#processor performance

HMI (Human-Machine Interface) is a medium for interaction and information exchange between systems and users. It is essential in fields involving human-machine communication, and can be seen in many industries. As technology advances, HMI continues to evolve. In addition to data collection, control, and display, future HMI will incorporate new interactive forms to enable machines to operate more intelligently and interact more efficiently with humans.

The increasing demand for more intelligent human-machine interactions also raises higher requirements for processors used in HMI applications. In order to assist engineers with terminal development requirements in selecting the main controller, in this article, the author will provide a detailed explanation of the three key elements that will influence the next generation of HMI.

Smarter Interaction

AI support will help the new generation of HMI achieve more powerful functions. For example, AI face recognition can be used to realize human access to devices, and AI gesture recognition can also be used to realize contactless control between people and devices. At the same time, it also allows the equipment to monitor and analyze the current system status more accurately. For example, in the medical field, intelligent HMI systems can allow doctors to interact with medical devices through gestures.

Balance of Power Consumption And Performance

AI function support puts forward higher requirements for the performance of processors, and the high integration and performance improvement of chips will inevitably increase power consumption and generate more heat. In order for devices with limited size to be able to adapt to a more diverse and complex environment, it is very important to have multiple power consumption mode options - the freedom to choose between high power consumption, low power consumption, and ultra-low power consumption modes. This not only allows performance to be properly optimized, but also helps to better control costs, achieving a balance between power consumption and performance.

Enhanced Communication Capabilities

The increase in real-time industrial communication protocols has also brought new challenges to the new generation of HMI applications. For example, the HMI applied in the smart factory not only needs to carry the task of exchanging information between people and equipment, but also needs to complete the function of communicating with other machines and equipments, which means that the HMI needs to have a stronger connection and control function.

FET6254-C SoM launched by Forlinx Embedded not only meets the traditional HMI's human-computer interaction needs but also can realize the three key elements mentioned above, empowering the new generation of HMI.

FET6254-C System on module is built on the TI Sitara™ AM6254 industrial-grade processor, featuring a quad-core Arm Cortex-A53 architecture with a maximum frequency of up to 1.4GHz. It enables edge AI capabilities, making the HMI smarter and more intelligent. During the development process, rigorous environmental temperature testing, pressure testing, and long-term stability testing were conducted to ensure that it can operate stably in harsh environments.

Not only the performance is guaranteed, but also the power consumption can be very low. Through a simplified power architecture design, the AM62x processor exhibits extremely low power consumption performance, with power as low as 5mW in deep sleep mode. With a core voltage of 0.75V, the operating power can be kept below 1.5W, greatly reducing system power consumption.

AM62x processor, as the next-generation MPU product in the TI Sitara™ product line, offers richer resource interfaces compared to the previous generation classic processor, the AM335x. It includes features such as 2 x Gigabit Ethernet with TSN support, 3 x CAN-FD, 9 x UART, 2 x USB 2.0 , 2 x LVDS interfaces, RGB, camera, audio, and more. This enhances the product's scalability and flexibility for various applications.

In addition to the advantages mentioned above, Forlinx Embedded has also ported a Chinese input method for the Linux system on the FET6254-C SoM. This makes it more convenient to invoke applications and helps users simplify their development workload. Moreover, the FET6254-C embedded board supports system burning via USB flash drive or TF card and can replace Uboot, Kernel, and device tree in the operating system, making it easy to achieve remote updates for products and helping users save on-site maintenance costs.

The combination of stable quality and rich functionality allows Forlinx Embedded's FET6254-C core board to demonstrate unique advantages in next-generation HMI applications, empowering HMI across industries such as industrial control, power, transportation, and healthcare. This enables machines to operate more intelligently and interact more efficiently with humans. The above is the HMI solution recommendation based on the Forlinx Embedded FET6254-C SoM. We hope it can be helpful for your product design.

Originally published at www.forlinx.net.

Why Indian companies does not manufacture processors

[ad_1]

Publish Date:Wed, 20 Jun 2018 04:00 PM (IST) नई दिल्ली (टेक डेस्क)।स्मार्टफोन खरीदते समय यूजर्स फोन के हर फीचर को बारिकी से परखते हैं। फोन में कैमरा क्वालिटी कैसी है, फोन में बैटरी कैपिसिटी कैसी है ये सब अच्छे से जांचने के बाद ही यूजर फोन लेते हैं। कैमरा और बैटरी के अलावा प्रोसेसर भी फोन का एक अहम कारक होता है। फोन की परफॉर्मेंस इसके प्रोसेसर पर निर्भर करती है। फोन का प्रोसेसर जितना…

Here in this review we have the APU from the Richland family named – AMD A8-6500T. There are two modifications of the APU in this model AMD A8-6500 and AMD A8-6500T; the difference between them is not so much but we have the full details of it in the table below:

Model APU AMD A8-6500 AMD A8-6500T Number of cores / threads 4/4 4/4 Frequency of processor cores (Base / Turbo Core) 3.5 / 4.1 – GHz 2.1 / 3.1 – GHz Graphics core HD 8570D HD 8550D Unified shader 256 256 GPU Frequency 800 – MHz 720 – MHz Memory Cache 4 – MB 4 – MB Maximum Memory Supported – DDR3 1866 – MHz 1600 – MHz TDP 65 – W 45 – W

The main difference in both models is energy-efficient, AMD A8-6500T reduced energy consumption due to lower clock frequencies, as well as a wide range of operating conditions.

