Nebius Group N.v. · NASDAQ:NBIS
Nebius — Investor Presentation (October 2024)
98 slides from Nebius's investor presentation, Oct 18, 2024
Slide 1: Nebius Group Investor Presentation
NEBIUS Nebius Group Investor Presentation —
Slide 2: Forward-looking statements
This presentation contains forward - looking statements within the meaning of the Private Securities Litigation Reform Act of 1995 , which involve risks and uncertainties. All statements contained in this presentation other than statements of historical facts, including, without limitation, state me
Slide 3: Table of contents
Table of contents 4 Introducing Nebius Group 1 3 Nebius : our core AI infrastructure business 1 4 Business overview 2 6 Competitive advantages 2 9 Clients 3 5 Growth plans 3 8 Financials 42 Toloka 5 7 TripleTen 7 1 Avride 8 6 Nebius Group Financials and Outlook 9 0 Appendices
Slide 4: Introducing Nebius Group
Introducing Nebius Group Nebius Group
Slide 5: Technology for Technologists
Technology for Technologists Strongly positioned to build one of the largest specialist AI infrastructure players
Slide 6: Nebius Group at a glance
Nebius Group at a glance Nebius Group (1) Refers to Nebius Group. As of September 30, 2024; (2) $2.3bn total cash as of September 30, 2024, without giving effect to any potential retur n of capital to shareholders or withholding tax. $2bn+ cash on Balance Sheet 2 and no debt NASDAQ listed with long
Slide 7: Building a global AI business integrating infrastructure, data and expertise
Building a global AI business which integrates essential components for successful AI development: infrastructure, data, and expertise Nebius Group (1) Approximately 28% stake on a fully diluted basis, upon exercise of penny warrants to acquire Common Stock and Series A and B sto ck; latest publicly
Slide 8: Illustrative value building blocks for Nebius Group
Illustrative value building blocks for Nebius Group Nebius Group value building blocks State - of - the - art owned Finnish data center with ~ 14 ,000 GPUs paid and deployed 1 Building specialized AI cloud Full stack & efficient infrastructure platform 2025 ARR $500 - 1,000m 2 Our core business: Neb
Slide 9: Total addressable market for Nebius
Over $ 26 0 bn total addressable market for Nebius Addressing a huge market opportunity across all of our businesses (1) Statista Market Insights, March 2024 ; (2) Nebius internal estimates based on a bottom - up analysis of publicly available disclosure on GPUs spending and GPU Cloud economics ; th
Slide 10: A well-synced team with a track record in tech and AI infrastructure
A well - synced team with a track - record of developing world - class tech and AI infrastructure Our R&D backbone ~850 in - demand tech professionals 1 AI / ML / LLM engineers, Data scientists & analysts , Cloud engineers, Security architects, and other ~ 1 , 3 00 Total headcount as of Sep - 24 Big
Slide 11: Significant opportunity to build a leading AI infrastructure player
Significant opportunity to build a leading AI infrastructure player Funding needs Substantial capital investments from the balance sheet, as well as public and private equity or debt The opportunity Multi - billion dollar annual revenue potential from pure - play in the fast - growing AI industry wi
Slide 12: Why do we think we can be successful?
