Connect with us

Business

Groq, an AI Chip Firm, raises $640 Million to take on Nvidia

Published

on

A fresh investment round led by Blackrock has secured $640 million for Groq, a business that is creating chips to run generative AI models faster than traditional processors. The company announced this information on Monday. Participating companies included Samsung Catalyst Fund, Cisco, KDDI, Neuberger Berman, and Type One Ventures.

Groq, which was allegedly hoping to raise $300 million at a somewhat lower ($2.5 billion) valuation, is celebrating a significant victory with this tranche, which puts the company’s total raised to over $1 billion and values it at $2.8 billion. Groq raised approximately $1 billion in April 2021 and $300 million in a funding round headed by D1 Capital Partners and Tiger Global Management. This valuation more than doubles that amount.

Groq also announced today that Stuart Pann, the former head of Intel’s foundry business and former CIO at HP, will join the startup as chief operational officer. Yann LeCun, the main AI scientist at Meta, will advise the company technically. Given Meta’s investments in its own AI chips, LeCun’s appointment is a little surprising, but it definitely provides Groq a strong ally in a competitive market.

Groq is developing an LPU (language processing unit) inference engine after coming out of stealth in 2016. The business asserts that their LPUs can operate generative AI models that are currently in use at 10 times the speed and 1/10th the energy of OpenAI’s ChatGPT and GPT-4o.

Jonathan Ross, the CEO of Groq, is renowned for his contribution to the creation of the tensor processing unit (TPU), Google’s proprietary AI accelerator processor used for model training and execution. Nearly ten years ago, Ross co-founded Groq with Douglas Wightman, an entrepreneur and former engineer at Google parent firm Alphabet’s X moonshot lab.

Groq offers an LPU-powered developer platform called GroqCloud, which includes an API that lets users use its chips in cloud instances, as well as “open” models like Google’s Gemma, OpenAI’s Whisper, Mistral’s Mixtral, and Meta’s Llama 3.1 series. (Groq also runs GroqChat, an AI-powered chatbot playground that it introduced at the end of last year.) More than 356,000 developers were using GroqCloud as of July. According to Groq, some of the money raised during this round will be utilized to expand the company’s capacity and introduce new models and features.

“Many of these developers are at large enterprises,” stated Stuart Pann, COO of Groq. “By our estimates, over 75% of the Fortune 100 are represented.”

As generative AI becomes more popular, Groq will have to contend with competition from rival AI chip startups as well as Nvidia, the industry leader in AI hardware.

Nvidia is attempting to preserve its dominance in the market for AI processors, which are needed to train and implement generative AI models. The company is expected to dominate between 70% and 95% of this industry.

As opposed to every other year, like in the past, Nvidia has pledged to release a new AI chip architecture annually. Additionally, it is apparently starting a new business unit dedicated to creating custom chips for cloud computing companies as well as other businesses, including AI devices.

Groq faces competition from Nvidia as well as Amazon, Google, and Microsoft, who now provide or plan to provide customized chips for artificial intelligence workloads on the cloud. Customers of Google Cloud can use the aforementioned TPUs as well as Google’s Axion chip in due course. Microsoft recently introduced Azure instances in preview for its Cobalt 100 CPU, with Maia 100 AI Accelerator instances to follow in the coming months. Amazon offers its Trainium, Inferentia, and Graviton processors through AWS.

Analysts predict that in the next five years, the AI chip market might exceed $400 billion in sales, and Groq may see competition from Arm, Intel, AMD, and an increasing number of startups. Due in large part to cloud vendors’ increasing capital expenditures to meet the capacity demand for generative AI, Arm and AMD in particular have seen their AI chip businesses flourish.

Late last year, D-Matrix secured $110 million to launch a platform for inference computation that it described as unique in the market. With $120 million, Etched came out of stealth in June to unveil a processor designed specifically to accelerate the transformer, the most popular generative AI model architecture available today. Masayoshi Son of SoftBank is purportedly trying to fund $100 billion for a chip business in order to take on Nvidia. Additionally, it’s said that OpenAI is in discussions to start a chip-making venture with investment firms.

Groq is significantly spending on industry and government outreach in an effort to carve out its niche.

To create Groq Systems, a new business unit, Groq bought Definitive Intelligence, a Palo Alto-based company that provides a variety of business-oriented AI solutions, in March. Serving entities, such as sovereign nations and U.S. government agencies, that want to construct new data centers employing Groq processors or integrate Groq chips into already-existing ones falls within stems’ jurisdiction.