Packaging and Cooling System:

Well box of the new model AMD A8-6500T is basically in white color, with the logo of the manufacturer and the stylized image processor with the inscription A Series. On the one side you can see the appearance of the processor, and below it you can see the recommended combinations of discrete graphics cards and APU, capable of operation AMD Dual Graphics. On the other side there is a label that contains the following information:

Model of Processor

Basic and Dynamic Maximum Clock Frequency

A Protective Hologram

Model Serial Number

In the box of the model you will find the APU AMD A8-6500T is quite standard, and:

Cooling System for the Processor

Quick Start Guide Manual

Computer Case Sticker

Well the cooling system of the model is normal, having small fan and aluminum radiator, made by Foxconn PVA070E12L with the diameter of 70 mm, which is the operating voltage of 12 V, and the current – 0.20 A Connection to the motherboard via a 4-pin connector that allows you to control the speed of rotation of the blades PWM method and implement monitoring.

Radiator which will contact with processor covered by thermal paste and has a very simple structure: it is completely made of aluminum and has a copper core. Given the relatively low level TDP of AMD A8-6500T, even such a simple cooling system for it to be enough.

Specifications Details:

Processor model is made in Germany and the final assembly was done in China. Set of contacts on the back of compatible processor sockets Socket FM2 and Socket FM2+.

Model AMD A8-6500T Labeling AD650TYHA44HL CPU Socket Socket FM2 (compatible with FM2+) Base Clock Frequency 2100 – Mhz Maximum Clock Frequency With AMD Turbo Core 3.0 3100 – Mhz Factor 31 Basic System Bus Frequency 100 – Mhz Cache Memory Of The First Level L1 4 x 16 – KB

2 x 64 (memory instructions)

Cache Memory Of The Second Level L2 2 x 2048 – Mhz Microarchitecture Piledriver Core Richland Number Of Cores / Threads 4/4 Maximum Rated Power (TDP) 45 – W Critical Temperature 71.3 – °C Tehprotsess 32- NM Technology Support Dual Graphics

UVD3

Turbo Core 3.0

PowerNow!

Eyefinity

Built-In Memory Controller Types Of Memory DDR3 (up to 1600MHz) Memory Channels 2 Integrated Graphics Core AMD Radeon HD8550D Stream Processors 256 GPU Clock Speed 720 – MHz Support Document DirectX 11

Shader Model 5.0

OpenGL 4.3

DirectCompute 11

OpenCL 1.2

According to the specifications table the main difference is frequency rate which is the minimum (2100 MHz) and highest (3100 MHz). This will allows it to choose optimal mode of operation, depending on the current load, which is beneficial to energy consumption.

If we look at the CPU-Z 1.69 results during testing utility LinX – it’s all good. In the basic mode of operation was included technology AMD Turbo Core 3.0, and clock speed most of the time kept at around 2794 MHz with a multiplier “x28″ and vcore 1.160 V.

During the test of AIDA64 – the CPU clock speed dropped below, up to the level of 2100 MHz, or to maximum of 3100 MHz. All changes to the clock in the process of passing the different stages of tests.

In idle mode the multiplier dropped to “x14″, and the frequency at the moment of reading was at around 1397 MHz. Supply voltage for this dropped to 0.856 V, and the core temperature – 23 °C.

Cache memory in AMD A8-6500T is allocated 16 KB cache per core with 4 channels associativity for caching data and 64KB for each of the two dual-core modules with 2 channels associativity for instructions. There is also a 2 MB L2 cache with 16 channels of associativity for each dual module.

Controller RAM running in dual channel mode and support DDR3 modules with up to 1600MHz.

