LOGIN | Register
Cooperation
Intel at Computex 2026
文章语言:
EN
Share
Minutes
原文
会议摘要
Intel's keynote at Computex 2026 highlights advancements in AI and silicon architecture, emphasizing collaboration with industry partners for improved performance and efficiency. The speech covers new Intel Core processors, AI innovations with Perplexity, and Xeon 6 plus for data centers, underscoring the importance of partnerships in driving technological progress and opening new business opportunities in various sectors.
会议速览
Celebrating 40 Years of Intel's Partnership with Taiwan and Vision for Future Growth
Reflects on the creation of Silicon Island, Intel's 40-year partnership with Taiwan, and the company's focus on engineering and execution for future success, highlighting the importance of the PC ecosystem and the shift towards purpose-built GPU and ASIC solutions.
A Journey of Growth and Win-Win Cooperation in Taiwan: 36 Years of Commitment
Reflecting on a first international trip to Taiwan in 1990, the speaker highlights the country's collaborative spirit and its lasting impact, returning after 36 years to continue fostering partnerships and growth.
Intel's Core Ultra and Core Series 3: Pioneering AI and Agentic Capabilities in PCs
Intel showcases its commitment to PC innovation with the Core Ultra Series 3, built on advanced process technology, offering premium mobile performance and battery life. With over 300 designs shipping, Intel extends its capabilities to the mainstream market through the Core Series 3, already scaled to 70+ designs, transforming PCs into agentic platforms.
Exploring Core Series 3: Battery Life and User Experience
The dialogue highlights the extensive lineup of over 390 designs and delves into the battery life capabilities of the Core Series 3, emphasizing its endurance to support long days of use, from 10 to 14 hours or more, ensuring great user experiences.
Revolutionizing Handheld Gaming with Arc G3: High Performance, Long Battery Life
The dialogue showcases the Arc G3's superior performance and efficiency in handheld gaming, highlighting its ability to run AAA games at 1080p resolution with over 120 frames per second, offering a significant improvement over competitors. This marks the beginning of a series of innovative designs aimed at the fast-growing handheld gaming market.
Intel's Latest Processors: Scaling Performance and Efficiency for Diverse Markets
Intel highlights its advanced processor lineup, leveraging versatile technology for premium, mainstream, and handheld gaming. The company is expanding its offerings into the Edge market with over 130 designs across various sectors, emphasizing best-in-class technology, user-friendly reference designs, and comprehensive software support to meet large-scale business demands.
Expanding Physical AI Markets with Leading IP and Chipsets
A dialogue highlights the potential of physical AI in various industries, emphasizing the company's strategy to lead with IP, chipsets, and complete reference platforms, aiming to capture a projected 25 trillion market by 2050.
Revolutionizing Hybrid Compute: On-Device AI for Enhanced Privacy and Efficiency
A collaboration between industry leaders introduces a groundbreaking AI operating system, orchestrating multiple models locally and on the cloud. This system, designed for privacy and efficiency, keeps sensitive data on-device while leveraging cloud resources for comprehensive analysis, marking a significant milestone in hybrid computing technology.
X86 Architecture: Powering Modern Data Centers and AI Workloads
Discusses the enduring significance of X86 architecture in powering foundational and emerging intelligent workloads, highlighting its role in data centers, AI, and mission-critical applications, with a focus on performance, efficiency, security, and resiliency.
Intel Xeon 6 Plus: Advancing Data Center Efficiency and Density for Next-Gen Cloud Infrastructure
Intel introduces Xeon 6 Plus at Computex, boasting 288 cores, 576 MB L3 cache, and Intel 18A tech. This innovation promises enhanced efficiency, compact server designs, and valuable real estate savings, supported by a robust ecosystem and ODM partners. Ideal for enterprises leveraging x86 architecture, it marks a significant leap in data center computing for cloud and network infrastructures.
Intel's Xeon 6 Plus: Bridging AI Workloads and Traditional Computing in Evolving Enterprise Infrastructures
Intel introduces Xeon 6 Plus to simplify large-scale intelligence deployment, addressing the growing AI demand in data centers. It unifies CPU-led enterprise infrastructures and GPU-heavy AI factories, crucial for agentic AI's iterative, tool-using, and concurrent agent-spawning nature, significantly boosting CPU demand for complex AI tasks.