More recently, the startup Groq signed an agreement to install tens of thousands of its LPUs in the Norway data center of the European company Earth Wind & Power. Additionally, Groq teamed with Carahsoft, a government IT contractor, to market its solutions to public sector clients through Carahsoft’s reseller partners.

Additionally, Groq and Aramco Digital, a Saudi Arabian consulting business, are working together to install LPUs in upcoming Middle Eastern data centers.

Groq, a company located in Mountain View, California, is developing new ties with customers while simultaneously moving forward with the development of its microprocessor. The company said in August of last year that it will be working with Samsung’s foundry division to produce 4nm LPUs, which are anticipated to outperform Groq’s initial 13nm chips in terms of efficiency and performance.

By the end of Q1 2025, Groq intends to deploy over 108,000 LPUs.

Business

Hyzon is the most recent startup backed by SPAC to fail

Published

on

Hyzon Motors, a hydrogen fuel cell developer, has shut down after struggling to sustain operations since going public during the 2020-2021 SPAC boom. Despite positive press, warning signs persisted, culminating in the company’s downfall.

A Rocky Start and SEC Troubles

Hyzon, a spinoff from Singapore’s Horizon Fuel Cell Technologies, raised $550 million in 2021 through a reverse merger with Decarbonization Plus Acquisition Corp. However, its operations were focused on Europe, Australia, and China, with no U.S. or North American business initially.

In 2021, short-seller Blue Orca Capital accused Hyzon of fabricating orders in China, leading to an SEC investigation. The company paid a $25 million fine, and CEO Craig Knight was replaced in 2022 by Parker Meeks, a former McKinsey & Co. partner.

Attempts to Revive the Business

Under Meeks, Hyzon closed its European and Australian operations and focused on specific markets like refuse trucks. The company also partnered with Fontaine Modification to retrofit Freightliner Cascadia trucks with 110-kilowatt fuel cell systems while developing a larger 200-kW system.

Despite technological progress, Hyzon struggled to generate sales. By the third quarter of 2023, it had only $100,000 in revenue. With just $14 million in cash, the board decided on December 19 to pay creditors and shut down operations. Remaining employees in Michigan and Illinois are set to lose their jobs by February 2024.

Optimism Faded

Until its third-quarter earnings call, Meeks expressed hope, citing potential fleet contracts and falling hydrogen prices, which were projected to drop to $10-$12 per kilogram by 2025. However, Hyzon’s high truck costs and inability to secure large orders sealed its fate.

Broader Industry Struggles

Hyzon’s collapse is part of a broader trend among hydrogen fuel cell and SPAC-funded startups. German company Quantron AG entered insolvency in late 2023, while Nikola Corporation faces funding challenges. Other SPAC-backed ventures like Lordstown Motors and Embark Trucks also failed due to financial difficulties.

Hyliion, however, has managed to thrive by pivoting to a fuel-agnostic stationary generator business, securing contracts, and achieving a significant stock price increase in 2023.

Continue Reading

Business

Japan’s efforts to create a dual-purpose defense startup environment

Published

on

To stay competitive in the global technological race, Japan must merge its defence and civilian innovation ecosystems, which involve diverse stakeholders. In September 2024, Japan’s Ministry of Defense and Ministry of Economy, Trade and Industry unveiled the concept of a “dual-use startup ecosystem.” This initiative seeks to integrate startups into research and development (R&D) to meet the technological demands of defence equipment.

Strengthening Defence Innovation

Prior to the announcement, the government identified approximately 200 startups in July 2023, outlining plans to support these companies with defence-related equipment and financial assistance to ease their entry into the market. The startups specialize in advanced fields such as drones, cyber defence, satellite communications, and electromagnetic wave technologies.

Leading this initiative is the Ministry of Defense’s Acquisition, Technology, and Logistics Agency through its newly established Defense Innovation Science and Technology Institute (October 2024). The aim is to efficiently incorporate civilian technologies into defence equipment, aligning with global trends where private-sector innovation plays a growing role in defence development. The model draws inspiration from the U.S. Defense Advanced Research Projects Agency (DARPA) and the Defense Innovation Unit, which rapidly integrate private-sector advancements into defence projects.

