Year: 2020

Intel has been doubling down on building chips and related architecture for the next generation of computing, and today it announced an acquisition that will bolster its expertise and work specifically in one area of future technology: artificial intelligence.

The semiconductor giant today announced that it has acquired SigOpt, a startup out of San Francisco that has built an optimization platform that can be used to run modeling and simulations (two key applications of AI tech) in a better way. Anthony described SigOpt as a startup built to “optimize everything” when we covered its Series A last year, but Intel specifically will be integrating the tech into its AI business, specifically into its AI Analytics Toolkit, a spokesperson tells me.

Terms of the deal were not disclosed but SigOp already counted a number of large enterprises — “SigOpt’s customer base includes Fortune 500 companies across industries, as well as leading research institutions, universities and consortiums using its products” — among its customers. The product was still in a closed beta, however. Notably, it had raised money from an interesting group of investors that included In-Q-Tel (the firm associated with the CIA that makes strategic investments) and Andreessen Horowitz, and Y Combinator, among others. It had raised less than $10 million.

The plan will be to continue providing services to existing users, and to continue building out the company’s platform — co-founders Scott Clark (CEO) and Patrick Hayes (CTO) and their team are joining Intel.

“We will continue to work with SigOpt’s existing customers and will also integrate the technology into our product roadmap,” a spokesperson confirmed.

While Intel is working hard on streamlining its business around next-generation chips to better compete against the likes of NVIDIA (which itself is growing substantially with the acquisition of ARM) and smaller players like GraphCore, in part by divesting more legacy operations, it seems a strong opportunity in providing services for its customers alongside those chips, and these services specifically will help customers with the compute loads that they will be running on those chips.

The focus for Intel has been on the next generation of computing to offset declines in its legacy operations. In the last quarter, even as it beat expectations, Intel reported a 3% decline in its revenues, led by a drop in its data center business. It said that it’s projecting the AI silicon market to be bigger than $25 billion by 2024, with AI silicon in the data center to be greater than $10 billion in that period.

In 2019, Intel reported some $3.8 billion in AI-driven revenue but it hopes that tools like SigOpt’s will help drive more activity in that business, dovetailing with the push for more AI applications in a wider range of businesses.

“In the new intelligence era, AI is driving the compute needs of the future. It is even more important for software to automatically extract the best compute performance while scaling AI models,” said Raja Koduri, Intel’s chief architect and senior vice president of its discrete graphics division. “SigOpt’s AI software platform and data science talent will augment Intel software, architecture, product offerings and teams, and provide us with valuable customer insights. We welcome the SigOpt team and its customers to the Intel family.”

While there could potentially be a number of applications for SigOpt’s tech, this is a signal of how bigger players will continue to consolidate specific services around their bigger business, giving the small startup a much bigger horizon in terms of potential business (even if it is all tied to customers that only use Intel hardware).

“We are excited to join Intel and supercharge our mission to accelerate and amplify the impact of modelers everywhere. By combining our AI optimization software with Intel’s decades-long leadership in AI computing and machine learning performance, we will be able to unlock entirely new AI capabilities for modelers,” said Clark in a statement.

Nutrium, a digital health startup which links dietitians and their patients via an app, has raised a €4.25 million Seed round led by Indico Capital Partners, alongside the the Social Innovation Fund in Portugal (SIF) and previous investors. It now offers professional nutrition software to 80,000 nutrition professionals and 800,000 patients in more than 40 countries.

With this investment round, Nutrium plans to double the team size in the next 24 months in order to focus on platform development and expand global sales in markets like Spain, France, Italy, USA and the UK where the company already has a strong customer base.

With the Nutrium platform, patients get integrated nutrition counseling which combines professional advice, continuous monitoring and access to commercial products.

André Santos, CEO and Co-founder of Nutrium commented: “We are moving closer to our vision of enabling the improvement of eating habits for millions of people globally.”

Stephan Morais, managing general partner at Indico said: “Nutrium will become a full-fledged platform bringing together nutritionists, patients, products and wellness data that will enable healthier and happier lives. We are pleased to back this jointly developed vision with capital and knowledge.“

Rui Ferreira, Vice President at Portugal Ventures said: “In 2017, when Portugal Ventures invested in Nutrium’s pre-seed round, the company was mainly present in two markets. Today, Nutrium operates in more than 40 markets, having increased its turnover exponentially.”

Nutrium’s competitors include NutriAdmin, AppointmentPlus, Evolution Nutrition which has raised $2.3M.

The global recovery in venture capital activity did not miss Europe, new data indicates.