Revolutionizing Compute Balance: CPU Dominance in AI Workloads and Agent Density
A detailed exploration of how modern AI systems, particularly agentic AI, are shifting compute requirements towards CPU dominance, showcasing a significant increase in CPU efficiency and density. The dialogue highlights the capabilities of Xeon 6 processors in handling complex AI tasks, achieving unprecedented core densities per rack, and enabling the operation of up to 150,000 agents. It emphasizes the importance of tailored solutions for various workloads, ensuring optimal performance and cost-effectiveness for CIOs and data center managers.
Expanding Intelligent Infrastructure: Red Scale Blueprints for Scalable Agentic Workloads
Emphasizing the need for system-level solutions, Intel introduces Red Scale Blueprints, collaborating with partners like Fox Comm and Samba Nova to scale intelligent infrastructure using open standards, exemplified by blueprints for AIC performance and agent density, ensuring customers can scale confidently without proprietary limitations.
Intel and Fox Com Unveil Next-Step Partnership for AI Infrastructure Solutions
Intel and Fox Com announce a collaboration to develop re-scale products using Intel Xeon processors, aiming to create differentiated AI infrastructure solutions. The partnership focuses on integrating, developing, and commercializing technologies to meet diverse AI workload needs, enhancing computing environments for joint customers. This marks a significant milestone in their continued partnership, with plans to unveil more innovations in the near future.
Intel's Ecosystem Approach for AI at Scale: Heterogeneous Computing and Partnership Expansion
The dialogue emphasizes the importance of a diverse ecosystem for AI scalability, highlighting Intel's collaborative efforts with partners like Sampa Nova. It underscores the need for tailored, heterogeneous computing solutions optimized for token consumption and generation, moving away from a one-size-fits-all approach. Intel's recent partnership with Sampa Nova is presented as a strategic move towards delivering advanced AI capabilities, showcasing a commitment to providing scalable, AI-ready infrastructure.
Intel and Samba Nova Collaborate on Advanced AI Inference Solutions and Heterogeneous Disaggregated Inference
Intel and Samba Nova announce a multi-year partnership focusing on high-performance, cost-efficient AI inference solutions using Xeon infrastructure. They introduce the SN 50 Sambar Act, a rack-scale AI infrastructure for agentic workloads, and showcase the world's first heterogeneous disaggregated inference leveraging RDU, Intel CPUs, and Nvidia GPUs, set to ship later this year.
Aggregated Inference: Combining GPUs, RDUs, and CPUs for 2-3x Faster AI Processing
Aggregated inference, utilizing a combination of GPUs, RDUs, and CPUs, has shown to be 2-3 times faster than using GPUs alone, significantly reducing end-to-end latency in AI tasks. This collaborative approach, where CPUs handle tooling execution, GPUs manage prefill and caching, and RDUs decode tokens, demonstrates a substantial performance improvement. The solution has garnered significant customer interest, highlighting its potential in accelerating AI processing.
Exploring Intel and Samba Nova's Rax Utilization Plans with Vista Equity Partners
The discussion focuses on how Vista Equity Partners plan to utilize Rax from Intel and Samba Nova, with an emphasis on strategic applications and potential benefits.
AI-Driven Demand for Computing: Reshaping Silicon and Software Systems
Discussion highlights enterprises' observations on AI's impact, driving demand for computing and transforming silicon, software, and system landscapes.
Revolutionizing Enterprise AI: Intel's Vector Compute 2 & Samba Nova's Air-Cooled Architecture
Intel introduces Vector Compute 2 with Samba Nova, offering the world's first commercially available architecture for disaggregated inference, aiming to transform enterprise AI with reliable, low-latency, low-cost inference at scale, supported by over 50 planned US deployments and partnerships for accelerated inference workloads.
Intel's Entry into Custom Silicon Market: A Journey of Purpose-Built Solutions
Intel highlights its venture into custom silicon, emphasizing partnerships with hyperscalers and telcos. Announces deployment of Intel Processing Unit with Google and next-gen infrastructure silicon for Ericsson, showcasing growth in purpose-built silicon sector. Calls for collaboration at Computex, aiming to align with industry needs and aspirations.
Expanding Partnerships for Custom Silicon Development
Excited about forging alliances to create custom silicon with leading-edge firms and dynamic startups, emphasizing the importance of purpose-built silicon in driving industry innovation.
Revolutionizing AI and Biomedical Engineering: Collaborative Innovations with Intel and Greenstone
A neurosurgeon and co-founder of a neurotech company discusses the development of brain-trained algorithms for speech with Intel, aiming to create AI more efficient and biologically inspired. Meanwhile, the head of cardiology at Stanford and founder of a biotech company partners with Intel to establish scalable reference architectures for life-size imaging, pushing the boundaries of biomedical engineering and AI integration.