Why do we think we can be successful? Expertise We have a strong team with extensive technical experience, having built 200MW+ of cutting - edge infrastructure and cloud business from the ground up Track record We built one of the largest tech companies 1 in Europe with a $30bn market cap and establ
Slide 13: Nebius: our core AI infrastructure business
Nebius : our core AI infrastructure business Visit website
Slide 14: A leading AI-centric cloud platform for intensive AI workloads
A leading AI - centric cloud platform built for intensive AI workloads Nebius
Slide 15: Nebius at a glance
AI - centric cloud platform building large, cost - efficient GPU clusters to service the explosive growth of the global AI industry Nebius at a glance (1) Primary driver in the AI industry, where inferencing workloads are expected to overtake training; (2) Refers to Nebius core business. Nebius Grou
Slide 16: Sustained demand growth in GenAI driving revenue growth
$33bn $260bn + 2023 2030 Sustained demand growth in GenAI driving revenue growth Market opportunity (1) Nebius internal estimates, Nebius internal estimates based on a bottom - up analysis of publicly available disclosure on GPU spending and GPU Cloud economics; the presented TAM includes estimates
Slide 17: Nebius business model: key pillars
4. AI expertise Dogfooding by in - house LLM team helps to adjust hardware and platform to the real needs of ML/AI practitioners 3. Cloud platform A proprietary cloud platform is essential for creating large GPU clusters optimized for extensive AI training and inferencing without performance bottlen
Slide 18: Core pillars deep dive: data centers (1/2)
Power usage effectiveness 1. Data centers: efficiency gains from supercomputing Core pillars deep dive: data centers (1/2) (1) Peer PUE as reported in the latest available full - year company reports as of October 2024, Nebius' figure represents PUE of Finnish data center under max. loads in 2020 -
Slide 19: Core pillars deep dive: data centers (2/2)
Optimized cooling 1. Data centers : pioneers in optimized cooling and heat recovery Core pillars deep dive: data centers (2/2) (1) Leveraging external ambient air to reduce air temperature in the data rooms; (2) As per ASHRAE standards for inlet air te mpe ratures; (3) Assuming that a household cons
Slide 20: Core pillars deep dive: hardware (1/2)
2 . Hardware: why do we design, develop and produce our own servers? Core pillars deep dive: hardware (1/2) Lower total cost of ownership via power efficient cooling and full control over the supply chain Faster time - to - market with on demand development, quick prototyping and production, deploym
Slide 21: Core pillars deep dive: hardware (2/2)
System architects Electronic engineers Embedded developers Mechanical engineers Thermal engineers 2 . Hardware: in - house expertise in R&D … (1) Compares electricity consumption of Nebius ’ custom - designed servers to off - the - shelf alternatives, showing ~20% lower usage. This allows Nebius to
Slide 22: Core pillars deep dive: data centers + hardware
In - house designed servers • Control over components purchases, design and production processes helps to reduce costs • More efficient operations due to optimized design and configuration Bringing data center ownership together with in - house hardware development drives additional shareholder valu
Slide 23: Nebius Platform: more than bare metal
3. Nebius Platform: more than 'bare metal', more flexible and wider product portfolio than GPU Cloud (1) Open - source object - rationale database system; (2) Open - source platform for managing the end - to - end machine learning lifecycle; (3) O pen - source data tools for batch processing, queryi
Slide 24: Nebius Studio: API SaaS for open-source models
Nebius Studio is a new product from Nebius designed to help foundation model users and app builders simplify the process of creating applications using these models with our first release, Inference Service, providing endpoints for the most popular large lan gua ge models 3. Nebius Studio: API SaaS
Slide 25: Core pillars deep dive: people and expertise
4 – 5. The most in - demand engineers are already on board Core pillars deep dive: people and expertise (1) Refers to Nebius core business. As of September 30, 2024. Software Development LLM team Hardware R&D Data center operations Product development Sales and marketing Other Team of 500+ professio
Slide 26: Nebius competitive advantages
Nebius competitive advantages Nebius
Slide 27: Nebius key competitive advantages
Nebius key competitive advantages differentiate it from other players Nebius differentiating factors Sole focus on AI infrastructure , allowing us to provide solutions precisely tailored to meet the specific needs of AI developers and businesses Benchmark - backed performance and cost efficiency Lon
Slide 28: Nebius proposition is differentiated vs. other players
Nebius proposition is differentiated vs. other players, allowing us to serve the broadest set of customer demands The Nebius proposition Our customer proposition that differentiates Nebius ... (1) Original Design Manufacturers (ODMs); (2) Part of DigitalOcean . Solutions tailored to meet the needs o
Slide 29: Our clients
Our clients Nebius
Slide 30: Nebius serves model builders and GenAI app builders