According to a PitchBook report, European venture capital activity rose in Q3 2020, putting the continent on pace to set a new yearly record for aggregate VC activity (as measured in Euros).

---

The Exchange explores startups, markets and money. Read it every morning on Extra Crunch, or get The Exchange newsletter every Saturday.

---

The strong results come in the wake of a cracking quarter for venture capital activity in the United States and a generally bullish period for the global VC market. Venture debt is also seeing something of a rebound from lulls seen earlier in the year.

Inside Europe’s Q3 however, was some less-than-good news: the amount of money that went to first-financings was weak, and much of the strong results from the continent were predicated on capital flowing into already-funded startups. There’s less pie for new companies than the top-line numbers might suggest.

Let’s get into the good and bad from Europe’s quarter, contrasting our new data with some prior numbers that we saw when looking into aggregate VC data from Q3.

We’re wrapping up our look at the post-summer venture rebound today, but there’s just a bit more we need to learn before we move on. Let’s get into it.

Europe’s third quarter

Starting with the good news: PitchBook reports that total European venture capital activity came to €10.6 billion in the third quarter of 2020. Per the financial and business data group, it was the third time in history that European venture capital activity crossed the €10 billion mark. (For the sake of comparison, United States-based startups raised around $37 billion, or about €31.5 billion, during the same period.)

Hermeus, a company seeking to build a Mach 5 aircraft that would be capable of making the trip from New York to London in just 90 minutes has raised a $16 million Series A round, led by Canaan Partners and including contributions from existing investors Khosla Ventures, Bling Capital, and the Rise of the Rest Seed Fund. The new funding will help the startup develop and ground test its first full-scale engine, the core component that will eventually power its debut Mach 5 aircraft.

Earlier this year, Hermeus was able to successfully demonstrate a sub-scale engine prototype, showing that the core design of its technology performed as intended. The company now plans to turn that into a version of the engine that matches its eventual production scale and power, while simultaneously expanding the footprint of its Atlanta-based test facility to also include some light in-house manufacturing capability. It’s also going to be working to continue the design of its debut aircraft, and says it will be sharing more info about that first plane over the course of the next few months.

Hermeus says that its target of Mach 5 flight is actually attainable using relatively mature technology already on market, and it cites a team with ample experience across a range of top-flight aerospace companies including SpaceX, Blue Origin, NASA, Boeing and more as another competitive advantage.

Mach 5 is nonetheless ambitious, however; the Concorde flew at speeds of just over Mach 2, and startup Boom Aerospace is targeting Mach 2.2 for its Overture commercial supersonic aircraft. NASA’s X-59 experimental supersonic jet, built by Lockheed Martin, will cruise at a speed of around Mach 1.42. Mach 5 obviously would be quite a bit faster than even the most ambitious of those projects, but Hermeus CEO AJ Piplica has said previously the company expects it to take around a decade of development before they produce a commercial passenger aircraft.

It’s been quite a time for chip industry consolidation, and today Marvell joined the acquisition parade when it announced it is acquiring Inphi in a combination of stock and cash valued at approximately $10 billion, according to the company.

Marvell CEO Matt Murphy believes that by adding Inphi, a chip maker that helps connect internal servers in cloud data centers, and then between data centers, using fibre cabling, it will complement Marvell’s copper-based chip portfolio and give it an edge in developing more future-looking use cases where Inphi shines.

“Our acquisition of Inphi will fuel Marvell’s leadership in the cloud and extend our 5G position over the next decade,” Murphy said in a statement.

In the classic buy versus build calculus, this acquisition uses the company’s cash to push it in new directions without having to build all this new technology. “This highly complementary transaction expands Marvell’s addressable market, strengthens customer base and accelerates Marvell’s leadership in hyperscale cloud data centers and 5G wireless infrastructure,” the company said in a statement.

It’s been a busy time for the chip industry as multiple players are combining hoping for a similar kind of lift that Marvell sees with this deal. In fact, today’s announcement comes in the same week AMD announced it was acquiring Xilinx for $35 billion and follows Nvidia acquiring ARM for $40 billion last month. The three deals combined come to a whopping $85 billion.

There appears to be prevailing wisdom in the industry that by combining forces and using the power of the checkbook, these companies can do more together than they can by themselves.

Certainly Marvell and Inphi are suggesting that. As they highlighted, their combined enterprise value will be more than $40 billion with hundreds of millions of dollars in market potential. All of this of course depends on how well these combined entities work together and we won’t know that for some time.

For what it’s worth, the stock market appears unimpressed with the deal with Marvell’s stock down over 7% in early trading, but Inphi stock is being bolstered in a big way by the announcement, up almost 23% this morning so far.

