Category: UNCATEGORIZED

Today Juniper Networks announced it was acquiring smart wide area networking startup 128 Technology for $450 million.

This marks the second AI-fueled networking company Juniper has acquired in the last year and a half after purchasing Mist Systems in March 2019 for $405 million. With 128 Technology, the company gets more AI SD-WAN technology. SD-WAN is short for software-defined wide area networks, which means networks that cover a wide geographical area such as satellite offices, rather than a network in a defined space.

Today, instead of having simply software-defined networking, the newer systems use artificial intelligence to help automate session and policy details as needed, rather than dealing with static policies, which might not fit every situation perfectly.

Writing in a company blog post announcing the deal, executive vice president and chief product officer Manoj Leelanivas sees 128 Technology adding great flexibility to the portfolio as it tries to transition from legacy networking approaches to modern ones driven by AI, especially in conjunction with the Mist purchase.

“Combining 128 Technology’s groundbreaking software with Juniper SD-WAN, WAN Assurance and Marvis Virtual Network Assistant (driven by Mist AI) gives customers the clearest and quickest path to full AI-driven WAN operations — from initial configuration to ongoing AIOps, including customizable service levels (down to the individual user), simple policy enforcement, proactive anomaly detection, fault isolation with recommended corrective actions, self-driving network operations and AI-driven support,” Leelanivas wrote in the blog post.

128 Technologies was founded in 2014 and raised over $97 million, according to Crunchbase data. Its most recent round was a $30 million Series D investment in September 2019 led by G20 Ventures and The Perkins Fund.

In addition to the $450 million, Juniper has asked 128 Technology to issue retention stock bonuses to encourage the startup’s employees to stay on during the transition to the new owners. Juniper has promised to honor this stock under the terms of the deal. The deal is expected to close in Juniper’s fiscal fourth quarter subject to normal regulatory review.

Relativity Space has bagged its first public government contract, and with a major defense contractor at that. The launch startup’s 3D-printed rockets are a great match for a particularly complex mission Lockheed is undertaking for NASA’s Tipping Point program.

The mission is a test of a dozen different cryogenic fluid management systems, including liquid hydrogen, which is a very difficult substance to work with indeed. The tests will take place on a single craft in orbit, which means it will be a particularly complicated one to design and accommodate.

The payload itself and its cryogenic systems will be designed and built by Lockheed and their partners at NASA, of course, but the company will need to work closely with its launch provider during development and especially in the leadup to the actual launch.

Relativity founder and CEO Tim Ellis explained that the company’s approach of 3D printing the entire rocket top to bottom is especially well suited for this.

“We’re building a custom payload fairing that has specific payload loading interfaces they need, custom fittings and adapters,” he said. “It still needs to be smooth, of course — to a lay person it will look like a normal rocket,” he added.

Every fairing (the external part of the launch vehicle covering the payload) is necessarily custom, but this one much more so. The delicacy of having a dozen cryogenic operations being loaded up and tested until moments before launch necessitates a number of modifications that, in other days, would result in a massive increase in manufacturing complexity.

“If you look at the manufacturing tools being used today, they’re not much different from the last 60 years,” Ellis explained. “It’s fixed tooling, giant machines that look impressive but only make one shape or one object that’s been designed by hand. And it’ll take 12-24 months to make it.”

Not so with Relativity.

“With our 3D printed approach we can print the entire fairing in under 30 days,” Ellis said. “It’s also software defined, so we can just change the file to change the dimensions and shape. For this particular object we have some custom features that we’re able to do more quickly and adapt. Even though the mission is three years out, there will always be last minute changes as you get closer to launch, and we can accommodate that. Otherwise you’d have to lock in the design now.”

Ellis was excited about the opportunity to publicly take on a mission with such a major contractor. These enormous companies field billions of government dollars and take part in many launches, so it’s important to be in their good books, or at least in their rolodexes. A mission like this, complex but comparatively low stakes (compared with a crewed launch or billion-dollar satellite) is a great chance for a company like Relativity to show its capabilities. (Having presold many of its launches already, there’s clearly no lack of interest in the 3D printed launch vehicles, but more is always better.)

