Samsung Exynos 5 Octa with Heterogeneous Multi-Processing and GPU Compute is the hidden gem in the Galaxy Note 3 and GALAXY Note 10.1, 2014 Edition, launched at ‘Samsung UNPACKED 2013 Episode 2’ event

but used for 3G / WiFi versions only … while for LTE versions Qualcomm Snapdragon quad-core SoCs at 2.3 GHz are used (probably the same one used in Xiaomi Mi3).

See also the The new Air Command S Pen User Experience making the Samsung Galaxy Note 3 phablet, and Galaxy Note 10.1, 2014 Edition tablet next-generation devices [‘Experiencing the Cloud’, Sept 12, 2013] post for another aspect of the advancement of the state-of-the-art, as well as the innovations in the Companion Device Computing as envisaged and implemented by Pranav Mistry and his TTT team from Samsung: the case of Galaxy Gear + Galaxy Note 3 [‘Experiencing the Cloud’, Sept 12, 2013] post.
Related posts:
SoC (System-on-Chip) [core information page on ‘Experiencing the Cloud’, Aug 28, 2013] for understanding the internal structure of a multi-core SoC with Heterogeneous Multi-Processing and GPU Compute as one of the most complex SoCs of 2013
Eight-core MT6592 for superphones and big.LITTLE MT8135 for tablets implemented in 28nm HKMG are coming from MediaTek to further disrupt the operations of Qualcomm and Samsung [‘Experiencing the Cloud’, July 20-29, 2013] for understanding a similar SoC from the competition, as well with “What is new vs. my earlier The state of big.LITTLE processing [‘Experiencing the Cloud’, April 7, 2013] report” section in the end of it
20 years of Samsung “New Management” as manifested by the latest, June 20th GALAXY & ATIV innovations [‘Experiencing the Cloud’, July 2, 2013] for understanding the whole Samsung phenomenon as well as for having a look at the quite similar ‘Samsung PREMIERE 2013’ event

What are the new Samsung Exynos 5 Octa (Exynos 5420) technological benefits over the competition?

Warren East, then CEO, ARM (before his July 2013 retirement) as quoted in my earlier post on Exynos 5 Octa [Exynos 5410], flexible display enhanced with Microsoft vision et al. from Samsung Components: the only valid future selling at CES 2013 [‘Experiencing the Cloud’, Jan 10, 2013]:

It is providing roughly twice the performance of today’s leading edge smartphones at half the power consumption when running common workloads.

Taehoon Kim, VP of marketing, System LSI Business, Device Solutions Division, Samsung Electronics in:

High processing performance based on multi-cores has become an essential factor for advanced mobile devices, and what users expect more is how long and seamlessly they can enjoy smart mobile computing experiences.

In an era when smartphones and tablets are evolving into the user’s primary compute device, Samsung’s Exynos 5 Octa [Exynos 5410], as the industry-first big.LITTLE enabled application processor, will drive innovation to bring outstanding user experience by handling diverse mobile workloads while also being optimized for power consumption

Simon Segars, then president, ARM (CEO from July 2013) in the same ARM press release:

big.LITTLE processor technology builds on ARM low-power leadership and sets a new standard for high performance and energy-efficient processing. By reducing processor energy consumption by up to 70 percent on common workloads, big.LITTLE technology enables users to do more with their smartphones for longer. As smartphones and tablets continue to evolve into users’ primary compute device, our partners are increasingly looking to ARM for innovations to help them deliver performance as well as the always-on, always-connected service their customers expect.

Noel Hurley, vice president, Strategy and Marketing, Processor Division, ARM in Samsung Primes [the new] Exynos 5 Octa [Exynos 5420] for ARM big.LITTLE Technology with Heterogeneous Multi-Processing Capability [press release, Sept 10, 2013]

We welcome Samsung’s continued commitment to deploying the leading-edge technology on their latest chips featuring the ARM Cortex™-A series of processors, ARM Mali™ GPUs and ARM Artisan™ physical IP.

Jem Davies, VP of Technology for ARM’s Media Processor Division (see in the details section later on):

Samsung is setting the way in terms of trend-setting devices, the new form factors like the phablets and the tablets that they’ve been producing. The Mali partners here want to see Mali being used in these really trend-setting devices, the things that are approaching new markets, and knowing that they can buy with confidence that there’s a whole variety of market segments now being addressed by our partners.

ARM Holdings work: big.LITTLE Optimization case #1
When the ARM Cortex-A7 cores are enough for an application: ARM® big.LITTLE™ Processing with Angry Birds game [ARMflix YouTube channel, Sept 11, 2013]

This demonstration shows how energy-efficient ARM® big.LITTLE™ multiprocessing (MP) is when running the Angry Birds application on Android with a big.LITTLE MP Kernel on ARM partner silicon This platform is built on the ARM Cortex™-A15 Quad 1.8GHz and Cortex™-A7 Quad 1.3GHz and ARM Mali™-T628 MP6. For more information please visit

ARM Holdings work: big.LITTLE Optimization case #2
When both types of the cores are needed for an application, depending on the situation: ARM® big.LITTLE™ Processing with QuickOffice [ARMflix YouTube channel, Sept 11, 2013]

This demonstration shows a real application, Quickoffice, running on ARM Lead Partner silicon. The ARM® big.LITTLE™ multiprocessing (MP) scheduler automatically uses the resources of all 8 cores to deliver the best performance. This is big.LITTLE processing in action — where we see the right core for the thread’s performance demands being used. This platform is built on the ARM Cortex™-A15 Quad 1.8GHz and Cortex™-A7 Quad 1.3GHz and ARM Mali™-T628 MP6. For more information please visit

Samsung Exynos Evolved messages:

  • The new Exynos 5 Octa’s Mobile Image Compression lowers total system power used for photo transfer
  • The ARM Mali™-T628 GPU in the new Exynos 5 Octa (5420) enhances graphics performance
  • The new Exynos 5 Octa (5420) has twice the 3D graphics processing capabilities of its predecessor
  • GPU Compute (GPGPU) in the new Exynos 5 Octa accelerates intensive operations, traditionally processed by the CPU
  • Exynos 5420 runs on 1.8GHz Cortex™-A15 & 1.3GHz A7 cores in big.LITTLE tech for 20% improved CPU processing over Exynos 5410
  • The Exynos 5 Octa (5420) has a memory bandwidth of 14.9 GB/sec for extremely fast data processing
  • The Exynos 5 Octa series with ARM big.LITTLE™ tech now supports Heterogeneous Multi-Processing!
  • Support for OpenGL® ES 3.0 & Full-Profile OpenCL 1.1 help the new Exynos 5 Octa (5420) run complex gaming graphics
  • With the industry’s widest memory bandwidth, the new Exynos 5 Octa supports a Full-HD 30fps WiFi display