  Testing Performance:

As for the testing the new model, we have used both motherboard form AMD and Intel of the same brand – ASUS, with the total memory of 8-GB – 2x4GB. Other details of the system is listed below:

AMD ASUS F1A75-V PRO (AMD A75, SFM1, DDR3, ATX) GIGABYTE GA-F2A75-D3H (AMD A75, sFM2, DDR3, ATX) AMD ASUS SABERTOOTH 990FX (AMD 990FX, sAM3 +, DDR3, ATX) Intel ASUS P8Z77-V PRO / THUNDERBOLT (Intel Z77, sLGA1155, DDR3, ATX) ASUS P9X79 PRO (Intel X79, sLGA2011, DDR3, ATX) Intel ASRock Z87M OC Formula (Intel Z87, sLGA1150, DDR3, mATX) Coolers Scythe Mugen 3 and ZALMAN CNPS12X (LGA 2011) Access Memory 2x 4GB DDR3-2400 TwinMOS TwiSTER 9DHCGN4B-HAWP Video Card AMD Radeon HD 7970 3GB GDDR5 Hard Disk Western Digital Caviar Blue WD10EALX, 1TB Power Supply Seasonic X-660, 660W, Active PFC, 80 PLUS Gold

Well, with the AMD Turbo Core 3.0, the performance of the processor is increased up to about 15%, which is good for this model and as well as other AMD APU.

Now if we compare it with the AMD A6-6400K, the performance is increased about 8% which is due to twice as many cores. And as compare to AMD A8-6600K, the performance of the tested model is as low as about 25%, due to higher clock frequencies, but with the higher level TDP (100 W), which is much complicates the assembly on the basis of its low noise computer in a compact package.

As compared to the Intel – the performance difference is about to 33% of the Intel Core i3-3220 and 52% of the Intel Core i3-4130. The main advantage of AMD A8-6500T on the data processors can be considered integrated graphics AMD Radeon HD 8550D, which shows better performance in gaming applications.

Overclocking:

During overclocking we have disabled  AMD Turbo Core 3.0, and the multiplier was put on the maximum mark – x31 at 37 MHz was raised FSB frequency, but the CPU voltage is not changed. It should be noted that the data manipulation did not affect the clock frequency, which during operation was constantly at 2883 MHz. During the stress test program LinX 0.6.4 CPU is working stable, and its core temperature did not exceed 34 °C.

The results that have been achieved as a result of manual overclocking, can be estimated as follows:

Normal Mode Overclocking Performance Futuremark PCMark 7 PCMark Score 2551 2632 3.18% Computation Suite 5687 5832 2.55% Futuremark 3DMark11 Score 5409 5902 9.11% Physics 2735 3326 21.61% Futuremark 3DMark Vantage CPU Score 6901 7751 12.32% SiSoft Sandra 2012 Arithmetic Overall performance GOPS 26.76 31.82 18.91% Dhrystone whole, GYPSUM 32.85 39.73 20.94% Whetstone double floating point Gflops 21.79 25.48 16.93% Multimedia Overall multimedia performance, megapixel / s 70.48 84.69 20.16% Multimedia integers Mpixel / s 84 101 20.24% Multimedia FP32/FP64 float megapixel / s 43.6 52.21 19.75% CINEBENCH R11.5 OpenGL 45.16 49.08 8.68% CPU 2.01 2.46 22.39% CPU (Single Core) 0.79 0.75 -5.06% WinRAR 4.20 3049 3175 4.13% Fritz Chess Benchmark 4.2 4434 5089 14.77% TrueCrypt 7.1a Encryption 73.2 89 21.58% Decryption 77.4 94.3 21.83% Batman Arkham City DirectX 11 (fps) 46.5 49 5.38% Rezident Evil 5 Benchmark DirectX 10 Smoothing x8 (fps) 62 72 16.13% RUSE DirectX 9, fps 17 21.51 26.53% Average: 14.81%

Average performance is increased up to 14.81% which is quite a good result.  And memory overclocking, the frequency is raise from 1600 MHz to 1831 MHz without voltage increase.

IGPU overclock AMD Radeon HD8550D conducted by raising its frequency. Thus value was obtained in 1108 MHz, which is simply an excellent result (an increase of 65% compared with the nominal index).

Final Thoughts:

According to the testing and overclocking results, this new model APU AMD A8-6500T is quite good for creating a home media computer. At comparable cost models Intel Core i3 lower price range it can boast the presence of more efficient graphics core (AMD Radeon HD8550D). TDP of the processor do not exceed 45 watts, and even allows you to install it in a compact housing HTPC.

With the overclocking potential increase the performance of processor cores by 14.81%, while the graphics core – 2.64%. If there is a desire to improve graphics performance without significant cash investments, you can use the proprietary technology, AMD Dual Graphics, which allows you to combine into a single configuration of the built-in graphics adapter and a discrete graphics card.

The next evolutionary step was the design of SoC (System-on-Chip). It involves the integration into the processor chipset, calculating functions CPU executes and coordination, ensuring the correct interaction of many internal interfaces.

First APU with SoC support 28-nm was AMD Temash and AMD Kabini, which replaced the 40-nm AMD Ontario model series and AMD Zacate. They are focused on the use of part of the budget tablets, nettops, notebooks and candy bars. After that AMD create a desktop version of APU: AMD Kabini, placing them in a PGA-shell, which makes it easy to change the processor if necessary. You should also add that AMD decided to use the name for the new APU familiar brands – AMD Athlon and AMD Sempron.

