本網站使用瀏覽器紀錄 (Cookies) 來提供您最好的使用體驗，我們使用的 Cookie 也包括了第三方 Cookie。相關資訊請訪問我們的隱私權與 Cookie 政策。如果您選擇繼續瀏覽或關閉這個提示，便表示您已接受我們的網站使用條款。 關閉
(11月10日，台北訊) 華碩與IBM舉辦《ASUS BUSINESS × IBM Security QRadar EDR聯合記者會》，宣布雙方合作，以內建AI與自動化技術的IBM Security QRadar EDR，作為華碩商用電腦供予企業客戶的端點偵測及回應解決方案。隨著資安議題升溫，企業部署資安設備的需求大增。華碩商用電腦名列世界前茅，IBM則在安全領域擁有全球領導地位，結合優質硬體產品、端點防護軟體、貼心在地服務，帶來史上最佳的商用資安解決方案，引領台灣企業邁向資安新時代！ 華碩代理的IBM Security QRadar EDR解決方案，使用AI與自動化技術、數百種機器學習和行為模型，偵測異常行為，並以幾乎即時的方式回應攻擊；以獨特的方式、從外部監控作業系統，能夠避免駭客的操縱或干擾，幫助企業保護端點、免受「零日攻擊」。 華碩全球副總裁林宗樑指出：「在商業科技的領域，華碩始終追求無與倫比，終極目標是為客戶提供最安全的解決方案。華碩商用與IBM的策略合作將強化彼此的核心競爭力，攜手開創全方位商務版圖，帶動創新和業務成長！」台灣IBM總經理李正屹表示：「我們很榮幸有機會與華碩在資安議題上攜手合作，將台灣作為起點，為企業提供前瞻且獨到的端點安全策略與解決方案。期待未來雙方合作的廣度與深度不斷增加。」 IBM 致力於協助客戶充分運用混合雲架構與AI科技，提升企業生產力與獲利能力。在生成式AI的能力越來越強大的今天，IBM希望讓人「聞駭色變」的資安成為企業積極應用AI的助力，而非阻力。同時，IBM 更積極地將AI技術嵌入資安解決方案中；IBM亞太區總經理譚保羅 (Paul Burton) 強調，「資安威脅當前，AI正是最佳的防禦工具。目前IBM所有安全解決方案與服務組合，例如威脅偵測與回應、身份識別、資料保護等，皆已具備AI能力。」 華碩於商用領域深耕多年，敏銳洞悉企業需求與政策發展動態，提供疾速、靈活、智慧的商務解決方案，深受各級別企業、公家機關信賴。華碩擁有最完整的商用產品線，涵蓋商用筆電、商用桌機、伺服器、工作站與螢幕、主機板等周邊零組件，所有產品經過嚴謹的耐久性測試，安全可靠，確保客戶在商業運作中無後顧之憂。華碩聯合科技系統事業總經理廖逸翔表示：「華碩商用電腦在台穩坐龍頭，累計拿下16年商用市場冠軍！透過與資安第一品牌IBM強強聯手，雙方優勢極大化，掌握趨勢，從台灣市場出發，與6000家經銷夥伴齊心建造安全可靠的IT環境，絕對是台灣企業最強後盾！」 網路攻擊的次數、頻率與複雜度與日俱增；企業需要採用最好的工具進行防禦，讓技術和企業現有環境、應用和業務能完美地結合。IBM以全球資安視角為基礎，提出「開放平台搭載AI與自動化工具」、涵括全部業務領域的宏觀資安營運機制，整合先進技術與服務，提供全球客戶端到端的資安解決方案。李正屹以今年稍早推出、IBM Security旗下的QRadar套件為例：全新的套件建構在開放平台之上，是專為滿足企業混合雲環境的需求而設計，內建先進的AI與自動化技術，具備威脅偵測、事件調查和回應能力；使用統一的現代化介面，有效地提升資安人員處理不斷演化的安全威脅與攻擊事件的速度、精準度與規模。AI技術可以明顯提升SOC的效率與準確性；IBM 資安託管服務已有實證可以自動關閉七成以上的告警，上線第一年即可減少告警分類所需時間達55%。 端點設備部份，華碩商用筆電(ASUS ExpertBook)支援TPM 2.0、指紋及紅外線攝影機臉部辨識、NFC登入，並擁有實體攝影機防窺閘設計；華碩商用桌機(ASUS ExpertCenter)則內建fTPM (韌體TPM)、Kensington防盜鎖槽，亦可選購dTPM 2.0 (獨立TPM晶片)…等，達到軟硬體兼具的全方位安全功能。今年ASUS Expert全面升級第13代Intel® Core™處理器機種，支援最新Intel vPro平台，以硬體協助抵禦勒索軟體、軟體供應鏈攻擊；更有遠近馳名的華碩商用小隊，協助企業進行提供售前規劃、售中諮詢、售後服務，幫助客戶風險評估與降低成本。透過華碩與IBM的合作，客戶可一站式採購頂尖硬體與資安軟體，輕鬆獲得最佳商用防護與服務。未來，雙方將並肩在資安藍海破浪前行，立足台灣、放眼全球！
New capabilities help enterprises, governments, and service providers operationalize digital sovereignty with continuous compliance and control across hybrid environments BOSTON, May 5, 2026 /PRNewswire/ -- At Think 2026, IBM (NYSE: IBM) today announced the general availability of IBM Sovereign Core, a new software platform designed to help organizations build and operate AI-ready sovereign environments and verify their control — giving enterprises and governments an end-to-end approach to digital sovereignty. IBM Corporation logo. As AI adoption accelerates, digital sovereignty has become a critical requirement, extending beyond data residency to include control over infrastructure, operations, and AI systems. Organizations must balance maintaining necessary authority with the pace of innovation, while facing increasing scrutiny from regulators, auditors, and boards. Yet most platforms can struggle to provide consistent, auditable answers to these requirements — creating a gap between policy and operational reality. "AI has made sovereignty a runtime requirement, not a policy statement. With IBM Sovereign Core, organizations don't have to choose between deploying AI at speed and verifying their control. Sovereignty shouldn't be a constraint on innovation — with the right software foundation, it's an enabler of it," said Dinesh Nirmal, SVP, IBM Software. Defining Digital Sovereignty for the AI EraIBM defines digital sovereignty across four pillars: Operational