Nebius is able to serve a large list of clients including model builders and Generative AI app builders Clients snapshot How we reach them? Client category Primary product Segments 2. Non AI - centric SaaS IaaS + PaaS Research and experiments in progress Community hackathons and events / Digital & S
Slide 31: Nebius client case studies (1/2)
Training scenarios Nebius client case studies (1/2) What we offer: Large resilient GPU clusters with Infiniband networking, orchestration tools (fully managed K8S service, Slurm operator based on K8S) and shared storage optimized for check pointing and data streaming Model types: • LLMs • Multimodal
Slide 32: Nebius client case studies (2/2)
Nebius client case studies (2/2) Inference scenarios What we offer: Large resilient GPU clusters with Infiniband network, orchestration tools with fully managed K8S service with auto - scaling and load balancing, storage services optimized for inference (managed and object storage) Model types: • LL
Slide 33: What our customers say: testimonials and recognition
What our customers say: testimonials and recognition Customer reviews “ Nebius is one of the best cloud platforms , providing excellent customer support and flexible offerings . Most other cloud providers don't provide a flexible offering of H100, which becomes a hurdle for early - stage startups. F
Slide 34: Nebius value proposition for clients
State - of - the - art hardware GPU clusters up to 4,000 H100 SXM5, interconnected via 3.2 Tb it/ s Infiniband Flexible capacity allocation and pricing On - demand and multiple term reserve consumption modes available User - centric approach • Web console, CLI, API, Terraform • Detailed documentatio
Slide 35: Growth plans
Nebius. f¢ tl —Alan srowth plar A pet 1A
Slide 36: Medium-term capacity expansion plans
Nebius has ambitious medium - term plans on capacity expansion of both data center and GPU capacity Capturing increasing market demand (1) In Nvidia H100 equivalent; (2) Finnish data center expansion Phase I commenced in Q2’24; (3) To be put into operation by the end of November 2024; (4) Depending
Slide 37: Huge potential for growth across multiple levers
External This is just the beginning: huge potential for growth across multiple levers … Roadmap of future growth (1) Nebius internal estimates based on a bottom - up analysis of publicly available disclosure on GPUs spending and GPU Cloud economic; the presented TAM includes estimates for two key ma
Slide 38: Financials
Financials Nebius
Slide 39: Historical and current trading
Historical and current trading Nebius historical financials (1) Annualized run - rate revenue by the end of the period (revenue for last month of the period multiplied by twelve); ARR for 3Q’24 is based on preliminary r eve nue for September 2024 as per management accounts. ARR 1 $m 21 4Q’23 32 1Q’2
Slide 40: Nebius financial outlook
Nebius expects the revenue to increase at least 5x and become Adj EBITDA positive in 2025 Nebius financial outlook (1) Annualized run - rate revenue by the end of the period (December revenue multiplied by twelve); ARR for 3Q’24 is based on prel iminary revenue for September 2024 as per management a
Slide 41: Nebius key investment highlights
Nebius key investment highlights Recap Pure play in attractive AI market i n its most in - demand segment: infrastructure Full stack with expertise across the whole AI infrastructure value chain; from data centers and in - house server design to AI cloud platform and expertise to run LLMs Strong tea
Slide 42: Toloka
Toloka Visit website
Slide 43: A data partner for all stages of AI development
A data partner for all stages of AI development from training to evaluation Toloka
Slide 44: Toloka at a glance (1/2)
(1 ) Cumulative amount of onboarded experts (domain experts and expert annotators) as of end of Q3 2024. Experts comprise highly skilled contributors from various industries ; (2 ) Monthly active users, for the period between January 1 and September 30, 2024 . Fast scaling platform with significant
Slide 45: Toloka at a glance (2/2)
(1) Supervised Fine - Tuning; (2) Reinforcement Learning from Human Feedback; (3) Direct Preference Optimisation. Toloka powers AI production with human intelligence Toloka at a glance (2/2) Data stage s RLHF 2 , DPO 3 , Preference collection SFT 1 Prompts generation Completions (responses) generati
Slide 46: Growth of Gen AI raises the demand for large volumes of data
(1 ) Toloka’s internal estimates based on our analysis on the client universe and assumptions on average spend by different categories on t he back of the company’s data and industry checks. Growth of Gen AI raises the demand for large volumes of data The market opportunity Gen AI Data User Profiles
Slide 47: Competitive positioning (1/4)
Toloka AI has both the expertise and technological infrastructure to build high quality human data operations at scale Competitive positioning (1/4) This makes Toloka one of only a few companies globally able to generate high - quality data at scale Deep & diversified expertise at different levels D
Slide 48: Competitive positioning (2/4)
Experts experience / satisfaction Customer experience / satisfaction Scalability Breadth Quality Intuitive onboarding process, user - friendly product, community - driven approach A true Data Partner approach providing expertise and acting as an extension of clients’ Data Science teams Ability to ra
