Why Microsoft wants to put data centers at the bottom of the ocean

Earlier this week, Microsoft announced the second phase of Project Natick, a research experiment that aims to understand the benefits and challenges of deploying large-scale data centers under water. In this second phase, the team sank a tank the size of a shipping container with numerous server racks off the coast of the Orkney islands and plans to keep it there for a few years to see if this is a viable way of deploying data centers in the future.

Computers and water famously don’t mix, as anyone who has ever spilled a cup of water over a laptop, so putting server racks under water sure seems like an odd idea. But as Microsoft Research’s Ben Cutler told me, there are good reasons for why the bottom of the ocean may be a good place for setting up servers.

The vast majority of people live within 200 kilometers of the ocean, Cutler noted, and Microsoft’s cloud strategy has long been about putting its data centers close to major population centers. So with large offshore wind farms potentially providing renewable power and the obvious cooling benefits of being under water (and cooling is a major cost factor for data centers), trying an experiment like this makes sense.

“Within Microsoft, we’ve spent an enormous amount of energy and time on cloud — and obviously money,” Cutler explained when I asked him about the genesis of this project. “So we’re always looking for new ways that we can innovate. And this idea sort of gelled originally with one of our employees who worked on a U.S. Navy submarine and knew something about this technology, and that this could maybe be applied to data centers.”

So back in 2013, the team launched phase one and dropped a small pressure vessel with a few servers into the waters of the Pacific Ocean. That experiment worked out pretty well. Even the local sea life seemed to appreciate it. The team found that the vessel didn’t heat up the water close to it by more than a few thousandths of a degree Celsius warmer than a few feet further away from it. The noise, too, was pretty much negligible. “We found that once we were a few meters away from the vessel, we were drowned out by background noise, which is things like snapping shrimp, which is actually the predominant sound of the ocean,” Cutler told me, and stressed that the team’s job is to measure all of this as the ocean is obviously a very sensitive environment. “What we found was that we’re very well received by wildlife and we’re very quickly colonized by crabs and octopus and other things that were in the area.”

For this second phase, the team decided on the location off the coast of Scotland because it’s also home to the European Marine Energy Center, so the infrastructure for powering the vessel from renewable energy from on- and off-shore sources was already in place.

Once the vessel is in the ocean, maintenance is pretty much impossible. The idea here is to accept that things will fail and can’t be replaced. Then, after a few years, the plan is to retrieve the vessel, refurbish it with new machines and deploy it again.

But as part of this experiment, the team also thought about how to best make these servers last as long as possible — and because nobody has to go replace a broken hard drive inside the vessel, the team decided to fill the atmosphere with nitrogen to prevent corrosion, for example. To measure the impact of that experiment, Microsoft also maintains a similar vessel on land so it can compare how well that system fares over time.

Cutler stressed that nothing here is cutting-edge technology. There are no exotic servers here and both underwater cabling and building vessels like this are well understood at this point.

Over time, Cutler envisions a factory that can prefabricate these vessels and ship them to where they are needed. That’s why the vessel is about the size of a shipping container and the team actually had it fabricated in France, loaded it on a truck and shipped it to England to test this logistics chain.

Whether that comes to pass remains to be seen, of course. The team is studying the economics of Natick for the time being, and then it’s up to Microsoft’s Azure team to take this out of the research labs and put it into more widespread production. “Our goal here is to drive this to a point where we understand that the economics make sense and that it has the characteristics that we wanted it to, and then it becomes a tool for that product group to decide whether and where to use it,” said Cutler.


Source: Tech Crunch

Uber and Lyft apply for electric scooter permits in SF

Uber and Lyft have officially put their respective names into the electric scooter competition. Uber and Lyft are among the 11 companies that applied to operate an electric scooter-sharing service within San Francisco city limits. The city, however, will only offer up to five companies permits to operate as part of a one-year test program.

Uber declined to comment, but confirmed that it has applied for a permit via JUMP, the bike-share startup Uber acquired for about $200 million in April. Once Uber is cleared to operate electric scooters, the plan is to integrate them into the Uber app and continue fleshing out Uber CEO Dara Khosrowshahi’s vision for a full-fledged multi-modal transportation platform.

