½ÃÀ庸°í¼­
»óǰÄÚµå
1648086

IDCÀÇ ¿ùµå¿ÍÀÌµå ÆÛÆ÷¸Õ½º ÀÎÅٽúê ÄÄÇ»ÆÃ(HPC, AI, ¾Ö³Î¸®Æ½½º) ÀÎÇÁ¶ó¿Í ¼­ºñ½º ºÐ·ù¹ý(2025³â)

IDC¢¥s Worldwide Performance-Intensive Computing (HPC, AI, and Analytics) Infrastructure and Services Taxonomy, 2025

¹ßÇàÀÏ: | ¸®¼­Ä¡»ç: IDC | ÆäÀÌÁö Á¤º¸: ¿µ¹® 36 Pages | ¹è¼Û¾È³» : Áï½Ã¹è¼Û

    
    
    



¡Ø º» »óǰÀº ¿µ¹® ÀÚ·á·Î Çѱ۰ú ¿µ¹® ¸ñÂ÷¿¡ ºÒÀÏÄ¡ÇÏ´Â ³»¿ëÀÌ ÀÖÀ» °æ¿ì ¿µ¹®À» ¿ì¼±ÇÕ´Ï´Ù. Á¤È®ÇÑ °ËÅ並 À§ÇØ ¿µ¹® ¸ñÂ÷¸¦ Âü°íÇØÁֽñ⠹ٶø´Ï´Ù.

ÀÌ IDCÀÇ Á¶»ç´Â ÆÛÆ÷¸Õ½º ÀÎÅٽúꡤÄÄÇ»ÆÃ(PIC) °æÀï ½ÃÀåÀÇ ±Ô¸ð¸¦ »êÁ¤ÇÏ°í ¿¹ÃøÇϱâ À§ÇÑ Åä´ë¸¦ ¸¶·ÃÇÕ´Ï´Ù. PIC´Â ºü¸£°Ô ¼ºÀåÇÏ´Â µÎ °¡Áö ÁÖ¿ä º¹ÇÕ ¿öÅ©·ÎµåÀÎ ¸ðµ¨¸µ ¹× ½Ã¹Ä·¹À̼Ç(M&S, Àϸí HPC)°ú Àΰø Áö´É(AI)¿¡ ´ëÇÑ ÀÎÇÁ¶ó Á¢±Ù ¹æ½ÄÀ» °áÇÕÇÕ´Ï´Ù. ¶ÇÇÑ ¿ÂÇÁ·¹¹Ì½º ¹× ¿ÀÇÁÇÁ·¹¹Ì½º ÀÎÇÁ¶ó, ÀÚü °ü¸®Çü ¹× ¼­ºñ½ºÇü ¹èÆ÷¸¦ °áÇÕÇÕ´Ï´Ù. ÄÄÇ»ÆÃ, ½ºÅ丮Áö ¹× ¿¬°á ÆÐ·¯´ÙÀÓÀÎ PIC´Â ¼öÇÐÀûÀ¸·Î ¸Å¿ì º¹ÀâÇÑ ¸í·ÉÀ» ½ÇÇàÇϰųª ´ë·®ÀÇ µ¥ÀÌÅÍ¿¡ ´ëÇØ ºñ±³Àû °£´ÜÇÑ ¸í·ÉÀ» ¼öÇàÇÏ´Â °¡Àå °­·ÂÇϰí È¿À²ÀûÀÎ ¹æ¹ýÀ» Á¦°øÇÕ´Ï´Ù. "Áö³­ ¸î ³â µ¿¾È µðÁöÅÐ Á¶Á÷ÀÇ ¸¹Àº »ç¿ë »ç·Ê¿¡¼­ ¼öÇÐÀûÀ¸·Î Áý¾àÀûÀÎ °è»êÀÌ Áß¿äÇØÁö¸é¼­ ÀϹÝÀûÀÎ ÀÎÇÁ¶ó ÆÐ·¯´ÙÀÓÀÌ µîÀåÇß½À´Ï´Ù."¶ó°í IDC ÀÎÇÁ¶ó ½Ã½ºÅÛ, Ç÷§Æû ¹× ±â¼ú ±×·ìÀÇ ¿¬±¸ ¸Å´ÏÀúÀÎ Madhumitha Sathish´Â ¸»ÇÕ´Ï´Ù. "IDC´Â ¼º´É Áý¾àÀû ÄÄÇ»ÆÃÀÌ °ø±Þ¾÷ü°¡ ÀÌ·¯ÇÑ »ç¿ë »ç·Ê¿¡ º¸´Ù ´É·üÀûÀÌ°í ¸ñÇ¥ ÁöÇâÀûÀÎ ¹æ½ÄÀ¸·Î Á¢±ÙÇÒ ¼ö ÀÖ´Â ±âȸ¸¦ Á¦°øÇÑ´Ù°í »ý°¢ÇÕ´Ï´Ù."