Slide 49: Trusted to provide expert data at scale
Trusted to provide expert data at scale Competitive positioning (3/4) Scaling is backed by robust infrastructure MS Azure as base infrastructure, with storage options in the US and EU, as well as private and on - premises data storage options Deep roots in the scientific community … … help Toloka to
Slide 50: Toloka AI is trusted by Big Tech and startups
Toloka AI is trusted by Big Tech, Fortune - 500 companies, and A - class startups given strong ML experience of the team and research partnerships Competitive positioning (4/4) “Thanks a lot to the Toloka team. Wonderful collaboration in the past two years.” Select customer feedback “Our collaborati
Slide 51: Multiturn Dialog Evaluation customer case study
Goal: Agent model for IT service management ticketing evaluation and fine - tuning Solution: customized pipelines for evaluating agent and user responses, ensuring criteria such as responsiveness, transparency, accuracy, groundedness , and helpfulness were met Result 1 : i ncreased throughput 250 ti
Slide 52: Toloka AI key competitive advantages
Toloka AI key competitive advantages differentiate it from other players Toloka AI differentiating factors (1) The number of experts engaged varies from 1,000s to 10,000s at different points throughout the year, depending on evolvin g n eeds of ongoing projects. 8,000 experts cumulatively onboarded
Slide 53: Significant potential for growth across multiple levers
This is just the beginning: significant potential for growth across multiple levers Roadmap of future growth One of the world’s largest open crowdsourcing platforms 2024+ Data partner for GenAI • Democratized access to high skilled experts • Gained expertise in building data pipelines for GenAI (RLH
Slide 54: Financial Highlights
Financial Highlights
Slide 55: Overview of key revenue drivers
Overview of key revenue drivers Toloka AI revenue Non - core non - Gen AI services Self - serve platform Evolved GenAI Classic GenAI Solutions Revenue categories Hybrid solutions , including ML use cases not effectively covered by GenAI or other state - of - the - art models at reasonable costs Laun
Slide 56: Toloka AI key investment highlights
Toloka AI key investment highlights Recap Delivering unmatched quality data at scale to satisfy broad needs of LLM development Large team of domain experts, operating a marketplace model to enable rapid and efficient scaling Premium technology backed by scientific research collaborations and extensi
Slide 57: TripleTen
TripleTen Visit website
Slide 58: A leading edtech platform for reskilling into tech careers
A leading edtech platform specializing in reskilling individuals for successful careers in tech TripleTen
Slide 59: TripleTen at a glance
TripleTen at a glance Addressing the widening skills gap created by rapid digitalization Expertise and results in STEM Long - standing collaboration across hardware and cloud as a Cloud and OEM partner Top - rated edtech 5 immersive B2C study tracks: Software engineering Quality assurance BI analyti
Slide 60: Market opportunity
Market opportunity $ 95 bn+ Global digital education market 1 202 4 2030 $ 19 bn + 26 % CAGR Capitalizing on growing demand for IT reskilling (1) Persistence Market Research, 2024 ; ( 2 ) BCG, 2023. Demand for tech talent is high and growing, including across TripleTen’s primary markets — US and LAT
Slide 61: TripleTen offers high-quality and immersive training
TripleTen offers high - quality and immersive training Immersive training & expert guidance … Flexible & personalized learning paths tailored to a student’s knowledge level Work - simulated environment featuring real projects and career services Expert guidance and detailed code reviews from IT prac
Slide 62: TripleTen offers affordable prices
TripleTen offers affordable prices backed by optimized costs and automation Competitive positioning (2/4) (1) Based on publicly available data on tuition fees for the comparable courses, as of October 2024. … enabled by optimized costs of training and processes automation Affordable offering vs peer
Slide 63: Best overall provider of software engineering bootcamps
Best overall provider of software engineering bootcamps 2 80+ employers among partners that offer job opportunities to graduates Note: Bootcamps listed in the table are those that TripleTen regards as its peers. Source of data: latest data reported by companies, e.g. to the Council on Integrity in R
Slide 64: What our customers say: testimonials and recognition
What our customers say: testimonials and recognition Competitive positioning (4/4) Across 800+ reviews on Career Karma, SwitchUp , and Course Report 1 4.9 “We hire TripleTen students because of their ability to work well, own the project, care about it and solve new and complex problems. Any softwar
Slide 65: TripleTen differentiating factors
Affordable pricing enabled by optimization of training costs through diversified sourcing of trainers, automation of e - learning workflows, and AI - enabled student support and guidance Top - rated edtech in the US by employment rate and student feedback with a solid median salary after graduation.