The deal, which has been approved by both companies’ boards, is expected to close by the second half of 2021 subject to shareholder and regulatory approval.

Walmart over the years has been working to turn its physical retail stores into online fulfillment centers, and now, with its latest set of test stores announced today, the retailer will try out ideas to make that transition more seamless. Walmart says it will deploy personnel to four test stores across the U.S., where they’ll prototype and iterate on new technology and tools that will serve the needs of Walmart’s in-store shoppers and online shoppers alike, including changes involving augmented reality, handheld mobile devices, new apps, in-store signage, omni-assortment, and revamped checkout stations.

The idea is to turn these four test locations into rapid prototyping environments, where teams can test solutions in real-time, make changes, scale what works and scrap what doesn’t. Some of the changes being put into place will be visible to the customer, while others will be more behind-the-scenes.

At launch, Walmart has identified four areas where it’s looking to test new ideas across assortment, inventory, picking and checkout process.

In one store, it will test moving the majority of the in-store apparel assortment online — meaning the same exact items can be found both in the store and online. This isn’t always the case today, as not everything stocked in the stores are also on the Walmart website, and vice versa. This test will focus on determining what has to take place to make all the eligible items in a store “omni-available,” Walmart says, a reference to its desire to be a true “omni-channel” retailer.

A second test will involve a new app that aims to speed up the time it takes to get items from the back room to the sales floor, using augmented reality (AR). In this test, instead of scanning the barcode on boxes that are ready to go, the app will use AR technology to highlight those boxes. The hope is that this will help to move the product to shelves, and in front of customers, faster than before.

Another experiment uses a combination of handheld devices and in-store signage to help associates better navigate to the right locations when picking items for online orders. In early tests, Walmart says the percentage of time it takes associates to find the items has already gone up by 20% in some of the categories that tend to be more difficult to find.

The fourth test will expand and build on an experimental checkout experience Walmart previously announced in June. In this store, Walmart does away with individual checkout lanes, and transitions cashiers into the role of “hosts” in a new area of the store that resembles a self-checkout destination. Here, customers can opt to check out themselves or have a “host” offer full-service checkout. In either case, store staff are around to help with any issues that arise.

The expectation is that checkout lanes will move more quickly than the old style of individual checkout lanes. With the latter layout, a surge of new customers coming to the registers could cause bottlenecks if there weren’t enough lanes staffed. In the long run, the new layout could free up cashiers to help with other tasks in store as a checkout station may not need as many “hosts” on hand to run things.

The four stores may test other technology and digital solutions in the future, as well, but Walmart didn’t expand on its roadmap plans. Two of the stores, including a Bentonville location, are up and running. Two more are planned to be up and running soon.

“We’re moving quickly to use our physical retail stores to not only serve in-store shoppers, but to flex to meet the needs of online shoppers, too, in ways that only Walmart can,” said John Crecelius, Walmart U.S. SVP of Associate Product and Next Generation Stores, in a statement. “That’s where our new test stores come in. Their purpose is to find solutions that continue to help our stores operate as both physical shopping destinations and online fulfillment centers in a way that has yet to be seen across the retail industry,” he added. 

CoreCare, a provider of revenue management services for healthcare companies dealing with public health benefit providers, has raised $3 million in a seed financing round.

The company, which uses machine learning, automates large swaths of billing and revenue cycle management to reduce the burden on hospitals, according to chief executive, Dennis Antonelos.

Already, companies like Creative Solutions in Healthcare, a nursing facility operator in Texas, which operates nearly 80 locations has signed up for the service.

Antonelos started the company in January, had the first product up by March and was accepted to Y Combinator in April. It now boasts over a dozen customers in Texas.

With the new $3 million in hand from investors including Primetime Partners, Goat Capital, Funders Club and Liquid2Ventures, Antonelos said the company would look to expand its sales and marketing and product capabilities.

CoreCare automates processing of billing and paperwork and clinical notes by linking electronic health records and medicare and medicaid information services and payers.

“We’re going through the organization and eliminating administrative waste so the organization can invest newly found resources into patient care,” Antonelos said.

The company uses a standard software as a service payment model and charges somewhere between $300 to $500 per-facility, per-month, according to Antonelos.

“These initial results are outstanding,” said Gary Blake, president, and co-founder of Creative Solutions in Healthcare, and one of CoreCare’s early customers. “In only a matter of months working with CoreCare’s CoreAccess software, we’ve seen a notable impact on our financial position. It has truly exceeded our expectations. CoreCare has changed the way we work with Managed Care, from top to bottom. We have been able to streamline our entire billing process, reduce admin costs, shorten the number of accounts receivable (AR) days and free up cash for growth. Every healthcare provider that works with managed care should work with CoreCare.”

