CtrlK
Lenovo Group Limited is a Chinese multinational technology company that designs, manufactures, and sells personal computers, smartphones, servers, workstations, storage systems, and a growing portfolio of artificial intelligence devices, infrastructure, and services. Lenovo has been the world's largest personal computer vendor by unit shipments every year since 2013, and it sits among the top providers of AI-optimized data center hardware, with a particular focus on what the company calls hybrid AI. The business is dual-headquartered in Beijing, China and Morrisville, North Carolina, with its registered office in Hong Kong, where shares trade on the Hong Kong Stock Exchange under ticker 992. As of fiscal year 2024/25, Lenovo reported revenue of about US$69.1 billion across roughly 72,000 employees worldwide.
Lenovo operates through three business groups. The Intelligent Devices Group (IDG) covers PCs, tablets, smartphones (sold under the Motorola brand), and smart accessories, and accounts for the majority of group revenue. The Infrastructure Solutions Group (ISG) builds the ThinkSystem and ThinkAgile families of servers and storage, including the ai_server products that have driven hyper-growth in 2024 and 2025 thanks to demand for nvidia HGX-class GPU systems. The Solutions and Services Group (SSG) wraps the hardware in managed services, vertical-industry solutions, and the TruScale as-a-service consumption model. AI is the unifying thread across all three groups: Lenovo positions itself as a one-stop shop covering personal AI on the device, enterprise AI in the data center, and public AI in the cloud, a triad that chairman and CEO Yang Yuanqing calls hybrid AI.
Lenovo was founded on 1 November 1984 in Beijing by Liu Chuanzhi and ten fellow engineers from the Institute of Computing Technology at the Chinese Academy of Sciences, with seed funding of 200,000 yuan (about US$25,000 at the time) from the Academy. The original company name was Legend, and its early business mixed importing foreign computer equipment with selling a Chinese-character add-in card that allowed Western PCs to display Mandarin text. Liu Chuanzhi remained the central figure for the company's first two decades, first as general manager and then as chairman, and is widely regarded as one of the architects of modern Chinese industry.
Throughout the 1990s, Legend captured the largest share of the rapidly growing Chinese PC market, often by undercutting foreign brands on price while building one of the country's deepest distribution networks. Hong Kong incorporation in 1988 gave the firm a vehicle to raise capital and to access global supply chains, and the eventual Hong Kong listing of Legend Holdings provided liquidity for further expansion.
The Legend name carried strong recognition inside China but was already trademarked elsewhere, so in April 2003 the company rebranded internationally as Lenovo, a portmanteau combining Le from Legend with the Latin novo (new). The new identity set the stage for the most consequential deal in the company's history. On 8 December 2004, Lenovo announced the purchase of IBM's Personal Computing Division for about US$1.25 billion in cash and stock plus assumption of around US$500 million in debt, a transaction that closed on 1 May 2005. The deal transferred the ThinkPad notebook line, the ThinkCentre desktop line, IBM's PC manufacturing assets, and a five-year license to use the IBM brand on PCs, instantly turning Lenovo into the world's third-largest PC maker and giving it a credible enterprise customer base in North America and Europe.
The IBM acquisition also brought a research and development center in Yamato, Japan that continues to lead ThinkPad design, and a sales office in Raleigh-Durham, North Carolina that became the seed of Lenovo's operational headquarters in Morrisville.
Lenovo executed two more transformative purchases in 2014. On 29 January 2014, Google announced it would sell Motorola Mobility to Lenovo for US$2.91 billion, a deal that closed on 30 October 2014. Lenovo received the Motorola smartphone brand, the Moto X and Moto G product lines, more than 2,000 patents, and roughly 3,500 employees. Google retained the Advanced Technologies and Projects (ATAP) group and the bulk of Motorola's broader patent portfolio. Motorola gave Lenovo a recognized smartphone presence in the Americas and Western Europe, complementing the strength of the existing Lenovo-branded mobile business in China.
Later the same year, on 1 October 2014, Lenovo completed its US$2.1 billion acquisition of IBM's x86 server business, including the System x rack and tower servers, BladeCenter and Flex System blade chassis, NeXtScale and iDataPlex hyperscale platforms, and the associated software and maintenance operations. The deal vaulted Lenovo from a marginal data center player into the world's third-largest x86 server vendor by revenue at the time, and it became the foundation of today's ThinkSystem and ThinkAgile portfolios that now anchor Lenovo's AI infrastructure business. Combined, the two 2014 acquisitions cost roughly US$5 billion and gave Lenovo the breadth needed to evolve from a PC pure-play into a diversified hardware and services company.
