CEO Anthony Mc Loughlin, who founded the company last year, has previously worked with Airbus, Mercedes and Rolls Royce, helping them to adopt analytics and AI.

Read more: AI start-up Better Futures raises €500,000 in pre-seed round

The company focused on making legacy chips is owned by Chinese electronics maker WingTech.

Read more: Dutch-based chipmaker Nexperia invests $200m in Germany

Eindhoven-based startup Axelera has raised $68mn as it looks to take its AI chip business global. One of the lead investors is Samsung Catalyst, the venture arm of the semiconductor giant Samsung Electronics.  

Axelera is developing AI chips that enable computer vision and generative AI in edge devices like robots and drones.

Because they’re located within the devices themselves, the AI chips can process data in milliseconds without the need to upload or download data to the cloud. This delivers high computing performance at a fraction of the cost and energy consumption of centralised AI models, said the startup.   

Axelera leverages what’s known as in-memory computing. That’s when data is stored in the main memory (RAM) instead of on traditional disk storage. This makes for even faster data processing and retrieval.  

“To truly harness the value of AI, organisations need a solution that delivers high-performance and efficiency while balancing cost,” said Fabrizio Del Maffeo, co-founder and CEO at Axelera AI. 

Axelera builds the chips, known as AI processing units (AIPUs), as well as the software that runs them. Dubbed Metis, the startup claims that the chip is the world’s most powerful AIPU for edge devices. 

Axelera claims it has built the world’s most powerful AI processing unit for edge devices. Credit: Axelera
Democratising AI 

Axelera’s chips make use of the instruction set architecture (ISA) RISC-V. An ISA acts like a bridge between the hardware and the software. It specifies both what the processor is capable of doing as well as how it gets done. 

RISC-V is a low-cost, efficient, and flexible ISA that can be customised to specific use cases. Crucially, unlike most ISAs, it is open source, which means no single entity controls it.

“Our mission is to democratise access to artificial intelligence,” said Del Maffeo. 

By specialising in edge AI, and developing both the software and hardware components, Axelera looks to give itself a competitive edge in a booming AI chip market dominated by the likes of Nvidia, Intel, and IBM. 

Speaking of IBM, Del Maffeo, who previously worked at Belgium-based tech lab Imec, co-founded Axelera alongside Evangelos Eleftheriou, a former veteran at the American tech giant.   

Axelera plans to put its AI processing units into full production in the latter half of this year. It looks to expand its presence in North America, where it already has an office, and into new industries such as automotive, digital healthcare, and surveillance. The startup is also exploring the development of high performance AI chips for data centres and supercomputers. 

Hailing from Eindhoven, Axelera exists in one of the most mature semiconductor tech hubs in the world. The city is home to Philips-founded NXP Semiconductors and ASML, which produces chip-making machines for almost every major semiconductor manufacturer on Earth. 

This latest funding round brings Axelera’s total raised to $120mn. New investors include the Samsung Catalyst Fund, European Innovation Council Fund, Innovation Industries Strategic Partnership Fund, and Invest-NL.

The EU’s supercomputer alliance yesterday announced the winners of its AI startup competition. Four companies will split €1bn — and precious uptime on world-class high-performance computer systems to train their models. 

Training and running AI requires massive amounts of computational resources. For example, as part of its deal with OpenAI, Microsoft built Eagle, one of the world’s most powerful supercomputers, for its new investment. Elon Musk is on track to build a massive Nvidia-based supercomputer for xAI, his new artificial intelligence startup.  

It is no secret that the EU has been pushing hard for what it calls “digital sovereignty.” Fearful of once again of being left behind in a global tech race, the bloc has devoted billions of euros to its supercomputer alliance, known as the EuroHPC. Moreover, it has made a vow to make its supercomputers available to European AI startups.

Naturally, these startups can’t just rock up and demand time on the nodes. To that end, EuroHPC last year launched a competition called the Large AI Grand Challenge. On Wednesday, it announced the four winners that will share the cash prize of €1bn.  

The <3 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!

They will also get 8 million computational hours on the LUMI and LEONARDO supercomputers (these rank among the fastest HPC systems in the world, and are situated in Kajaani, Finland, and Bologna, Italy, respectively) over the next 12 months. Having access to such powerful systems reduces large AI model training time from years to weeks. 

The four winners, selected from 94 submitted proposals, are: 

Lingua Custodia, France. Fintech startup offering AI and natural language processing for the finance sector. 
Unbabel, Portugal. Language technologies company combining human and machine translation for multilingual support for all 24 official EU languages. 
Tilde, Estonia. Also focuses on machine translation along with AI-powered chatbots for the Balto-Slavic languages. 
Textgain, Belgium. AI startup helping companies and organisations gain insights from unstructured data through predictive text analytics. Specifically, the company focuses on the analysis of hate speech.