Sovereignty — Control over how environments are operated Data Sovereignty — Control over data at rest, in use, and in motion Technology Sovereignty — Open, modular architecture that avoids vendor lock-in AI Sovereignty — Control over where models run and how inference is governed Together, these pillars form the foundation of IBM's unified approach to digital sovereignty—bringing control across operations, data, technology, and AI. A Unified Approach for Digital SovereigntyIBM Sovereign Core introduces a new model for operational sovereignty, where governance, compliance, and control are built into the system from the start to enable organizations to scale AI while maintaining their sovereignty, trust, and operational independence. IBM Sovereign Core delivers an integrated sovereign software platform that combines control plane, identity, security, compliance, and AI execution functions operate within a single deployment model. Key capabilities include: Customer-operated control plane enabling full authority over configuration, operations, and lifecycle management In-boundary identity, encryption, and data services, ensuring all access, secrets, keys, logs, and audit evidence remain under customer control Continuous compliance monitoring and evidence generation, providing real-time audit readiness Preloaded regulatory frameworks to accelerate company defined compliance postures across regions and industries Governed AI execution, ensuring models, inference, and agent operations run within defined sovereign boundaries Open, modular architecture built on open standards, supporting portability and avoiding vendor lock-in Together, these capabilities establish a sovereign control plane that enables organizations to operate environments and verify their control across data, operations, and AI. Continuous, Verifiable ComplianceIBM Sovereign Core enables organizations to move from static compliance models to dynamic continuous, verifiable compliance models that are verifiable. Integrated monitoring, drift detection, and automated evidence generation allow organizations to: Validate compliance in real time Maintain audit-ready evidence within the sovereign boundary Reduce reliance on manual validation and point-in-time audits This ensures sovereignty is not only defined, but observable, enforceable, and provable at scale. AI Governance Within Your Sovereign BoundaryAs AI systems become central to enterprise operations, governance must extend beyond data to include models, inference, and agent behavior. IBM Sovereign Core enables organizations to deploy and operate AI models, agents, and inference workloads entirely within the sovereign boundary, ensuring their: Control over where AI processing occurs Traceability of model execution and decisions Governance over access, updates, and lifecycle management This ensures that AI systems operate with accountability, transparency, and control, even in highly regulated environments. Extensible Ecosystem and Deployment Flexibility Built on open, enterprise-grade technologies like Red Hat OpenShift and Red Hat AI, IBM Sovereign Core enables organizations to extend existing investments across hybrid and partner environments. IBM Sovereign Core includes an extensible catalog that organizations can curate for their own users, with their own applications, or populated with pre-vetted IBM, third-party