The company will be going to space before then, though, if all continues to go according to plan. The first orbital test flight is scheduled for late 2021. “We’re actually printing the launch hardware right now, the last few weeks,” Ellis mentioned.

The NASA Tipping Point program that is funding Lockheed with an $89.7 million contract for this experiment is one intended to, as its name indicates, help tip promising technologies over the edge into commercial viability. With hundreds of millions awarded yearly for companies pursuing things like lunar hoppers and robotic arms, it’s a bit like the agency’s venture fund.

Disney announced earlier this month that it’s going all-in on streaming media.

As part of this new strategy, the company is undergoing a major reorganisation of its media and entertainment business that will focus on developing productions that will debut on its streaming and broadcast services.

This will include merging the company’s media businesses, ads and distribution, and Disney+ divisions so that they’ll now operate under the same business unit.

As TechCrunch’s Jonathan Shieber reports, Disney’s announcement follows a significant change to its release schedule to address new realities, including a collapsing theatrical release business; production issues; and the runaway success of its Disney+ streaming service — all caused or accelerated by the national failure to effectively address the COVID-19 pandemic.

So what better time than now to give Disney+ the Extra Crunch user experience teardown treatment. With the help of Built for Mars founder and UX expert Peter Ramsey, we highlight some of the things Disney+ gets right and things that should be fixed. They include zero distractions while signing up, “the power of percentages,” and the importance of designing for trackpad, mouse and touch outside of native applications.

Zero distractions while signing up

If the user is trying to complete a very specific task — such as making a payment — don’t distract them. They’re experiencing event-driven behaviour.

The win: Disney have almost entirely removed any kind of distractions when signing up. This includes the header and footer. They want you to stay on-task.

Image Credits: Disney+

Steve O’Hear: This seems like a very easy win but one we don’t see as often as perhaps we should. Am I right that most sign-up flows aren’t this distraction-free and why do you think that is?

Peter Ramsey: Yeah, it’s such an easy win. Sometimes you see sign-up screens that have Google Adwords on it, and I think, “You’re risking the user getting distracted and leaving for what, half a penny?” If I had to guess why more companies don’t utilise this technique, it’s probably just because they don’t want to deal with the technical hassle of hiding a bunch of elements.

The power of percentages

Only use percentages when it makes sense. 80% off sounds like a lot, but 3% doesn’t. Percentages can be a great way of making a discount seem larger than it actually is, but sometimes it can have the reverse effect. This is because people are generally bad at accurately estimating discounts. “What’s 13% off £78?”

The fail: If you sign up to a year of Disney+, then you’re offered 16% free. But 16% of a £60 bundle isn’t easy to calculate in your head — so people guess. And sometimes, their guesses may be less than the actual value of the discount.

The fix: In this instance, it would be far more compelling (and require less mental arithmetic), if it was marketed as “60 days free.” Sixty days is both easy to understand and easy to assign value to.

Image Credits: Disney+

Percentages may be harder to process or evaluate in isolation as an end user but they are easy to compare with each other i.e., we all know 25% off is better than 10% off. Aren’t you advocating obscuring the actual saving in favour of what sounds better on a case-by-case basis and therefore actually working against the end user? Of course I’m playing devils advocate a little here.

So, it’s actually a really complex dilemma, and there’s no “easy” answer — this would probably make a great dinner time conversation. Yes, if you’re offering two discounts, then a percentage may be the easiest way for people to compare them.

Six Russian intelligence officers accused of launching some of the “world’s most destructive malware” — including an attack that took down the Ukraine power grid in December 2015 and the NotPetya gloibal ransomware attack in 2017 — have been charged by the U.S. Justice Department.

Prosecutors said the group of hackers, who work for the Russian GRU and reside in Russia, are behind the “most disruptive and destructive series of computer attacks ever attributed to a single group.”

“No country has weaponized its cyber capabilities as maliciously or irresponsibly as Russia, wantonly causing unprecedented damage to pursue small tactical advantages and to satisfy fits of spite,” said John Demers, U.S. U.S. assistant attorney general for national security. “Today the Department has charged these Russian officers with conducting the most disruptive and destructive series of computer attacks ever attributed to a single group, including by unleashing the NotPetya malware. No nation will recapture greatness while behaving in this way.”

