Leading Independent Investment Research, Morningstar, Inc.

Morningstar

Leading Independent Investment Research

Morningstar, Inc.

Morningstar, Inc. is a leading provider of independent investment research in North America, Europe, Australia, and Asia. We offer an extensive line of products and services for individual investors, financial advisors, asset managers, and retirement plan providers and sponsors. 

Morningstar provides data on more than 500,000 investment offerings, including stocks, mutual funds, and similar vehicles, along with real-time global market data on more than 16 million equities, indexes, futures, options, commodities, and precious metals, in addition to foreign exchange and Treasury markets. Morningstar also offers investment management services through its investment advisory subsidiaries, with more than $180 billion in assets under advisement or management as of June 30, 2015. We have operations in 27 countries.

Download the Morningstar Fact Sheet : http://corporate.morningstar.com/US/documents/MarketingFactSheets/AboutMorningstarFactsheet.pdf

Thematic (Sector/Industry)

Estonia Semiconductors and semiconductor equipment / Thematic (Sector/Industry)

<blockquote>There is nothing more deceptive than an obvious fact.</blockquote>Taken from Sir Arthur Conan Doyle’s The Adventures of Sherlock Holmes, this quote aptly describes our view on <a href="https://www.smartkarma.com/entities/5891" target="_blank" data-smartkarma-citation="entities:5891:1">Nvidia Corp (NVDA US)</a>'s implied dominance in artificial intelligence, as illustrated by the exponential rise in its stock price over the past 18 months. Although we believe Nvidia is well positioned to continue benefiting from the burgeoning AI phenomenon, we think <a href="https://www.smartkarma.com/entities/16764" target="_blank" data-smartkarma-citation="entities:16764:1">Intel Corp (INTC US)</a> and other peers also stand to gain as each firm brings to market its own custom-tailored AI solutions. By 2021, we project the AI accelerator chip market to be $20 billion.The information technology space is undergoing a monumental shift. Historically, rapid hardware innovations preceded comparable software advances. First came the faster computer, then the pertinent programs to take advantage of it. Over the past decade, chipmakers have found it increasingly onerous to maintain the cadence set by Moore’s law. The process of shrinking transistors to make faster chips is no longer clockwork. Meanwhile, the rise of AI, machine learning, and deep learning has put chip suppliers on their heels as they attempt to address these dynamic requirements with specialized chips to accelerate said workloads. Companies with seemingly infinite data are looking to leverage AI to create algorithms that make data understandable and actionable. Nvidia's GPUs have emerged as an initial frontrunner to satisfy these needs, but we foresee alternative solutions, especially from Intel. Both firms boast economic moats stemming from cost advantages and intangible assets related to chip design.<h2>Key Takeaways</h2><ul><li>We believe the artificial intelligence trend, led by deep learning, remains in the very early stages, as we estimate that only 7% of server workloads in 2016 were AI-related (machine and deep learning).</li><li>The chips used to enable AI, dubbed "accelerators," include an assortment of offerings from Intel, Nvidia, <a href="https://www.smartkarma.com/entities/11600" target="_blank" data-smartkarma-citation="entities:11600:1">Xilinx Inc (XLNX US)</a>, and other chipmakers. These processors work alongside CPUs to speed up AI workloads.</li><li>Deep learning is the iteration of AI that we think holds the most promise. Training and inferencing are the two main components of deep learning. Training involves teaching a computer to complete a task, while inferencing shows the trained computer a brand-new instance on which to perform the task.</li><li>We foresee the majority of training tasks occurring in the data center, while inferencing will take place in both the data center and in edge devices (smartphones, smart homes, vehicles, and so on).</li><li>Cloud vendors have led the adoption of deep-learning techniques to take advantage of the large amounts of data gained from their normal operations. Early examples include <a href="https://www.smartkarma.com/entities/409" target="_blank" data-smartkarma-citation="entities:409:1">Amazon.com Inc (AMZN US)</a>'s Alexa, <a href="https://www.smartkarma.com/entities/11369" target="_blank" data-smartkarma-citation="entities:11369:1">Alphabet Inc (GOOG US)</a>'s Google Now, and <a href="https://www.smartkarma.com/entities/15273" target="_blank" data-smartkarma-citation="entities:15273:1">Microsoft Corp (MSFT US)</a>'s Cortana, all of which require accelerators to function.</li><li>We expect the accelerator chip market to reach a CAGR of 93% over the next five years, due to the diminishing returns of Moore's-law-driven CPU performance improvements. By 2021, we forecast the accelerator market will be $20 billion. We suspect the market is even more optimistic than us, as Nvidia's stock price has risen exponentially in the past 18 months, thanks to its AI prospects.</li><li>Nvidia is today's leader in the accelerator space, as the firm has successfully leveraged its GPUs for PC graphics into AI. However, we believe the market has priced in overly ambitious expectations for the firm. While we expect Nvidia's GPUs to remain dominant in training, we think other solutions are more suitable for the inferencing portion of deep learning, which we estimate to be a larger opportunity.</li><li>Intel is the dominant player in server CPUs, and the firm has built a broad portfolio of accelerators via acquisition and internal development. These include FPGAs, co-processors, and ASICs, all of which we project to carve out share in the accelerator market, especially on the inference side. Shares are modestly overvalued today, but we'd be avid buyers on any meaningful pullback.</li></ul>Analyst: Abhinav Davuluri