and open source software and services from an ecosystem of software and infrastructure partners that includes: AMD, ATOS, Cegeka, Cloudera, Dell, Elastic, HCL, Intel, Mistral, MongoDB, and Palo Alto Networks. CPU, GPU, and AI inference environments can be provisioned using standardized templates and automated configuration profiles, enabling teams to deploy and manage workloads consistently across sovereign regions while maintaining alignment to their compliance and sovereignty requirements. Designed for Regulated Regions, Enterprises and GovernmentsIBM Sovereign Core is designed for organizations that require greater control, flexibility, and compliance across sensitive workloads. Enterprises can run regulated applications and AI workloads within controlled environments Governments and public sector organizations can support sovereign operations for critical services Service providers and regional cloud operators can deliver sovereign cloud and AI services at scale Across these use cases, IBM Sovereign Core enables organizations to innovate with AI while maintaining demonstrable authority over their systems, data, and operations. IBM Sovereign Core is now generally available. Supporting Quotes"Control and compliance have long been barriers to enterprise AI adoption," said Marjorie Janiewicz, Chief Revenue Officer at Mistral AI. "IBM Sovereign Core delivers a ready-to-deploy foundation that allows our models to operate within trusted boundaries from day one, enabling organizations to accelerate AI adoption while maintaining full control over their data. We're pleased to be the first model provider partner for Sovereign Core to certify our frontier models and support enterprise customers globally." "AI is reshaping how nations, governments and enterprises operate, making digital sovereignty not just a policy discussion, but a leadership priority," said Philip Guido, EVP & Chief Commercial Officer, AMD. "As organizations deploy AI at scale, control over technology becomes increasingly critical. Together with IBM and Red Hat, we're enabling clients to exercise greater autonomy and flexibility in their digital sovereignty choices across the full IT stack through an open, secure and transparent approach." "As organizations navigate increasingly complex compliance and regulatory requirements, we're seeing strong demand for digital platforms and software that allows sensitive data to remain within controlled, compliant boundaries," said Gaetan Willems, VP Cloud & Digital Platforms, Cegeka. "Partnering with IBM to offer a pre-architected solution through our in-country environment enables us to deliver enterprise-ready software to our clients, while allowing them to address local compliance standards." "As part of our strategic alliance with IBM, Deloitte is focused on enabling a Sovereign Core technology stack that supports clients in building secure, scalable, and compliant cloud ecosystems. Leveraging IBM's platform capabilities alongside Deloitte's deep regulatory and transformation expertise, we help organizations address critical imperatives such as data localization, security, regulatory compliance, and India's DPDP requirements." — Vinay Prabhakar Chief Commercial Officer Deloitte AP & National Sales & Alliances Leader Deloitte South Asia "India deserves cloud infrastructure that is sovereign by design. At NxtGen, that has always