EuroHPC also decided to award an additional 800,000 hours on the Mare Nostrum supercomputer in Barcelona to a fifth entry, namely quantum computing and AI startup Multiverse Computing from Spain. 

Following the 12 months, the winning startups are expected to release them under open-source and non-commercial licences, or publish research findings. 

US Internet providers must now display clear pricing and product information. HP Ink controversy continues to stain the company’s reputation with consumers. Is the EU’s competition legislation working? Early numbers seem to show it might be. And there’s a 10 million dollar bet that Elon Musk is wrong about AI.

All this and more on the “all bets are off” edition of Hashtag Trending. I’m your host, Jim Love. Let’s get into it.

New regulations from the Federal Communications Commission have taken effect yesterday, mandating that all broadband internet service providers clearly display labels detailing the prices, speeds, data caps and other key information about their service plans.

The rules are aimed at helping consumers make more informed choices by requiring ISPs to disclose this data in a simple, standardized format akin to nutrition labels on food products.

In addition to fees charged, the labels must also now list any monthly data caps or overage fees, upfront costs like equipment rental fees, a provider’s customer service contact information, and any other plan limitations like throttling policies.

Despite this progress, consumer advocacy group Next Century Cities   continued to push for even more information, saying that these broadband “nutrition labels” often overstate the real-world speeds customers can expect or obscures caps and fees.

Speaking to the FCC last month, one group urged that in addition to maximum speeds, the labels should show the average speeds users actually experience, as estimates of “typical” speeds are frequently overly optimistic.

While comprehensive, some experts warn the amount of required disclosures could overburden smaller ISPs with limited resources compared to industry giants. For now, only providers with more than 100,000 subscribers must comply and smaller providers have been given an additional year to comply.

The FCC is still gathering feedback on whether to mandate the display of promotional pricing periods and expiration dates, as well as taxes and fees beyond the base rate.

Next Century Cities is further advocating for a streamlined complaint process to report issues like digital discrimination in broadband deployment to the commission.

With the labeling rules now in place, the hope is that customers will be better equipped to comparison shop for broadband and avoid being misled about the true costs and capabilities of different internet packages.

Sources include: ArsTechica, Engadget, and BroadbandBreakfast

Of all of the emails I get about stories, the HP printer issue is near the top of the list. People write me, with their frustrations. And it turns out, they take these to the courts as well.

Printer owners are pushing back against HP Inc. in an ongoing class action lawsuit over firmware updates that allegedly disabled their devices from using third-party ink cartridges.

In a filing this week in an Illinois court, the plaintiffs accused HP of using software changes to monopolize the replacement ink market and “take advantage of customers’ sunk costs” in HP printers.

The consumers claim that despite never agreeing to only use HP-branded ink, recent firmware updates prevented their printers from accepting more affordable third-party cartridges.

They allege HP violated several anti-competitive statutes through this “tying scheme” accomplished via unauthorized software changes solely aimed at blocking rival ink suppliers.

The plaintiffs are seeking damages covering the cost of now-useless non-HP cartridges, as well as an injunction forcing HP to undo the firmware lockout.

For its part, HP insists it went to “great lengths” to inform buyers that its printers are designed to exclusively use HP cartridges containing security chips.

The company says the updates represent legitimate “dynamic security” measures to combat counterfeit ink, and that it does not conceal or block remanufactured cartridges reusing official HP chips.

HP also argues the plaintiffs cannot claim overcharge damages from the manufacturer under federal antitrust laws when they purchased through intermediaries.

As printer makers increasingly push subscription models, the controversy highlights long-standing tensions over the high costs of proprietary ink replacements versus third-party alternatives.

The bitter legal battle seems primed to further antagonize HP’s customer base over what critics condemn as anti-competitive practices designed to sustain lucrative ink sales.

Sources include: The Register

We’ve done a number of stories on legisltation and regulation from the EU that is aimed at increasing customer choice and promoting real competition. Is it working? In one case it seems to have had an impact.

It turns out that some alternative web browsers are reporting an uplift in user interest and downloads in the European Union following the recent enforcement of a new digital regulation called the Digital Markets Act or DMA.

The landmark rules, which took effect last month, require dominant tech gatekeepers like Apple and Google to present mobile users with choice screens displaying alternative browsers and other core apps.

The goal is to shake up competition against pre-installed defaults and make users more aware of their options beyond Safari on iOS or Chrome on Android.

While it’s still very early days, several smaller browser makers have already shared positive metrics pointing to increased attention from EU users.

