NEWS: 公告在東京證券交易所JASDAQ標準市場新上市

市場調查報告書

軟體定義,硬體設備支援:資料中心的超越CPU的多樣化

Software Defined, Hardware Assisted: Diversifying Beyond the CPU in the Datacenter

出版商 IDC 商品編碼 977437
出版日期 內容資訊 英文 13 Pages
訂單完成後即時交付
價格
軟體定義,硬體設備支援:資料中心的超越CPU的多樣化 Software Defined, Hardware Assisted: Diversifying Beyond the CPU in the Datacenter
出版日期: 2020年12月18日內容資訊: 英文 13 Pages
簡介

本報告介紹了資料壓縮和加密之類的傳統元素,以CPU,GPU,FPGA和ASIC格式加速資料中心計算的新興方法,例如AI,安全性,媒體編碼和加密貨幣挖掘。 我們分析計算加速器,主要供應商,加速器可以採用的各種格式,以及使軟體開發人員更輕鬆,更一致地實現這些計算加速器的解決方案。

執行概述

新的市場發展及動態

  • IDC加速器的分類
  • 資料中心工作負載的加速器
    • 中央處理器(CPU)
    • 圖形處理單位(GPU)
    • 現場可程式化閘陣列(FPGA)
    • 特定用途的積體電路(ASIC)
  • 主要的資料中心處理器供應商
    • AMD
    • IBM
    • Intel
    • Nvidia
  • 資料中心加速器的實行
    • 廣告界內面板
    • 具備附加功能塊的擴張CPU
    • 多晶片CPU包裝
    • 功能加速器卡
  • 資料中心運算加速器的軟體開發
    • AMD
    • Intel
    • Nvidia

對技術供應商的建議

參考資料

  • 相關調查
  • 摘要
目錄
Product Code: EUR145760520

This IDC Market Perspective analyzes the emerging approaches to accelerating datacenter compute around key workloads such as AI, security, media encoding, and cryptocurrency mining as well as long-established elements such as data compression and encryption. It looks across CPU, GPU, FPGA, and ASIC forms of compute acceleration, the key vendors supplying the market, the different forms such accelerators can take, and the solutions that enable software developers to target these compute accelerators more easily and consistently."As the world increasingly moves to digital solutions, many emerging data driven workloads are finding themselves performance constrained using traditional server architectures and CPUs," said Andrew Buss, research director, IDC European Enterprise Infrastructure. "Increasingly, software is targeting the deployment to datacenter compute accelerators such as GPUs, FPGAs, and ASICs to reach levels of performance in a rack that would otherwise require an entire data hall to achieve."

Executive Snapshot

New Market Developments and Dynamics

  • IDC Accelerator Taxonomy
  • Datacenter Workload Accelerators
    • Central Processing Units (CPUs)
    • Graphical Processing Units (GPUs)
    • Field-Programmable Gate Arrays (FPGAs)
    • Application-Specific Integrated Circuits (ASICs)
  • Leading Datacenter Processor Vendors
    • AMD
    • IBM
    • Intel
    • Nvidia
  • Implementations of datacenter accelerators
    • Add-in Boards
    • Enhanced CPUs with additional functional blocks
    • Multichip CPU Packages
    • Function Accelerator Cards
  • Software Development for Datacenter Compute Accelerators
    • AMD
    • Intel
    • Nvidia

Advice for the TECHNOLOGY SUPPLIERS

Learn More

  • Related Research
  • Synopsis