been our operating principle and our promise to customers: superior performance, lower TCO, and zero trade-offs on compliance. The IBM Sovereign Core collaboration is a direct expression of that commitment, bringing secure, compliant cloud capabilities built for India's regulatory reality." — A S Rajgopal, MD & CEO, NxtGen "The ability to establish—and prove—control over data, models and operations is quickly becoming a differentiator for trust, resilience and innovation in AI adoption," said Bill Pearson, vice president of data center and AI software, Intel Data Center Group. "Establishing sovereign AI frameworks, spanning open hybrid architectures and transparent governance, requires a foundation built for performance, security, and control. Intel continues to innovate with technologies such as Intel Xeon 6 processors and Intel TDX to provide in‑use protection for data and AI workloads across environments with CPU and GPUs to support sovereign infrastructure initiatives like IBM Sovereign Core." Supporting MaterialsBlog: IBM Sovereign Core: The new end to end system for Sovereign AIIBM Sovereign Core Product PageIBM Sovereign Core Ecosystem CatalogIBM Sovereign Core Compliance CenterStatement of Direction for IBM Sovereign Core About IBMIBM is a leading provider of global hybrid cloud and AI, and consulting expertise. We help clients in more than 175 countries capitalize on insights from their data, streamline business processes, reduce costs and gain the competitive edge in their industries. Thousands of governments and corporate entities in critical infrastructure areas such as financial services, telecommunications and healthcare rely on IBM's hybrid cloud platform and Red Hat OpenShift to affect their digital transformations quickly, efficiently and securely. IBM's breakthrough innovations in AI, quantum computing, industry-specific cloud solutions and consulting deliver open and flexible options to our clients. All of this is backed by IBM's long-standing commitment to trust, transparency, responsibility, inclusivity and service.Visit www.ibm.com for more information. Media contact: Kate LehmanIBMKate.Lehman@ibm.com
AI賦能的全新諮詢服務提供安全穩健的技術平台、統一共享標準及可重複應用的AI資產，助企業加速增長、驅動創新。 香港2026年1月22日 /美通社/ -- IBM 近日正式推出IBM Enterprise Advantage服務。這項業界首創的資產型諮詢服務，整合了經實戰驗證的AI工具與專業能力，協助企業快速構建、合規管治並大規模量產運營專屬定制的企業內部AI平台。 IBM推出IBM Enterprise Advantage服務 現時，企業可憑藉IBM Enterprise Advantage重構工作流程、打通AI與現有系統的對接環節、大規模擴展新型智能體應用，而且無須調整旗下雲服務供應商、AI模型或核心基礎設施——不論採用的是Amazon Web Services、Google Cloud、Microsoft Azure、IBM watsonx等主流雲平台，還是開源或閉源模型，企業均可基於現有技術投資無縫拓展業務。 IBM Enterprise Advantage匯集了IBM諮詢顧問的技術實力與行業洞察，其底層架構源自IBM內部AI驅動的交付平台——IBM Consulting Advantage。該平台已構建起日漸豐富的垂直行業AI智能體與應用生態，服務超過150個客戶項目，經實戰驗證最高可提升顧問工作效率達50%，助客戶更快收穫業務成果。 如今，IBM透過Enterprise Advantage服務，將這套經過實戰檢驗的方法論和技術能力開放予客戶，助其打造專屬AI平台，縱橫複雜的AI市場，釋放企業核心價值。 IBM Enterprise Advantage 以終身學習教育龍頭企業培生集團（Pearson）為例，該集團運用此項服務打造定制化AI驅動平台，透過「人類專業知識能力+智能體助手」的協同模式，提升日常工作效率，並強化數據驅動的決策能力。 某製造業企業依托Enterprise Advantage服務推進生成式AI戰略落實，具體舉措包括挖掘高價值應用場景、開展定向原型驗證，推動管理層就「大規模量產、平台優先」的核心戰略達成共識。基於這項解決方案，該客戶已透過多技術融合，在安全合規、可控可管的環境中完成AI助手的部署上線，為生成式AI在全企業範圍內的大規模量產推廣與深度應用奠定堅實基礎。 「眾多企業正積極佈局人工智能領域，但如何實現大規模量產的商業價值，仍是一大核心挑戰。」IBM高級副總裁兼IBM諮詢業務主管Mohamad Ali表示：「IBM早已在內部攻克諸多類似難題——憑藉人工智能變革內部運營模式，取得可量化的顯著成效，並由此形成一套經實戰驗證的實戰指南，助客戶邁向成功。Enterprise Advantage解決方案正是將這一框架賦能予客戶，透過整合專業人才能力、數碼員工及即開即用型AI資產，協助客戶信心滿滿地實現AI大規模量產部署，創造可見的實質效益。」 Enterprise Advantage服務現已正式上線，點擊此處可体验。 關於IBM IBM 是全球領先的混合雲、人工智能及企業服務提供商，幫助超過 175 個國家和地區的客戶，從其擁有的數據中獲取商業洞察，簡化業務流程，降低成本，並獲得行業競爭優勢。金融服務、電信和醫療健康等關鍵基礎設施領域的數千家政府和企業實體依靠 IBM 混合雲平台和紅帽 OpenShift 快速、高效、安全地實現數字化轉型。IBM 在人工智能、量子計算、行業雲解決方案和企業服務方面的突破性創新為我們的客戶提供了開放和靈活的選擇。對企業誠信、透明治理、社會責任、包容文化和服務精神的長期承諾是 IBM 業務發展的基石。瞭解更多信息，請訪問： www.ibm.com/查詢更多本地資訊，請訪問IBM香港新聞間：https://hongkong.newsroom.ibm.com/ 傳媒查詢：崔守峯 shou.feng.cui@ibm.com