Revolutionizing Biomedicine: Intel, Greenstone, and Hitachi Unite for AI-Driven Drug Development
Intel collaborates with Greenstone Biosciences, leveraging AI computing and large-scale biobank data to expedite drug development. Greenstone's biobank of human-induced pluripotent stem cells enables personalized organoid testing, promising quicker and cost-effective medication trials. Hitachi's contributions in foundry tools and quantum computing further accelerate advancements, shaping the future of biomedicine through cutting-edge partnerships.
Expanding Intel-Siemens Partnership for Advanced Computing and Industrial Automation
Intel and Siemens deepen their collaboration, spanning design, manufacturing, and chip applications, integrating AI for enhanced efficiency and sustainability, aiming to revolutionize industrial automation and computing solutions.
Intel's Vision for AI-Driven Future: Collaborative Ecosystem and Technological Advancements
Intel emphasizes its commitment to innovation and collaboration, highlighting achievements in advanced packaging, customer engagement, and the launch of new SoCs. The company underscores its role in shaping the future of computing, particularly in the AI era, and its dedication to continuous engineering and performance improvements. Intel invites partners and customers to join in creating a new era of technology, setting the stage for future events and advancements.
要点回答
Q:What historical event does the speaker reference when talking about the foundation of the semiconductor industry?
A:The speaker references the founding of companies like Intel and Apple, which broadly set the motion for what became known as Silicon Valley, in the context of the historical event nearly six decades ago when a group of engineers and venture capitalists founded these companies.
Q:What role did the speaker play in the semiconductor industry in Taiwan?
A:The speaker was invited by Minister Kat Li to lay the foundation of venture capital concept in Taiwan 40 years ago and started a venture fund with the help of Minister Kat Lee and the Development Fund, contributing to the creation of the semiconductor industry in Taiwan.
Q:How has the Taiwan PC ecosystem contributed to Intel's growth and success?
A:The Taiwan PC ecosystem has played a critical role in Intel's growth and success, as evidenced by the speaker's attendance at Intel's 40th anniversary in Taiwan and the long-standing partnership with suppliers, partners, and customers there.
Q:What is the speaker's approach as the CEO of Intel?
A:As the CEO of Intel, the speaker focuses on execution and brings focus back to the core of the company. Intel is perceived as an engineering company, and the CEO ensures engineering-driven success by reviewing engineering reports and aligning with customers and partners.
Q:What does the speaker indicate about the importance of the PC ecosystem?
A:The speaker indicates that the PC ecosystem is crucial, as Intel has continuously increased the pace of progress across all PC segments with Intel system solutions. This ecosystem is important for driving advancements in various computing segments including workstations, desktops, creators' tools, gamers' setups, and premium and mainstream laptops.
Q:What new product was launched at CES and what are its capabilities?
A:At the Consumer Electronics Show (CES), the speaker launched the Core Ultra Series 3, which is the first product built on an Ana process technology. This product sets a new standard for premium mobile performance and battery life, offering great user experiences across tasks with a fast CPU, improved GPU, low power processing, NPU, and multimedia capabilities. It supports AI and agentic experiences and is designed to transform every PC into an agentic capable platform.
Q:How many designs have been created using the latest Core Ultra IP, and what is the significance of this number?
A:More than 300 designs have been created using the latest Core Ultra IP across consumer and commercial segments. This demonstrates the extensive scaling of capabilities, with the Intel Core Series 3, which was introduced in April, already having scaled up to 70 plus designs. This brings the total series lineup to nearly 400 designs in a few short months, showcasing the massive scale of Intel's product offerings.
Q:What battery life performance does the Intel Core Series 3 offer?
A:The Intel Core Series 3 offers substantial battery life performance, designed to last longer than a typical workday, which can be 10 hours, 12 hours, 14 hours, or more. The exact battery life is not specified, but the emphasis is on providing ample battery life for all user needs.
Q:What is the goal of Series 3 and what market does it aim to impact?
A:The goal of Series 3 is to bring premium feel and experiences to incredibly thin form factors for mainstream PCs, aiming to impact the market with its sleek designs and high performance.
Q:What performance improvements does the Arc G3 provide compared to competition?
A:The Arc G3 is more than 40% faster than the competition and can run all AAA games at 1080p resolution with many at over 120 frames per second.
