a dual core proc will rival a dual proc system. An example is solving a times table from 1 to 10 square. The only operation is simple and short (multiplication) and the data is short (single integer). Also, each part of the square can be solved independently of the others. So the entire program can be placed in the cache and sped through very quickly. This will execute in about 2.0 times as fast as a single proc.
The problem comes when the code can’t fit in that little cache memory. If it has to move in and out of the RAM it will be slower because each proc will be vying for that same bus time to get the data.
Okay, so you wanted to know what the speed of the processor is. It is 2.16 ghz max. It can do about twice as many instructions each clock cycle as a single processor, but those instructions can only be helpful in optimal cases.
Also, your idea about 1GB = 1024MB is completely off topic. That is just naming issues, the processor really does run right about 2,160,000,000 cycles per second.
Oh, and so no one feels the need to point this out, I do know that executable commands, such as in assembly, take more or less than one cycle to complete. Often one to load the value, a few to manipulate it, and another to send it back to memory. Some procs can do several a cycle. That depends on the architecture of the processor. In this case, the core 2 duo is a very efficient architecture, making it very good for its raw speed.
Even more important is that, as AMD has been telling us for years now, architecture is more important than clock speed. In this case your processor is much faster than any pentium 4 processor, even running at 3.5 Ghz.
So shove than in your friend’s face.
also remember that clock speed isn’t the only way, or even a decent way, to measure a processor’s performance. Your CPU can outperform most any Pentium 4s, even the 3.2 gHz ones.
Core Duo and Core 2 Duo have a shared L2 cache, Pentium D and Athlon 64 don’t share any cache. Upcoming quad core AMD CPU will have a shared L3 cache…
Well that’s really a matter of memory/data path architecture, a Pentium D is also “dual core” (you could say a bit too literal even), yet the cores communicate via FSB just like two (single-core “Netburst”) Xeons do. On AMDs side the multiprocessor boards can have NUMA which has its own bag of pros and cons etc…shared caches do not only have advantages either…
In modern CPUs all instructions need several cycles to complete, that’s why they became pipelined so you can start executing new ones each cycle without the previous instructions already finished. And to feed more that one instruction per cycle to the CPU, they became superscalar (the original Pentium was the first superscalar x86 CPU IIRC), and to not have the execution units mostly empty due to data dependencies things like branch prediction and out-of-order execution have become more and more sophisticated.
Pentium 4 has a super deep pipeline which allows very high frequency, but little parallelism and horrible misprediction penalty. Though the concept didn’t really fail due to bad logic design alone, the power density just went through the roof after several iterations of structural shrinking…it physically could easily clock at 6 or 8GHz (which intel really planned, and has also been proven to be possible by enthusiasts), wouldn’t it go up in smoke without liquid nitrogen 
have you seen that? And that’s an old-school, 90nm opteron.
K10 core will supposedly outperform K8 core by ~70%. And there are 8core K10 parts planned. Holy shit, that would make a nice little workstation. Single socket!
Well, in short, your friend is right. Your core 2 duo runs at 2.16ghz. Even if it had 10 core, it would still run at the same clock speed, its just multi-threaded not multi clock speed 