Now this time AMD new platform AM1 was introduced, according to IDC in fourth quarter of 2013, most of the market of desktop systems (38%) occupy exactly entry-level solutions. PC mainstream class is 30%, and productive desktops – 32%.

A number of supports controllers that traditional systems are part of the chipset:

SD memory cards up to 2 TB

2x USB 3.0

8x USB 2.0

PS / 2 interface and a variety of internal sensors (temperature, fan speed, etc.)

video ports eDP, DisplayPort / HDMI and VGA

4x PCI Express x16 interface for connecting discrete graphics card

2x SATA 6 Gb / s

4x PCI Express x1, one of which is used to connect Gigabit Ethernet Controller

APU Platform AMD AM1 has very important advantage of graphics performance, maybe that’s enough for fast and high-quality processing of the operating system interface, high-definition video (4K Ultra HD), wireless video transmission (Miracast), run undemanding games, quick photo editing and other similar tasks.

Summary table of the technical specifications of the first APU Platform AMD AM1 is as follows:

Model

AMD Sempron 2650

AMD Sempron 3850

AMD Athlon 5150

AMD Athlon 5350

Market Segment

Desktop system

CPU Socket

Socket AM1

Processor Core

AMD Kabini

Microarchitecture

AMD Jaguar

Tehprotsess

28-NM

Design

SoC

Number Of Cores

2

4

4

4

Clock Frequency, Ghz

1.45

1.3

1.6

2.05

Cache Level L1, KB

Instructions

2 x 32

4 x 32

4 x 32

4 x 32

Data

2 x 32

4 x 32

4 x 32

4 x 32

Cache Level L2

1 – MB

2 – MB

2 – MB

2 – MB

Graphics Core

Type

AMD Radeon R3

Number Of Stream Processors

128

Clock

400 – MHz

450 – MHz

600 – MHz

600 – MHz

Controller RAM

Number Of Channels Supported

1

Maximum Number Of Modules

2

Type

DDR3-1333

DDR3-1600 / DDR3L-1600

DDR3-1600 / DDR3L-1600

DDR3-1600 / DDR3L-1600

The Maximum Volume GB

16

Additional Controllers

PCI Express 2.0, HD Audio, SD, USB 3.0, SATA 6 Gb / s, LPC and other

Supported Ports

2 x USB 3.0 8 x USB 2.0 2 x SATA 6Gb / s HDMI DisplayPort PS / 2

Indicator TDP

25 – W

For the Review and testing performance of the new APU AMD AM1 Platform - AMD Athlon 5350, used the full system (CPU + Motherboard + RAM). This will give us the opportunity to fully appreciate the possibilities of the whole platform AMD AM1, and also allow to understand for what tasks it is best suited.

Motherboard: AMD AM1 Platform

As for the motherboard we have used the model from ASRock – ASRock AM1B, made ​​in the format Mini-ITX. There are also some other model available for AMD AM1. Motherboard has quite standard layout for these compact solutions: CPU socket is in the middle; the left edge of the PCB interfaces are located, and from the opposite – slots for RAM; the lower part is reserved for the connector PCI Express x16. Recall that it uses only 4 lines of standard PCIe 2.0.

AMD AM1 platform primarily positioned as a base for office PCs, nettop or HTPC, not gaming configurations. Therefore, most likely, a PCI Express slot is occupied by some board to enhance the multimedia capabilities of the system, such as an external audio card or TV tuner. Memory supported is up to a 16 GB – 1600 MHz. There is also no support for dual-channel mode.

ASRock may not use additional controller ASMedia ASM1061, which implements support for two more SATA 6 Gb/s. Exactly the same pattern is observed with connectors USB 3.0: 2 x USB 3.0, located on the interface board, operate under control of the processor, and the work is 2, which may be connected to the terminal on the motherboard controller provided ASMedia ASM1042A.

For the display ASRock AM1B, have a VGA, DVI and HDMI port on the panel, support for resolution 4096 x 2160 at 24 Hz frequency sweep. Also support LAN connector, LPT port, 3x audio jacks, a pair of USB 2.0 and PS/2. Audio path is based on a chip Realtek ALC662, and network interface operates under gigabit chip Realtek RTL 8111GR.

TDP of AMD Kabini processors is declared at 25 watts, so its power subsystem are presented fairly low expectations. Is quite enough for 2-phase module VRM, which motehrboard ASRock AM1B-ITX had. The system is powered via the 24-pin ATX. Although, given the low power consumption of processors AMD Kabini, it is possible that we will see models of motherboards and powered from an external adapter (DC 19V)

Memory:

Memory subsystem obtained in testing configuration consists of one module AMD AE34G1609U1S, which belongs to the original AMD Radeon Memory series. According to the marking and the inscriptions on the label, it has a capacity of 4 GB and can operate at a nominal frequency of 1600 MHz with 9-9-9-28 latencies and voltage of 1.5 V.