Historical Roots and Persistent Challenges

Japan’s push for dual-use technologies is not entirely new. Efforts began with the 2013 National Security Strategy and the 2014 Strategy on Defense Production and Technological Bases, emphasizing public-private partnerships. These policies responded to challenges like globalized supply chains, Japan’s deteriorating security environment, the shrinking defence industry, and the need for technological cooperation with allies.

However, gaps between policy and implementation have hindered progress. A major issue is the low profitability of the defence industry, which has driven many private companies out of the sector. Reform efforts must offer stronger incentives for startups to participate. While increased defence spending has benefited traditional firms, smaller companies and startups face uncertain gains.

Another obstacle is the private sector’s cautious stance on defence R&D, rooted in Japan’s post-war anti-militarist norms. Many academic and industrial players perceive military involvement as a reputational risk in the predominantly civilian-focused business landscape.

For instance, the Ministry of Defense’s 2015 research funding initiative faced strong opposition from the academic community, including the Science Council of Japan, which criticized it for potentially restricting free scientific inquiry. This resistance has limited the impact of defence-related reforms, and startups entering the sector may encounter similar challenges.

Emerging Opportunities in a Changing Context

Despite these hurdles, Japan’s new dual-use startup ecosystem reflects an evolving political and social landscape. Since the 2010s, Japan’s national security policies have shifted to address growing security threats and fiscal constraints. Public opinion has gradually become more open to pragmatic national security measures, although resistance persists in some sectors.

Startups, particularly those led by younger entrepreneurs who are less tied to traditional business norms, are poised to play a pivotal role in this policy’s success.

Economic Security as a Catalyst

Economic security policies are further driving changes in Japan’s defence innovation ecosystem. The 2022 Economic Security Promotion Act has marked the beginning of “economic securitisation,” incorporating critical and emerging technologies into national policy. Initiatives like the “Key and Advanced Technology R&D through Cross Community Collaboration Program” have expanded R&D budgets, with applications spanning both civilian and military domains under the label of “multi-use” technologies.

By framing defence-related R&D as part of economic security, the government is addressing concerns within Japan’s political culture. This approach may reduce normative barriers for companies and universities to engage in defence-related activities.

A Defining Moment for Japan’s Innovation Ecosystem

As economic securitisation gains traction, Japan faces an opportunity to establish a robust defence innovation ecosystem. However, this moment also demands tough decisions from the private sector about their involvement in defence projects. Balancing commercial interests with normative considerations will shape the future of Japan’s defence and civilian innovation integration.

Continue Reading

Business

Microsoft plans to incorporate non-OpenAI AI models into its 365 Copilot products

Published

on

Microsoft is expanding its flagship AI product, Microsoft 365 Copilot, by integrating both internal and third-party AI models to diversify beyond its reliance on OpenAI’s technology and reduce operational costs, according to sources familiar with the matter.

This marks a shift for Microsoft, a major investor in OpenAI, which previously highlighted its exclusive access to OpenAI’s models as a key advantage. When Microsoft introduced 365 Copilot in March 2023, its use of OpenAI’s GPT-4 model was a primary selling point.

The company now seeks to reduce its dependence on OpenAI due to concerns about cost and performance for enterprise users, the sources said. A Microsoft spokesperson confirmed that OpenAI remains a key partner for advanced AI models, but Microsoft also customizes OpenAI’s models as part of their agreement.

“We incorporate various models from OpenAI and Microsoft depending on the product and experience,” Microsoft stated. OpenAI declined to comment.

Microsoft is training its own smaller models, including the latest Phi-4, and customizing open-weight models to make 365 Copilot faster and more cost-effective. These efforts aim to lower operational expenses and potentially reduce costs for customers, sources said.

Microsoft’s leadership, including CEO Satya Nadella, is closely monitoring these developments.

This strategy aligns with changes in other Microsoft units, such as GitHub, which added models from Anthropic and Google in October alongside OpenAI’s GPT-4. Similarly, its consumer chatbot Copilot now integrates both in-house and OpenAI models.

Microsoft 365 Copilot, an AI assistant for enterprise applications like Word and PowerPoint, is still proving its value to businesses. While adoption among Fortune 500 companies has reached 70%, many enterprises remain in the pilot phase, according to Gartner. Pricing and utility have been cited as potential hurdles.

Despite these challenges, adoption is accelerating. Analysts at BNP Paribas Exane predict that Microsoft will sell 365 Copilot to over 10 million paid users this year. In a November blog post, Microsoft highlighted its growing traction within large enterprises.

Continue Reading

Trending

error: Content is protected !!