Fintech startup Wise has raised a $12 million Series A round. The company offers business bank accounts with an interesting go-to-market strategy. Wise partners with other companies so that they can offer bank accounts to their own customers.

For instance, if you’re running a marketplace or an e-commerce platform that matches companies with individual customers, you can leverage Wise to offer bank accounts to your partner companies. RemoteTeam is using Wise to improve its payroll experience for… remote teams.

e.ventures is leading today’s funding round with Grishin Robotics also participating. Seed investors Base10 Partners and Techstars are also investing again.

Wise isn’t a classic bank-as-a-service company as it doesn’t want to power neobanks and help them get started. Instead, the startup targets other companies that touch on financial services but can’t offer those services because it’s such a big investment.

Integrating Wise in your product doesn’t require significant development or regulation efforts. You don’t have to develop an entire banking user interface as you can just redirect your customers to Wise. The fintech startup also handles know-your-customer and know-your-business (KYC and KYB) processes.

When your clients have their own Wise accounts, it lets them do all the basic things you’d expect from a business bank account. You can hold money, pay with bank transfers, a debit card, a virtual card or checks, and get paid using card payments, ACH and checks.

Behind the scene, BBVA provides banking services, which means that your deposits are FDIC insured up to $250,000. The company also uses Stripe for some features and other infrastructure companies.

Wise co-founder and CEO Arjun Thyagarajan describes those partners as building blocks. The company can swap those partners and integrate with other APIs to launch in new countries for instance.

Interestingly, if you choose to offer Wise bank accounts to your partners, you’ll share the revenue on deposits and interchange fees.

Up next, the company plans to expand to other countries, such as Canada. It’ll also try to tackle specific verticals, such as marketplaces for telemedicine and healthcare startups in general. It could require adding different features for different types of customers.

Wise is also negotiating some partnerships with high-profile companies, which should bring new customers to the platform.

European lawmakers are taking aim at big tech’s ability to push its own services in search results at the expense of rivals, with Commission EVP Margrethe Vestager confirming today that a legislative proposal due in a few weeks will aim to ban what she called “unfair self-preferencing”.

The concern is that so-called gatekeeper platforms have the ability to manipulate the way that they rank different businesses — and “show their own services more visibly than their rivals”, she said in a speech.

The Commission is expected to propose a package of legislative measures next month to update long-standing EU ecommerce rules and propose new strictures for platforms with significant market power (aka gatekeepers) — making good on its earlier pledge to reboot digital regulation.

In her speech to the EPC Digital Clearinghouse today, Vestager confirmed that the Digital Services Act (DSA) and Digital Markets Act (DMA) will be introduced in a few weeks’ time.

The Commission is surely enjoying its timing, here, with grumblings of political discontent against big tech over the pond and the US Department of Justice having just filed an antitrust case against Google. Although the EU executive’s proposals for reworking digital rules have been years in the making.

Vestager said the DSA will update the existing E-Commerce Directive — by requiring digital services to “take more responsibility for dealing with illegal content and dangerous products”, including by standardizing processes for reporting illegal content and dealing with content reports and complaints.

“Those new responsibilities will help to keep Europeans just as safe online as they are in the physical world. They’ll protect legitimate businesses, which follow the rules, from being undercut by others who sell cheap, dangerous products. And by applying the same standards, all over Europe, they’ll make sure every European can rely on the same protection – and that digital businesses of all sizes can easily operate throughout Europe, without having to meet the costs of complying with different rules in different EU countries,” said Vestager.

She also confirmed increased transparency requirements would be in the package — such as related to content takedowns and recommendations; and also disclosures for online ads, including both who’s paying for an ad and “why we’ve been targeted by a certain ad”.

The DMA proposal will have two components, per Vestager: A “clear list of dos and don’ts” for “big digital gatekeepers”, which she said “will be based on our experience with the sorts of behaviour that can stop markets working well”; and a “harmonised market investigation framework” that will span the EU’s single market — giving the executive the power to preemptively intervene in digital markets to address structural problems before they become entrenched and lead to baked in Internet monopolies.

Recent press reports have suggested that the list of dos and don’ts that’s coming down the pipe for big tech could be lengthy — although the final detail remains to be seen.

But a ban on some forms of self-preferencing will certainly be on that list.