Q:How is Intel scaling its processors and what markets are they targeting?
A:Intel is scaling its processors, such as the atna process technology, to bring powerful performance and efficiency to the premium, mainstream, and handheld gaming segments. They are targeting the Edge business with their latest series 3 products for large scale business in multiple verticals including manufacturing, robotics, and retail.
Q:What technology and support does Intel provide for edge computing?
A:Intel provides the latest products, technology, chipsets, easy-to-use reference designs, and appropriate software stacks for edge computing.
Q:What is the projected growth of the physical AI market by 2050?
A:The projected growth of the physical AI market is expected to be a 25 trillion market by 2050.
Q:How is the Perplexity Computer's AI operating system beneficial for privacy and cost?
A:The Perplexity Computer's AI operating system creates a team of agents that uses up to 20 different AI models to balance intelligence, accuracy, privacy, and cost. This allows users to run smaller models locally on the Intel Core Ultra Series 3 GPU, maximizing value per watt and per user while maintaining privacy.
Q:What is the significance of the hybrid inference technology developed by Perplexity and Intel?
A:The hybrid inference technology developed by Perplexity and Intel, such as in the case of 'Project Falcon', allows confidential work to be done locally on the device while maintaining privacy, with additional agents spinning up as necessary. This creates a hybrid system that brings all inputs and outputs together while keeping sensitive information on the device, thus providing true hybrid inference orchestration.
Q:What is the future vision for computing according to the speakers?
A:The future vision for computing includes more computing in the data center and more computing on the local machine, as both speakers believe in this architecture and see it as a big milestone for engineering.
Q:What are the two flagship CPU cores offered by Intel, and what are their optimization focuses?
A:The two flagship CPU cores offered by Intel are Pcos and Ecos, with Pcos optimized for performance and Ecos optimized for efficiency.
Q:What is the expected growth in data center power demand for AI inference workloads by 2030?
A:The expected growth in data center power demand for AI inference workloads by 2030 is from 80 GW to about 100 GW.
Q:What is the role of Intel Xeon 6 plus in data centers?
A:The role of Intel Xeon 6 plus in data centers is to deliver efficiency and density, enabling partners to save on real estate, have more compact servers and racks, and support the next era of cloud and network infrastructure.
Q:How does Intel's Xeon 6 plus lineup of data center processors fit into the broader market?
A:Intel's Xeon 6 plus joins the lineup of data center processors, which also includes Xeon 6 based on P cores. These solutions deliver new performance and choice for enterprises whose infrastructure is built on x86 and Xeons.
Q:Why is balancing traditional mission-critical applications and AI workloads crucial for enterprises?
A:Balancing traditional mission-critical applications and AI workloads is crucial for enterprises because they need to prepare for AI while running their day-to-day critical applications.
Q:What is the change in the balance between CPU-led enterprise infrastructure and GPU-heavy AI factories?
A:The change in the balance between CPU-led enterprise infrastructure and GPU-heavy AI factories is becoming more parity as a result of the increasing demand for CPU due to AI.
Q:How does agentic AI differ from traditional AI in terms of compute requirements and infrastructure impact?
A:Agentic AI differs from traditional AI in that it operates by given goals rather than prompts, is iterative and automated, and uses a variety of tools and processes that traditionally fall within the realm of CPUs and x86. This results in a rapid increase in CPU demand for AI and a shift in the ratio of CPU to GPU usage.
Q:What are the findings from the comparison of traditional AI inference setup versus an AI system using Xeon 6 plus processors?
A:The comparison shows that an AI system using Xeon 6 plus processors has a near parity balance between CPU and GPU work, with the CPU being more dominant. Such a system can support up to 150,000 agents per 32 rack units of compute space, showcasing the high density and efficiency of the Xeon 6 plus processors.
Q:What are the characteristics of the workloads that are driving the need for high-performance and high-density, power-efficient cores?
A:Workloads that require very high-performance cores are pushing the frequencies, while there's also a need for very high-density, power-efficient cores to meet the varied demands of different workloads.
Q:What is Red Scale Blueprints, and which companies are involved in this initiative?
A:Red Scale Blueprints is an initiative started by Intel to work with ecosystem partners to develop scale-out solutions built on open standards. It's designed to help customers rapidly scale their intelligent infrastructure with confidence, without relying on proprietary workarounds.
Q:What is the purpose of the partnership between Intel and Foxconn?