Lyft also confirmed to TechCrunch that the company applied for a permit, but declined to share any further details. Here’s the full list of companies that applied, via the SFMTA:

  1. Bird
  2. CycleHop
  3. JUMP via Uber
  4. Lime
  5. Lyft
  6. ofo
  7. Razor (yes, *that* Razor)
  8. Ridecell
  9. Scoot
  10.  Spin
  11.  USSCooter

San Francisco’s permit process came as a result of Bird, Lime and Spin deploying their electric scooters without permission in the city in March. As part of a new city law, which went into effect June 4, scooter companies are not able to operate their services in San Francisco without a permit. The SFMTA said it’s aiming to notify companies of their permit status by the end of June.

For more information about electric scooter regulation in San Francisco, be sure to check out my previous coverage.


Source: Tech Crunch

IBM and the DoE launch the world’s fastest supercomputer

IBM and the U.S. Department of Energy’s Oak Ridge National Laboratory (ORNL) today unveiled Summit, the department’s newest supercomputer. IBM claims that Summit is currently the world’s “most powerful and smartest scientific supercomputer” with a peak performance of a whopping 200,000 trillion calculations per second. That performance should put it comfortably at the top of the Top 500 supercomputer ranking when the new list is published later this month. That would also mark the first time since 2012 that a U.S.-based supercomputer holds the top spot on that list.

Summit, which has been in the works for a few years now, features 4,608 compute servers with two 22-core IBM Power9 chips and six Nvidia Tesla V100 GPUs each. In total, the system also features over 10 petabytes of memory. Given the presence of the Nvidia GPUs, it’s no surprise that the system is meant to be used for machine learning and deep learning applications, as well as the usual high performance computing workloads for research in energy and advanced materials that you would expect to happen at Oak Ridge.

IBM was the general contractor for Summit and the company collaborated with Nvidia, RedHat and InfiniBand networking specialists Mellanox on delivering the new machine.

Image credit: Oak Ridge National Laboratory

“Summit’s AI-optimized hardware also gives researchers an incredible platform for analyzing massive datasets and creating intelligent software to accelerate the pace of discovery,” said Jeff Nichols, ORNL associate laboratory director for computing and computational sciences, in today’s announcement.

Summit is one of two of these next-generation supercomputers that IBM is building for the DEO. The second one is Sierra, which will be housed at the Lawrence Livermore National Laboratory. Sierra, which is also scheduled to go online this year, is less powerful at an expected 125 petaflops, but both systems are significantly more powerful than any other machine in the DoE’s arsenal right now.


Source: Tech Crunch

Airbnb creates $10M fund to cover cancelled reservations in Japan after regulatory shift

Airbnb has been one of the breakthrough stories in the wave of shared-economy startups that have emerged out of Silicon Valley, with a valuation of $30 billion for its travellers platform that lets people book private homes as accommodations, as well as other services. But even so, it’s not immune to the force of regulation and the impact it can have on its business.

Airbnb has had to cancel a swathe of reservations in Japan, after a change in local laws required hosts to have specific licenses, but some have failed to get these ahead of the deadline set by regulators.

It’s not clear how many people or hosts have been impacted — the numbers are shifting as hosts receive their licenses — but Airbnb says that it has set up a fund of $10 million to cover any travellers who might get put out as a result of the rules. Some have estimated that as much as 80 percent of bookings have been impacted by the changes.

As Airbnb notes, the cancellations and its resulting moves are a result of changes to the country’s Japanese Hotels and Inns Act. Modified last year to include people using private homes for tourist accommodation for up to 180 days/year, those hosting now have to register and display a license number alongside their listings. The tourism authority (JTA) had set a deadline of June 15 to do this, and those who hadn’t received a license by June 1 had to cancel reservations booked before June 15, and Airbnb has extended this to cover the gap of other travellers so that they have time to make alternative arrangements:

“Any reservation scheduled for guest arrival between June 15 and June 19 at a listing in Japan that does not currently have a license has been cancelled,” Airbnb writes. “Going forward, unless the government reverses its position, we will automatically cancel and fully refund any reservations at listings in Japan that have not been licensed within 10 days of guest arrival.”

