It has always been a frequent question -- "Will I benefit from multiple processors?"

With the growing popularity of dual core processors, the topic is more important than

ever! Will multiple processors or a dual core processor be beneficial to you, and what

are the differences between them? These are the questions this which always comes in

mind before purchacing a machine.

A major question for some people getting ready to buy a high-end system is whether they

want or need to have two processors available to them. For anyone doing video editing,

multi-threaded applications, or a lot of multitasking the answer is a very clear 'yes'.

Then the question becomes whether two separate processors (as in a dual Xeon or Opteron

system) is the way to go, or whether a single dual core processor (like a Pentium D or

Athlon64 X2) will do just as well. Dual CPU vs dual core -- which is better?!

Intel did not increase the speed of their front-side-bus (the connection between the CPU

and the motherboard) when they switched to dual-core, meaning that though the processing

power doubled, the amount of bandwidth for each core did not. This puts a bit of a

strain on the Intel design, and likely prevents it from being as powerful as it could

be. To counteract this effect, Intel continues to use faster system memory to keep

information supplied to the processor cores. As a side note, the highest-end Intel chip,

the Pentium Extreme Edition 955, has a higher front-side-bus speed, as well as having a

larger (2MB per core) cache memory and the ability to use Hyperthreading (which all

non-Extreme Edition Pentium D processors lack). This makes it a very tempting choice for

those wanting to overcome some of the design handicaps of Intel's dual-core solution.

AMD, on the other hand, does not use a front-side-bus in the traditional sense. They use

a technology called HyperTransport to communicate with the chipset and system memory,

and they have also moved the memory controller from the chipset to the CPU. By having

the memory controller directly on the processor, AMD has given their platform a large

advantage, especially with the move to dual-core. The latest generation of AMD

single-core processors can use single- or dual-channel PC3200 memory, but it is

interesting to note that even though dual-channel operation doubles the memory speed, it

does not double the actual memory performance for single-core processors. It appears

that dual-channel memory just provides significanly more bandwidth than a single

processor core can use. However, with dual-core processors all that extra bandwidth can

be put to good use, allowing the same technology already present in single-core chips to

remain unchanged without causing the same sort of bottleneck Intel suffers from.

there are also multicor-multiprocessor boards :-)

As a fraction of total system cost, the price difference between single and dual core is low enough that you don't really need much benefit.

If you run major computational tasks, you would like to get the low cost performance boost of splitting the work between two cores. But whether the work will split depends on the specific program.

Even if you run no major computational tasks or the ones you run will use only one processor, you may find a second core is valuable just to make the system more responsive. I have more experience with this in XP, but I think some of the same effects occur in Linux. When the system is busy with network traffic or other activities, the scheduler in a single core system may be far less responsive to new mouse and keyboard events than scheduler documentation would lead you to expect. It should immediately steal a few milliseconds away from whatever it is busy with to respond to the user, but it often doesn't. With two cores the system is far less likely to become scheduled so badly that it is unresponsive.

A multiple core design is harder to cool well and a multiple processor design is harder to design/build well. For the ordinary end user, two single core CPUs will cost so much more than one dual core that any subtle differences in performance (after normalizing to the same speed and L2 cache size) will be overridden by cost. (I'm talking primariy about the big price difference in comperable performance motherboards for single vs. dual CPU packages. For the CPU itself, a dual core CPU might not cost much less than two comperable single core CPUs).

For the very high end, where you need more high speed cores than the manufacturer can build/cool in one package, you need multiple packages with multiple cores per package. But for an ordinary user who wants two processing units, only dual core in one package makes sense.

What matters is the actual latency and bandwidth of the memory accesses. The font side bus may not be the limiting factor. If you had two processors with seperate front side busses sharing the same memory, you might have less total bandwidth than two core sharing one front side bus.

Where did you get that from?

The balance between CPU speed, L2 cache size and memory speed is heavily dependent on the type of processing you are doing.

Some kinds of processing depend heavily on memory so an increase in effective memory speed will translate directly to an increase in performance, an increase in CPU speed would make no difference, and splitting the work between two processors would slow it down a little vs. using just one.

Other kinds of processing fit almost all their memory needs into L2 cache so the speed of the actual memory barely matters at all.

In between, I'm sure you could find problems for which the L2 cache miss rate of one CPU is light enough that memory bandwidth hardly matters while the L2 cache miss rate of a dual core would be heavy enough that memory bandwidth is very significant.

If you're going to quote someone word for word, at least include the url for the original article, otherwise it's plagiarism.

Original article:
http://www.pugetsystems.com/articles.php?id=23

I should learn to check for things like that before typing a real reply!

The starting post of this thread was not only plagiarism, but also plagiarism of content that was a bit out of date a year ago when it was written.

Even if you plagiarize, it is still wrong.

They did increase to DDR2-667, but they include a hidden feature that used in Pentium 4 which is quad-pumped. Also they did not need to increase it because the fetching and caching algorithms a very, very efficient. They also increase the FSB to 800 MHz and now it is up to 1333 MHz. Ouch, have fun getting that amount of heat dissipated.

Yes it is true that AMD uses HyperTransport to interconnect to the south bridge chip and other devices. Also it is correct that the K8 processors and above contains built-in memory controller. What is not true is they did not use HyperTransport for system memory. If they did, latency will increase. They use a special direct connection to the internal memory controller of the K8 processor and above.


A single non-microarchitecture core does a better job processing one instruction at a time than a single microarchitecture processor. Just think that microarchitecture processor is a general purpose DSP. A dual core processor is slower than two physical processors because the dual core processor is specialized to processing a lot of instructions at once at a cost of data bandwidth compared to physical processors. Physical non-microarchitecture processors can handle large or small loads of data but at a cost of executing multiple instructions at one time. Though other hardware can aid in processing a lot of data several times faster than any processor. This type of processor is a general purpose GPU.


BTW, this forum is for Linux hardware. Nothing in this thread is for Linux.

I think, having reviewed the strange posts by "rockaway", that this is another forum-bot.
This thread is probably not worth replying to. I have reported it and will let the mods decide.