Since the nettop and HTPC gather in compact cases where, as a rule, difficult to organize a good cooling, the presence of additional heatsinks on the memory chips definitely will not be superfluous.

AMD Athlon 5350:

AMD Kabini, which is presented as AMD Athlon 5350 is the system, the box of AMD Athlon 5350 is in white color with the logo of AMD Athlon on the front and some features listed.

Cooling system for the AMD Athlon 5350 is little bit different other coolers, which are equipped with processor family AMD Trinity / Richland / Kaveri / Zambezi / Vishera, primarily for its compact dimensions. The length and width of the cooling system is 55 mm (excluding fasteners), and the height – 40 mm only. And it shows the size of already established with a fan.

AMD has changed the mounting system: instead of the usual latches cooler is fastened to the board by means of two spring-loaded plastic clips latches. As a result, the cooling system with mounts for boards equipped with processor sockets Socket AM3 / AM3 + / FM2 / FM2 +, here is no longer fit.

The radiator is made with aluminum core, which runs four sections of thin aluminum fins. For their low profile blower fan used FOXCONN PVA050E12L 50 mm and a capacity of 1.92 watts. Power is supplied via a 3-pin connector with support for monitoring the rotational speed of blades.

In idle mode the CPU temperature was 36 °C and at maximum load CPU temperature was – 43 °C. at maximum fan speed during the experiment reached 2950 rev / min. All measurements were carried out on the open bench.

Specifications Details:

AMD Athlon 5350 model made in micro-PGA package and looks similar to other processors sold under the brand AMD. On the heat-spreading lid is labeling and country of production (in this case Taiwan).

Model

AMD Athlon 5350

Labeling

AD5350JAH44HM

CPU Socket

Socket AM1

Clock Frequency

2050 – MHz

Factor

20.5

Base Frequency

100 – MHz

Cache Memory Of The First Level L1

4 x 32 – KB

Cache Memory Of Second Level L2

2048 – KB

Microarchitecture

AMD Jaguar + AMD GCN

Codename

AMD Kabini

Number Of Cores / Threads

4/4

Support Document

MMX (+), SSE, SSE2, SSE3, SSSE3, SSE4A, SSE4.1, SSE4.2, x86-64, AMD-V, AES, AVX

Maximum Rated Power (TDP)

25 – W

Tehprotsess

28 – NM

Technology Support

AMD 64-bit

AMD Virtualization

AMD UVD (Universal Video Decoder)

AMD VCE (Video Codec Engine)

Built-In Memory Controller

Max Memory Gb

16

Types Of Memory

DDR3 / DDR3L

Maximum Frequency

1600 – MHz

Number Of Memory Channels

1

Integrated Graphics Core AMD Radeon R3 Graphics (AMD Radeon HD 8400)

Stream Processors

128

Texture Units

8

Clock Speed GPU

600 – MHz

Support Tech

DirectX 11.2

OpenGL 4.3

OpenCL 1.2

Shader Model 5.0

During normal operation, the speed of AMD Athlon 5350 was 2055 – MHz and multiplier “x20, 5″, while the bus speed was almost  100 – MHz. And was voltage was 1.288 V. In idle factor decreases to “x8″, thus the frequency drops to almost 800 MHz. The measurement voltage is 1.024 V.

Cache of AMD Athlon 5350 is distributed as follows:

First level L1 – for each of the 4 cores allocated to 32 KB data with 8 channels of associativity and 32 KB for instructions with 2 channels associativity;

Second level L2 – 2 MB for all nuclei with 16 channels associativity;

DDR3 memory controller works in single channel mode and guaranteed support modules with frequency up to 1600 MHz.

The model uses AMD Athlon 5350 series graphics core AMD Radeon R3 Graphics, based on advanced microarchitecture AMD GCN. Also GPU-Z utility was able to identify and codenamed GPU - AMD Radeon HD 8400.

Whose performance graphics core AMD Radeon R3 resembles a graphics card AMD Radeon R5 230 1GB GDDR3, with the only difference being that the latter has 32 ​​stream processors more.