In April 2021, Lenovo restructured into the three-group model that still governs the company today: IDG for devices, ISG for infrastructure, and SSG for solutions and services. The reorganization recognized that hardware margins alone could not sustain the growth ambitions Yang Yuanqing had set, and that recurring services revenue and AI-enabled solutions would have to drive the next era. In August 2023, Lenovo committed an additional US$1 billion over three years to AI innovation, on top of existing R&D spending, with the money earmarked for AI devices, AI-optimized infrastructure, and embedded generative AI in vertical solutions. By the close of fiscal 2024/25, the strategy had paid off: ISG revenue jumped 63 percent year on year to a record US$14.5 billion as enterprises raced to install AI training and inference clusters, while SSG continued an unbroken streak of double-digit growth quarters thanks to demand for managed and as-a-service offerings.
Yang Yuanqing serves as chairman and CEO of Lenovo Group, a dual role he has held since assuming the chairmanship from founder Liu Chuanzhi in 2011. Yang joined Lenovo in 1989 as a sales representative and rose through the ranks of the China PC business before being named CEO of the parent company. He has been the chief architect of the ai_pc strategy and the hybrid AI thesis, and he frequently shares the keynote stage with nvidia CEO Jensen Huang at industry events.
The Infrastructure Solutions Group is led by Ashley Gorakhpurwalla, who was appointed in late 2024 after Kirk Skaugen, a long-tenured Intel data center executive who had grown ISG from about US$4 billion to roughly US$10 billion in annual revenue, departed in mid-2024. Other key executives include Luca Rossi as president of IDG, Ken Wong as president of SSG, and Wong Wai Ming as chief financial officer. Sergio Buniac heads the Motorola Mobility business, which retains its own brand identity and product roadmap inside Lenovo.
Lenovo maintains principal operating centers in Beijing, Morrisville, Hong Kong, Singapore, and Slovakia, with research and development sites that span more than 12 cities including Yamato (Japan), Bengaluru, Shanghai, Hefei, Stuttgart, and London. The company opened new AI Innovation Centers in 2024 and 2025 in cities including Riyadh, Saudi Arabia and London, the latter built in partnership with Imperial College.
Yang Yuanqing has labelled the second half of the 2020s as Lenovo's AI decade and has argued that the most useful AI deployments will be hybrid by design, blending on-device inference, on-premises enterprise infrastructure, and public cloud services. The company structures its AI message around three layers:
The hybrid framing matters commercially because it lets Lenovo sell across every layer of an AI deployment rather than competing in only one. It also matters technically because Qira, AI Now, and the Learning Zone all combine local small language models with selective cloud calls, an architecture that has become more common across the industry in 2025 and 2026.
Lenovo was one of the first vendors to ship Copilot+ PCs in mid-2024 alongside microsoft's launch of the certification, and it has since extended the lineup across all three of the qualifying silicon platforms: Qualcomm Snapdragon X Elite and X Plus on Arm, intel Core Ultra Series 2 (Lunar Lake) on x86, and AMD Ryzen AI 300 Series on x86. To qualify as a Copilot+ PC, a system must include a neural processing unit (NPU) that delivers at least 40 trillion operations per second (TOPS), and Lenovo's Snapdragon X Elite models advertise more than 45 TOPS of dedicated NPU performance.
The ThinkPad T14s Gen 6 and the Yoga Slim 7x were among the early Snapdragon flagships, and the IdeaCentre Mini x became the first consumer desktop based on Snapdragon X silicon. On the Intel side, Lenovo introduced the Aura Edition premium tier, a multi-year collaboration that bundles exclusive AI features (including a unified Smart Modes panel, a Smart Care service experience, and a co-engineered set of camera and audio enhancements) on top of the base Lunar Lake hardware. AMD Ryzen AI variants of the ThinkPad and Yoga families round out the trio.