SINGAPORE, March 21, 2025 /PRNewswire/ -- Today, IBM (NYSE: IBM) announced the Asia Pacific geography winners of the 2025 IBM Partner Plus Awards, which celebrate IBM partners who advance businesses, reshape industries and create positive impact through partnership. 2025 IBM Partner Plus Awards -- Asia Pacific Winners "The award nominations we received in APAC were truly inspiring, reflecting the commitment, impact, and innovation our partners are delivering across industries. The winners exemplify the power of combining deep expertise with IBM technology to drive meaningful transformation and deliver exceptional outcomes for clients," said Chetan Krishnamurthy, General Manager, Partner Ecosystem, IBM APAC. The APAC geography winners of this regional round in the IBM Partner Plus Awards for each category include: AI for Business in APACTech Mahindra Automation in APACOctane Software Solutions Hybrid by Design in APACTruis Modernization in APACGS Lab | GAVS Sustainability in APACBeijing Shuto Technology Workforce Productivity in APACAppolo Computers The APAC geography winners have also been named as finalists for the global awards in each category, winners of which will be announced at IBM Partner Plus Day on May 5, 2025 during Think 2025 in Boston. Partners self-nominated for the IBM Partner Plus Awards and entries were judged on Expertise, Partnership, Impact, Innovation, and Storytelling. Winners were selected across six categories in five geographies: APAC, EMEA, Japan, Latin America and North America. For more information on the IBM Partner Plus Awards, including a list of all geography award recipients, visit https://www.ibm.com/partnerplus/awards. Partners eligible to self-nominate for the awards are part of IBM's Partner Plus, a program designed to help deepen partners' technical expertise, accelerate time to market and win with clients with AI and hybrid cloud. For more information on IBM Partner Plus, please visit www.ibm.com/partnerplus. About IBMIBM is a leading provider of global hybrid cloud, artificial intelligence, and consulting expertise. We help clients in over 175 countries harness insights from their data, optimize business processes, reduce costs and gain a competitive advantage in their industries. More than 4,000 government and corporate entities in critical infrastructure areas such as financial services, telecommunications and healthcare rely on the IBM hybrid cloud platform and Red Hat OpenShift to achieve their digital transformations quickly, efficiently and securely. IBM's breakthrough innovations in artificial intelligence, quantum computing, industry-specific cloud solutions and consulting offer open and flexible options to our clients. All of this is backed by IBM's long-standing commitment to trust, transparency, accountability, inclusion and service. Visit www.ibm.com for more information. Media Contact:Paranee ReymondonExternal Relations Leader, IBM Asia PacificEmail: paranee@th.ibm.com IBM Corporation logo.