The $10 million fund, Airbnb said, will cover “additional expenses for guests who are scheduled to travel to Japan and have had their plans interrupted due to a cancellation.” Those whose reservations are cancelled on or after June 15 because of the license situation will get a full refund and a coupon worth “at least 100% of the booking value” to use on a future Airbnb trip. They will also receive a $100 coupon for an Airbnb Experience. 

Those who are unable to find alternative Airbnb places to stay for their trip will be put in touch with a travel agency in Japan, JTB, to find alternatives.

For those who are impacted by this news, Airbnb will be sending you step-by-step instructions of what to do next, or you can find them here.

This is not the first time that Airbnb has had a stumble on the heels of regulatory changes. In Amsterdam, regulators are preparing to halve the number of nights a property can be let out to 30 nights per year starting in 2019, from 60 nights currently. Berlin and Barcelona have also tried to limit the platform’s growth with their own regulations.


Source: Tech Crunch

Workday acquires Rallyteam to fuel machine learning efforts

Sometimes you acquire a company for the assets and sometimes you do it for the talent. Today Workday announced it was buying Rallyteam, a San Francisco startup that helps companies keep talented employees by matching them with more challenging opportunities in-house.

The companies did not share the purchase price or the number of Rallyteam employees who would be joining Workday .

In this case, Workday appears to be acquiring the talent. It wants to take the Rallyteam team and incorporate it into the company’s engineering unit to beef up its machine learning efforts, while taking advantage of the expertise it has built up over the years connecting employees with interesting internal projects.

“With Rallyteam, we gain incredible team members who created a talent mobility platform that uses machine learning to help companies better understand and optimize their workforces by matching a worker’s interests, skills and connections with relevant jobs, projects, tasks and people,” Workday’s Cristina Goldt wrote in a blog post announcing the acquisition.

Rallyteam, which was founded in 2013, and launched at TechCrunch Disrupt San Francisco in September 2014, helps employees find interesting internal projects that might otherwise get outsourced. “I knew there were opportunities that existed [internally] because as a manager, I was constantly outsourcing projects even though I knew there had to be people in the company that could solve this problem,” Rallyteam’s Huan Ho told TechCrunch’s Frederic Lardinois at the launch. Rallyteam was a service designed to solve this issue.

Last fall the company raised $8.6 million led by Norwest Ventures with participation from Storm Ventures, Cornerstone OnDemand and Wilson Sonsini.

Workday provides a SaaS platform for human resources and finance, so the Rallyteam approach fits nicely within the scope of the Workday business. This is the 10th acquisition for Workday and the second this year.

Chart: Crunchbase

Workday raised over $230 million before going public in 2012.


Source: Tech Crunch

Facebook launches Fb.gg gaming video hub to compete with Twitch

Facebook wants a cut of the 3+ hours per week that young adult video gamers spend watching other people play. So today it’s launch Fb.gg – as in the post-competition courtesy of saying “good game” — a destination where viewers can find a collection of all the video game streaming on Facebook. Fb.gg will show video based on the games and streaming celebrities they follow, their Liked Pages and Groups, plus it will display featured creators, esports competitions, and gaming conference events.

Aggregating gaming content could make sure it doesn’t get lost in the fast-moving News Feed. It could be especially helpful for people whose Facebook friends aren’t into the gaming niche. The personalized recommendations based on Facebook activity could help the social network out-curate video-only sites like YouTube and Twitch. And if game streamers feel like they can build a big audience on Facebook, they’ll share there.

Meanwhile, Facebook is opening up its new monetization option to more gaming broadcasters. Facebook is launching the Level Up Program for emerging gaming content creators. Available in the next few months, those with access will be able to take monetary tips from their stream viewers in the form of the Facebook Stars virtual currency that they can redeem for cash. They’ll also get early access to new Facebook livestreaming features and tips on how to build their fan base. Facebook is also rolling out its monthly subscription fan patronage feature test to more gamers in the coming weeks. Gamers can check out the eligibility requirements for these programs here.