GPU Compute Offload Balances Performance, Power, and Cost [ARMflix YouTube channel, July 15, 2013]

GPU Compute Offload Balances Performance, Power, and Cost – Patrick Moorhead, President and Principal Analyst, Moor Insight and Strategy. Whitepaper:

Epic Citadel Benchmark on New Exynos 5 Octa (5420) Reference Platform [SamsungExynos YouTube channel, Aug 27, 2013]

Samsung demoed the new Exynos 5 Octa (5420) processor at the ARM booth during SIGGRAPH 2013, using the Epic Citadel benchmark to show what the new mobile AP is capable of on a reference platform. The benchmark shows significant output in graphics performance, thanks in part to the system’s support for OpenGL ES 2.0. By supporting this API, among others, the new Exynos 5 Octa offers the horsepower needed for multi-layer rendering of high-end, complex gaming scenarios, post-processing and sharing of photos and video, and high-function multi-tasking operations. Find out more about Samsung Exynos processors on our website ( or on Twitter ( To learn more about the new Exynos 5 Octa (5420), visit the official product page: For information about ARM’s Mali-T628 GPU with GPU Compute, visit the ARM® Mali™ website:

Samsung Exynos blog (see in the details section later on):

  • In combination with big.LITTLE architecture, GPU Computing significantly increases power efficiency for noticeably better battery life.
  • By combining GPU Compute technology with ARM® big.LITTLE™ processing architecture, the new Exynos 5 Octa benefits from two layers of energy efficiency.

ARM Holdings work: big.LITTLE Optimization case #3
When in addition to the both types of the cores the GPU Compute is also needed for an for an application: ARM® big.LITTLE™ Processing with ARM® Mali GPUs Demonstrating GPU Compute [ARMflix YouTube channel, Sept 11, 2013]

This demo shows how ARM® big.LITTLE™ multiprocessing (MP) can work alongside GPU Compute on ARM lead partner silicon. This application makes full use of RenderScript on the GPU to deliver the high performance of GPU compute from the ARM Mali™ GPU and the big.LITTLE CPU system to quickly and seamlessly improve the image quality of all the images on the page. This platform is built on the ARM Cortex™-A15 Quad 1.8GHz and Cortex™-A7 Quad 1.3GHz and ARM Mali™-T628 MP6. For more information please visit

The New Samsung GALAXY Note 10.1 Delivers Unparalleled Tablet Viewing, Productivity and Mobility [press release, Sept 4, 2013]

imageSamsung Electronics Co., Ltd., a global leader in digital media and convergence technologies, today unveiled the GALAXY Note 10.1, 2014 Edition, an original approach to balancing productivity, powerful content creation and consumption in one portable tablet device. Equipped with WQXGA Super clear LCD (2560×1600) resolution in a stunning 10-inch display, 1.9 GHz [Samsung Exynos] Octa Core processor (for 3G / WiFi only version) and 3GB RAM, the GALAXY Note 10.1 (2014 Edition) demonstrates Samsung’s innovation leadership by delivering ultimate productivity capabilities while remaining extremely thin and light.
“The new GALAXY Note 10.1 is the most progressive 10-inch tablet, delivering the best viewing and multitasking experiences. It is the most recent demonstration of Samsung Mobile’s focus on constant product innovation to stay aligned with shifting consumer interests,” said JK Shin, CEO and President of IT & Mobile Division at Samsung Electronics. “The GALAXY Note 10.1 (2014 Edition) unites a range of features that will consistently surprise consumers as they realize how much easier and more enjoyable it makes their everyday lives.”
The new GALAXY Note 10.1 expands on the advanced productivity and creativity leadership delivered by the original Samsung GALAXY Note 10.1. In addition to enabling productivity, the device has been redesigned with a sleek, light, slim frame that is both fashionable and portable.
The GALAXY Note 10.1 (2014 Edition) will come in three connectivity options: WiFi Only, WiFi and 3G, WiFi and LTE, available in 16/32/64GB + Micro SD. Comes with two color options, Jet Black and Classic White, the GALAXY Note 10.1 (2014 Edition) will be offered globally and will be available starting from Q3, 2013.
GALAXY Note 10.1 (2014 Edition) will be on display at the Samsung booth # 20 at IFA 2013, September 6 through September 11, 2013. Full details and product images are available at  or
[GALAXY Note 10.1 (2014 Edition): Related articles – Gallery – Videos – Features – Tech Specs]

Samsung UNPACKED 2013 Episode 2 livestream (full length) [SAMSUNGmobile YouTube channel, Sept 7, 2013], the event starts at [8:20]

Live at Samsung UNPACKED 2013 Episode 2 from the Tempodrom in Berlin where we unveil the GALAXY Note 3, GALAXY Gear and GALAXY Note 10.1 (2014 Edition).

[0:19:55]: GALAXY Note 3 and GALAXY Gear will be launched starting from September 25th in more than 140 countries. And all of these products will be promotionally available in October worldwide.

