Cloud Computing is Disrupting the AI Chips MarketFeb23

MRP

MCALINDEN RESEARCH PARTNERS  | DIBS

SITE  MAIL INFO

DATA

ECON

POL

FIN

REAL

LABOR

MFG

TECH

ENERGY

END

 

We bring you our Daily Intelligence Briefing courtesy of McAlinden Research Partners. The report is provided to Hedge Connection members for free. Below is snapshot, login to view the full report. Not a member? Join today.

McAlinden Research Partners is currently offering a complimentary full month subscription of the DIB. Activate yours today – http://www.mcalindenresearchpartners.com/hc-trial.html

Daily Intelligence Briefing – February 13, 2018

 

FEATURED TOPIC: Cloud Computing is Disrupting the AI Chips Market

 

The Artificial Intelligence race continues to heat up as AI technologies will have created a $1.2 trillion insights-driven market and filtered into almost all new software product and services by 2020. However, the landscape of the chips sector may be set for a huge shift as traditional chips and the way they learn are becoming increasingly obsolete compared to new cloud-based platforms of neural networks for machine learning.

 

In an unprecedented move, Google has made its advanced AI chips, originally designed to handle the company’s own processes, available to the public. On Monday, the internet giant said it would allow other companies to buy access to those chips through its cloud-computing service. The cloud allows outside users to access server networks connected to the AI chips, housed in Google’s data centers. Google hopes to build a new business around the chips, called tensor processing units, or TPUs. TPU chips have helped accelerate the development of everything from Google Assistant, the service that recognizes voice commands on Android phones, to Google Translate, the internet app that translates one language into another.

 

Likewise, Amazon has begun to build its own AI chips through its deep learning framework, MXNet, to improve its Alexa-enabled devices. Amazon’s first piece of silicon is understood to be aimed at shortening response times and making interactions with it more intuitive and lifelike. Amazon Web Services, which already offers a number of AI services via the cloud, may also be working on an AI chip for data centers.

 

Generating even more competition are China’s cloud providers, Alibaba, Tencent, and Baidu. As Chinese companies seek to expand their reach, propelled by massive state-backed investment, they may increasingly aim their cloud services at US companies and developers, and vice versa.

 

Amazon’s project is being spearheaded by Annapurna Labs, an Israeli semiconductor company acquired by Amazon in 2015 for $350 million, helped along by tech from security camera maker Blink, acquired late last year for $90 million. This type of M&A activity could become a more integral part of the semiconductor sector now that there are at least 45 startupsworking on AI computer chips. As of late, raising money for AI computing has become much easier. For many years, processor development has been out of reach for start-ups, as established players built up barriers to competition using both business and legal frameworks. Recently, though, investors look at these startup AI companies not only as firms that could actually end up shipping an ‘AI product’ but as ones that will be gobbled up for millions by the likes of Intel, Google, Apple, Microsoft, Facebook etc.

 

Lightmatter is a prime example of startups developing the future of AI, as the company makes photonic chips that essentially perform calculations at the speed of light, leaving transistors in the dust. One such complex type of math problem common in AI applications is a matrix vector product. Doing these quickly is important for comparing large sets of data with one another, for instance if a voice recognition system wants to see if a certain sound wave is sufficiently similar to “OK Google” to initiate a response. As demand increases for AI-based products, these calculations need to be done more frequently and faster, but current technologies are reaching the limits of just how quickly and efficiently they can be accomplished and relayed back to the user. Lightmatter’s photonic chips could be one answer to fixing this issue, following the company’s $11 million Series A. 

 

Innovative startups and internet-based companies are beginning to make and acquire their own hardware cut costs and pose a significant threat to the business of traditional hardware players like Nvidia, Intel and Qualcomm. A new wave of artificial intelligence, including services like Google Assistant, are driven by “neural networks,” which are complex algorithms that can learn tasks on their own by analyzing vast amounts of data. Typically, engineers train these algorithms using graphic processing units, or GPUs, mostly supplied by Nvidia. With Google transferring its AI via the cloud, gains in efficiency can be virtually limitless.