The six accused Russian intelligence officers. (Image: FBI/supplied)

In charges laid out Monday, the hackers are accused of developing and launching attacks using the KillDisk and Industroyer (also known as Crash Override) to target and disrupt the power supply in Ukraine, which left hundreds of thousands of customers without electricity two days before Christmas. The prosecutors also said the hackers were behind the NotPetya attack, a ransomware attack that spread across the world in 2017, causing billions of dollars in damages.

The hackers are also said to have used Olympic Destroyer, designed to knock out internet connections during the opening ceremony of the 2018 PyeongChang Winter Olympics in South Korea.

Prosecutors also blamed the six hackers for trying to disrupt the 2017 French elections by launching a “hack and leak” operation to discredit the then-presidential frontrunner, Emmanuel Macron, as well as launching targeted spearphishing attacks against the Organisation for the Prohibition of Chemical Weapons and the U.K.’s Defence Science and Technology Laboratory, tasked with investigating the use of the Russian nerve agent Novichok in Salisbury, U.K. in 2018.

The six Russians are all charged with seven counts of conspiracy to hack, commit wire fraud, and causing computer damage.

Remember back in March when the VC game was done for the year, checkbooks were snapping shut and startup layoffs led the headlines? So much for all that. Q3’s venture capital numbers are in and they are anything but weak.

In retrospect, the Q2 VC slowdown looks more like a short-lived recharge ahead of a big push in Q3 than anything existential. We can see this today through the lens of data concerning what happened after June concluded and we moved into Q3.

According to data from PitchBook (data source) and CBInsights (data source), there was a lot to like about the third quarter if you were a U.S.-based startup.

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

I want to dig into the data and pull out most important data points for you. We’ll get you informed and out the door in around 900 words.

If you want a more global look at the venture capital world in Q3, don’t worry. We’re doing that tomorrow right here at The Exchange. Ready? This should be both fun and informative. Let’s go!

A massive third quarter

To get a clear look at the U.S. venture capital market, we’ll start from the top down. So, the biggest numbers first, followed by increasingly narrow slices of data so we can drill down into smaller startups.

First, the top-line numbers:

• How much money was raised by U.S.-based startups in Q3 2020? $36.5 billion, according to CBInsights, $37.8 billion according to PitchBook. Those numbers are effectively the same for purposes. CBInsights calls the number a seven-quarter high, up 22% from the Q3 2019 number and 30% from the Q2 2020 result. PitchBook agrees that Q3 2020 was strong, but has its count just under Q2 2020’s own.
• How many deals was that money spread between? CBInsights counts 1,461 VC deals in Q3 2020 for U.S.-based startups. Per its numbers, that figure is up 1% from Q2 2020 and down 11% from Q3 2019. PitchBook, in contrast, counts 2,990 total deals, inclusive of rounds that it expects to be added as information about the quarter fills in. That tally “held steady” compared to Q3 2019, per the company.

What to make of all this information? Simple: Q3 2020 U.S.-based startup venture capital dollar volume was very strong, with deal counts coming in slightly weaker.

This means that we saw fewer, larger deals in the quarter on average, right? Let’s see:

The Raspberry Pi Foundation is launching a new product today — the Compute Module 4. If you’ve been keeping an eye on the Raspberry Pi releases, you know that the flagship Raspberry Pi 4 was released in June 2019. The Compute Module 4 features the same processor, but packed in a compute module for industrial use cases.

A traditional Raspberry Pi is a single-board computer with a ton of ports sticking out. Compute Modules are somewhat different. Those system-on-module variants are more compact single-board computers without any traditional port.

It lets you create a prototype using a traditional Raspberry Pi, and then order a bunch of Compute Modules to embed in your commercial products. “Over half of the seven million Raspberry Pi units we sell each year go into industrial and commercial applications, from digital signage to thin clients to process automation,” Eben Upton wrote on the Raspberry Pi blog.

Some things are strictly similar between the Raspberry Pi 4 and the Compute Module 4, such as the 64-bit ARM-based processor with VideoCore VI graphics. This is going to represent a huge upgrade for previous Compute Module customers.