IBM Guardium Data Security Center provides unified, SaaS-first data security capabilities amid hybrid cloud, AI, and quantum transformation New software part of IBM's comprehensive approach for protecting hybrid cloud and AI ARMONK, N.Y., Oct. 22, 2024 /PRNewswire/ -- As hybrid cloud-, AI-, and quantum-related risks upend the traditional data security paradigm, IBM (NYSE: IBM) is launching IBM Guardium Data Security Center – allowing organizations to protect data in any environment, throughout its full lifecycle, and with unified controls. IBM Guardium Data Security Center IBM Guardium Data Security Center provides a common view of organizations' data assets, empowering security teams to integrate workflows and address data monitoring and governance, data detection and response, data and AI security posture management, and cryptography management together in a single dashboard. IBM Guardium Data Security Center includes generative AI capabilities to help generate risk summaries and boost security professionals' productivity. The center features IBM Guardium AI Security, software to help protect organizations' AI deployments from security vulnerabilities and data governance policy violations at a time when generative AI adoption – and the risk of "shadow AI," the presence of unsanctioned models – is surging. IBM Guardium Data Security Center also features IBM Guardium Quantum Safe, software that helps clients to protect encrypted data from the potential risk of future cyberattacks driven by bad actors who gain access to cryptographically relevant quantum computers. IBM Guardium Quantum Safe builds upon expertise from IBM Research – including IBM's post-quantum cryptography algorithms – and IBM Consulting. "Generative AI and quantum computing provide immense opportunities, but they also bring new risks," says Akiba Saeedi, Vice President, IBM Security Product Management. "During this transformative time, organizations need to improve their crypto-agility and carefully monitor their AI models, training data, and usage. IBM Guardium Data Security Center – with its AI Security, Quantum Safe, and other integrated capabilities – provides comprehensive risk visibility." IBM Guardium Quantum Safe helps organizations gain visibility and manage enterprise cryptographic security posture to address vulnerabilities and guide remediation. It allows organizations to enforce policies based on external, internal, and government regulations by pulling crypto algorithms used in code, vulnerabilities detected in code, and network usages into a single dashboard for security analysts to monitor policy violations and track progress – without having to piece together information distributed across various systems, tools, and departments. Guardium Quantum Safe offers customizable metadata and flexible reporting so that critical vulnerabilities can be prioritized for remediation. Learn more about IBM Guardium Quantum Safe IBM Guardium AI Security manages security risk and data governance requirements for sensitive AI data and AI models. It helps discover AI deployments, address compliance, mitigate vulnerabilities, and protect sensitive data in AI models through a common view of data assets. IBM Guardium AI Security integrates with IBM watsonx and other generative AI SaaS providers. For example, IBM Guardium AI Security helps discover "shadow AI" models and then shares them with IBM watsonx.governance, so they no longer elude governance. Learn more about IBM Guardium AI Security An integrated approach for a transformative time Hybrid cloud-, AI-, and quantum-era risks mean key data – from medical records and financial transactions to IP and critical infrastructure – require new forms of protection. During this transformative time, organizations urgently need a trusted partner and an integrated approach to data security – not a patchwork of point solutions. IBM is pioneering this integrated approach. IBM Guardium Quantum Safe dovetails with broader Quantum Safe offerings from IBM Consulting and Research. The software is powered by technology and research developed by IBM Research. Several of IBM Research's post-quantum cryptography algorithms were recently standardized by the U.S. National Institute of Standards and Technology (NIST), marking a crucial milestone to protect the world's encrypted data from the risk of bad actors who may gain access to cryptographically relevant quantum computers to carry out cyberattacks in the future. IBM Consulting's Quantum Safe Transformation Services leverage these technologies to help organizations define risks, inventory and prioritize them, confront them – and then scale the process. IBM Consulting's cybersecurity bench includes scores of professionals with experience in cryptography and quantum safe technology. Dozens of clients in telecommunications, finance, government, and other industries leverage IBM Quantum Safe Transformation Services to help safeguard their organizations against future risks and present risks like harvest now, decrypt later. IBM today is also adding decentralized identity features to its Verify portfolio: IBM Verify Digital Credentials enables users to store and manage their own credentials. The feature digitizes physical credentials like drivers' licenses, insurance cards, loyalty cards, and employee badges, which can then be standardized, stored, and shared with comprehensive security, privacy protection, and control. IBM Verify is an IAM (identity access management) solution that protects identity across the hybrid cloud. Statements regarding IBM's future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only. About IBMIBM is a leading provider of global hybrid cloud and AI, and consulting expertise. We help clients in more than 175 countries capitalize on insights from their data, streamline business processes, reduce costs and gain the competitive edge in their industries. More than 4,000 government and corporate entities in critical infrastructure areas such as financial services, telecommunications and healthcare rely on IBM's hybrid cloud platform and Red Hat OpenShift to affect their digital transformations quickly, efficiently and securely. IBM's breakthrough innovations in AI, quantum computing, industry-specific cloud solutions and consulting deliver open and flexible options to our clients. All of this is backed by IBM's long-standing commitment to trust, transparency, responsibility, inclusivity and service. For more information, visit www.ibm.com. Media contact: Kevin Zawacki Communications, IBM Software kevin.zawacki@ibm.com 914-837-4333 Learn more about IBM Guardium Quantum Safe Learn more about IBM Guardium AI Security IBM Corporation logo.
