Romain Lavault’s Post

The whole future and progress of artificial intelligence currently rests on one company. #Nvidia. Today, almost every tech startup as well as large companies are pushing towards ai. In this rush, access to the GPU processors are limited where demand is outstripping supply. Besides, supplies are being stockpiled by the bigtech firms, which have the money to purchase these components. It seems like the whole ai world is at the mercy of one company while the rest of the semicon companies like Intel and ARM are pouring billions into creating alternatives. A comparable competitor could be years away. #gpu #ai #artificialintelligence #llm #cpu #semiconductors #tech #startup #cloud

Nvidia is investing heavily in AI startups, according to the FT article linked below. Takeaways: 1) According to estimates by Dealroom, Nvidia participated in 35 deals in 2023, making the chipmaker the most active large-scale investor in AI. 2) Nvidia’s latest quarterly filings reveal that it made $872M worth of investments in “non affiliates” in the first three quarters of 2023, more than 10 times as much as during the same period last year. 3) Nvidia’s investments now include Inflection AI and Cohere, two of the biggest rivals to ChatGPT maker OpenAI; Hugging Face, a provider of data and tools for AI developers; CoreWeave, a cloud infrastructure company and French-based startup Mistral. 4) All the companies receiving these investments are Nvidia customers, either for chips or software. 5) Nvidia denies that it conditions investment on use of its chips, or that those companies receive preferential treatment in terms of obtaining chips. Dave’s take: Nvidia is already the preeminent AI chipmaker in the world, though they may eventually face significant competition from AMD, Intel and Huawei, among others. The strategic bets they are placing in AI startups are quite intriguing-do they plan vertical integration well beyond chipmaking and software? What kind of regulatory hurdles will they face if they do? https://lnkd.in/gGsT-Q8e

There is so much stupid money chasing #AI! It makes me want to go back to do my MBA again. What was that I learned about venture-funding portfolio strategy? How about totally irrational exuberance? Basically.. more $$$ than brains and pure arrogance. Throwing money at totally inexperienced Harvard dropouts. How exciting! I lost count years ago of all the briefings I've had with AI-chip startups comparing their designs to NVIDIA. Wanna know why it's "the world's first transformer ASIC"? Although likely not..really more an ASSP..but because it's a dumb idea! You're a hyperscaler looking to gain some efficiency for specialized workloads on some of your compute nodes. OK.. build an ASIC. Just don't think you can build a company on today's latest fad. But hey... it's Silicon Valley. They need $120M just for the housing subsidies. 😁

Iliad and Xavier Niel Launch Paris AI Research Center with $105 Million Investment-Romain Dillet Xavier Niel , CEO of telecom company iliad,CTO Alexi Karalis brings , plans to invest up to $210 million in AI projects. Iliad will establish a Paris-based AI research center with an initial investment of $105 million, focusing on hiring AI researchers and supporting PhD students. Additionally, Iliad subsidiary Scaleway is bolstering its GPU and AI cloud capabilities by acquiring 1,016 NVIDIA H100 GPUs, making it a major European GPU cloud provider. Scaleway intends to collaborate with Nvidia on AI tools. Furthermore, Scaleway will host a European AI conference at Station F, aiming to attract engineers and researchers to discuss AI innovations. Niel aims to position Europe as a significant player in AI development. #Iliad #XavierNiel #AIResearchCenter #AIProjects #AI #ArtificialIntelligence #TechInvestment #ParisAI #AIResearch #Scaleway #GPU #CloudComputing #Nvidia #AIInnovation #EuropeanTech

Elon Musk's xAI Raises $6 Billion to Accelerate AI Development Elon Musk's AI startup, xAI, has successfully raised $6 billion in funding, signifying a major achievement for the emerging company. This substantial financial boost aims to launch xAI’s first products, upgrade its infrastructure, and support continuous research and development of advanced technologies. Developing AI technology is a costly endeavor, partly due to the high price of necessary hardware. For instance, Nvidia’s upcoming Blackwell B200 AI graphics cards are priced between $30,000 and $40,000 each. Reports indicate that xAI plans to acquire 100,000 of Nvidia's H100 chips to power a supercomputer aimed at enhancing Grok. Musk has outlined a goal to have this advanced data center operational by fall 2025. read more here: https://lnkd.in/eMX_Je97 #xai #elonmusk #ai #nvidia