Testing Performance:

AMD

ASUS F1A75-V PRO (AMD A75, SFM1, DDR3, ATX) GIGABYTE GA-F2A75-D3H (AMD A75, sFM2, DDR3, ATX)

AMD

ASUS SABERTOOTH 990FX (AMD 990FX, sAM3 +, DDR3, ATX)

Intel

ASUS P8Z77-V PRO / THUNDERBOLT (Intel Z77, sLGA1155, DDR3, ATX) ASUS P9X79 PRO (Intel X79, sLGA2011, DDR3, ATX)

Intel

ASRock Z87M OC Formula (Intel Z87, sLGA1150, DDR3, mATX)

Coolers

Scythe Mugen 3 (LGA 1155/1366, AMD Socket AM3 + / FM1 / FM2) ZALMAN CNPS12X (LGA 2011)

Memory

2x 4GB DDR3-2400 TwinMOS TwiSTER 9DHCGN4B-HAWP

Video Card

AMD Radeon HD 7970 3GB GDDR5

Hard Disk

Western Digital Caviar Blue WD10EALX, 1TB, SATA 6 Gb / s, NCQ

Power Supply

Seasonic X-660, 660W

To compare the performance of the tested model, we have some picked cheap solutions for desktop PCs, because the real competitor of the tested model, AMD Athlon 5350, are Intel Pentium J2900 and Intel Celeron G1610, which are not yet released.

Tthe average performance of AMD Athlon 5350 is almost equal to the AMD A4-6300, that is not only comparable performance indicators, but also the recommended prices. But because of its low power consumption, in this pair certainly looks better AMD Athlon 5350. As compare to AMD FX-4130 and Intel Pentium G3220, latter have a distinct advantage in the form 72% and 54% respectively. But to be honest, we thought that the gap will be more. From this we can conclude that, in practice, AMD Athlon 5350 is not so bad and it can cope with everyday tasks.

Performance of the GPU AMD Radeon R3, is 9% better than the to the GPU AMD Radeon HD 7480D used in AMD A4-4000, and about the same ahead of Intel HD Graphics (Intel Pentium G3220).

Final Thoughts about AMD AM1:

AMD managed to create an even more attractive in terms of cost and energy efficiency with APU AMD Kabini. Its positive side is the possibility of buying a CPU and motherboard separately, whereas before this bunch actually represent a whole, and as a result the user can get a platform, not quite meet his needs.

4-core model AMD Athlon 5350 performance is comparable to APU AMD A4-6300. That is, its capacity will be enough for everyday tasks such as surfing the Internet, word processing, watching movies, etc.

It has low power consumption (TDP of 25 W), AMD Athlon 5350 is perfect as a base for HTPC. In addition, users have a real opportunity to create a completely silent system for little money, because its power adapter and conventional enough, but for CPU cooling – only one large radiator.

We liked and opportunities iGPU. It can even run some modern games at low graphics settings or undemanding arcade projects.

In this article we are going to review the model A10-7700K, also testing its performance and overclocking potential of this new model. But first start with the packaging and specifications.

Packaging and Specifications:

AMD A10-7700K comes in a small box. Packaging is in black color, on the box there is logo (Black Edition), it’s processor with an unlocked multiplier. Also the frequency of the processor 3.8 GHz, cache L2 (4 MB), the number of cores (10: 6 4 CPU + GPU), as well as the presence of the standard cooler supplied.

On the sides briefly describes the capabilities of technology and AMD hUMA microarchitecture AMD GCN.

In the box you will find the following:

User Guide

Processor Cooler

APU AMD Logo sticker

AMD A10-7700K has no different to the earlier APU, Richland / Trinity. But the extra two pins will not let you install it onto the CPU socket Socket FM2, therefore, that why you need to buy a Socket FM2+ platform.

The processor is made in Malaysia and crystal grown in Germany, as reported by the inscription «Diffused in Germany».

Cooling system has the familiar design and consists of low aluminum radiator, which has been blown, low-profile 70-mm fan. At first glance, such a design may not seem very effective, especially if we consider the TDP of AMD A10-7700K, reaching 95 watts. In idle mode the processor temperature was 32 °C, at a 50% load – 47° C at a 100% load – 60° C. Maximum fan speed during the experiment reached 3500 rev / min.

Specifications Details:

Model

AMD A10-7700K

CPU Socket

Socket FM2 +

Clock Frequency

3400 – MHz

Maximum Clock Frequency With Turbo Core 3.0

3800 – MHz

Base Frequency

100 – MHz

Cache Memory Of The First Level L1, KB

2 x 96 (memory instructions)

4 x 16 (data memory)

Cache Memory Of The Second Level L2, KB

2 x 2048 – KB

Microarchitecture

AMD Steamroller + AMD GCN

Codename

AMD Kaveri

Number Of Cores (CPU + GPU)

10 (4 + 6)

Support Document

MMX (+), SSE, SSE2, SSE3, SSSE3, SSE4A, SSE4.1, SSE4.2, x86-64, AMD-V, AES, AVX, XOP, FMA3, FMA4

Maximum Rated Power (TDP)

95 -W

Critical Temperature,

72.4 – °C

Tehprotsess

28 – NM

Technology Support

AMD Dual Graphics

AMD TrueAudio

AMD Mantle

AMD Turbo Core 3.0

AMD PowerNow!

