Which Processor to Buy for Your PC: all you have to keep in mind
If you are upgrading your PC or buying a new computer, the processor is one of the components that need more attention. The investment is important and it’s therefore, essential to know the specifications of each model in order to choose the best processor.
Choosing Processor: What should I look for?
The details and features around a processor go far beyond the manufacturer, cores or operating frequency. Knowing what each number and associated technology means is key so that, within the investment, we want to get the most out of it and not end up getting the worse, which does not offer us the ability to process what we need to play, work or for our day to day leisure.
Number of Cores and Frequency of work
The two main and most referenced data when we speak of a processor are cores and frequency. The combination of both elements and their individual figures globally mark the performance of this component of our Personal computer.
The figures that first appear when comparing processors are the number of cores and the frequency of work. But do not be fooled by the figures because, except for specific cases, they are not data that we can directly compare between models of processors.
The cores of a processor define the number of independent central processing units with which the chip counts. Currently, we can find processors with dozens of cores, although if we limit ourselves to models designed for the consumer market, it is common to have processors with 2, 4, 6, 8 and up to 10 cores.
Associated with the number of cores we find another key figure to evaluate the performance of a processor: the frequency of work. It is the speed of opening and closing of the transistors that make up the processor, and the one offered by the manufacturer is the operating frequency in which the cores will work most of the time inside the TDP of the chip.
Regarding this value, in reality, what marks a better performance is the IPC or instructions per clock cycle. Here Intel takes the lead in its latest generations of processors equal to GHz, so do not go insane by the frequency associated with each processor. They are not figures that directly compare an increase of yield with the same relation, except in generations and similar models.
In some models of Intel as well as AMD, in addition to the base frequency, we find another data of maximum frequency or Turbo to which it can increase its operation a nucleus in certain circumstances of temperature and consumption. The operation of this technology depends on whether the chip is from Intel (and whether it is Turbo Boost 2.0 or the new generation 3.0) or AMD, but it is not especially critical for most consumers.
Overclocking and Multithreading: Getting More out of the Processors
In some ranges of processors, we find the so-called yarns as a number associated with cores. It is a technology that Intel admit as Hyper-threading and divides the work of a core into two threads. In the market, not all processors are multithreaded, so depending on the task we want to perform with our PC, we should opt for those that allow two threads per core.
If we refer to AMD processors, the division of labour of each core in two threads is similar to that made by Intel.
Another improvement that we can take into consideration when looking at cores and working frequency of a processor is the possibility of overclocking, i.e adjusting power, voltage, cores and other values to raise the performance of the equipment. The usual is to increase the maximum working frequency that marks the manufacturer. That task can be done on both Intel and AMD processors, which are ready for it.
In the case of Intel models, we can detect them because they are marked by “K” in their name and therefore they are prepared to perform this action thanks to that they are unlocked (with their associated software) and are recommended for those who wish to squeeze their equipment to the maximum even knowing the risks.
Voltage and Cache
If we continue to look at the datasheet, there are two values that we should not let go when talking about processors. The first is the cache memory that includes the chip. It is a very fast but volatile type of memory located next to the processor and it tries to ensure that the CPU has the information it will need to list when it needs it, and not reduce its processing capacity.
The most forgotten of the specifications of the processors is the cache. If you can afford it, you do not get on that figure when it comes to choosing chip.
In theory, the cache (of different levels L1, L2, L3, L4 or Smart caché which is what it calls Intel) if better, then the greater, something that happens in processors of higher range. In the range of non-exclusive consumption is usually between 2 and 8 MB.
Regarding the TDP (Thermal Design Power), we are talking about the average power in watts that dissipates the processor working on the base frequency with all its active cores. The lower that figure, the less warming the chip suffers and the less consumption we have in the equipment. This value is closely related to the lithography of the manufacturing process, expressed in nm.
Memory Socket and Memory
Whether we are going to buy a new processor with which to mount a PC from scratch or update a current model, two are the main compatibilities to which we must attend to motherboard and RAM.
Each processor model usually has support for one or several types of RAM (as well as maximum amount of RAM supported), so make sure that if you want future compatibility or with current components, that type of memory (DDR3, DDR4 …) is considered in the technical file of the processor that you are going to choose.
Although you can look at the processor as a separate component, it actually has a very close relationship with motherboard and RAM, especially when upgrading a computer.
You should also pay attention to the type of socket (mechanical and electrical connections between the processor and the motherboard) with which the processor to which you aspire is compatible. Maybe you are thinking of upgrading your PC without buying a new motherboard and therefore you have to limit the search for the new processor. If it is a new computer, the processor choice will largely determine which type of motherboard (or RAM) is the most convenient.
Other features that depend on the type of processor you have chosen but have considerably less weight except specific cases could be PCI-Express buses or technologies of protection or proprietary of each manufacturer.
Lastly, do not forget the type of suitable heat sink, as well as the cooling system that each processor model may require for proper and safe operation.
Integrated Graphics on the Processor
Depending on the range of processor you choose, the generation and manufacturer, the processor can carry the integrated graphics (GPU). It’s the strength of AMD chips in the mid-range, at least until very recently.
If you are not going to need dedicated graphics you should look for a processor model that takes the integrated graphics and gives you the performance that requires the use you are going to give the PC.
Depending on whether you plan to purchase or mount a standalone graphics solution regardless of whether or not the processor is integrated, you will have to opt for a series or another chip.
For office tasks and use of games not very demanding/basic edition, the GPU associated with the processor in certain ranges can be more than enough. And even with new AMD APU solutions from superior ranges or the new Intel Core Iris, thinking about more serious gaming or working without the need for dedicated elements is not a problem.
In the case of integrated solutions, there is important data to consider such as operating frequency, maximum resolution support, DirectX and OpenGL compatibility, support for combined dedicated solutions, proprietary technologies of each manufacturer or the maximum memory supported, which it shares with the system. And that also must be taking into consideration.
Any corrections or suggestions from your end?