See also: Exclusive: List of countries receiving Galaxy Note III’s Exynos 5420 or Snapdragon 800 variant, SM-N9005 specs confirmation [SamMobile, Aug 15, 2013]

The other product launched at IFA 2013 with 1.9 GHz Samsung Exynos Octa Core processor (for 3G only version): GALAXY Note 3

Introducing Samsung GALAXY Note 3 [SAMSUNGmobile YouTube channel, Sept 4, 2013]
Hardware & Design
The GALAXY Note 3 comes with a wider (5.7″) Full HD Super AMOLED display, yet slimmer (8.3mm) and lighter (168g) hardware design, and is powered by a 2.3 GHz Quad-Core [Qualcomm Snapdragon] processor [for LTE version, and Samsung Exynos 5 Octa for 3G only version], 3GB of RAM, 32/64 GB or user memory, 3,200 mAh battery and runs on Google’s Android 4.3 Jelly Bean OS. Also, equipped with a 13MP rear camera with Smart Stabilization and high CRI LED flash, you can take crisp photos, even in low light and active situations. Plus, it can record and play in full HD (1080p), and record in UHD (*may differ by market).
The GALAXY Note 3 also features a textured back cover and delicate stitching that delivers a premium look and feel.
Everyday Made Easier with S Pen
Air Command, a palette of features and commands which you can activate on any screen simply by hovering and clicking the S Pen button, lets you access useful features for everyday tasks such as Action Memo, Scrapbook, and S Finder.
Smarter Large Screen Experience
With a larger Super AMOLED screen, the GALAXY Note 3 not only provides a stunning and defined viewing experience but also enables multitasking capabilities that allow users to fully utilize the larger screen. The enhanced new Multi Window allows you to easily move content between applications in one quick step with the Drag and Drop mode.


ARM big.LITTLE Hangout with the Experts [ARMflix YouTube channel, Aug 14, 2013]

A technical discussion about ARM big.LITTLE processing technology from some of ARM’s experts.
Robin Randhawa, Principal Software Engineer, ARM
Charles Garcia-Tobin, Advanced Product Design Group, ARM
Brian Jeff, Senior Product Manager, ARM
Ian Smythe, Director of Marketing, ARM (Moderator)
Discussion questions with timecodes:
  1. What is big.LITTLE? 1:25
  2. How can the system switch between cores? 5:51
  3. How does big.LITTLE save energy, compared to running one set of cores? 8:40
  4. Why even use the big.LITTLE configuration? 13:03
  5. How does the Software work? 14:58 (Software models 22:15)
  6. How does this effect applications programmers? 25:46
  7. What are the performance benefits of big.LITTLE? 28:47
  8. Where can silicon partners get code for this? 34:48 (Software links 36:45)
  9. Does this run on Android? 37:25

Samsung Primes Exynos 5 Octa for ARM big.LITTLE Technology with Heterogeneous Multi-Processing Capability [press release, Sept 10, 2013]

Samsung Electronics announced its Heterogeneous Multi-Processing (HMP) solution for the Exynos 5 Octa to fully maximize the benefits of the ARM® big.LITTLE™ technology. The HMP capability gives system-level designers the ability to develop solutions that deliver the right combination of high-performance and low-power to carry out tasks such as 3D gaming, complex augmented reality and advanced web browsing.
Samsung Exynos OCTA-pella: Performance + Efficiency in Perfect Harmony [SamsungExynos YouTube channel, Sept 9, 2013]
Samsung’s Exynos 5 Octa series of mobile processors with ARM© big.LITTLE™ technology now supports heterogeneous multi-processing (HMP)! This advanced technology allows the Exynos 5 Octa processors to provide exceptional performance and increased power efficiency. In the OCTA-pella video, you’ll see how the Exynos 5 Octa uses ARM big.LITTLE processing to balance workloads across CPU cores, using the right core for the right task. Find out what else Exynos has in store on our Facebook page: and on Twitter: If you want to learn more about the Exynos 5 Octa series of processors, visit our website:… Discover the advantages of ARM© big.LITTLE™ processing:
“It’s usually assumed that the big CPU will do all the performance-critical work, however, power-efficient little cores can handle many significant workloads all on their own, so the workload is balanced within the system,” said Taehoon Kim, vice president of System LSI marketing, Samsung Electronics.
“We welcome Samsung’s continued commitment to deploying the leading-edge technology on their latest chips featuring the ARM Cortex™-A series of processors, ARM Mali™ GPUs and ARM Artisan™ physical IP.”  said Noel Hurley, vice president, Strategy and Marketing, Processor Division, ARM.
HMP is the most powerful use model for ARM big.LITTLE technology, as it enables the use of all physical cores at the same time. Software threads with high priority or high computational intensity can be allocated to the ‘big’ Cortex-A15 cores while threads with less priority or are less computationally intensive,  can be performed by the ‘LITTLE’ Cortex-A7 cores, enabling a highly responsive, low-energy system to be built.
Software implementation is essential to maximizing the benefits of big.LITTLE technology. Multi-processing software controls the scheduling of threads of execution to the appropriate core. In earlier versions of the big.LITTLE software, the whole processor context is moved up to the ‘big’ core or down to the ‘LITTLE’ core based on the measured work load. In-depth study and analysis of diverse use case scenarios enable Samsung to achieve efficiency and high-performance, while managing power levels to deliver optimal user environments.    

The HMP solution for Samsung’s Exynos 5 Octa application processors will be available to customers in 4Q of 2013.

Exynos 5 Octa: Heterogeneous Multi-Processing Capability [Samsung Exynos blog, Sept 10, 2013]

If you’ve been paying attention, you know that the Exynos 5 Octa packs a serious punch when it comes to processing power and energy efficiency. Now, the team at Samsung has made the Exynos 5 Octa even better with the introduction of a new Heterogeneous Multi-Processing (HMP) solution.
Overview of big.LITTLE Technology
Before we jump into the benefits of HMP, let’s take a step back and go over the basics ofARM® big.LITTLE technology. In the Exynos 5 Octa, eight CPU cores are responsible for everything from browsing the web to playing your favorite game on your 5 Octa-powered mobile device. Four “big” 1.8GHz ARM®Cortex™-A15 cores handle intensive tasks like graphically rich gaming or HD video playback. Less intensive tasks like e-mail or text functions are tackled by four “LITTLE” 1.3GHz Cortex™-A7 cores. By dividing and conquering tasks and assigning them to the proper CPU cores, big.LITTLE technology maximizes performance while minimizing power loss.
HMP Makes big.LITTLE Technology Even Better
Now this is where HMP comes into play. Like a sports team, big.LITTLE technology relies on a software “coach” to call the plays and assign tasks to each core. In a basic implementation of big.LITTLE technology, this “coach” would alternate between using “big” and “LITTLE” CPU cores based on the computational intensity of any given task, and one core or cluster of cores would remain inactive while its counterpart was engaged.
With HMP, all eight of the CPU cores in the Exynos 5 Octa can be utilized at the same time. This provides users with an unlimited mobile experience in the current mobile environment and also paves the way for more advanced and complex functionality in the future. HMP is extremely versatile. Using a global load balancing scheduler, HMP can assign a single core to handle a task with low computational intensity in order to maximize power efficiency. On the flipside, HMP can also simultaneously utilize each of the eight individual cores in the 5 Octa to run multiple tasks in real time. The global load balancing scheduler pays attention to user workloads and will pull in the necessary available resources for the system to run flexibly and efficiently. By analyzing and assigning tasks,this highly complex software system maximizes efficiency by balancing CPU workload.
The result is the most advanced use of big.LITTLE technology to date and a huge leap forward for multi-processing capability in mobile devices. By allowing for the simultaneous operation of both “big” and “LITTLE” cores in the Exynos 5 Octa, Samsung offers an optimized HMP solution to the balancing act of maximizing mobile device capability while minimizing power loss.
Samsung has always been a leader in big.LITTLE technology, and this new Octa-core HMP solution is an industry first. HMP sets the stage for the future as mobile devices are increasingly called upon to handle complex and graphically rich tasks. Through this innovative solution, the benefits of big.LITTLE technology are maximized to their full potential. Get ready, because the future of mobile processing is evolving, and the Exynos 5 Octa with HMP is leading the way.

