Processor frequency: clock, maximum

A whole generation of computerusers who did not find the famous "megahertz race" that unfolded between two leading manufacturers of CPUs for desktop computers (who are not aware of Intel and AMD) at the turn of the millennium. Its end came about in 2004, when it became obvious that the frequency of the processor - not the only characteristic that affects its performance. Extremely "voracious" and extremely high-frequency Pentium IV processors on the Prescott core were closely matched to 4 GHz, and at the same time they hardly competed with the K8 architecture, on which new "stones" from AMD were built, which had a frequency of no more than 2.6-2, 8 GHz.

After that, both manufacturers synchronously departed fromthe practice of identifying their products at the operating frequency and moved to abstract model indices. This decision was justified by the unwillingness to mislead the end user about the performance of the processor, focusing only on one of its characteristics. Indeed, there is also the frequency of the processor bus, and the size of the cache memory, and the technological process by which the core is made, and much more. But the frequency of the processor is still one of the most intuitive for most people measuring the "quality" of the CPU.

The processor clock speed, indeed,affects its performance, characterizing the number of operations performed per second. But the fact is that processors, built on different cores, spend a different number of cycles on performing one operation, and from generation to generation this parameter can differ at times. It is thanks to this that the current processor with a nominal frequency of 2.0 GHz will leave far behind the flagship seven years ago with a clock speed of 3.8 GHz. In addition, the speed of the processor, as already mentioned above, affects the size of the cache (the more it is, the less the processor will have to resort to relatively slow RAM), and the bus speed of the processor (the higher it is, the faster the data will be exchanged between "stone" and RAM), and many other, not so noticeable, but from no less important, characteristics.

Recently, the notion of the maximum frequency of the processor begins to enter into use.

Gradually, both Intel and AMD are implementing in theirproducts such function as auto-acceleration. The technology is essentially the same, one manufacturer calls Turbo Boost, the other is Turbo Core, but this does not change its essence: the processor frequency can dynamically change, and automatically, without user intervention. The need for this technology is due to the fact that the multi-core of modern processors has become, in fact, the norm, but the multithreading of modern applications, unfortunately, not yet. The operating system, seeing that one of the processor cores is loaded significantly more than others, independently increases the frequency of this kernel, while trying to leave the processor within its "native" heat pack (ie, the system tries to hedge against overheating of the equipment). And, depending on the processor model and on specific conditions, such a frequency increase can be from 100 to 600-700 MHz, and this, you will agree, is a significant increase in performance. This technology supports most of the latest processors from both manufacturers. Intel has, in particular, all the CPUs in the Core i5 and Core i7 series, AMD has all the processors on the AM3 + connector, the processors on the FM1 connector (except the processors with the disabled graphics core), and some "stones" to the AM3 platform (six-core Tuban and quad-core Zosma). And for Intel processors based on the Socket 1155 connector, this auto-acceleration is all the more relevant, given that due to some architectural features a full "overclocking" by increasing the frequency of the processor bus is practically impossible. However, this topic is a completely different article ...