Review of mobile processors : Test | Hashrate | Specs | CPU | Config

The heart of any laptop is the central processing unit. And if in the class of processors for desktop computers only price and performance play a role, then in the case of mobile processors, these characteristics

another equally important parameter is added – power consumption, and not so critical (but in some cases playing a role) – the size of the processor. There is an opinion among computer users that the performance of modern laptops directly depends on the power of the processor. This is not entirely true, in fact, the performance of laptops, as well as in other desktop computers, depends primarily on the configuration of the entire system, or rather the platform on the basis of which the laptop is created. The processor is the main part of the entire platform, so to speak, its heart. Therefore, when buying a laptop, first of all, you should pay attention to the central processor. Very often, when choosing a processor, the first thing they are interested in is the clock frequency – it is considered that the higher it is, the more productive the machine itself. This is not true. And even more so, the clock frequency of the processor in no way characterizes the performance of the entire system. The performance of the processor itself is determined by architecture, clock speed, and cache size. A higher frequency is needed, in fact, only for complex mathematical calculations. Therefore, you should not focus on this parameter. Of course, if you will not be doing math calculations or processing 3D graphics on your laptop. In all other cases, the speed of work will be determined primarily by the balance of the entire system as a whole: the amount of RAM, the speed of the hard disk, the processor. After you decide on the type of processor and its performance, you should start choosing a specific model. Let’s start with the manufacturer.

INTEL mobile processors

At the moment, a large part of the market for mobile central processors is occupied by the Taiwanese corporation Intel with a family of Core i3 / i5 / i7 chips, but it should be noted that in modern laptops completely different chips are hiding under this marking: at the beginning of 2011 they were transferred to another new microarchitecture Sandy Bridge. The new S andy Bridge microarchitecture is the official successor to Nehalem.

The difference between the new Sandy Bridge architecture and the old Nehalem is that the so-called “system agent” is located in Sandy Bridge on the same die with the computing one, which includes a new generation graphics core, DDR3 1333 MHz RAM controllers, PCI Express 2.0 controller and blocks power management and display output. Other major innovations include L0 microinstruction buffer, shared L3 cache (L3 or LLC, Last Level Cache), next-generation Turbo Boost automatic overclocking technology, and SIMD AVX instruction set support. All chips are manufactured according to 32nm technological standards.

The new processor models, keeping the old Core i3 / i5 / i7 markings, received other indices that clearly distinguish them from the previous generation chips. All indexes of Sandy Bridge mobile processors are four-digit and start with the number “2”. There can be one or two letters at the end of the numeric index. The letter “M” stands for a standard mobile processor with a heat dissipation of 17 to 35 W, “E” stands for a standard embedded chip, the letters “QM” and “QE” stand for quad-core mobile processors with increased performance with a heat dissipation of 45 W, and the index “XM” is assigned extreme “models with an unlocked multiplier. By the index, one can also distinguish between standard chips with normal and reduced heat dissipation: for the first it ends with “9M”, for the second – with “7M”. Sandy Bridge mobile models differ from desktop models only in lower frequencies, reduced power consumption and some system features.

Intel currently produces 22 Core family mobile processors based on the Sandy Bridge microarchitecture: one dual-core Core i3 chip, seven dual-core Core i5, seven dual-core Core i7, six quad-core Core i7 and one extreme Core i7 XM. All models are equipped with multithreading technology Hyper Threading and automatic overclocking Turbo Boost except for the younger model Core i3.

The amount of shared L3 cache, available if necessary via the ring bus to all cores, including the graphics and the system agent, is 3 MB for the dual-core Core i3 and Core i5, 4 MB for the dual-core Core i7, 6 MB for the quad-core Core i7 and 8 MB for the flagship quad-core Core i7 2820QM and Core i7 Extreme 2920XM.

Along with high-performance models with the Core iх index, Intel also produces two budget mobile models under the old Celeron brand, built on Sandy Bridge: B 810 (1.6 GHz) and B847 (1.1 GHz). These dual-core chips feature legacy integrated graphics, no Hyper Threading, no Turbo Boost, and 2MB L3 shared cache.