Norway’s Opera says new user growth was up 63% from February to late March, while fellow Norwegian browser Vivaldi reports a 36.7% jump in EU downloads, rising to nearly 70% in the eight countries where it appears on Apple’s choice screen.

The privacy-focused Brave browser also cited a doubling of daily iOS installs in the EU compared to pre-choice screen levels.

And little-known Cyprus-based rival Aloha claimed to have seen 250% growth in new users as it jumped from the 4th to 2nd biggest EU market.

However, not all alternative browsers are seeing clear gains yet. Veteran players like Mozilla’s Firefox, DuckDuckGo and Ecosia say it’s too early to accurately assess the DMA’s impact as choice screen rollouts are still ongoing, although some claim that these browsers are purposely holding back from reporting success because they want to keep the pressure on to make the choices even more clear and easier to adopt.

For example, the are complaints that Apple’s iOS implementation in particular has significant design flaws hampering users’ ability to make meaningful choices about switching browsers.

The European Commission has open investigations into suspected cases of improper compliance by the tech giants, including Apple’s choice screen methodology.

With this continued pressure from the largest alternative browsers, and given the EUs track record, it is likely they will be monitoring closely to ensure dominant gatekeepers are genuinely opening their platforms to greater competition and consumer choice as intended.

Sources include: TechCrunch

Some tech industry CEOs are putting their money where their skepticism is when it comes to Elon Musk’s ambitious predictions about artificial intelligence surpassing human intelligence in the next few years.

During a recent interview, the billionaire claimed AI will likely exceed the cognitive capabilities of any single human by the end of 2024, with AI as a whole outstripping the combined intelligence of all humans within just five years.

But those bold forecasts are being met with raised eyebrows and big bets from some AI experts who view Musk’s timeline as wildly unrealistic.

Gary Marcus, CEO of machine learning startup Geometric Intelligence, publicly offered up $1 million to anyone, including Musk, who can prove him wrong.

That prompted Damion Hankejh, CEO of ingk.com, to raise the stakes even further, saying he’d cover a $10 million wager against Musk’s AI predictions coming true.

Marcus said Musk has not responded to the million-dollar challenges yet, but added the Tesla CEO has previously ignored Marcus’ smaller $100,000 bet that artificial general intelligence was not actually imminent, as Musk claimed.

For Marcus, the bets are about more than just money. He wants to spark a public discussion with Musk about what artificial intelligence can realistically achieve in the near-term versus the almost utopian promises that have become common from tech leaders.

Marcus argues many in the industry have a track record of making scientifically implausible claims and missing self-imposed deadlines, pointing to the ongoing challenges with self-driving cars as one example.

While large language models have made rapid advances, Marcus contends the notion they could exceed human-level general intelligence within just a couple of years is fanciful, estimating that milestone may still be decades away.

As CEOs literally gamble over contrasting AI outlooks, the high-stakes bets underscore an intensifying debate over whether too much hype is obscuring the real state and timeline of artificial intelligence development.

I don’t know. Just this once and only once. I’m putting my money on Elon being right.

As always, love to hear what you might think.

And that’s our show for today…

Thanks for those who’ve written in with comments including the person who wrote me about their trials and tribulations HP printers and ink purchases.

Keep it coming.  And don’t forget, you can find us on YouTube now. If you check us out there, please give us a like or even a subscribe as we try to build and audience there as well.

I’m your host Jim Love, have a Fantastic Friday.

Vidyard, a popular video messaging company, has recently announced its entry into the virtual digital avatar market and secured new funding of $15M from Export Development Canada (EDC), BMO Capital Partners, and existing investors Battery Ventures, Bessemer Venture Partners, and iNovia Capital. This brings the company’s total funding to $90.7M.

According to Gartner Research, AI avatars using generative AI technology will support 70% of digital and marketing communications by 2025, up from less than 5% in 2022. The AI avatar market is valued at USD 14.34B in 2022 and is expected to grow at a compound annual growth rate (CAGR) of 47.1% from 2023 to 2030. The growth is driven by increased demand for remote collaboration tools, customer preferences for video content in corporate communications,  hybrid work environments, and advancements in video streaming technology.

Jonathan Lister, Chief Operating Officer at Vidyard says “the increasing demand for authentic and personalized characters for online interactions by companies from various industries is opening up a future where AI-powered video communications can become the cornerstone of productive connections between sellers and buyers.”

Vidyard’s advantage over leading AI video and avatar providers, such as Synthesia, HeyGen, and Tavus, is that its products are embedded in a sales workflow, making it easy to create personalized videos and send them to buyers all within the sales ecosystem. Market leaders like Microsoft, Marketo, and Hubspot are already using Vidyard.