Samsung Brings Enhanced Mobile Graphics Performance Capabilities to New Exynos 5 Octa Processor [press release, July 23, 2013]

Samsung Electronics Co., Ltd., a world leader in advanced semiconductor solutions, today introduced the latest addition to the Exynos product family with top level of graphic performance driven by a six-core ARM® Mali™-T628 GPU processor for the first time in the industry. With mobile use case scenarios becoming increasingly complex, Samsung’s newest eight-core ARM Cortex™ application processor gives designers a powerful, energy efficient tool to build multifaceted user interface capabilities directly into the system architecture. Samsung will demonstrate the new Exynos 5 family at SIGGRAPH 2013 in the ARM booth, #357; Exhibit Hall C at the Anaheim Convention Center.
Samsung’s new Exynos 5 Octa (product code: Exynos 5420), based on ARM Mali™-T628 MP6 cores, boosts 3D graphic processing capabilities that are over two times greater than the Exynos 5 Octa predecessor. The newest member of the Exynos family is able to perform General-Purpose computing on Graphics Processing Units (GPGPU) accelerating complex and computationally intensive algorithms or operations, traditionally processed by the CPU. This product also supports OpenGL® ES 3.0 and Full Profile Open CL 1.1, which enables the horsepower needed in multi-layer rendering of high-end, complex gaming scenarios, post-processing and sharing of photos and video, as well as general high-function multi-tasking operations.
“ARM welcomes the latest addition to the successful Exynos Octa 5 series, which uses ARM’s Mali GPU solution to dramatically improve graphics performance,” said Pete Hutton, executive vice president & general manager, Media Processing Division, ARM. “ARM big.LITTLE™ and ARM Artisan® Physical IP technologies continue to be at the heart of the Octa series and now complement the new functionality brought by ARM GPU Compute. This combination enables unprecedented capabilities in areas such as facial detection and gesture control, and brings desktop-quality editing of images and video to mobile devices.”
“Demand for richer graphic experiences is growing rapidly nowadays,” said Taehoon Kim, vice president of System LSI marketing, Samsung Electronics. “In order to meet that demand from both OEMs and end users, we developed this processor which enables superb graphical performance without compromising power consumption.”
The newest Exynos processor is powered by four ARM Cortex®-A15™ processors at 1.8GHz with four additional Cortex-A7™cores at 1.3 GHz in a big.LITTLE processing implementation. This improves the CPU processing capability by 20 percent over the predecessor by optimizing the power-saving design.
In addition, the mobile image compression (MIC) IP block inside this System-on-Chip successfully lowers the total system power when bringing pictures or multimedia from memory to display panel. This feature results in maximizing the usage hours of mobile devices with a high-resolution display such as WQXGA (2500×1600), in particular when browsing the web or doing multimedia application requiring the frequent screen refresh.
The new Exynos 5 Octa processor also features a memory bandwidth of 14.9 gigabytes per second paired with a dual-channel LPDDR3 at 933MHz, enabling an industry-leading fast data processing and support for full HD Wifi display. This new processor also incorporates a variety of full HD 60 frames per second video hardware codec engines for 1080p video recording and playback.
The new family of Exynos 5 Octa is currently sampling to customers and is scheduled for mass-production in August.
For Further information :

Transforming your mobile and TV experience with GPU Compute [Trina Watt on Multimedia blog of ARM, July 22, 2013]

As a mother of three young children I am very aware of their approach to technology, and how unaccepting they are of the established ways of interacting with devices that anyone over the age of 18 is used to. The first time I gave my 5 year old daughter a mouse she looked at me as if I was mad – why use this when you can touch the screen? Gesture, touch and voice are much more natural not only to elementary kids but to us all.
ARM has been working with our partners for several years on how GPUs can improve user experience, initially through bringing improved graphics to a wider range of devices, but more recently, improving interaction with your devices through GPU Compute support in the Mali-T600 series of GPUs. ARM have just created a video that gives more background on the potential use cases for GPU Compute and how we can expect them to change the way we interact with even more of our electronic devices.
ARM brings GPU Compute to mobile devices [ARMflix YouTube channel, July 22, 2013]
GPU Compute is becoming reality. Its advanced computational abilities and energy efficiencies are inspiring innovation in the mobile industry, innovation which will drive novel, exciting and intuitive user experiences for consumers. In this video ARM explains the key features of GPU Compute and sets out a vision of what GPU Compute will enable for consumers. Discover GPU Compute with the ARM® Mali™-T600 series
Gesture control is becoming increasingly popular in premium devices. What GPU Compute does is bring those features to a wider range of devices – you can expect to see it being built into more DTVs, tablets, and smartphones. GPU Compute enables smaller gestures to be supported even in low lighting, so searching for the remote down the back of the sofa will become a thing of the past! I am interested to see how we solve the “who is in control” issue once there is no longer a remote control to fight over…
Facial detection can also benefit from GPU Compute. Facial detection brings features like “smile detection” when taking photos, or having a camera only take the picture when everyone is looking in the same direction and has their eyes open. These types of services can link to and enhance other features, such as your tablet or DTV being aware of who is looking at a screen so it can make sure the content is suitable to the audience, or powering down when no one is watching.
More of our daily lives are being captured through smartphones and tablets. I have numerous nearly unusable videos that have been shot in motion – having video stabilization built into my mobile devices will help me capture more of the dynamic moments in my kids’ early years, and being able to edit those videos on my tablet (to crop out the inevitable shot of my feet before posting on Facebook for the grandparents to see!) is just another one of the benefits to come.
So next time I fall asleep in front of the television watching the latest episode of Game of Thrones, GPU Compute will help me, a tired parent, know how far I got through the last episode – then pick up where I left off! Got to love technology.
imageTrina Watt, Director of Channel Marketing, Media Processing Division, ARM.
I like to think of myself as a “geek in marketers clothing”. Gadgets and technology have been a passion for me as long I can remember – from dismantling my first radio when I was about 8 to now running around regularly with 3 phones, a tablet and laptop to feed my tech thirst. I started in the tech industry nearly 20 years ago in Motorola and I have never ventured far from it. I am currently focused on promoting the visually exciting Mali graphics processors. I get to work with a wide range of partners who are creating the innovative devices of the future. For a geek it doesn’t get much better than that!