Gamers have plenty of options to earn money from YouTube ad revenue shares and Twitch’s tipping options. Facebook needs to ramp these monetization efforts up quickly if it wants to capitalize in the sudden surge in game streaming.


Source: Tech Crunch

Lime brings electric scooters to LA

While electric scooter startups are at a standstill in San Francisco, Lime is taking its scooter service to Santa Monica, Calif. — competitor Bird’s home turf. Lime was planning to launch its new model of scooter that it built in partnership with Segway in San Francisco last month, it’s now debuting them in the Los Angeles area first.

These Segway-powered Lime scooters are designed to be safer, longer-lasting via battery power and more durable for what the sharing economy requires, Lime CEO Toby Sun told TechCrunch in May. Now, instead of a maximum distance of 23 miles or so, Lime scooters can go up to 35 miles.

“A lot of the features in the past on scooters were made for the consumer market,” Sun said. “Not for the shared, heavy-duty markets.”

On the safety side, Lime enhanced its night-light on both the front and back of the scooter, and has added a light to flash below the deck. Lime has also added an additional brake, to have one on both the front and rear wheels.

Lime, which also has its pedal-assist electric bikes out and about in the LA area, says this is the first multimodal transportation service in LA. This news comes following reports of Lime raising a $250 million round led by GV.


Source: Tech Crunch

Facebook alerts 14M to privacy bug that changed status composer to public

Facebook has another privacy screwup on its hands. A bug in May accidentally changed the suggested privacy setting for status updates to public from whatever users had set it to last, potentially causing them to post sensitive friends-only content to the whole world. Facebook is now notifying 14 million people around the world who were potentially impacted by the bug to review their status updates and lock them down tighter if need be.

Facebook’s Chief Privacy Officer Erin Egan wrote to TechCrunch in a statement:

We recently found a bug that automatically suggested posting publicly when some people were creating their Facebook posts. We have fixed this issue and starting today we are letting everyone affected know and asking them to review any posts they made during that time. To be clear, this bug did not impact anything people had posted before – and they could still choose their audience just as they always have. We’d like to apologize for this mistake”.

The bug was active from May 18th to May 27th, with Facebook able start rolling out a fix on May 22nd. It happened because Facebook was building a ‘featured items’ option on your profile that highlights photos and other content. These featured items are publicly visible, but Facebook inadvertently extended that setting to all new posts from those users.

The issue has now been fixed, and everyone’s status composer has been changed back to default to the privacy setting they had before the bug. The notifications about the bug leads to a page of info about the issue, with a link to review affected posts.

Facebook tells TechCrunch that it hears loud and clear that it must be more transparent about its product and privacy settings, especially when it messes up. And it plans to show more of these types of alerts to be forthcoming about any other privacy issues it discovers in the future.

Facebook depends on trust in its privacy features to keep people sharing. If users are worried their personal photos, sensitive status updates, or other content could leak out to the public and embarrass them or damage their reputation, they’ll stay silent.

With all the other issues swirling after the Cambridge Analytica scandal, this bug shows that Facebook’s privacy issues span both poorly thought-out policies and technical oversights. It moved too fast, and it broke something.


Source: Tech Crunch

Bad things happen when you train AI using ‘the darkest corners of Reddit’

MIT researchers trained an artificial intelligence using Reddit and you won’t believe what happened next. Just kidding. Of course you will. The worst things happened.

Norman, who naturally gets his name from the guest-murdering proprietor of the Bates Motel, is the “world’s first psychopath AI,” according to its creators at MIT. First, possibly, but certainly not the last. The creation is a kind of thought experiment designed to explore how the data we use to train machine learning algorithms ultimately influences its behavior.

In this case, the deep, endless well of human misery that is the internet was used to teach poor psychotic Norman the ways of the world.