Lenovo AI Now is the company's flagship on-device assistant for commercial AI PCs. It is built on Meta's Llama 3 family of models, runs locally using the NPU and integrated GPU, and supports document summarization, knowledge-base retrieval, drafting assistance, and workflow automation. Because the assistant operates locally, it can be pointed at sensitive corporate documents without those files leaving the device. AI Now first appeared on the ThinkPad X1 Carbon Aura Edition and ThinkPad X9 Aura Edition and has expanded to additional commercial Yoga and ThinkBook lines in 2025 and 2026.
Lenovo Learning Zone targets students and educators on AI PCs sold through the company's education channel. The application offers real-time transcription of lectures, AI-generated summaries and study notes, automatically generated quizzes drawn from class material, and adaptive learning features that adjust difficulty as the student progresses. Lenovo positions the Learning Zone as part of a wider Lenovo Education AI initiative that pairs hardware with classroom management tools and teacher training, with the goal of making personalized learning more practical for schools that lack dedicated edtech budgets.
Lenovo has used the ThinkBook Plus line as a public testbed for unusual AI PC form factors. The ThinkBook Plus Gen 5 Hybrid, introduced at CES 2024, pairs a Windows 11 laptop base built around an Intel Core Ultra 7 155H processor with a detachable 14-inch 2.8K OLED Android tablet that runs on a Qualcomm Snapdragon 8+ Gen 1 chip. The two halves can dual-boot or operate as fully independent devices, with a starting price near US$1,999. CES 2025 introduced the ThinkBook Plus Gen 6 Rollable, whose vertically expanding display unfurls from 14 inches to 16.7 inches at the press of a button, and CES 2026 added the ThinkBook Modular AI PC concept with a swappable secondary screen and removable keyboard. These models are sometimes called dual-screen or AI Twin devices because the secondary display surface gives a built-in canvas for an AI assistant or for AI-generated outputs to live alongside the user's primary task.
| Family | Representative model | Silicon | Notes |
|---|---|---|---|
| ThinkPad business | ThinkPad X1 Carbon Aura Edition | intel Core Ultra 9 (Lunar Lake) | Aura Edition AI features, AI Now |
| ThinkPad business | ThinkPad T14s Gen 6 | Qualcomm Snapdragon X Elite | Copilot+ PC, 45+ TOPS NPU |
| Yoga consumer premium | Yoga Slim 7x | Qualcomm Snapdragon X Elite | First Yoga on Snapdragon X |
| Yoga creator | Yoga Pro 9i Aura Edition | Intel Core Ultra 9 HX (Arrow Lake H) | Creator Zone AI apps, optional NVIDIA dGPU |
| ThinkBook commercial | ThinkBook Plus Gen 5 Hybrid | Intel Core Ultra 7 + Qualcomm Snapdragon 8+ Gen 1 | Detachable Android tablet |
| ThinkBook commercial | ThinkBook Plus Gen 6 Rollable | Intel Core Ultra | Rollable 14 to 16.7 inch display |
| IdeaPad consumer | IdeaPad Slim 3x | Qualcomm Snapdragon X | Mainstream Copilot+ PC |
| IdeaCentre desktop | IdeaCentre Mini x | Qualcomm Snapdragon X | First Snapdragon consumer desktop |
| Legion gaming | Legion Pro 7i Gen 10 | Intel Core Ultra 9 275HX + RTX 5080 or 5090 | AI engine for game tuning |
| Legion gaming | Legion 5i Gen 10 | Intel Core Ultra 7 + RTX 50-series | Mainstream AI gaming |
At the company's Tech World event during CES 2026, Yang Yuanqing and Jensen Huang appeared together to introduce Lenovo and Motorola Qira, billed as a personal AI super agent and a personal ambient intelligence system. Qira is designed to span PCs, smartphones, tablets, and wearables sold under both the Lenovo and Motorola brands, allowing a user to start a task on one device and finish it on another while a single agent maintains context. Capabilities at launch include cross-device notification triage, real-time transcription and translation, calendar and travel coordination, and the orchestration of more specialized third-party AI agents.
Qira mixes large and small models. The default cloud-based reasoning calls OpenAI's GPT family, while on-device tasks run on Microsoft's Phi-4 mini model so that the agent works offline and protects local data. Lenovo plans to expand the model roster over time and has identified Saudi Arabia's HUMAIN as a potential collaborator for region-specific large language models. Initial rollout began on select Lenovo and Motorola devices in the first calendar quarter of 2026.