In particular, you get much better video performance with 4Kp60 hardware decode for H.265 videos, 1080p60 hardware decode for H.264 videos, 1080p30 hardware encode of H.264 videos. You can also take advantage of the dual HDMI interfaces to connect up to two 4K displays at 60 frames per second.

Another big change with the Compute Module 4 is that there are a ton of options. You can choose compute modules with or without wireless technologies (Wi-Fi and Bluetooth), with 1GB, 2GB, 4GB or 8GB of RAM, with 8GB, 16GB or 32GB of eMMC flash storage. There’s also a model without any eMMC flash storage in case you want to use external eMMC or the SD card interface.

You can mix-and-match those specs to keep your costs down at scale. The result is that there are 32 different versions of the Compute Module 4 ranging from $25 (no wireless, 1GB of RAM, ‘Lite’ eMMC) to $90 (wireless, 8GB of RAM, 32GB of eMMC).

The form factor has changed compared to the previous Compute Module, which means that you’ll need a new Compute Module IO Board to take advantage of all the interfaces and start developing. It costs $35.

Image Credits: Raspberry Pi Foundation

SiriusXM today completed its previously announced $325 million acquisition of podcast platform Stitcher from E.W. Scripps, and has now launched Stitcher’s podcasts on Pandora across all tiers of the streaming service. The deal brings top Stitcher titles to Pandora, including Freakonomics Radio, My Favorite Murder, SuperSoul Conversations from the Oprah Winfrey Network, Office Ladies, Conan O’Brien Needs a Friend, Literally! with Rob Lowe, LeVar Burton Reads, and WTF with Marc Maron, among others.

On Pandora, the podcasts will be indexed using the company’s proprietary Podcast Genome Project technology. This system leverages  automated technology — like natural language processing, collaborative filtering, and other machine learning approaches — then combines that with human curation to make personalized recommendations to podcast listeners on Pandora’s app.

The podcasts will also continue to be available in the Stitcher app in North America, the company says.

The Stitcher acquisition brought with it several key assets, including its own mobile listening app, which includes a premium tier of exclusives, and the Midroll Media network for podcast advertising. Stitcher also creates its own original programs and runs multiple content networks, via Earwolf.

That means SirusXM gained thousands of top podcasts with the deal’s closure. The company also now claims it has the “largest addressable audience in North America” across all categories of digital audio, including music, sports, talk, and podcasts thanks to the combination of satellite radio service SiriusXM, streaming app Pandora, and now Stitcher.

The company believes the deal will help it to attract more creators to its platform, thanks to the enhanced production, marketing, and distribution capabilities it offers, following the deal’s close. Advertisers, meanwhile, will be able to more precisely target podcasts for better ad efficiency, and will gain access to improved measurements, says SiriusXM.

In terms of Stitcher’s execs, CEO Erik Diehn will now report to Scott Greenstein, President and Chief Content Officer of SiriusXM, who also oversees content at Pandora. Stitcher’s Chief Revenue Officer, Sarah van Mosel, will report directly to John Trimble, Chief Advertising Revenue Officer of SiriusXM.

“We are deepening our position in podcasting, the fastest-growing sector in digital audio, and with completion of this transaction, our vision is taking shape,” said SiriusXM CEO Jim Meyer, in a statement about the deal’s completion. “With Stitcher and its varied assets, we are now a one-stop shop able to meet the needs of podcast creators, publishers and advertisers, while also providing listeners with access to great shows, series and programming.”

Despite the coronavirus pandemic, which disrupted many consumer trends and accelerated others, podcasting still remains one of the fast-growing digital audio industries. Podcast downloads returned to pre-COVID levels this summer, and Spotify reported that podcast consumption more doubled in Q2 and nearly a quarter (21%) of its active users now listen to podcasts.

Stitcher was not SiriusXM’s first acquisition focused on podcasts or ad technologies. It also bought podcast management platform Simplecast this June, and before that, it acquired AdsWizz for $66.3 million to power Pandora’s advertising efforts.

The European Union has switched on cross-border interoperability for a first batch of COVID-19 contacts tracing apps that use Bluetooth proximity to calculate the exposure risk of smartphone users after a pilot of the system last month.