 

Charles Bergan, vice president of engineering at Qualcomm, believes that, if neural networks were to become capable of “one-shot learning”, the cumbersome process of feeding reams of data into algorithms to train them would be rendered obsolete. This could have serious consequences for the hardware industry, as both existing tech giants and startups are currently focused on developing more powerful processors designed to run today’s data-intensive AI algorithms. Winners in AI and machine learning will be internet and cloud-based companies whose products continue to evolve into distributed networks, as opposed to being sold in individual physical units. Investors can gain exposure to companies developing AI chips via the Semiconductor ETF (SOXX) or the Internet Index ETF (FDN).

 

HERE IN THE MEANTIME are some related articles on AI(the stories are summarized in the TECHNOLOGY section of today’s report):

• AI – Google Makes Its Special A.I. Chips Available to Others 
• AI – Lightmatter aims to reinvent AI-specific chips with photonic computing and $11M in funding 
• AI – 45 startups now working on AI computer chips 
• AI – Amazon Developing AI Chip To Speed Up Alexa Devices

 

CHART: Semiconductors (SOCC) vs Internet/Cloud Companies (FDN) vs S&P 500 (SPY)

OTHER STORIES HIGHLIGHTED IN TODAY’S DIBS:

• Markets: 

• Cryptocurrency – A World First: Gibraltar Plans to Regulate and Legalize Initial Coin Offerings 

• Economics and Trade: 

• ASEAN – Singapore May Start Taxing Amazon and Lazada 
• France – 3.5 Million Jobless and French Companies Still Can’t Find Staff

• Services: 

• Supply Chain – Amazon reportedly to launch challenge to FedEx, UPS

• Technology: 

• Batteries – Batteries, Not The Model 3, Are The Real Keys To Tesla’s Future Success 
• 3DP – New process allows 3-D printing of nanoscale metal structures

• Transportation:  

• Aerospace – Donald Trump wants to shut down the International Space Station and get ready for private space 
• Aviation – Trump infrastructure plan could sell off Reagan, Dulles airports 
• Autos – European Autos Rally May Be Running Out of Gas as Risks Loom 
• Trucking – There Aren’t Enough Truckers, and That’s Pinching US Profits

• Commodities: 

• Oil – Demand for oil surges in India 
• Oil – OPEC President Says Shale Won’t Thwart Plan to Clear Glut 

• Endnote: 

• CHART: Asia’s Central Banks Can Do Little About Their Biggest Challenge

JOE MAC’S MARKET VIEWPOINT

• Joe Mac’s Market Viewpoint: The Coming Value Rotation
• Joe Mac’s Market Viewpoint: Beyond the BOND BUBBLE
• Joe Mac’s Market Viewpoint: A Review of MRP’s Latest Change-Driven Investment Themes
• Joe Mac’s Market Viewpoint: The Gathering Storm
• Joe Mac’s Market Viewpoint: Contrarian Crude Call

CURRENT MRP THEMES

Autos  (S)	Electric Utilities  (L)	TIPS  (L) / Long-Dated UST (S)
Defense  (L)	Industrials & Materials (L)	U.S. Financials & Regional Banks (L)
Emerging Markets (L)	Oil & U.S. Energy  (L)	U.S. Homebuilders & Construction (L)
France  (L), Greece (L)	Palladium  (L)	U.S. Healthcare Providers (S)
Gold & Gold Miners (L)	Robotics & Automation (L)	Video Gaming (L)
Lithium (L)	Steel  (L)	Value over Growth  (L)

About the DIBs: MRP focuses on identifying transformational change in the global economy and offering an investment thesis whenever an opportunity arises that has not yet been recognized by the market. The DIBs are MRP’s compilation of articles and data from multiple sources on subjects reflecting disruptive change that have potential investment implications for an industry or group of securities. We share these with our clients who may already have or may be considering exposure in the industries affected. The subjects change daily and constitute an excellent update on featured topics. 

MAJOR DATA POINTS

Top

• United states, 3-Month Bill Auction, 2/12/2018: 1.570% from prior 1.500%
• Denmark, Inflation Rate, YoY, JAN: 0.7% from prior 1%
• Switzerland, Inflation Rate, YoY, JAN: 0.7% from prior 0.8%
• China, Vehicle Sales, YoY, JAN: 11.6% from prior 0.1%
• China, Outstanding Loan Growth, YoY, JAN: 13.2% from prior 12.7%
• India, Inflation Rate, YoY, JAN: 5.07% from prior 5.21%

MARKETS

Top

Cryptocurrency – A World First: Gibraltar Plans to Regulate and Legalize Initial Coin Offerings 

The Government of Gibraltar is weighing up legislation to regulate initial coin offerings (ICOs) and their secondary markets. In an announcement by the Gibraltar Financial Services Commission (GFSC) today, the financial watchdog for the British Overseas Territory said the government and the authority are jointly developing legislation for blockchain-based token offerings in the territory. The decision follows a discussion by the government with local stakeholders in December after an initial September statement outlined authorities’ intent to introduce a ‘complementary regulatory framework covering the promotion and sale of tokens.’