Lenovo's Infrastructure Solutions Group has become one of the most visible beneficiaries of the global AI build-out. The ThinkSystem GPU server portfolio supports the leading accelerators from both nvidia and AMD, and Lenovo's Neptune liquid cooling technology underpins the largest configurations. ISG revenue grew 63 percent year on year in fiscal 2024/25 to a record US$14.5 billion, and the AI server subset of that business grew at triple-digit rates, according to the company's quarterly disclosures.
The flagship NVIDIA-based platform is the ThinkSystem SR680a V3, an 8U air-cooled or liquid-cooled server that pairs two 5th Generation Intel Xeon Scalable processors with eight NVIDIA HGX H100, H200, or B200 GPUs interconnected by NVLink and NVSwitch. The SR680a V3 with B200-SXM-180GB modules is positioned for large-scale generative AI training, and the H200 variant offers a cost-optimized option for organizations whose workloads do not yet require Blackwell-class memory bandwidth. The SR680a V4, announced in 2025, upgrades to NVIDIA HGX B300 SXM modules and is engineered around FP4 inference for next-generation reasoning models.
For smaller deployments, Lenovo offers a range of ThinkSystem 4-GPU and 2-GPU configurations using the H100, H200, and L40S, plus the ThinkSystem SC777 V4 Neptune blade and the ThinkSystem N1380 Neptune chassis for liquid-cooled NVIDIA Grace Hopper and Grace Blackwell deployments. The Lenovo Hybrid AI 289 platform, endorsed by NVIDIA as a turnkey enterprise reference architecture, bundles HGX H200 or B200 or B300 nodes with NVIDIA Spectrum-X networking, NVIDIA AI Enterprise software, and NVIDIA Blueprints into a fully validated stack.
Lenovo also offers the ThinkSystem SR685a V3, an 8-GPU server based on 4th Generation AMD EPYC Genoa CPUs and eight AMD Instinct MI300X accelerators. Each MI300X carries 192 GB of HBM3 memory, giving the server about 1.5 TB of total HBM3 capacity per node and very large model fits without sharding. AMD Infinity Fabric provides up to 1 TB/s of aggregate inter-GPU bandwidth, and Lenovo positions the system for generative AI training and inference, financial services workloads such as fraud detection and KYC, and scientific HPC use cases.
Lenovo's Neptune brand covers six generations of direct water cooling, with the current sixth generation introduced in 2024. The technology pipes warm water (up to about 45 C) directly to cold plates on CPUs, GPUs, memory, and voltage regulators, removing roughly 98 percent of system heat through the water loop and eliminating the internal fans found in air-cooled servers. Lenovo claims up to 40 percent lower power consumption versus comparable air-cooled designs and a 3.5x improvement in thermal efficiency. The flagship ThinkSystem N1380 Neptune chassis enables 100 kW server racks to operate without dedicated computer-room air conditioning, and the ThinkSystem SR780a V3 has reached a power usage effectiveness (PUE) of 1.1 in customer deployments, meaning only 0.1 watts of cooling is consumed for every watt of compute. Neptune underpins the company's recent collaboration with nvidia on Blackwell-class deployments.
| Server | Form factor | GPU configuration | Cooling | Primary use |
|---|---|---|---|---|
| ThinkSystem SR680a V3 | 8U rack | 8x nvidia HGX H100 or H200 SXM | Air or liquid | Generative AI training, large LLMs |
| ThinkSystem SR680a V3 (B200) | 8U rack | 8x NVIDIA HGX B200 180GB SXM | Air or liquid | Frontier-class training and inference |
| ThinkSystem SR680a V4 | 8U rack | 8x NVIDIA HGX B300 SXM | Liquid (Neptune) | FP4 reasoning models, multimodal training |
| ThinkSystem SR685a V3 | 8U rack | 8x AMD Instinct MI300X 192GB | Air | Generative AI, finance, scientific HPC |
| ThinkSystem SC777 V4 Neptune | High-density blade | NVIDIA Grace Hopper or Grace Blackwell | Liquid (Neptune) | High-performance AI and HPC clusters |
| ThinkSystem N1380 Neptune | Vertical chassis | Up to 100 kW per rack | Liquid (Neptune) | Hyperscale AI factories |
| Lenovo Hybrid AI 289 platform | Validated reference architecture | NVIDIA HGX H200, B200, or B300 | Air or liquid | Turnkey enterprise AI deployments |
At CES 2026, Lenovo and NVIDIA jointly announced the Lenovo AI Cloud Gigafactory, a gigawatt-scale program for delivering full AI factories to enterprise and sovereign customers. The program packages Lenovo Neptune liquid-cooled hybrid AI infrastructure, NVIDIA accelerated computing platforms, Lenovo's manufacturing footprint, and the full lifecycle Lenovo Hybrid AI Factory Services into a single integrated framework. The first gigafactory deployments target enterprise customers and national AI initiatives, including possible collaboration with Saudi Arabia's HUMAIN, which had previously announced plans for AI factories of up to 500 megawatts powered by hundreds of thousands of NVIDIA GPUs over the following five years.