AMD Eyefinity

AMD Fluid Motion Video

Built-In Memory Controller

Types Of Memory

DDR3

The Maximum Frequency

2133 – MHz

Number Of Memory Channels

2

Integrated Graphics Core AMD Radeon R7 Graphics

Stream Processors

384

Texture Units

24

Modules Rasterization

8

Clock Speed GPU

720 – MHz

Support Document

DirectX 11.2

OpenGL 4.3

DirectCompute 11

OpenCL 1.2

Shader Model 5.0

Without AMD Turbo Core 3.0 enable, AMD A10-7700K speed is equal to 3400 MHz with a reference frequency of 100 MHz and multiplier X34. At the time of reading of the core voltage was 1.284 V. If you engage the dynamic increase in the frequency or just the automatic overclocking using proprietary Turbo Core 3.0, the factor is increased by four points to a value of X38. CPU speed increased to 3800 MHz, and the voltage – up to 1,425 V.

Structure of the cache of AMD A10-7700K is slight changed. Now for the instructions allocated 96 KB instead of 64 KB for each 2-core module with 3 instead of 2 channels. Data is still assigned to 16 KB per processor core with 4 channels of associativity. Cache second level L2: 2 MB per 2-core processor module with 16 channels of associativity. Cache in the third level L3 is absent. DDR3 memory controller run in dual channel mode and supported frequency is up to 2133 MHz.

For graphic, AMD A10-7700K used AMD Radeon R7 Graphics series. The exact name of GPU-Z utility failed to determine, it is not surprising, since even the company itself is not voiced marking AMD integrated APU graphics processors. Graphical part of AMD A10-7700K 6 cores include GPU, has 384 shader processors and 24 texture units. Operating frequency is 720 MHz.

Testing Performance:

Motherboard (AMD)

ASUS F1A75-V PRO (AMD A75, SFM1, DDR3, ATX) GIGABYTE GA-F2A75-D3H (AMD A75, sFM2, DDR3, ATX)

Motherboard (AMD)

ASUS SABERTOOTH 990FX (AMD 990FX, sAM3 +, DDR3, ATX)

Motherboard (Intel)

ASUS P8Z77-V PRO / THUNDERBOLT (Intel Z77, sLGA1155, DDR3, ATX) ASUS P9X79 PRO (Intel X79, sLGA2011, DDR3, ATX)

Motherboard (Intel)

ASRock Z87M OC Formula (Intel Z87, sLGA1150, DDR3, mATX)

Coolers

Scythe Mugen 3 (LGA 1155/1366, AMD Socket AM3 + / FM1 / FM2) ZALMAN CNPS12X (LGA 2011)

Memory

2x 4GB DDR3-2400 TwinMOS TwiSTER 9DHCGN4B-HAWP

Video Card

AMD Radeon HD 7970 3GB GDDR5

Hard Disk

Western Digital Caviar Blue WD10EALX, 1TB, SATA 6 Gb / s, NCQ

Power Supply

Seasonic X-660, 660W, Active PFC, 80 PLUS Gold

Using Turbo Core in AMD A10-7700K gives boost up to 5.5%. Such a result is not that significant, but for maximum performance, this feature may be useful in most cases it does not make sense to turn off.

As compared to previous AMD A10-6800K the results was surprising in most tests APU AMD Richland passed the new generations with an average of 3.5%. it is because of the higher clock speeds AMD A10-6800K (4.1 / 4.4 GHz vs. 3.4 / 3.8 GHz).

This is manifested essence microarchitecture enhancements AMD Steamroller: it allows to achieve almost similar results at much lower clock speeds (700 – 600 MHz). But in terms of price/performance model AMD A10-6800K is 15 – 20 dollars cheaper than the AMD A10-7700K.

AMD A10-7700K performance comparable with Intel Core i5-4670K, Intel Core i5-4670K is much faster than AMD A10-7700K. Its advantage is the average was 76%, and in some cases it even more. Therefore, the comparison of any APU and 4-core CPU from Intel cannot be considered.

AMD A10-7700K still somehow cannot compete with the 2-core Intel Core i3-4130, but on the side of the latter would be an advantage in the form of 18%. This proves once again that even the embodiment of revolutionary technology does not guarantee instant performance boost, and still requires optimization and the corresponding code.

We are dealing with a hybrid processor, the half of the crystal takes graphic part. From the perspective of the integrated video core of Intel’s solutions from AMD A10-7700K is simply no competition. For comparison, the graphics core AMD Radeon HD 8670D (APU AMD A10-6800K) also includes a 384 shader processors and runs at a higher frequency (844 MHz), but inferior performance AMD Radeon R7 Graphics on average by 15%. As they say – progress in the face.

Overclocking:

Uplift factor to the value of x44, increased the frequency of new items from 3400 MHz to 4400 MHz, with the base rate has been fixed at around 100 MHz, and had to raise the voltage to 1.400 V. In this mode, the processor correctly passed the stress test program LinX 0.6.4. During the experiment, the maximum temperature was recorded 66 °C.