Google’s preferencing of its own services in search results has been on the European Commission’s antitrust radar for years — with a multi-year investigation into its Shopping search comparison service culminating in a $2.7BN fine in 2017 and an order to Google to cease abusive self-preferencing. Despite that action rival price comparison services have continued to complain it’s still not playing fair. Hence the Commission deciding more needs to be done now.

Another restriction Vestager confirmed affected major dual marketplaces — which are set to face future EU controls is on how they can use third party sellers’ data. She argued that the asymmetry of platforms both having access to sellers’ data and competing against those third parties in other markets “can seriously damage fairness” — saying the proposal “aims to ban big gatekeepers from misusing their business users’ data in that way”.

Again it’s an issue that’s been on the Vestager’s radar for some time. Last year, for example, the Commission opened a formal investigation into ecommerce giant Amazon’s use of merchant data (although that probe remains ongoing).

The other core plank of the DMA involves reform of digital competition rules, as EU lawmakers look to evolve the regulatory toolbox to keep pace with digital business.

“We face a constant risk that big companies will succeed in pushing markets to a tipping point, sending them on a rapid, unstoppable slide towards monopoly — and creating yet another powerful gatekeeper,” said Vestager, explaining the push for a harmonised set of rules to tackle structural problems in digital markets across the EU.

The risk of leaving it to EU Member States’ national competition authorities to tackle such issues is “a fragmented system, with different rules in different EU countries”, she went on, adding: “We’ve come to a point where we have to take action. A point where the power of digital businesses – especially the biggest gatekeepers – threatens our freedoms, our opportunities, even our democracy. And where the trust that successful digitisation relies on is becoming seriously frayed.”

The message to tech giants from the EU’s executive is an unwavering “things are going to have to change” — with enforced responsibility coming down the pipe to replace patchy self-regulation.

Vestager also made it clear the Commission is paying attention to how the future rebooted digital rules will be enforced — which is a key point given how a lack of uniformly vigorous enforcement has taken some of the shine off the EU’s rebooted data protection framework (because decision powers are held at the Member State level).

The commissioner said “effective enforcement” will be a vital component of the DSA package, arguing that: “To really give people trust in the digital world, having the right rules in place isn’t enough. People also need to know that those rules really work – that even the biggest companies will actually do what they’re supposed to. And to make sure that happens, there’s no substitute for effective enforcement.”

This means the package will include measures aimed at improving the way national authorities cooperate — “to make sure the rules are properly enforced, throughout the EU”, as she put it.

“Our proposal won’t change the fundamental principle, that digital services should be regulated by their home country. But it will set up a permanent system of cooperation that will help those regulators work more effectively, to protect consumers all across Europe. And it will give the EU power to step in, when we need to, to enforce the rules against some very large platforms,” she added.

The Commission is also clearly banking on the DMA as its key enforcement lever against big tech’s market-denting bulk — by being able to intervene proactively as a way to foster and sustain competition.

And with anger at big tech riding high across Europe the Commission likely feels confident in getting bu-in from EU Member States’ representatives on the EU Council and the elected members of the European Parliament — support that it’ll need to get its legislation proposals across the line.

 

Honeywell, which was a bit of a surprise entrant into the quantum computing space when it announced its efforts to build the world’s most powerful quantum computer earlier this year, today announced its newest system: the Model H1. The H1 uses trapped-ion technology and features 10 fully connected qubits that allow it to reach a quantum volume of 128 (where quantum volume (QV) is a metric of the overall compute power of a quantum computer, no matter the underlying technology). That’s higher than comparable efforts by IBM, but also well behind the QV 4,000,000 machine IonQ says it was able to achieve with 32 qubits.

The H1 will be available to enterprises through the Azure Quantum platform and the company says that it is partnering with Zapata Computing and Cambridge Quantum Computing on this project.

When it first announced its efforts, Honeywell said that its experience in building control systems allowed it to build an advanced ion trap and more uniform qubits that hence make error correction easier.

In addition to the next generation of its quantum computer, the company also today announced its overall quantum roadmap for the next ten years. The plan here is to go from 10 to 40 qubits with all-to-all connectivity as it moves toward a next generation of devices that are fault tolerant and can be deployed at a larger scale.

“Honeywell’s aggressive quantum computing roadmap reflects our commitment to achieving commercial scale for our quantum business. Our subscription-based model provides enterprise customers with access to Honeywell’s most advanced system available,” said Tony Uttley, President of Honeywell Quantum Solutions. “Honeywell’s unique methodology enables us to systematically and continuously ‘upgrade’ the H1 generation of systems through increased qubit count, even higher fidelities and unique feature modifications.”