 

As a result, the regulations will mandate ICO issuers to follow disclosure rules that includes ‘adequate accurate and balanced disclosure of information’ to all prospective token buyers. Further, the regulations will also establish mechanisms to prevent financial crime. CCN

ECONOMICS AND TRADE

Top

ASEAN – Singapore May Start Taxing Amazon and Lazada 

Singapore may unveil an e-commerce tax in next week’s budget, setting the tone for a region that’s grappling with online retail’s assault on brick-and-mortar vendors. Eight of the 12 economists in a Bloomberg survey said the Feb. 19 budget will contain a new tax on online vendors, with another betting that cross-border digital transactions will now be included in the goods-and-services tax regime. Thailand, Indonesia, and Malaysia — where governments are funding ambitious infrastructure programs — are also considering similar plans.

 

Southeast Asian governments are seeking to level the playing field for traditional vendors given the rapid growth of online retailing on platforms such as Lazada, controlled by Alibaba Group Holding Ltd., and Amazon.com Inc. BMI Research projects the region’s six biggest economies will boost e-commerce to $64.8 billion in 2021 from $37.7 billion last year, while Credit Suisse Group AG estimates that online shopping growth could outpace that of traditional retailers by six to 10 times over the next few years. B 

 

*

 

France – 3.5 Million Jobless and French Companies Still Can’t Find Staff 

Figeac-Aero has a problem that says a lot about the French economy: the Airbus supplier can’t find enough qualified people to staff its factories. Even with one in 10 workers out of a job, companies cite difficulty in hiring as one of their biggest problems as a long-awaited economic recovery takes hold.

 

The crunch is particularly acute for Patrice Parisot, who runs a Figeac-Aero plant in Auxerre. Despite running ads, organizing job dating events and offering training, few applicants call, fewer turn up, and the factory is short of operators for its brand new machines. That’s left deliveries running late and the company saddled with 75 million euros ($92 million) of orders, which none of its seven French sites can take on. 

 

The issue faced by Parisot and others captures why France needs to revamp expensive worker-training programs and improve incentives to work. In fact, President Emmanuel Macron wants to go even further, overhauling the cherished welfare system. B

 

*

SERVICES

Top

Supply Chain – Amazon reportedly to launch challenge to FedEx, UPS 

Amazon is poised to launch a shipping service, “Shipping with Amazon” or “SWA,” that would challenge FedEx, United Parcel Service, DHL and other players, the Wall Street Journal reported Friday. The new service will start out in Los Angeles “in coming weeks,” delivering for sellers on Amazon’s marketplace but slated to expand almost immediately, according to the report. 

 

Bypassing third-party shippers like UPS and FedEx could save Amazon $1.1 billion annually, Citigroup analysts said in 2016. That translates to savings of $3 or more on a typical delivery, which now costs $7.81 on average. Amazon’s shipping and fulfillment costs until its most recent quarter grew faster than its revenue, but so far the company has focused on making moves to optimize those operations. RDive

TECH

Top

AI – Google Makes Its Special A.I. Chips Available to Others 

A few years ago, Google created a new kind of computer chip to help power its giant artificial intelligence systems. These chips were designed to handle the complex processes that some believe will be a key to the future of the computer industry. On Monday, the internet giant said it would allow other companies to buy access to those chips through its cloud-computing service. Google hopes to build a new business around the chips, called tensor processing units, or T.P.U.s. 

Google is in the vanguard of a movement to design chips specifically for artificial intelligence, a worldwide push that includes dozens of start-ups as well as familiar names like Intel, Qualcomm and Nvidia. And these days, companies like Google, Amazon and Microsoft are not just big internet companies. They are big hardware makers.