“Norman suffered from extended exposure to the darkest corners of Reddit,” the researchers state, “and represents a case study on the dangers of artificial intelligence gone wrong when biased data is used in machine learning algorithms.”

In particular, the team used “an infamous subreddit […]that is dedicated to documenting and observing the disturbing reality of death.” That information had a fairly profound impact on Norman’s gig captioning photos (inkblots, in this case), when compared to neural networks trained on more standard data.

Here are a couple of those responses:

You get the gist, right? The rest can be found over here.

“When people say that AI algorithms can be biased and unfair,” the team explains, “the culprit is often not the algorithm itself, but the biased data that was fed to it.”


Source: Tech Crunch

M17 delays IPO debut after pricing this morning on NYSE

M17 Entertainment, a Taipei-based live streaming and dating app group, priced its IPO this morning on the NYSE and was expected to open trading today according to their final press release. But with just a little more than two hours to go before market closing, it’s still not trading, and no one seems to know why.

An interview I had scheduled with the CEO earlier this afternoon was canceled at the last minute, with the company’s representative saying that M17 couldn’t comment since its shares were not yet actively trading, and thus the company remains under an SEC-mandated quiet period.

M17 has had a rocky non-debut so far. Originally targeting a fundraise of $115 million of American Depository Receipts (shares of foreign companies listed domestically on the NYSE), the company concluded its roadshow raising less than half of its target, for a final investment of $60.1 million. The company priced its ADR shares at $8 each, with each ADR representing 8 shares of the stock’s Class A security.

My colleague Jon Russell has covered the company’s rapid growth over the past three years. It was formed from the merger of dating app company Paktor and live-streaming business 17 Media. Joseph Phua, who was CEO of Paktor, became CEO of the joint M17 company following the merger. Together, the two halves have raised tens of millions in venture capital.

M17 provides live-streaming and dating apps throughout “Developed Asia”

The company’s main product is a live-streaming product where creators can build their fan bases and brands. Fans can purchase virtual gifts to send to their favorite artists, and those points are proving to be extraordinarily lucrative for the company. The company, according to its amended F-1 statement, has seen tremendous revenue growth, netting $37.9 million of revenue in the first three months of this year. The company has also been able to attract more live-streaming talent, increasing its contracted artists from 999 at the end of December 2016 to 7,719 at the end of March this year.

That’s where the good news ends for the company. Despite that revenue growth, operating losses are torrential, with the company losing $24.8 million in the first three months of this year. The company in its statement says that it has $31.4 million in cash and cash equivalents, giving it limited runway to continue operations without a strong IPO debut.

User growth has been mostly stagnant. Active monthly users has increased from 1.5 million to 1.7 million between March 31 of 2017 and 2018. What the company has succeeded in doing is monetizing those users much better. The percentage of users paying on the platform has more than doubled over the same time period, and the value of those users has increased more than 40 percent to $355 per user per month.

The big challenge for M17 is revenue quality. Live streaming represents 91.4 percent of the company’s revenues, but those revenues are concentrated on a handful of “whales” who buy a freakishly high number of virtual gifts. The company’s top 10 users represent 11.8 percent of all revenues (that’s $447,220 per user in the first three months this year!), and its top 500 users accounted for almost a majority of total revenues. That concentration on the demand side is just as heavy on the supply side. M17’s top 100 artists accounted for more than a third of the company’s revenue.

That concentration has improved over the past few months, according to the company’s filing. But Wall Street investors have learned after Zynga and other whale-based revenue models that the sustainability of these businesses can be tough.

Finally, one complication for many investors wary of the increasing use of dual-class stock issues is the governance of the company. Phua, the CEO, will have 56.3 percent of the voting rights of the company, and M17 will be a controlled company under NYSE rules according to the company’s amended filing. Class B shares vote at a 20:1 ratio with Class A share voting rights.

All of this is to say that while the company has had some dizzying growth in its revenue numbers over the past 24 months, that success is moderated by some significant challenges in revenue concentration that will have to be a top priority for M17 going forward. Why the company priced and hasn’t traded remains a mystery, and we have reached out for more comments.


Source: Tech Crunch