National apps whose backends are now linked through the gateway service are Germany’s Corona-Warn-App, the Republic of Ireland’s COVID tracker, and Italy’s immuni app.

This means a user of one of those apps who travels to any of the other countries can expect their national app to send relevant exposure notifications in the same way it should if they had not travelled — without the need to download any additional software.

Collectively, the three national COVID-19 apps have been downloaded by around 30 million people which the EU said corresponds to two-thirds of such downloads in the region.

Image credit: EU Publications Office

Other national apps are expected to gain interoperability as they are added to the service in the coming weeks — with at least 18 more compatible national apps identified at this stage.

A second batch of national apps is expected to be added next week after a period of testing — namely: Czechia’s eRouška, Denmark’s smitte stop, Latvia’s Apturi COVID and Spain’s Radar Covid (although the latter still doesn’t have full coverage in Spain with the Catalonia region yet to integrate it with its regional healthcare system). Further compatible apps are slated to be added in November.

The gateway has been designed to work, in the first instance, with official coronavirus apps that have a decentralized architecture — meaning any that use a centalized architecture, such as France’s StopCovid app, aren’t currently compatible.

The UK’s collection of apps, meanwhile — for England & Wales, Scotland and Northern Ireland — are unlikely to get plugged in, despite having a technically compatible app architecture, as the country is due to exit the trading bloc at the end of this year. (So interoperability would require a separate agreement between the UK and the EU.)

“About two third of EU Member States have developed compatible tracing and warning apps, and the gateway is open to all of them, once they are ready to connect. The connection will gradually take place during October and November, however apps can also connect at a later stage if national authorities wish so. An ‘onboarding protocol’ has been developed, setting out the necessary steps,” the Commission notes in an Q&A.

The cross-border system for the EU’s apps works via the use of a gateway server, developed and set up by T-Systems and SAP and operated from the Commission’s data centre in Luxembourg, which receives and passes on arbitrary identifiers between national apps.

“No other information than arbitrary keys, generated by the apps, will be handled by the gateway,” the EU notes in a press release. “The information is pseudonymised, encrypted, kept to the minimium, and only stored as long as necessary to trace back infections. It does not allow the identification of individual persons, nor to track location or movement of devices.”

Getting a cross-border system up and running so swiftly across a patchwork of national COVID-19 apps is an achievement for the EU, even as there are ongoing questions about the utility of Bluetooth-based coronavirus exposure notifications in the fight against the spread of the novel coronavirus — with much of Europe now experiencing a second wave of the pandemic.

However EU commissioners suggested today that such apps can be a useful complement to other measures, such as manual contact tracing.

Commenting in a statement, StellaKyriakides, EU commissioner for health and food safety, said: “Coronavirus tracing and warning apps can effectively complement other measures like increased testing and manual contact tracing. With cases on the rise again, they can play an important role to help us break the transmission chains. When working across borders these apps are even more powerful tools. Our gateway system going live today is an important step in our work, and I would call on citizens to make use of such apps, to help protecting each other.”

“Free movement is an integral part of the Single Market — the gateway is facilitating this while helping save lives,” added Thierry Breton, commissioner for the internal market.

Before the 2016 election, Vice Ventures founder and general partner Catharine Dockery was bullish about the future of recreational cannabis in the United States.

“We saw quite a bit more optimism around national legalization, with the feeling that a wave of states legalizing recreational use would be the final push needed” to see drug reform, she said. It was good news for Dockery, who was planning to launch a firm investing in categories like cannabis, CBD, psychedelics and sextech.

She announced a $25 million fund in June 2019, but the national policy landscape had shifted considerably.

“The vitriol and division around the election really haven’t left room for substantive discussions. I think this will eventually change, but don’t have high hopes for much policy debate until the election is complete, if at all,” she said. “In a time of uncertainty, we’re taking a small step back.”

Along with many VC firms, Vice Ventures has raised the bar regarding which startups it will fund, but several investors told TechCrunch they were split about how they’re making decisions in the closing days of the presidential campaign. After a booming summer, some said momentum is increasing, while others told us that expectations have never been higher for startups.