A separate category of mobile processors manufactured by Intel is the Atom: these chips are designed for installation in netbooks and nettops – inexpensive laptop and desktop computers focused primarily on the Internet. The Atom design has not undergone any fundamental changes over the year, only the model range has been expanded, and they are still produced using the rapidly aging 45nm technology. The only significant addition to the Atom family is the “system on a chip” Oak Trail, presented in April 2011 and designed for smartphones and tablet computers. This chip is very similar to the 2010 Moorestown, designed for portable devices based on Linux. The new chip features a PCI controller, a graphics core that supports 1080p HD video and HDMI, and most importantly, the ability to work with a variety of operating systems, including Windows, Android and MeeGo. So far, there are only two chips in this series: Atom Z650 (1.2 GHz) and Atom Z670 (1.5 GHz).

AMD Mobile Processors

In June 2011, AMD introduced the new A-Series mobile APUs, which are part of the Llano family of chips based on the Fusion Architecture that combine one or more AMD X86 processor cores and hundreds of Radeon cores on a single die. In other words, we are talking about the merger of the central (CPU) and graphics processing unit (GPU) into a single unit – APU, Accelerated Processor Unit (accelerated processing unit). The Liano is the successor to the Zacate processor, which has given Intel a lot of headaches by competing with their Atom chip in the laptop segment over the past few months.

The whole idea behind APUs is to move from general-purpose but low-efficiency to specialized and easily scalable high-efficiency chips. Such chips can consist of various types of computing cores, selected in different combinations to solve certain problems. The first models of this type were supposed to be multi-core processors with integrated computing and graphics cores.

Despite the fact that the concept of the APU itself was unveiled at the end of 2006, AMD is clearly late in bringing these chips to the market. Intel has been successfully selling processors with integrated graphics for several years, and the first mass-produced hybrid chips from AMD of the Fusion family, codenamed Brazos, were introduced only in early 2011. The processors consisted of one or two Bobcat computing cores, a single-channel DDR3 1066 MHz memory controller, and a Mobility Radeon HD 6xxx (Cedar) graphics core with eighty universal stream processors. The integrated graphics fully supported DirectX11, DirectCompute and OpenGL APIs and were hardware accelerated for HD video playback up to 1080p.

A-Series includes two or four Stars (K10.5) -based computing cores with 1MB L2 cache each, integrated graphics, up to 32GB dual-channel DDR3 1600MHz memory controller and PCI Express 2.0 controller on 24 lines, sixteen of which can be used to connect discrete graphics accelerators. The processors are manufactured using 32nm technology.

The most important advantage of the A-Series, perhaps, over all modern central processors is the graphics built into the processor, which can compete on equal terms with discrete graphics cards. For the first time in history, the integrated video core is capable of “pulling” even very resource-intensive computer games. In the older A-Series models, the Radeon HD 6620G graphics accelerator is installed with four hundred universal stream processors, that is, at the level of the discrete mobile graphics of the middle class Radeon HD 5650/573/5750/5770 or HD 6530/6550/6570. The junior chips use video cores Radeon HD 6620G with 320 or Radeon HD 6480G with 240 universal stream processors.

Another important advantage of the A-Series is the possibility of simultaneous operation of the on-chip and discrete graphics accelerator, in contrast to Intel’s version of switchable graphics, which supports image output either through integrated or discrete video. According to AMD, this “pooling of forces” can, under favorable circumstances, improve graphics performance by up to 75 percent over standard graphics performance when using only AMD Radeon discrete graphics cards.

To process high-definition video, the chip implements a unified third-generation video decoder module (UVD3) capable of hardware decoding of all common formats, including H.264, MPEG-2, MPEG-4 (DivX / XviD) and Blu-ray, including Blu- ray 3D. Proprietary technologies Perfect Picture HD and Steady Video are designed to ensure the display of video of flawless quality.

The new Turbo Core technology, similar to Intel’s Turbo Boost, automatically overclocks both individual computing and graphics cores. The redesigned power management system, according to AMD estimates, promises up to ten and a half hours of laptop battery life based on A-Series chips.

The AMD A-Series processor line currently consists of seven models: two dual-core and five quad-core. The heat dissipation of modifications with the MX index is 45 W, with the M – 35 W.