Spotlight on the New Exynos 5 Octa [Samsung Exynos blog, Aug 8, 2013]

imageMeet the new and improved Exynos 5 Octa processor, designed to bring enhanced graphics performance and energy efficiency to the next generation of high-end smartphones and tablets. Advanced ARM® GPU Compute technology and a suite of power-saving features allow our latest processor to run complex applications while conserving battery life – because a mobile device isn’t really mobile if you have to worry about charging it all of the time. We’ll fill you in on the most important specs for this new SoC and give you an idea of what to expect from future devices that run on it.
Enhanced Graphics with New ARM® Mali™-T628 GPU
imageThe new Exynos 5 Octa (5420) packs an ARM® Mali™-T628 GPU, which is powerful enough to support 3D graphics processing that’s more than twice as fast as the Exynos 5410, for smoother rendering and diminished lag times. The Mali-T628 is also capable of GPU Compute functionality, meaning it can perform general-purpose computing and intensive tasks to help alleviate the CPU’s workload and speed up processing times. In combination with big.LITTLE architecture, GPU Computing significantly increases power efficiency for noticeably better battery life. OpenGL® ES 3.0 and Full profile OpenCL™ 1.1 support provide the additional power that’s needed to render high-quality, complex gaming scenarios and handle comprehensive photo and video editing.
Making Advanced Applications Possible
GPU Compute technology was once limited to PCs and other desktop devices, but ARM Mali GPUs are bringing this functionality to mobile. There are a lot of advantages associated with GPU Computing that will open up opportunities for innovative applications on mobile devices that use the new Exynos 5 Octa (5420).
  • Video and image stabilization and editing – Get creative! Edit photos right after you take them or apply filters to live video directly on your smartphone.
  • Facial detection – Refined facial recognition allows for enhanced security features and content restrictions for children.
  • “Open eye” detection – You won’t have to retake nearly as many photos if your smartphone’s camera only captures an image when everyone is looking at the camera.
Increased Power Efficiency for Extended Use
imageThe new Exynos 5 Octa (5420) features four 1.8GHz ARM® Cortex™-A15 cores and four 1.3GHz Cortex™-A7 cores in an optimized big.LITTLE™ configuration. This power-saving design provides 20% increased CPU performance over the previous version of the processor, but with 70 percent greater energy efficiency than Cortex-A15 cores alone. The system switches amongst eight cores of Cortex-A15 and Cortex-A7 processors depending on workload, selecting the right core for the right task and saving power in the process.

If you’re an avid mobile gamer who’s into FPS games with complex graphics or fast-paced, high-frame-rate racing games, the “big” Cortex-A15 cores are there to do the heavy lifting. Even if gaming isn’t your thing, you’ll notice faster page loading times and better graphics quality when browsing websites with high-quality media. Big.LITTLE technology is all about balance, so the “LITTLE” Cortex-A7 cores step in to take care of lighter workloads, like pulling up work emails, playing music and texting. The Cortex-A7 series is the most energy-efficient processor that ARM has to offer, helping your device go longer between charges so you can stay connected while you’re busy living life.

All of you Photoshoppers and Instagram addicts will be happy to hear that the Exynos 5 Octa (5420 ) also features Mobile Image Compression (MIC), which lowers the system power required to bring photos from memory to your screen for processing. This means photo-editing doesn’t unduly impact battery life, and you can get the maximum use out of your mobile devices with high-resolution displays.
To learn more about the flexibility that GPU Compute technology brings to the new Exynos 5 Octa, check out this guest blog post by Trina Watt, Director of Channel Marketing for ARM.
More specs and details can also be found on the new Exynos 5 Octa (5420) product page.

New Exynos 5 Octa with GPU Compute Enables Cool New Applications [Samsung Exynos blog, Aug 1, 2013]