There is nothing more deceptive than an obvious fact.Taken from Sir Arthur Conan Doyle’s The Adventures of Sherlock Holmes, this quote aptly describes our view on Nvidia Corp (NVDA US)'s implied dominance in artificial intelligence, as illustrated by the exponential rise in its stock price over the past 18 months. Although we believe Nvidia is well positioned to continue benefiting from the burgeoning AI phenomenon, we think Intel Corp (INTC US) and other peers also stand to gain as each firm brings to market its own custom-tailored AI solutions. By 2021, we project the AI accelerator chip market to be $20 billion.The information technology space is undergoing a monumental shift. Historically, rapid hardware innovations preceded comparable software advances. First came the faster computer, then the pertinent programs to take advantage of it. Over the past decade, chipmakers have found it increasingly onerous to maintain the cadence set by Moore’s law. The process of shrinking transistors to make faster chips is no longer clockwork. Meanwhile, the rise of AI, machine learning, and deep learning has put chip suppliers on their heels as they attempt to address these dynamic requirements with specialized chips to accelerate said workloads. Companies with seemingly infinite data are looking to leverage AI to create algorithms that make data understandable and actionable. Nvidia's GPUs have emerged as an initial frontrunner to satisfy these needs, but we foresee alternative solutions, especially from Intel. Both firms boast economic moats stemming from cost advantages and intangible assets related to chip design.Key TakeawaysWe believe the artificial intelligence trend, led by deep learning, remains in the very early stages, as we estimate that only 7% of server workloads in 2016 were AI-related (machine and deep learning).The chips used to enable AI, dubbed "accelerators," include an assortment of offerings from Intel, Nvidia, Xilinx Inc (XLNX US), and other chipmakers. These processors work alongside CPUs to speed up AI workloads.Deep learning is the iteration of AI that we think holds the most promise. Training and inferencing are the two main components of deep learning. Training involves teaching a computer to complete a task, while inferencing shows the trained computer a brand-new instance on which to perform the task.We foresee the majority of training tasks occurring in the data center, while inferencing will take place in both the data center and in edge devices (smartphones, smart homes, vehicles, and so on).Cloud vendors have led the adoption of deep-learning techniques to take advantage of the large amounts of data gained from their normal operations. Early examples include Amazon.com Inc (AMZN US)'s Alexa, Alphabet Inc (GOOG US)'s Google Now, and Microsoft Corp (MSFT US)'s Cortana, all of which require accelerators to function.We expect the accelerator chip market to reach a CAGR of 93% over the next five years, due to the diminishing returns of Moore's-law-driven CPU performance improvements. By 2021, we forecast the accelerator market will be $20 billion. We suspect the market is even more optimistic than us, as Nvidia's stock price has risen exponentially in the past 18 months, thanks to its AI prospects.Nvidia is today's leader in the accelerator space, as the firm has successfully leveraged its GPUs for PC graphics into AI. However, we believe the market has priced in overly ambitious expectations for the firm. While we expect Nvidia's GPUs to remain dominant in training, we think other solutions are more suitable for the inferencing portion of deep learning, which we estimate to be a larger opportunity.Intel is the dominant player in server CPUs, and the firm has built a broad portfolio of accelerators via acquisition and internal development. These include FPGAs, co-processors, and ASICs, all of which we project to carve out share in the accelerator market, especially on the inference side. Shares are modestly overvalued today, but we'd be avid buyers on any meaningful pullback.Analyst: Abhinav Davuluri

Accelerator: Rise of the Machine (And Deep) Learning Phenomenon