Slide 66: Significant potential for TripleTen growth
This is just the beginning: significant potential for TripleTen growth across multiple levers Roadmap of future growth Continue to expand B2C offering Launch new programs : • Cyber Analyst (recently launched) • UI/UX designer (near future launch) • Continue to expand in the US & gain share from comp
Slide 67: Financial highlights
Financial highlights
Slide 68: TripleTen historical financials (1/2)
2.2 2.1 3.4 6.1 9.8 H2'22 H1'23 H2'23 H1'24 9M'24 ARR 1 , $m Historical and current trading TripleTen historical financials (1 /2) New Students, thousands (1) Annualized run - rate revenue as of the end of the period (calculated as revenue as of the last month of the period multiplied by twelve); Q3
Slide 69: TripleTen historical financials (2/2)
Efficient business model with optimized costs TripleTen historical financials (2/2) Initial Customer Payments CAC 1 CAC fully covered by initial customer payments as of Q3’23 (1) Customer acquisition costs comprised principally of marketing and advertising expenses; per one student. Q4’22 Q2’23 Q4’2
Slide 70: TripleTen's key investment highlights
TripleTen’s key investment highlights Recap Top - rated US bootcamp characterized by high employment rate, salaries and student feedback Automated flexible e - learning platform simplifies launching and scaling on new markets and enables affordable pricing for customers and cost savings for the busi
Slide 71: Avride
Avride Visit website
Slide 72: Developer of autonomous driving technology for cars and delivery robots
Developer of autonomous driving technology for self - driving cars and delivery robots
Slide 73: Avride at a glance: who we are
Who we are? Avride at a glance 200+ self - driving tech engineers & software developers 1 7+ years experience of the team in developing autonomous solutions Offices and R&D hubs Headquarters Austin, TX US - based executive team with additional offices in Tel Aviv, Belgrade, Seoul USA & South Korea U
Slide 74: Avride at a glance
Return of capital to shareholders via share repurchase Avride is uniquely positioned in the industry by simultaneously developing autonomous driving technology for cars and delivery robots Avride at a glance (1) The number of passenger rides, driven kilometers and delivered orders is cumulative for
Slide 75: Autonomous vehicles and delivery robots share technologies
Autonomous vehicles and delivery robots leverage shared and mutually enhancing technologies Avride at a glance Common elements Code deployment and data collection Simulator, offline algorithms and more Localization Perception Models training system Delivery robots’ faster deployment and monetization
Slide 76: Scalable business model focused on large end markets
Scalable business model focused on large and attractive end markets Market opportunity $46bn TAM in 2030 1 $6bn TAM in 2030 2 (1) Markets&Markets Forecast, 2023; (2) Grand View Research, 2023; (3) Management calculations based on third party data; calculated as an averag e of forecasts ranging from
Slide 77: Autonomous cars: snapshot (1/2)
Avride is working on the highest level of automation, developing fully autonomous vehicles Autonomous cars: snapshot (1/2) Advanced Driver - Assistance Systems (ADAS) Adaptive Cruise Control & Lane Keep Assist constantly monitored by a human driver Self - Driving Systems Driver - out - of - the - lo
Slide 78: Autonomous cars: snapshot (2/2)
Rapid fleet progression, reaching first passenger rides in under two years Autonomous cars: snapshot (2/2) (1) 0 incidents with human injuries or significant damage to the vehicle. Enabled by the team's expertise across multiple domains, including ML, computer vision, hardware engineering, sensor te
Slide 79: Delivery robots: snapshot (1/2)
Avride delivery robots have been successfully tested and can operate in diverse conditions Delivery robots: snapshot (1/2) (1) Software - enforced speed limit to ensure pedestrian safety; the robots’ design allows to reach a maximum speed of 24 kmh; (2) Delivery robots can charge wirelessly or have
Slide 80: Delivery robots: snapshot (2/2)
Successful and quick rollout will enable further commercial partnerships globally Delivery robots: snapshot (2/2) ~200,000 orders delivered by robots designed by the team since the launch of the service 2 – 3 weeks for robo - delivery roll - out in a new location: from mapping & testing to integrati
Slide 81: Delivery robots: customer and partner case studies
Human - less delivery robots operations 1 High - load delivery robots operations 2 Goal: Meet the delivery demands of one of the largest university campuses in the U.S. Solution: Deploy advanced robot dispatch algorithms to maximize efficiency and ensure smooth operations, avoiding traffic jams or c
Slide 82: Autonomous cars and delivery robots: efficiency and key value drivers