 

This does not mean that Google will stop buying chips from Nvidia and other chip makers. But it is altering the market. Over the years, Google has even flirted with the possibility of designing its own version of the chips it buys from Intel. NYT

 

AI – Lightmatter aims to reinvent AI-specific chips with photonic computing and $11M in funding 

It takes an immense amount of processing power to create and operate the “AI” features we all use so often, from playlist generation to voice recognition. Lightmatter is a startup that is looking to change the way all that computation is done — and not in a small way. The company makes photonic chips that essentially perform calculations at the speed of light, leaving transistors in the dust. It just closed an $11 million Series A. 

 

So what exactly does Lightmatter’s hardware do? Instead of breaking that matrix calculation down to a series of basic operations with cascades of logic gates and transistors, Lightmatter’s photonic chips essentially solve the entire problem at once by running a beam of light through a gauntlet of tiny, configurable lenses (if that’s the right word at this scale) and sensors. TC

 

AI – 45 startups now working on AI computer chips 

There’s a new frontier in computer processor development giving the minnows a chance to compete with the likes of Intel, Nvidia and Qualcomm. According to an article this weekend in the New York Times this computer processor market segment targets artificial intelligence. NYT contributor Cade Metz reports that there are now at least 45 start-ups working on AI chips and at least five of them have raised more than $100 million from investors. 

 

For many years processor development has been out of reach for start-ups, as established players built up barriers to competition using both business (partnerships, vertical integration, massive R&D budgets) and legal (copyright etc) frameworks. However the AI processor niche players often tackle processor designs from a completely different angle. As it is such a new field, innovative left-field thinking can have a big impact. 

 

Investors look at these startup AI companies not only as firms that could actually end up shipping an ‘AI product’ but as one that will be gobbled up for millions by the likes of Intel, Google, Apple, Microsoft, Facebook etc . Hexus

 

AI – Amazon Developing AI Chip To Speed Up Alexa Devices 

Amazon is developing an artificial intelligence chip meant to speed up its future Alexa-enabled devices, The Seattle, Washington-based e-commerce giant’s first piece of silicon is understood to be aimed at placing a larger emphasis on offline processing so as to reduce Alexa’s reliance on the cloud, consequently shortening its response times and making interactions with it more intuitive and lifelike, i.e. more similar to talking to a human being. 

 

The project is understood to be spearheaded by Annapurna Labs, an Israeli semiconductor company acquired by Amazon in 2015 for $350 million. The tech giant has yet to share any details on Annapurna’s post-acquisition endeavors in an official capacity. The exact scope of its capabilities is also unclear, with insiders suggesting a number of complex tasks such as music playback would still be handled entirely in the cloud and hence continue featuring a small delay between the moment they receive a command and the one in which they comply with it. AH

 

Batteries – Batteries, Not The Model 3, Are The Real Keys To Tesla’s Future Success 

Batteries are really what Tesla is all about — not the media-loving, mass market Model 3, not Roadsters, not falcon-wing doors, and certainly not the hyperloop. Batteries are the keys to completing a worldwide transition to a clean energy economy, and Tesla, with all its entrés to date into increased battery storage capacity, will really have to step up its game to dominate the market share as CEO Elon Musk has publicly envisioned. 

 

Tesla has had to reduce its Model 3 production targets, in part because of issues with battery manufacturing. Tesla’s enormous battery factory — about 5.5 million square feet — in Nevada, called the Gigafactory or Gigafactory 1, is key to Tesla’s future. Once at full capacity, Tesla reportedly expects the Gigafactory to introduce production cost cuts for lithium-ion electric vehicle batteries and energy storage products by some 30%, compared to pre-Gigafactory batteries.CT

 

3DP – New process allows 3-D printing of nanoscale metal structures 

For the first time, it is possible to create complex nanoscale metal structures using 3-D printing, thanks to a new technique developed at Caltech. The process, once scaled up, could be used in a wide variety of applications, from building tiny medical implants to creating 3-D logic circuits on computer chips to engineering ultralightweight aircraft components. It also opens the door to the creation of a new class of materials with unusual properties that are based on their internal structure. 

 

Greer’s group 3-D prints structures out of a variety of materials, from ceramics to organic compounds. Metals, however, have been difficult to print, especially when trying to create structures with dimensions smaller than around 50 microns, or about half the width of a human hair. Greer and Vyatskikh are still refining their technique; right now, the structure reported on in their paper includes some voids left behind by the vaporized organic materials as well as some minor impurities. SD