Hubspot’s VP of Platform Ecosystem, Scott Brinker, who has been testing Vidyard’s AI Avatar platform, says that “Virtual avatars will address a lot of critical pain points for HubSpot users, including the stage fright associated with traditional video recording, the massive amount of time it takes to get them ‘right,’ the difficulty of finding the right place and time to do it, and the fact that it’s rarely their top priority.”

In a recent customer survey by GTM Partners, Vidyard found that when video was added, sales funnel performance increased by 85%, four to five times more meetings were booked. In addition, two to four times more sales-qualified opportunities were generated. This resulted in a 25% increase in close rates.

According to Gartner, AI’s capability to help sellers engage more effectively with prospects and customers and to do so at scale by automating labor-intensive tasks is one of the most valuable contributions to using AI in sales.  As communications continue to shift in customer preferences, virtual avatars are making it extremely easy for sales and marketers to get messages out highly personalized and improve customer and buyer interactions.

Kelvin Beachum, a National Football Athlete of the Arizona Cardinals, has found Vidyard’s virtual avatar technology useful. He says “AI technology is just beginning to crack the surface of its capabilities. For me, time is money. As an organization, we are always trying to be efficient with our time to accomplish as much as possible at a high level. Using AI for efficiency allows for a high-velocity workflow of my daily operations, and I am excited to continue implementing it.”

Lister says “B2B buying and selling is fundamentally broken. Buyers want timely and relevant information and automation that helps simplify their complex buying journey. While the sellers struggle to build meaningful relationships, communicate with an ever-increasing number of stakeholders, and deliver high-quality insights given the limitations of current sales technology.”

As this market continues to unfold, there is a need to keep a vigilant eye on deep fakes and the increasing risks of video communication. Stanford has released guidelines on digital communication ethics to help mitigate deep fake risks.

Swiss startup Neustark has raised $69mn to take its carbon capture technology global. Decarbonization Partners, a partnership between the world’s largest asset manager BlackRock and Singaporean investment firm Temasek, led the funding round.

Neustark captures waste CO2 from biogas production. It injects it into recycled concrete, which then gets reused to make new roads and buildings. The concrete stores the carbon for at least 100,000 years, even if it gets demolished.

“We turn the world’s largest waste stream — demolition concrete — into a carbon sink,” said Johannes Tiefenthaler, co-CEO and co-founder at Neustark.

Neustark was founded in 2019 by Johannes-Tiefenthaler and Valentin Gutknecht. Credit: Neustark

Neustark has two kinds of machines. One captures CO2 at biogas plants and liquifies it. The company transports this pressurised carbon dioxide to its second device, located at a nearby concrete recycling facility. This machine mineralises the CO2 into limestone, where it is used to create new concrete.

The <3 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!

The startup currently has 19 plants. They operate at biogas factories and concrete recycling facilities in Switzerland, Austria, Liechtenstein, and Germany. Some 40 new plants are currently under construction in Europe, it said. 

Neustark will use the fresh funding to expand in Europe and “enter new markets in North America and Asia Pacific,” said Tiefenthaler.

The business of carbon removal    

Neustark sells carbon removal credits to businesses looking to offset their emissions. The company has sold 120,000 tons worth of these credits to date. Its clients include Microsoft, UBS, and NextGen. 

However, Neustark has only actually captured 1,819 tons of CO2, according to its website. Such a discrepancy is commonplace in the world of carbon credits.

Most startups today sell carbon removal credits through offtake agreements. Buyers purchase credits based on a guarantee that the carbon removal provider, in this case Neustark, will remove a certain quantity of carbon over the next, say, 6-10 years.  

For major backers of the technology like Google, Meta, and Salesforce, securing a long term supply of carbon removal credits is an attractive proposition as they look to tackle the hard-to-decarbonise parts of their operations.  

For Neustark, offtake agreements provide a key early revenue source, alongside private and public investment. “It allows us to scale up at speed,” a Neustark spokesperson told TNW via email. 

However, from giant carbon-sucking machines and planting trees to throwing dust on soil, carbon removal technology is still unproven on a large scale. Some scientists fear that the lure of buying carbon removal credits will distract companies from investing in reducing their emissions at source.   

Despite the uncertainty, billions of dollars in funding are flowing into the budding industry, spurred on by the IPCC’s Sixth Assessment Report. The landmark report cites carbon removal as key to limit global warming to 1.5C or below.

“We need to exponentially accelerate the removal of CO2 if we want to reach net zero goals by 2050,” said Tiefenthaler. “This target will only be possible by globally deploying highly scalable, measurable and commercially viable carbon removal solutions at the scale of millions of tons per year.”

Neustark aims to remove 1 million tons of CO2 by 2030.