新聞要點： 新一代IBM Telum II 處理器和 IBM Spyre 加速器可釋放更多企業級 AI 能力，包括大語言模型和生成式 AI 先進的 IO 技術實現並簡化可擴展的 IO 子系統，進一步降低能耗和數據中心佔地面積 香港2024年9月2日 /美通社/ -- 近日，IBM（紐約證券交易所：IBM）在 Hot Chips 2024大會上公佈了即將推出的 IBM Telum® II 處理器和 IBM Spyre™ 加速器的架構細節。這些新技術旨在大幅擴展下一代 IBM Z 大型主機系統的處理能力，通過新的 AI 集成方法，加速企業對傳統 AI 模型和大語言 AI 模型的協同使用。 IBM 重磅推出全新Telum處理器，依託新一代 IBM Z 大型主機加速 AI應用 隨著基於大語言模型的 AI 項目從概念驗證階段進入生產階段，企業對高能效、高安全性和高度可擴展解決方案的需求日益迫切。摩根士丹利最近發佈的一份研究報告預測，在未來幾年，生成式 AI 的電力需求將以每年 75% 的速度激增，其 2026 年的能耗或將與西班牙 2022 年的全年能耗相當。許多 IBM 客戶表示，支持適當規模的基礎模型和針對 AI 工作負載的混合架構越來越重要。 此次IBM發佈的主要創新技術包括： IBM Telum II 處理器：這一全新芯片將搭載於下一代 IBM Z 系列主機，與第一代 Telum 芯片相比，其頻率和內存容量均有提升，高速緩存提升40%；集成 AI 加速器內核和數據處理單元 (DPU) 的性能也得到改善。IBM Telum II處理器將支持大語言模型驅動的企業計算解決方案，滿足金融等行業的複雜交易需求。 IO 加速單元：Telum II 處理器芯片上的全新數據處理單元 (DPU) 旨在加速大型主機上用於聯網和數據存儲的複雜 IO 協議，可簡化系統操作，提高關鍵組件性能。 IBM Spyre加速器：可提供額外的AI計算能力，與 Telum II 處理器相得益彰。Telum II 和 Spyre 芯片共同構成了一個可擴展的架構，可支持AI集成建模方法，即將多個機器學習或深度學習的AI模型與基於編碼器的大語言模型相結合。通過利用每個模型架構的優勢，AI集成的方法可以生成比單個模型更準確、更穩健的結果。Spyre 加速器芯片在 Hot Chips 2024 大會期間進行了預覽，並將作為Telum II 處理器的附加選件提供。每個加速器芯片均與IBM 研究院合作開發，通過一個 75 瓦 PCIe 適配器連接。與其他 PCIe 卡一樣，Spyre 加速器可根據客戶需求進行擴展。 IBM主機和 LinuxONE 產品管理副總裁 Tina Tarquinio 表示：「得益於IBM強大的多代並行的開發路線圖，我們得以在保持技術領先的同時，滿足企業不斷升級的 AI 需求。Telum II 處理器和 Spyre 加速器旨在提供安全、節能、高性能的企業計算解決方案。這些多年研發的創新成果將被引入下一代 IBM Z 平台，幫助客戶大規模利用大語言模型和生成式 AI技術。」 Telum II 處理器和 IBM Spyre 加速器將由 IBM 的長期合作夥伴三星晶圓代工（Samsung Foundry）生產，採用其高性能、高能效的 5 納米工藝節點。二者將共同支持企業的先進AI 用例，釋放業務價值，從而創造新的競爭優勢。利用AI集成的方法，客戶可以更快、更準確地獲得預測結果。適用的生成式 AI用例包括： 保險理賠欺詐檢測：通過AI集成方法將大語言模型與傳統神經網絡相結合，以提高性能和準確性，增強對保險理賠欺詐的檢測。 