Recently Samsung announced their latest addition to the Exynos 5 Octa family. With the ARM® Mali™-T628 as its GPU, this is the first silicon to reach the market which contains a second generation processor from the ARM Mali-T600 series. Improved GPU support is the key differentiating focus of this version of the Exynos Octa. In the past GPU support was solely about graphics performance; however, with the inclusion of the ARM Mali-T628 in Exynos 5 Octa, Samsung have brought market-leading GPU Compute support to mobile devices.
The ARM Mali-T628 comes with support for a wide range of APIs including OpenGL® ES 3.0 and 2.0, OpenCL™ andRenderscript™. This breadth of API support opens up a wider range of use cases for consumers. ARM has been spearheading activities for OpenGL ES 3.0 since its launch and this momentum continued with the ARM Mali-T600 series becoming one of the first to gain conformance. ARM’s continuous market driving is now expanding to the GPU Compute area. Previously, GPU Compute has only been in PC- or desktop-style devices. Now, ARM Mali GPUs are enabling GPU Compute within the mobile power boundary. Samsung started shipping the ARM Mali-T604 GPU back in back in October 2012 and since that point the ecosystem around GPU Compute has been growing with more and more partners seeing the advantages that GPU Compute can bring.
These advantages include video and image stabilization and editing (meaning users don’t have to wait to get home until they edit and upload), facial detection (enabling enhanced security, access to suitable content and smile detection) and “eyes open” detection (so that the photo is taken only when everyone is looking at the camera or smiling.)
GPU Compute also enables the application of filters to images and live video, opening up new forms of creativity. To see more of what GPU Compute can provide across smartphones, tablets and DTVs, watch this video.
ARM brings GPU Compute to mobile devices [ARMflix YouTube channel, July 22, 2013]
Since this is the second generation of the ARM Mali-T600 series, additional architectural refinements have been made which mean that the ARM Mali-T628, when at the same performance point as an ARM Mali-T604, provides a 50% energy-efficiency improvement, alternatively, when consuming the same level of power, offers substantial increased performance. GPU Compute is about making current use cases more efficient. The combination of GPU Compute and ARM’s big.LITTLE™ technology in one SoC opens up new opportunities for task management. Certain tasks can be handled more quickly and using less energy on a GPU then on a CPU – math-intensive activities in particular often run better on the parallelized GPU architecture. This means you get twice the energy efficiency benefit when you combine GPU Compute and big.LITTLE, as Samsung have done in the Exynos 5 Octa. The GPU takes suitable tasks off the CPU, allowing the CPU to work more often in LITTLE mode and ultimately increases energy efficiency on the GPU and energy savings on the CPU. It also frees up the CPU to run other, more latency-sensitive tasks.
At the end of the day, GPU Compute provides more flexibility in what consumers are able to do with their devices – meaning more end devices will be available which are both energy efficient and enablers of cool new applications, so consumers will no longer have to sacrifice one feature for the other.
imageTrina Watt, Director of Channel Marketing, Media Processing Division, ARM.
I like to think of myself as a “geek in marketers clothing”. Gadgets and technology have been a passion for me as long I can remember – from dismantling my first radio when I was about 8 to now running around regularly with 3 phones, a tablet and laptop to feed my tech thirst. I started in the tech industry nearly 20 years ago in Motorola and I have never ventured far from it. I am currently focused on promoting the visually exciting Mali graphics processors. I get to work with a wide range of partners who are creating the innovative devices of the future. For a geek it doesn’t get much better than that!

GPU Compute Technology: Benefits in the New Exynos 5 Octa [Samsung Exynos blog, Aug 19, 2013]

Mobile processors have advanced rapidly over the past several years due to the development of new technologies and manufacturing processes that allow for significantly increased performance and power efficiency. A more recent development from ARM®, GPU Compute technology harnesses the power of a component that was traditionally reserved for graphical processing and uses it to improve system-level performance and workload distribution. We’ll review the advantages and applications of this technology in the new Exynos 5 Octa (5420) to give you a better understanding of its importance.
What is GPU Compute Technology?
Over time, the market has come to expect higher and higher resolution displays, which has been a key driver in GPU advancements. Both resolution and power increasingly dominate processing requirements for the next generation of smartphones and tablets. As we demand more and more of our mobile devices, we have to find ways to meet these evolving needs without sacrificing power efficiency. Enter ARM’s general-purpose computing on graphics processing units (GPGPU), also known as GPU Compute technology.GPU computing is best defined by the ARM experts who developed it:
“…The computational performance of the GPU, historically used for graphics, is harnessed to augment the main processor (CPU) for certain applications where the GPU architecture will be more effective. The result is improved performance and energy-efficiency and a more efficient use of the system as a whole, making computational photography, computer vision, advanced imaging, point-of-interest extraction and augmented reality possible because of the extended processing capacity.”
Long story short, GPU Compute allows a system to match workloads to specific compute devices for optimal performance and power savings. ARM GPUs with GPU Compute are more efficient and better suited than the CPU when it comes to executing certain tasks such as math-intensive operations. By taking on these tasks, the GPU alleviates some of the CPU’s workload and lowers overall power consumption.
Advantages of GPU Compute
GPU Compute technology comes with a host of benefits, from increased performance and energy savings to ramped-up application support. Our most recent processor, the new Exynos 5 Octa (5420), packs an ARM® Mali™-T628 with GPU Compute that significantly enhances graphics performance. Compared to the Mali-T604 GPU in the Exynos 5 Dual, the Mali-T628 can reach the same performance levels while providing a 50% improvement in energy efficiency. The newer generation of the Mali-T600 series also shows increased performance levels when consuming the same amount of power. By combining GPU Compute technology with ARM® big.LITTLE™ processing architecture, the new Exynos 5 Octa benefits from two layers of energy efficiency.
Trina Watt, Director of Channel Marketing at ARM, recently wrote a guest blog post on the advantages associated with GPU Compute functionality. Depending on the mobile device, GPU Compute enables video and image stabilization and editing, advanced facial detection, “eyes open” detection for photo-taking and filter application to images and live video. All of these capabilities open up a wide range of possibilities for advanced apps and features that aren’t currently supported on many phones and tablets. Better still, overall optimization of system workload means you can run those cutting-edge apps without unduly draining battery power – a win-win outcome for any mobile system.
GPU Compute in Action
imageGPU Compute has already made its way into mobile devices through Exynos 5 Dual -equipped products like the Google Nexus 10. You’ll see the benefits of this technology in action once you experience this tablet’s super high-resolution display. With a 2560×1600 (WQXGA) screen resolution, text is sharper and colors are more vivid than anything you’ve experienced on mobile.
You’ll also find this functionality on advanced development boards like the Exynos 5 Dual-powered Arndale board, which includes a number of common peripherals. With this full-fledged prototyping platform, developers can get a handle on working with GPU Compute-equipped systems in preparation for the first round of mobile devices to implement the new Exynos 5 Octa (5420).
imageBy pulling together high-end ARM CPUs, ARM Mali graphics with GPU Compute and advanced technologies like ARM big.LITTLE processing, Samsung Exynos has fully realized the ideal GPU Compute system. Stay tuned for news about upcoming devices that will run on the new Exynos 5 Octa (5420)!
Need to know more? Head over to ARM’s blog to find out how GPU Compute technology can transform your mobile and TV experience.