“If anything, the pace is increasing,” said Alexa Von Tobel of Inspired Capital. Traditionally, she said founders scale back on fundraising efforts close to the winter holidays because investors’ vacation mentality is kicking in. This year, “I think we’ll continue to see founders taking advantage of the ample flow of capital right now and shore up resources so they can enter 2021 on strong footing,” she said.

While that may be good news for founders, Von Tobel said Inspired Capital is not giving too much weight to the election internally.

“We think of ourselves as patient capital, focused on looking for the best companies no matter the timing,” she said. “While we know the election will create noise and have an impact on businesses long-term, it does not have a place in our process right now.”

Inspired Capital invests more broadly in the early-stage environment, which plays a part in its ability to invest through crises and turbulence. It seems that firms that have more niche investment theses have been more likely to change their pace ahead of the election.

This has been a long time coming, but the OpenStack foundation today announced that it is changing its name to ‘Open Infrastructure Foundation,” starting in 2021.

The announcement, which the foundation made at its virtual developer conference, doesn’t exactly come as a surprise. Over the course of the last few years, the organization started adding new projects that went well beyond the core OpenStack project and renamed its conference to the ‘Open Infrastructure Summit.’ The organization actually filed for the ‘Open Infrastructure Foundation’ trademark back in April.

Image Credits: OpenStack Foundation

After years of hype, the open-source OpenStack project hit a bit of a wall in 2016, as the market started to consolidate. The project itself, which helps enterprises run their private cloud, found its niche in the telecom space, though, and continues to thrive as one of the world’s most active open-source projects. Indeed, I regularly hear from OpenStack vendors that they are now seeing record sales numbers — despite the lack of hype. With the project being stable, though, the Foundation started casting a wider net and added additional projects like the popular Kata Containers runtime and CI/CD platform Zuul.

“We are officially transitioning and becoming the Open Infrastructure Foundation,” long-term OpenStack Foundation executive president Jonathan Bryce told me. “That is something that I think is an awesome step that’s built on the success that our community has spawned both within projects like OpenStack, but also as a movement […], which is [about] how do you give people choice and control as they build out digital infrastructure? And that is, I think, an awesome mission to have. And that’s what we are recognizing and acknowledging and setting up for another decade of doing that together with our great community.”

In many ways, it’s been more of a surprise that the organization waited as long as it did. As the foundation’s COO Mark Collier told me, the team waited because it wanted to sure that it did this right.

“We really just wanted to make sure that all the stuff we learned when we were building the OpenStack community and with the community — that started with a simple idea of ‘open source should be part of cloud, for infrastructure.’ That idea has just spawned so much more open source than we could have imagined. Of course, OpenStack itself has gotten bigger and more diverse than we could have imagined,” Collier said.

As part of today’s announcement, the group is also adding four new members at Platinum tier, its highest membership level: Ant Group, the Alibaba affiliate behind Alipay, embedded systems specialist Wind River, China’s Fiberhome (which was previously a Gold member) and Facebook Connectivity. To become a Platinum member, companies have to contribute $350,000 per year to the foundation and must have at least 2 full-time employees contributing to its projects.

“If you look at those companies that we have as Platinum members, it’s a pretty broad set of organizations,” Bryce noted. “AT&T, the largest carrier in the world. And then you also have a company Ant, who’s the largest payment processor in the world and a massive financial services company overall — over to Ericsson, that does telco, Wind River, that does defense and manufacturing. And I think that speaks to that everybody needs infrastructure. If we build a community — and we successfully structure these communities to write software with a goal of getting all of that software out into production, I think that creates so much value for so many people: for an ecosystem of vendors and for a great group of users and a lot of developers love working in open source because we work with smart people from all over the world.”

The OpenStack Foundation’s existing members are also on board and Bryce and Collier hinted at several new members who will join soon but didn’t quite get everything in place for today’s announcement.

We can probably expect the new foundation to start adding new projects next year, but it’s worth noting that the OpenStack project continues apace. The latest of the project’s bi-annual releases, dubbed ‘Victoria,’ launched last week, with additional Kubernetes integrations, improved support for various accelerators and more. Nothing will really change for the project now that the foundation is changing its name — though it may end up benefitting from a reenergized and more diverse community that will build out projects at its periphery.