As a result, growth rate reached 29.4%, which can be considered an excellent result. System performance is affected as follows:

In nominal mode

In the dispersal

Increase productivity

Futuremark PCMark 7

PCMark Score

3022

3239

7.18%

Computation Suite

6836

7480

9.42%

Futuremark 3DMark11

Score

6476

6830

5.47%

Physics

8058

8859

9.94%

Futuremark 3DMark Vantage

CPU Score

10935

12554

14.81%

SiSoft Sandra 2012

Arithmetic

Overall performance GOPS

50

60.28

20.56%

Dhrystone whole, GYPSUM

76.1

91.4

20.11%

Whetstone double floating point Gflops

32.91

39.76

20.81%

Multimedia

Overall multimedia performance, megapixel / s

106.11

129.66

22.19%

Multimedia integers Mpixel / s

124

145.46

17.31%

Multimedia FP32/FP64 float megapixel / s

66.7

83.32

24.92%

CINEBENCH R11.5

OpenGL, fps

61.64

70

13.56%

CPU, pts

3.4

4.18

22.94%

CPU (Single Core), pts

0.96

1.14

18.75%

WinRAR 4.20

3148

3440

9.28%

Fritz Chess Benchmark 4.2, knodes / s

7348

8722

18.70%

TrueCrypt 7.1a (Serpent-Twofish-AES, MB / s)

Encryption

139

173

24.46%

Decryption

146

178

21.92%

x264

1 pass, fps

35.38

39.91

12.80%

2 pass, fps

8.2

10.09

23.05%

Batman Arkham City

DirectX 11 (fps)

86

88

2.33%

Rezident Evil 5 Benchmark

DirectX 10 Smoothing x8 (fps)

83.9

100

19.19%

RUSE

DirectX 9, fps

26.7

28.9

8.24%

Average

15.60%

The average overclocking performance was 15.6%. Also checked overclocking capabilities of iGPU AMD Radeon R7. Managed to raise its frequency 720 MHz to 1108 MHz (53.9% increased value), the voltage did not change. It looks pretty good but see the final result in table below:

In nominal mode

In the dispersal

Increase productivity

Futuremark 3DMark11

Score

2163

2826

30.65%

Graphiks score

1968

2704

37.40%

Physics score

4297

4252

-1.05%

Combined score

2165

2433

12.38%

Futuremark 3DMark Vantage

Score

7547

9214

22.09%

GPU score

6865

8798

28.16%

CPU Score

10744

10739

-0.05%

SiSoft Sandra 2012

Arithmetic

Overall performance GOPS

184.65

284.28

53.96%

Native floating shaders megapixel / s

649.18

999.29

53.93%

Emulated double shaders Mpixel / s

52.52

80.88

54.00%

Cryptographic

Speed ​​cryptography MB / s

5490

8200

49.36%

Speed ​​encryption / decryption

4260

6340

48.83%

Speed ​​hashing MB / s

7000

10620

51.71%

Warhammer 40,000: Dawn of War II – Retribution

DirectX 10, fps

59.8

61.7

3.18%

Rezident Evil 5 Benchmark

DirectX 10, fps

61.3

65.7

7.18%

DiRT: Showdown

DirectX 11, fps

60.4

68.9

14.07%

Avarage

29.11%

Equally impressive and helpful real productivity gains averaged 29.11%. Therefore, optimization of parameters GPU will be very even desirable, especially because it does not need to raise the voltage.

Last Words For the Hybrid Processor:

Let’s say that APU AMD Kaveri is a worthy continuation of the series of hybrid processors. AMD continues to further improve the line APU. Despite the introduction of new microarchitecture AMD Steamroller, problems with poor performance in single-threaded applications have not gone away.

As tests showed, 2-core models of Intel easily pass-away 4-core AMD A10-7700K. As for the graphic part, then there is definitely a noticeable improvement over previous APU generations. Been run on two series of AMD Radeon microarchitecture AMD GCN even with an equal number of shader processors and a lower frequency shows great performance.

AMD A10-7700K takes the advantage over AMD A10-6800K with an average of 15%. Moreover, in practice it turned out that the APU AMD Kaveri have excellent overclocking potential. Moreover, optimization of parameters gives a real performance boost. Especially overclocking graphics core AMD Radeon R7 Graphics after raising the frequency from 720 to 1108 MHz, its performance has increased by almost 30%.

Intel Pentium G3220 Review and Testing Performance

Intel Pentium G3220 primarily made for the office and classroom computers. Intel Pentium G3220 has sufficient capacity and low power consumption also an integrated graphics core and with affordable cost. (more…)

Intel Core i3-4130 Review and Testing Performance

Intel Core i3-4130 belongs to the fourth generation (Haswell) line Intel Core. With the release of this series, the main focus was shifted to the sphere of energy optimization for the possibility of creating compact desktop or mobile computers productive enough. (more…)