The Lenovo AI Innovators program is the company's curated ecosystem of independent software vendors (ISVs) and integration partners. Launched in 2023 and significantly expanded since, it had grown to more than 50 ISV partners offering 165+ validated solutions by 2025. Partners are vetted for technical capability, vertical expertise, and adherence to Lenovo's Responsible AI guidelines, and they progress through a three-stage maturity model: Explore, Deploy, and Scale. The program is supported by Lenovo AI Discover Labs, which give partners and customers a place to test solutions and run proofs of concept on production-grade hardware.
The ecosystem covers vertical use cases across financial services (fraud detection, KYC, algorithmic trading), manufacturing (predictive maintenance, computer vision quality control), healthcare (clinical decision support, medical imaging), retail (autonomous shopping, demand forecasting), and smart city deployments. AI Innovators is the primary route by which Lenovo's hardware finds its way into specific industry workloads without the company itself having to build every application.
| Partner | Domain | Lenovo collaboration |
|---|---|---|
| nvidia | GPUs, AI software | ThinkSystem GPU servers, Hybrid AI 289 platform, Lenovo AI Cloud Gigafactory program |
| AMD | GPUs and CPUs | ThinkSystem SR685a V3 with Instinct MI300X, AMD EPYC server platforms |
| intel | x86 CPUs and Lunar Lake | Aura Edition AI PC features, Xeon-based ThinkSystem servers |
| Qualcomm | Arm CPUs and NPUs | Snapdragon X Copilot+ PC lineup, IdeaCentre Mini x desktop |
| microsoft | Operating system, AI models | Copilot+ PC certification, microsoft_copilot integration, Phi-4 mini model in Qira |
| Meta | Open foundation models | Llama 3 powering Lenovo AI Now, Llama models in Moto AI |
| OpenAI | Frontier LLMs | GPT models powering Lenovo and Motorola Qira super agent |
| Mobile and AI | Gemini integration in Motorola Razr Moto AI suite | |
| HUMAIN (Saudi Arabia) | Sovereign AI | Potential AI factory collaboration, regional model partnership |
| Imperial College London | Research | Joint AI Technology Centre |
| Cyberport (Hong Kong) | Startups | Strategic partnership announced September 2025 to support local AI startups |
Motorola Mobility has remained a separately branded business inside Lenovo since the 2014 acquisition. Its current product strategy centers on the Razr family of foldable flip phones and the Edge line of slab-style flagships. The 2024 and 2025 Razr generations introduced Moto AI, a feature suite that includes Catch Me Up (a notification summarizer), Pay Attention (long-form audio capture), Remember This (a personal memory store), Next Move (contextual on-screen suggestions), and Look & Talk (voice interaction triggered when the phone is in stand mode). Moto AI draws on a mix of partner models including Google Gemini, Meta Llama, and microsoft Copilot, along with Motorola's own on-device processing.
The Razr Ultra adds a dedicated AI Key on the side of the phone for instant agent invocation and uses an AI-powered photo enhancement engine and Signature Style image processing to tailor camera output to individual user preferences. CES 2026 saw Motorola brought into the Lenovo Qira agent program so that a single conversation could move between a Motorola phone, a Lenovo PC, and a Motorola wearable.