Up close and personal with the latest Mali demos [Trina Watt on Multimedia blog of ARM, Aug 27, 2013]

The buzz at this year’s SIGGRAPH was fantastic and you have already heard from a number of my colleagues including Jem Davies with “Back from SIGGRAPH 2013 – Mali Rocks” and Akshay Agarwal with “From Advanced Graphics to Casual Gaming in the Cali Summer – Mali Everywhere.” But would you like to see more for yourself? Well now you can!
Check out our latest demos on ARMFlix and hear from our partners including Samsung, Unity and GameStick. I particularly recommend the ARM Trollheim demo as a great place to start, with Phill Smith explaining how OpenCL™ can vastly improve procedural terrain generation.
ARM Mali Trollheim demo: in-depth overview [ARMflix YouTube channel, Aug 27, 2013]
Phill Smith, Demo Manager at ARM, compares two methods of texture generation with the Trollheim demo. The first demonstration, on Exynos 5 Octa hardware with an ARM® Mali™-T628 GPU, generates terrain using GPU Shaders; the second, on the Arndale Development Board with an ARM Mali-T604 GPU, uses OpenCL™.
An immediate hardware comparison (call it obvious if you will, but we could hardly start with anything else) is first on our list to share with you. Jae-Uck Ahn, Marketing Manager for Samsung and representative for the new Exynos 5 Octa processor, analyzes the performance of our first and second generation ARM® Mali™-T600 series GPUs and clearly points out the advantages which the ARM Mali-T628 has over its predecessor, the ARM Mali-T604; to name but a few these include brighter colours, sharper images, higher frame rates – and that’s not even mentioning the massive increase in performance efficiency.
Samsung Exynos 5 Octa on ARM Mali – Siggraph 2013 [ARMflix YouTube channel, Aug 27, 2013]
Jaeuck Ahn, Marketing Manager, Samsung Electronics, describes the features and benefits of the recently announced Samsung Exynos 5 Octa based on ARM Mali-T628 MP6 GPUs, 4 ARM Cortex-A15 processors and 4 ARM Cortex-A7 processors with ARM big.LITTLE technology. Jaeuck then demos the Exynos 5420 platform versus the Exynos 5 Dual based on the Mali-T604 GPU based Nexus 10 platform showing a 2x improvement in graphics performance
And it’s not just regular mobile devices, such as smartphones and tablets, which Mali is involved in. Our vision is to bring stunning visual computing to all mobile and consumer devices and the GameStick provides a fantastic example of Mali powering alternative gadgets. John Vega, Games Relationship Manager for GameStick, explains here how simply converting mobile games, which are now frequently designed for touch screens, to a new format can generate a hugely different, more interactive and responsive user experience – all still centred around great graphics supplied by the Mali-400 GPU.
Improving the mobile gaming experience with GameStick [ARMflix YouTube channel, Aug 27, 2013]
John Vega, Games Relationship Manager at GameStick, demonstrates the touch-based iOS mobile application The Other Brothers running on a big screen with the GameStick.
The extent of API support in GPUs is a critical factor on the quality of graphics of the resultant product. As APIs evolve each generation is enabling more and better features. Android™’s recent upgrade to 4.3and its inclusion of support for OpenGL® ES 3.0 will go a long way in enhancing the look and experience of Android games (and as we’re in >50% of Android tablets and >20% of Android smartphones this is a big deal for the Mali ecosystem – especially as our GPUs already had support for all the latest APIs). A direct comparison of the capabilities of two different APIs is an excellent way of showing this and here Unity demonstrates the differences between OpenGL ES 2.0 and 3.0 on a Nexus 10.
Unity discuss the benefits of OpenGL® ES 3.0 [ARMflix YouTube channel, Aug 27, 2013]
Renaldas Zioma of Unity discusses the experience of employing OpenGL ES 3.0 in the latest version of Unity Chase. Includes a visual comparison with the OpenGL ES 2.0 version.
One final video to share with you is not of a demo, but is a great wrap up to this blog and to the first half of ARM Mali’s year as a whole. Jem Davies, VP of Mali Technology, gives an overview of where Mali is and where he sees Mali GPUs going in the future. The general conclusion that can be made is that many more first-rate demos showing off new technological advancements can be expected from the Mali team in the very near future.
Interview with ARM’s VP of Mali Technology [ARMflix YouTube channel, Aug 27, 2013]
Jem Davies, VP of Technology for ARM’s Media Processor Division, discusses the growth in ARM® Mali™ GPU popularity and the opportunities for innovation that the ARM Mali-T600 series with GPU Compute is opening up. [1:34 Samsung is setting the way in terms of trend-setting devices 1:38 the new form factors like the phablets and 1:41 the tablets that they’ve been producing and the 1:44 the Mali partners here want to see Mali being used in these really 1:48 trend-setting devices, the things that are 1:52 approaching new markets, and knowing that they can buy with confidence that 1:56 there’s a 1:57 whole variety of market segments now being addressed ny our partners.]

ARM’s predictions for the future electronic media & entertainment industry [Matt Spencer on Multimedia blog of ARM, Sept 5, 2013]