A:The purpose of the partnership between Intel and Foxconn is to develop scale-out products built upon Intel Xeon processors. They aim to explore the development, integration, and commercialization of differentiated RA scale AI infrastructure solutions to address diverse AI workload requirements.
Q:What was announced regarding the collaboration between Intel and SambaNova?
A:Intel and SambaNova announced a partnership to provide high-performance, cost-efficient AI inference solutions and demonstrated the world's first heterogeneous disaggregated inference using SambaNova's RDU with Intel's CPU and Nvidia GPUs.
Q:What are the results of the initial tests on the disaggregated inference stack?
A:The initial results of the disaggregated inference tests found that aggregated inference using the disaggregated stack, with Xeon 6 processors, Someone Nova's SM 50 RDS, and Nvidia GPUs, is 2 to 3 times faster than just using GPUs alone.
Q:What customer interest has been observed for the new AI solutions developed?
A:There is tremendous customer interest in the new AI solutions developed by Intel and its partners, indicating strong demand from enterprises for these computing solutions.
Q:What technology has been launched to offer the world's first disaggregated inference architecture?
A:The technology launched is Vector Core, developed with partners at Cambium Capital, offering the world's first commercially available disaggregated inference architecture.
Q:What is the significance of the Samba Nova stack in the new architecture?
A:The Samba Nova stack is significant because it is an air-cooled data center that will deliver reliable, low latency, and low cost inference at scale, which is what enterprise customers and communities are asking for.
Q:Why is selecting the right silicon architecture important for enterprises?
A:Selecting the right silicon architecture is critical for enterprises because large workloads are increasingly becoming strategic assets for companies, and they are seeking silicon that meets their exact needs.
Q:Who are some of the partners involved in developing purpose-built silicon at Intel?
A:Partners involved in developing purpose-built silicon at Intel include Google, Ericsson, and Greenstone Technologies, among others.
Q:What is the collaboration between Intel and Google regarding infrastructure processing units?
A:Intel and Google have entered into a partnership wherein Intel is delivering the infrastructure processing unit (IPU), which is a piece of silicon vital for hyperscaler performance.
Q:How is purpose-built silicon being applied in the biomedical engineering field?
A:In the biomedical engineering field, purpose-built silicon is being applied to emulate the functionality of the human brain, with companies like Echo Neuro Technologies developing neuromorphic computing solutions that are trained on brain activity itself.
Q:What is the relationship between Echo Neuro Technologies and Intel?
A:Echo Neuro Technologies is collaborating with Intel to develop brain-trained algorithms for streaming speech that are efficient and can approach the efficiency of biological computation.
Q:Who is Joseph Wu and what is his connection with Greenstone Technologies?
A:Joseph Wu is a professor of Medicine, director of the Stanford Cardiovascular Institute, and co-founder of Greenstone Biosciences. He is also involved with establishing scalable reference architectures with Greenstone Technologies.
Q:What are Intel and Greenstone doing to accelerate the development of new medicines?
A:Intel and Greenstone are working together by combining state of the art human genetics, biology from Greenstone with advanced AI computing from Intel. This collaboration aims to scale data processing, storage, and analysis to speed up the development of new medicines.
Q:What capabilities does the partnership with Hitachi bring to Intel's efforts?
A:The partnership with Hitachi is enhancing Intel's work in foundry tools and quantum computing systems, which will help accelerate their work and plans in these areas.
Q:How is the collaboration between Intel and Siemens in industrial automation?
A:Intel and Siemens have an expanded partnership that now includes the entire value chain from design to manufacturing to chip applications in Siemens products. They are improving design quality, managing product lifecycle, and focusing on automation, electrification, quality, and sustainability.
Q:What is the significance of the Intel-PC Edge opportunity?
A:The Intel-PC Edge opportunity is a strategic direction for Intel focusing on developing a physical AI data center and emerging intelligence center from silicon to system and applications. This initiative is made possible by a large ecosystem of partners, suppliers, and customers.
Q:What changes are being made at Intel under the leadership of the new CEO?
A:The new CEO at Intel has challenged their team to build a new Intel that is not resting on past glories but is instead working towards the future. This involves being forward-thinking and creating new business opportunities across existing and emerging domains.
Q:What progress has Intel achieved in the past year?
A:Intel has made significant progress in the past year including shipping 18A to high volume with multiple products, executing on advanced packaging milestones, engaging customers and building their foundry business, introducing new SoCs for major compute platforms, and rebuilding and strengthening partnerships across the ecosystem.
play
English
English
进入会议
1.0
0.5
0.75
1.0
1.5
2.0