The Solutions and Services Group has become one of Lenovo's most strategically important units. SSG packages devices and infrastructure with managed services, vertical-industry solutions, and as-a-service offerings, and it has produced 18 consecutive quarters of double-digit revenue growth as of late 2025, with quarterly revenue of about US$2.6 billion and an operating margin above 22 percent. The flagship as-a-service brands are TruScale Device-as-a-Service for client devices and TruScale Infrastructure-as-a-Service for servers, storage, and AI infrastructure. Customers pay per used unit of compute, storage, or device-month rather than capitalizing the assets upfront, and Lenovo retains responsibility for installation, maintenance, refresh, and end-of-life recovery.
Lenovo's AI services portfolio sits inside SSG and covers AI-optimized infrastructure design, model deployment, custom development, and managed operations. Lenovo Hybrid AI Factory Services is the corresponding wrapper for the Lenovo AI Cloud Gigafactory program, providing the planning, build, and lifecycle management work that turns a hardware order into a running AI environment.
Lenovo reported full-year fiscal 2024/25 revenue of US$69.077 billion, up 21 percent year on year and the second-highest annual revenue in the company's history, with non-HKFRS net income of US$1.462 billion (up 36 percent). The PC business remained the volume engine, with full calendar year 2025 PC shipments of about 70.8 million units representing a global market share of about 24.9 percent according to IDC, ahead of HP at about 57.5 million units and Dell at about 41.1 million units. In Q4 calendar 2025, Lenovo's share rose to about 25.3 percent. The company has held the global PC unit-shipment crown every year since 2013.
The largest growth contribution in fiscal 2024/25 came from ISG, with revenue up 63 percent to a record US$14.5 billion, driven by AI server demand and accompanied by a return to operating profitability in the second half. SSG continued its long streak of double-digit growth quarters with about 13 percent annual growth and an operating margin around 21 to 22 percent, and IDG revenue grew at high single digits with operating margins resilient near 7 percent. In the early quarters of fiscal 2025/26, group revenue reached an all-time fiscal-quarter high of about US$22.2 billion (up 18 percent year on year) with all three groups posting double-digit revenue growth, and Lenovo highlighted high-double-digit revenue growth in AI servers and triple-digit revenue growth in AI PCs, AI smartphones, and AI services.
Lenovo operates more than 30 manufacturing facilities across China, Mexico, Brazil, India, Hungary, Germany, Japan, and the United States, including the Whitsett, North Carolina server factory near its Morrisville operational headquarters. The geographic spread is intentional: it lets Lenovo localize production to comply with sovereignty requirements (including local-content rules in countries such as Saudi Arabia and India) and to absorb tariff or trade shocks. Lenovo's CFO has described the manufacturing footprint as an AI sovereignty asset that lets customers buy AI infrastructure that is built within their own legal jurisdiction.
Lenovo's research and development workforce of more than 20,000 engineers spans more than 12 cities, with major centers in Beijing, Shanghai, Hefei, Yamato (Japan), Bengaluru, Stuttgart, Morrisville, and London. New AI-focused R&D sites have come online in Riyadh and Belgrade in 2024 and 2025, and a Lenovo and Imperial College London AI Technology Centre opened in 2025 with a focus on industrial AI and digital trust.
Industry analysts generally rate Lenovo's hybrid AI strategy as one of the most coherent in the PC and server industry. The Futurum Group, TBR, and IDC have credited the company with successfully diversifying from a hardware pure-play into a balanced device, infrastructure, and services business, and have highlighted the AI Innovators program as one of the more developed partner ecosystems among large server vendors. Reviewers at Tom's Hardware, PCWorld, Engadget, and Notebookcheck have praised the Snapdragon X-based ThinkPad and Yoga lines for battery life and on-device AI performance, while remaining skeptical of some of the more experimental form factors such as the rollable ThinkBook Plus Gen 6 and the modular concept shown at MWC 2026.
Criticism centers on three areas. First, gross margins on AI servers remain thin: the ISG turn to profitability in fiscal 2024/25 was hard-won and is sensitive to GPU supply allocation from nvidia and AMD. Second, the Qira super agent depends heavily on third-party models from OpenAI and microsoft, exposing Lenovo to model-pricing and policy changes outside its control. Third, the Motorola brand has struggled to regain premium-tier share in North America despite the AI feature push, and the AI Key on the Razr 60 Ultra has received mixed reviews for usefulness in everyday workflows.