Key efficiencies achieved Avride’s autonomous vehicle offering has been highly efficient since launch Efficiency levers Launch From the outset in 2017, we leveraged cutting - edge technologies and solutions to address early - stage challenges more swiftly and effectively than competitors Autonomous
Slide 83: Financial Highlights
Financial Highlights
Slide 84: Avride mid-term outlook
Unsupervised public launch in multiple cities Supervised public launch for closed list of users Rollout of vehicles for R&D deployments Deployment Vehicles 200+ vehicles 100+ vehicles (R&D and production) 10 - 20 vehicles (R&D) Fleet size ~3,000 robots 1 ,000 + robots 100+ robots Deployment Delivery
Slide 85: Avride key investment highlights
Avride key investment highlights Recap One of the most experienced teams in the autonomous vehicles industry Fast track to commercial deployment at a fraction of costs Partners hips with leading automakers and delivery platforms Technology tested in diverse road conditions with outstanding safety re
Slide 86: Nebius Group Financials and Outlook
Nebius Group Financials and Outlook Nebius Group
Slide 87: Nebius Group financial outlook: Nebius core business
Nebius future growth driven by significant expansion in TAM and further penetration … Nebius Group financial outlook: Nebius core business (1) GPUs available in H100 equivalent; (2) Based on the planned capacity expansion and assumed utilization and pricing as per the company’s medium - term financi
Slide 88: Financial outlook: additional businesses
… supported by growth across our other businesses Nebius Group financial outlook: additional businesses Expand revenue across all Toloka revenue categories and continue estab - lishing presence in the growing market Secure new customers and expand wallet share with existing clients for Classic GenAI
Slide 89: Nebius Group expects $500-700m revenue in 2025
Nebius Group financial outlook Nebius Group expects to deliver $500 - 700m revenue in 2025 and push towards Adj EBITDA breakeven (1) Adjusted EBITDA/(loss) means net income/(loss) from continuing operations plus depreciation and amortization, certain SBC ex pense, interest expense, income tax expens
Slide 90: Appendices
Appendices 91 Historical trading 9 2 Management overview 9 3 BoD overview 9 4 Employee snapshot 9 5 Shareholder structure 96 Sustainability overview 97 Reconciliation of Adjusted EBITDA to US GAAP Net income / (loss)
Slide 91: Nebius Group historical financials
Historical and current trading 1 Nebius Group historical financials Revenue $m Adjusted EBITDA 2 $m Capex $m (1) Q3 2024 results are based on preliminary management accounts; (2) Adjusted EBITDA/(loss) means U.S. GAAP net income/(loss) from continuing operations plus depreciation and amortization, c
Slide 92: Nebius Group management team
Nebius Group Nebius Group management team Business units Arkady Volozh CEO and Founder Visionary founder; previously co - founded successful Tech companies including Yandex and CompTek International C - Suite Andrey Korolenko Chief Product & Infrastructure Officer 20 years supervising IT infrastruct
Slide 93: Nebius Group Board of Directors
Nebius Group Board of Directors — based in John Boynton Chairman Founder, investor and board member of tech ventures with 30+ years in IT and tech - related fields Non - executive Independent Ophir Nave COO Proven leader and legal and corporate finance expert, having previously led Corporate M&A at
Slide 94: Nebius Group employee snapshot
Nebius Group employee snapshot 1 (1) Data as of Q3 2024. Total headcount, people Share of women is in line with industry trends for IT companies Diversity Geographical presence 70% m ale 26% female 4% non - binary / prefer not to state 1,303 43 nationalities 86% of employees are based in the Netherl
Slide 95: Nebius Group ownership structure
Nebius Group ownership structure (1) Family trust established by Mr . Volozh in December 2019. Total shares issued & outstanding Economic ownership Class A 163, 641 , 444 Class B 35,698,674 Excluding 16 2 , 700 , 826 Class A shares held in treasury pending use under our equity incentive plans and fo
Slide 96: Sustainability overview
Committed to doing business responsibly and sustainably at every stage of the journey, to becoming a leading player in AI, and to ensuring that products and solutions have a positive impact on society Sustainability overview (1) Uptime Institute Global Data Center Survey Results 2023; Nebius' figure
Slide 97: Reconciliation of Adjusted EBITDA to US GAAP net income/(loss)
2023 2024 FY Q4 Q3 Q2 Q1 Q3 1 H1 Q2 Q1 247.1 (72.0) 84.3 176.3 58.5 ( 91.1 ) (414.0) (97.5) (316.5) Net income / (loss) (582.8) (15.4) (181.1) (250.6) (135.7) - 295.1 58.2 236.9 Less: Net income from discontinued operations (335.7) (87.4) (96.8) (74.3) (77.2) ( 91.1 ) (118.9) (39.3) (79.6) Net loss
Slide 98: Find more information on our website
Find more information on our website : https://group.nebius.com/ Main Office Gustav Mahlerlaan 300 Amsterdam, Netherlands Investor Relations Contact askIR@nebius.com