The International Broadcasting Convention (IBC) is rapidly approaching. This is one of the leading conferences in the electronic media and entertainment sectors. It is a great place to find out about the future trends in media for the broadcast and mobile markets.
In this blog post, I am going to make a few predictions on what will be big at IBC, and as a result will be coming to your living room in the near future.
Natural UI
A Natural UI (NUI) is a man-machine interface that should be imperceptible to the user. A simple gesture with your hand, a smile, a conversation – “Hey TV, what’s on now… Have I seen that one before? OK, let’s watch it then.” Creating a more seamless interaction between man and machine is going to be one of the next big driving factors in consumer electronics. However, saying that, nobody has yet managed to create one that is a brand USP – has anyone bought a TV because you can wave at it?
At IBC, we will be seeing a raft of new improvements in the fields of gesture detection, face and speech recognition plus many others. All of these can be used to improve the realisation of NUI, and all of which can benefit from use of OpenCL™ in the ARM® Mali™ T6xx GPUs.Great examples of this can be found with some of our Mali partners such as eyeSight™, SoftKinetic™, CrunchFish and Ittiam®, all of whom are making great use of OpenCL to improve performance in both speed and energy consumption whilst performing what would traditionally be a highly CPU compute intensive task.
But what does this mean in terms of an actual on-screen UI? An implementation of a NUI is going to have to be as subtle as the interaction model that is driving it. This is the part of the NUI story that I believe will be in its infancy on the show floor at IBC.
To implement a UI that feels natural requires a lot of visual processing and this is where an ARM Mali GPU comes in. Imagine the use case where we want to show the results of the query “What’s on TV tonight”. The current trend in the UI space would be to de-focus the full screen video by either shrinking it to allow room for the results to be shown or overlaying the results over the video with a simple alpha effect. This is neither a natural nor unobtrusive way of pulling my focus to the results of the query. To do this unobtrusively required changing my physical focus point to the new information whilst not taking it fully away from the video. A simple way of achieving this would use a full screen Gaussian Blur effect to the video and overlaying the results in sharp focus over the top. This kind of process requires use of a GPU.
So, even though the NUI principles require that the UI is mostly invisible, we can see that the harder we try to achieve this, the harder the GPU is going to have to work.
Having seen UltraHD content for the first time last year at IBC, I have to say that the impact is immediate. There is no need to wear glasses as with current 3D technologies, the pictures seem to come alive with huge levels of detail and vibrant colours. There will be a lot of technology at IBC to help deliver broadcast UltraHD content to consumers from cameras and HEVC encoders through to STBs and TVs.
In terms of hardware accelerating HEVC content, the current generation of hardware does not have native capability to decode this content. So a traditional approach would be to use the CPU to decode the video, which would be a battery and CPU intensive task. ARM has recently been working with Ittiam to offload this compute intensive task to the GPU to improve decode performance for frame rate and battery consumption.
But what about the UI? Rendering a flowing, responsive UI at UltraHD resolutions is not an easy task. There are many more pixels that will need to be flung around the screen and the desire for browsers and UIs to hit a jank-free 60fps requires a top-of-the-range GPU combined with sufficient memory with high bandwidth. When the subtle complexities of NUI are added to the equation, the next generation of performance-efficient GPUs will be earning their keep in the consumer TV and set-top box (STB) space.
So the prediction for IBC in the UltraHD arena is that there will be a lot of new hardware, both consumer and professional, but the user experiences on these devices will be in their early stages. We will start to see chipsets designed with the specific GPU needs of UltraHD resolutions, but implementations of User Interfaces on this new class of devices will not be very mature.
Rise of the Companion
The use of a companion device to augment live and recorded content will also be a big part of the show. This will partly be delivered by APIs that allow the synchronization of the content on the main screen and the application running on the companion device. From a UI perspective, I would expect to see a drive towards cross platform application frameworks.
To ease integration into the companion device and to allow rapid deployment of new applications, I would expect to see this implementation of the companion experience realised using HTML5. HTML5 is a great platform to enable this cross connectivity and interface design, but there can be difficulties getting the required performance due to the underlying browser and JavaScript engine implementations.
I would therefore expect to see a number of companies at IBC showing both bottom up and top down approaches to solving the problem. The bottom up approach would require bespoke rendering and JavaScript engines that solve some of these questions, whilst the top down approach will give you performance debugging tools and frameworks to help identify and solve performance bottlenecks in the HTML5 applications.
I see HTML everywhere
HTML5 will be the de-facto standard for implementing User Interfaces at the show. The problem with HTML5, however, is that a lot of UI implementations end up looking very similar. You can often tell an HTML5 UI just by looking at it.
Now that HTML5 is the standard for UIs I would expect to see more differentiation in the implementations. Vendors will be more experimental and start to push the bounds of standard web design when approaching their interfaces. More animation, more effects… more pazazz! It will no longer be acceptable to put a few buttons on the screen with a simple highlight and expect consumers to accept this as a good UI design.
ARM’s position
It is an exciting time to be in the market, with a lot of great new technologies maturing and filtering through to consumer products. ARM is in a great position at the heart of this process, and Mali technology is going to be key to the success of some of these initiatives with its adoption of technologies such as OpenGL® ES 3.0, OpenCL 1.1 and its work with key technology partners.
We are already seeing great performance from next generation Mali-based devices and the Samsung Exynos 5 Octa with Mali-T628 MP6 shows class-leading performance. I see a bright future for Mali-based devices in this exciting and fast paced industry.
imageMatt Spencer, UI and Browser Marketing Manager, Media Processing Division, ARM. Matt is a technologist and innovator at heart, working for 15 years on embedded platforms in every manner of language from assembler through to JavaScript. He recently made the move to the ‘dark side’ (Marketing), for which he hopes he hasn’t lost the respect of his technical peers and, more importantly, where he hopes to help educate and bring like-minded engineers together with the goal of making embedded systems rock.

About Nacsa Sándor

Lazure Kft. • infokommunikációs felhő szakértés • high-tech marketing • elérhetőség: Okleveles villamos és automatizálási mérnök (1971) Munkahelyek: Microsoft, EMC, Compaq és Digital veterán. Korábban magyar cégek (GDS Szoftver, Computrend, SzáMOK, OLAJTERV). Jelenleg Lazure Kft. Amire szakmailag büszke vagyok (időrendben visszafelé): – Microsoft .NET 1.0 … .NET 3.5 és Visual Studio Team System bevezetések Magyarországon (2000 — 2008) – Digital Alpha technológia vezető adatközponti és vállalati szerver platformmá tétele (másokkal együttes csapat tagjaként) Magyarországon (1993 — 1998) – Koncepcionális modellezés (ma használatos elnevezéssel: domain-driven design) az objektum-orientált programozással kombinált módon (1985 — 1993) – Poszt-graduális képzés a miniszámítógépes szoftverfejlesztés, konkurrens (párhuzamos) programozás és más témákban (1973 — 1984) Az utóbbi időben általam művelt területek: ld. (Experiencing the Cloud) – Predictive strategies based on the cyclical nature of the ICT development (also based on my previous findings during the period of 1978 — 1990) – User Experience Design for the Cloud – Marketing Communications based on the Cloud
This entry was posted in consumer computing, consumer devices, smartphones, SoC, tablets and tagged , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , . Bookmark the permalink.

2 Responses to Samsung Exynos 5 Octa with Heterogeneous Multi-Processing and GPU Compute is the hidden gem in the Galaxy Note 3 and GALAXY Note 10.1, 2014 Edition, launched at ‘Samsung UNPACKED 2013 Episode 2’ event

  1. Pingback: The new Air Command S Pen User Experience making the Samsung Galaxy Note 3 phablet, and Galaxy Note 10.1, 2014 Edition tablet next-generation devices | Experiencing the Cloud

  2. Pingback: The question mark over Wintel’s future will hang in the air for two more years | Experiencing the Cloud

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s