反洗錢高級監測：對可疑金融活動進行高級檢測，支持遵守監管要求並降低金融犯罪風險。 AI 助理：加速應用生命週期、知識和專業技能的傳授、代碼解釋和轉換等。 規格和性能指標： Telum II 處理器：配備八個運行頻率達 5.5GHz的高性能內核，每個內核配備 36MB二級高速緩存，片上高速緩存容量增加 40%（總容量達 360MB）。每個處理器抽屜的虛擬 L4 高速緩存為 2.88GB，相比上一代增加 40%。集成的 AI 加速器可實現低延遲、高吞吐量的交易中 AI 推理，例如增強金融交易期間的欺詐檢測，並且每塊芯片的計算能力是上一代的四倍。Telum II 芯片中集成了最新的 IO 加速單元 DPU。在設計上，其IO 密度提高 50%，可大幅提高數據處理能力，進一步提高 IBM Z 的整體效率和可擴展性，使其成為處理大規模AI工作負載和數據密集型應用的不二之選。 Spyre 加速器：這是一款專為複雜 AI 模型和生成式 AI 用例提供可擴展功能的企業級加速器。它有高達 1TB 的內存，可在普通 IO 抽屜的八塊卡上串聯工作，以支持大型主機的整體 AI 工作負載，同時每塊卡的功耗不超過 75W。每塊芯片由 32 個計算內核組成，支持 int4、int8、fp8 和 fp16 數據類型，適用於低延遲和高吞吐量的 AI 應用。 產品時間表 作為 IBM 下一代 IBM Z 和 IBM LinuxONE 平台的中央處理器，Telum II 處理器預計在 2025 年向 IBM Z 和 LinuxONE 客戶提供。IBM Spyre 加速器仍在技術預覽階段，預計也將於 2025 年推出。 關於 IBM 未來方向和意向的聲明可能會隨時更改或撤銷，恕不另行通知。 訪問下方鏈接瞭解更多： Telum II：https://www.ibm.com/blog/announcement/telum-ii IBM Spyre：https://research.ibm.com/blog/spyre-for-z IO 加速器：https://community.ibm.com/community/user/ibmz-and-linuxone/blogs/michael-becht/2024/08/23/io-accelerator?communityKey=e7b7d299-8509-4572-8cf1-c1112684644f 關於 IBM IBM 是全球領先的混合雲與人工智能、以及企業服務提供商，為全球175個國家和地區的客戶服務，幫助企業把握其數據洞察、簡化業務流程、降本增效，獲得行業競爭優勢。 IBM 混合雲平台和紅帽OpenShift 為全球超過4,000家政府和企業機構的關鍵性基礎設施提供有力支撐，例如來自金融服務、電訊和醫療健康等行業的客戶，幫助他們快速、高效、安全地實現數碼轉型。 IBM 在人工智能、量子運算、特定行業的雲解決方案以及企業服務等方面的突破性創新，使其可以為客戶提供開放和靈活的選擇。 IBM 對信任、透明、責任、包容和服務的歷久彌新的承諾，是我們業務發展的基石。查詢更多資料，請瀏覽：www.ibm.com/ 關於 IBM 香港，請登入Facebook 頁面 www.facebook.com/IBMHongKong 傳媒查詢：郭韜 gguotao@cn.ibm.com
A12 藝術空間
IBM
請先登入後才能發佈新聞。
還不是會員嗎？立即 加入台灣產經新聞網會員 ，使用免費新聞發佈服務。 （服務項目） （投稿規範）