Elon Musk’s AI startup, xAI, has ambitious plans to build a supercomputer, which is being referred to as a ‘gigafactory of compute’. This supercomputer is intended to power Grok, xAI’s AI model12. The supercomputer is planned to be operational by fall 20252. The supercomputer will be powered by 100,000 Nvidia GPUs, making it four times larger than any existing GPU clusters2. This massive scale of operation is expected to pose challenges, particularly in terms of power requirements2. xAI has recently raised $6 billion in funding, showing that the hype around AI is not dying down2. Musk personally invested $750 million in xAI2. Massive Scale: The supercomputer will utilize Nvidia’s H100 GPUs and aims to be four times larger than any existing GPU clusters. Timeline: Musk intends to have the supercomputer operational by fall 2025 and is personally accountable for its timely completion. Investor Presentation: In May, Musk presented his ambitious plan to investors, highlighting the significance of this advancement for AI development. Grok is a large language model (LLM) built and trained by xAI. It’s designed to have a broad general knowledge and be able to have natural conversations1. The ultimate goal for Grok is to act as an accessible research assistant for anyone, aiding in quick information retrieval, data processing, and idea generation. Grok is modeled after the Hitchhiker’s Guide character in the sci-fi series, with the goal of being able to answer almost anything and suggest relevant questions to ask. It’s also part of xAI’s mission to advance scientific discovery and gain a deeper understanding of our universe. Moreover, Grok stands at the forefront of explainable AI (XAI), revolutionizing how AI systems are developed, understood, and trusted. Grok’s unique approach to AI transparency enables users to comprehend the decision-making processes of complex algorithms, fostering trust and reliability critical for widespread adoption. In summary, Grok is poised to redefine our interactions with AI, with its unique set of capabilities and its commitment to transparency and accessibility.

"For every $1 spent on a GPU, roughly $1 needs to be spent on energy costs to run the GPU in a data center. So if Nvidia sells $50B in run-rate GPU revenue by the end of the year, that implies approximately $100B in data center expenditures." 'During a gold rush, sell shovels'. Great time to be in the data center business. "The end user of the GPU needs to earn a margin too. Let’s assume they need to earn a 50% margin. This implies that for each year of current GPU CapEx, $200B of lifetime revenue would need to be generated by these GPUs to pay back the upfront capital investment. This does not include any margin for the cloud vendors". Good read from David Cahn. 👍🏻 "During historical technology cycles, overbuilding of infrastructure has often incinerated capital, while at the same time unleashing future innovation by bringing down the marginal cost of new product development. We expect this pattern will repeat itself in AI." 🎯 https://lnkd.in/epvgxq5H #vc #venturecapital #startups #entrepreneurs #entrepreneurship #funding #tech #technology #innovation #deeptech #ai #artificialintelligence #

🌟 Spotlight on Inference.ai 🌟 They're breaking barriers in AI, making GPU computing affordable for everyone! By offering affordable access to advanced GPU resources, they're empowering startups and SMBs to harness the power of AI without the prohibitive costs. Their global partnership with data centers enables a diverse GPU fleet, making cutting-edge technology accessible to all. Dive into their journey:

🚀💡 According to 𝐅𝐢𝐧𝐚𝐧𝐜𝐢𝐚𝐥 𝐓𝐢𝐦𝐞𝐬 (https://lnkd.in/dk5USHHV), Nvidia has emerged as one the most prolific investor in AI companies in 2023, participating in a whopping 35 deals! 🎉 This marks a significant increase from previous years, demonstrating Nvidia's commitment to fostering innovation in the AI sector. 📈🚀 Based on my research with the use of crunchbase (https://lnkd.in/dUedpZDb) here's a breakdown of some of the most recent investments: 📅 💰 𝐄𝐬𝐬𝐞𝐧𝐭𝐢𝐚𝐥 𝐀𝐈, which is building software for enterprises to use large language models. (Announced on 12th December) 📅 💰 𝐌𝐢𝐬𝐭𝐫𝐚𝐥 𝐀𝐈, an open-source platform developer assembling teams to develop generative AI models. (Announced on 11th December) 📅 💰 𝐓𝐨𝐠���𝐭𝐡𝐞𝐫 𝐀𝐈, a cloud-based platform designed for constructing open-source generative AI and infrastructure for developing AI models. (Announced on 30th November) 📅 💰 𝐁𝐏-𝐅𝐋𝐀𝐂, an AI and a vast computing pool, providing limitless expansion possibilities for web3. (Announced on 18th November) 📅 💰 𝐂𝐞𝐧𝐭𝐌𝐋, which accelerates Machine Learning workloads by maximizing training and inference efficiency leading to compute cost reduction. (Announced on 25th October) Investing in AI startups and companies is more important than ever in 2023. 🌐💡 As AI continues to revolutionise industries from healthcare to finance, supporting these companies is crucial for driving innovation, creating jobs, and solving complex problems. It is not only fueling technological advancement but also shaping the future of AI ❗ 🌟🚀 #Nvidia #AI #Startups #Investment #Innovation #FutureTech #Dealroom

Elon Musk's xAI Raises $6 Billion to Accelerate AI Development Elon Musk's AI startup, xAI, has successfully raised $6 billion in funding, signifying a major achievement for the emerging company. This substantial financial boost aims to launch xAI’s first products, upgrade its infrastructure, and support continuous research and development of advanced technologies. Developing AI technology is a costly endeavor, partly due to the high price of necessary hardware. For instance, Nvidia’s upcoming Blackwell B200 AI graphics cards are priced between $30,000 and $40,000 each. Reports indicate that xAI plans to acquire 100,000 of Nvidia's H100 chips to power a supercomputer aimed at enhancing Grok. Musk has outlined a goal to have this advanced data center operational by fall 2025. read more here: https://lnkd.in/enRrPbbH #xai #elonmusk #ai #nvidia