Mohacsy's Safe Portfolio Member Forum

[dexprs]

Tech Trader Daily: Intel To Ride the Rise in Artificial Intelligence Chips, Says Daiwa -- Barron's Blog
DOW JONES & COMPANY, INC. 12:43 PM ET 11/17/2017
Symbol Last Price Change
INTC 44.9158up -0.7342 (-1.61%)
NVDA 211.32up -0.29 (-0.14%)
XLNX 71.9up -0.77 (-1.06%)
AMD 11.51down +0.26 (+2.31%)
QUOTES AS OF 01:59:41 PM ET 11/17/2017
Intel (INTC) is going to be transformed by its push into chips for artificial intelligence, according to a report today by Deepak Sitaraman, who raised his rating on the stock to Buy from Neutral, as he assumes coverage of the stock from his predecessor, Yoko Yamada, who left the firm earlier in the year.

Interestingly, Sitaraman also took over coverage of Intel(INTC) competitor Nvidia(NVDA) , and also boosted that stock to Buy from Neutral. Sitaraman has a $55 price target on Intel(INTC) shares, and a $260 price target on Nvidia(NVDA) shares.

" We see Intel(INTC) as a beneficiary of the anticipated robust growth in Datacenter and Automotive end-markets driven by AI," writes Sitaraman.

Investors, writes Sitaraman, are underestimating what that means in terms of the opportunity. "With AI driven compute at an inflection, we see Datacenter and Automotive end-markets growing to $53.1bn long-term from $11.8bn in 2017 (21% CAGR)," he writes.

Sitaraman also thinks Intel(INTC) isn't getting proper credit for its abilities in A.I.. He explains three things Investors need to understand:

1) its early days for AI, as we expect the addressable market to grow by a factor of four long-term, 2) recognizing changes in datacenter architectures for AI training, the company has shown a willingness to adapt, and has assembled a portfolio of assets to address new AI workloads, and 3) based on our scorecard to evaluate AI chip vendors across five key metrics (see page 21), we think the company's strengths in ecosystem and their ability to drive the AI roadmap, bode well for future success. Lastly, expectations around AI remain low, so an investment in Intel(INTC) currently includes a free option on any future AI related upside.

Sitaraman offers up the following comparison of how Intel(INTC), Nvidia(NVDA), and Xilinx(XLNX) and Advanced Micro Devices(AMD) stack up on various aspects of A.I.:

Note that Nvidia(NVDA) is number one, by his ranking:

NVIDIA (NVDA) has a commanding lead in datacenter solutions for deep learning training, and is well positioned in the Automotive market when the market for self-driving cars starts to ramp. The company's key attributes are price/ performance per watt based on customer traction, software and ecosystem. We think the company leads on price performance per watt based on the adoption of its technology by leading customers. In software, the company has built on its proprietary CUDA language, a key differentiator that was first introduced in 2006 to gain an early lead in GPGPU computing. On ecosystem, given that management sees the biggest competition as the "lack of adoption", it is doing everything possible to develop a robust ecosystem of developers and applications. Some of the initiatives here include the Deep Learning Institute (DLI), GPU Tech Conferences held through the world, and the open sourcing of its Deep Learning Accelerator (DLA). Currently, NVIDIA(NVDA) dominates the training market and is starting to make in-roads into inferencing. NVIDIA(NVDA) ranks #1 on our scorecard with a score of 4.1/5.0.

But don't count out Intel's(INTC) ability to invest:

Intel's (INTC) key attributes are on ecosystem and ability to invest in the roadmap. In ecosystem, the company has a commanding mind share with datacenter application developers, customers and its efforts to broaden the portfolio to address AI in the datacenter and the self-driving car market, bode well for the company. While Intel's(INTC) current datacenter dominance gives it a strong starting point with regard to the ecosystem, we believe the company will have to accelerate software efforts related to its AI portfolio to ultimately be a winner. While some believe the company has been late in reacting to growth in this market, we think it's still early days. And while admittedly, Intel's(INTC) strategy may not be as focused as NVIDIA's(NVDA) we believe the portfolio approach to address the one-size-does-not-fit-all view when it comes to AI workloads could be successful at the end of the day. Critical to this, however, is traction with its Nervana NNP products that were recently announced and that will start shipping in 2018. Here, all indications appear to support the notion that AI remains one of the focus areas for the company. This, together with the company's ability to invest (and attract talent), Intel's(INTC) R&D spend relative to peers and cash pile are positives. Intel(INTC) ranks #2 on our scorecard with a score of 3.0/5.0.

Xilinx (XLNX) is "well positioned," he notes, but "has much work to do." AMD, which he ranks fourth, has its "R&D efforts spread quite thin, and as such, we question the impact they can have without more of a focus on this opportunity."

Intel (INTC) shares today are down 69 cents, or 1.5%, at $44.96, while Nvidia(NVDA) stock is up 92 cents at $212.53.

More at Barron's Tech Trader Daily blog, http://www.barrons.com/tech-trader


(END) Dow Jones Newswires
11-17-171243ET
Copyright (c) 2017 Dow Jones & Company, Inc.