IDCÀÇ ¿ùµå¿ÍÀ̵å PIC(HPC, AI, ºÐ¼®) ÀÎÇÁ¶ó¿Í ¼­ºñ½º ºÐ·ù

PIC(HPC, AI, ºÐ¼®) ÀÎÇÁ¶ó¿Í ¼­ºñ½º ºÐ·ù º¯°æ(2025³â)

ºÐ·ù °³¿ä

  • Å×Å©³î·¯Áö °ø±Þ¾÷ü¿¡ ´ëÇÑ ¾îµå¹ÙÀ̽º
    • ÆÛÆ÷¸Õ½º¸¦ Àü·ÂÀ¸·Î Ãß±¸
    • ÆÛÆ÷¸Õ½º Á߽à ÄÄÇ»ÆÃ ¼ÒÇÁÆ®¿þ¾î ½ºÅÃÀ» ÆÐŰÁöÈ­
    • ÀúÁö¿¬ »óÈ£ Á¢¼Ó°ú ³×Æ®¿öÅ© Áö¿ø
    • ÇÏÀ̺긮µå ¾îÇÁ·ÎÄ¡

Á¤ÀÇ

  • ÆÛÆ÷¸Õ½º Á߽à ÄÄÇ»ÆÃÀ̶õ ¹«¾ùÀΰ¡?
    • ÆÛÆ÷¸Õ½º Á߽à ÄÄÇ»ÆÃ ¿öÅ©·Îµå
      • º¹ÇÕ ¿öÅ©·Îµå
    • IDCÀÇ ±â¾÷ ¿öÅ©·Îµå¿ÍÀÇ Á¤ÇÕ
  • ÆÛÆ÷¸Õ½º Á߽à ÄÄÇ»ÆÃ ÀÎÇÁ¶ó ºÎ¹®
  • ÀÎÇÁ¶ó Á¶´Þ°ú µµÀÔ
    • º¥´õ À¯Çü
    • ±¸ÀÔÀÚ À¯Çü
    • ÄÁÆ®·Ñ Ç÷¹ÀÎ
    • µµÀÔ Àå¼Ò
    • °ü¸® À¯Çü
      • Á¶´Þ À¯Çü
  • ÀÎÇÁ¶ó Çϵå¿þ¾î
    • ÀÎÇÁ¶ó Çϵå¿þ¾î : ÄÄÇ»ÆÃ
      • ÄÚÇÁ·Î¼¼¼­(°¡¼Ó ÄÄÇ»ÆÃ)
    • ÀÎÇÁ¶ó Çϵå¿þ¾î : ½ºÅ丮Áö
      • ³»ºÎ ½ºÅ丮Áö(¼­¹ö ±â¹Ý ½ºÅ丮Áö Ç÷§Æû ¶Ç´Â ½ºÅ丮Áö Áý¾àÇü ¼­¹ö)
    • ÅëÇÕ ÀÎÇÁ¶ó
      • ÅëÇÕ ÀÎÇÁ¶ó
      • ÇÏÀÌÆÛÄÁ¹öÁöµå ÀÎÇÁ¶ó
      • ÄÄÆ÷Àúºí ÀÎÇÁ¶ó
    • ÀÎÇÁ¶ó ¼ÒÇÁÆ®¿þ¾î
      • ¹°¸® ¹× °¡»ó ÄÄÇ»ÆÃ ¼ÒÇÁÆ®¿þ¾î
      • ½ºÅ丮Áö ¼ÒÇÁÆ®¿þ¾î
      • ½Ã½ºÅÛ °ü¸® ¼ÒÇÁÆ®¿þ¾î
      • ±âŸ ÀÎÇÁ¶ó ¼ÒÇÁÆ®¿þ¾î
      • PIC Àü¿ë ¹× ÆÛºí¸¯ Ŭ¶ó¿ìµå ¼­ºñ½º
      • PIC ÀÎÇÁ¶ó ȯ°æ ¼Ó¼º
      • ¿öÅ©·Îµå Ãß»óÈ­
        • º£¾î¸ÞÅ»
        • °¡»óÈ­
        • ÄÁÅ×À̳ÊÈ­
      • ¿öÅ©·Îµå °³¿ä
        • °è»ê Áý¾àÇü
        • ¸Þ¸ð¸® Áý¾àÇü
        • µ¥ÀÌÅÍ Áý¾àÇü
        • ³×Æ®¿öÅ© Áý¾àÇü
        • ÇÏÀ̺긮µå
      • º´·ÄÈ­(Ŭ·¯½ºÅÍ)
        • ÀüÅëÀû
        • Ŭ¶ó¿ìµå ³×ÀÌÆ¼ºê
      • ½ºÄÉÀϸµ(Ŭ·¯½ºÅÍ)
        • ·¹º§ 1 : ³ëµå³»(ÇÁ·Î¼¼¼­)
        • ·¹º§ 2 : ³ëµå³»(°¡¼Ó±â)
        • ·¹º§ 3 : ³ëµå³»(Ŭ·¯½ºÅÍ)
        • ·¹º§ 4 : ³ëµå³»(µ¥ÀÌÅͼ¾ÅÍ)
        • ·¹º§ 5 : ³ëµå³»(Áö¸®ÀûÀ¸·Î ºÐ»ê)
      • »çÀÌÁî(Ŭ·¯½ºÅÍ)
      • ³ëµå³» Åë½Å ÇÁ·ÎÅäÄÝ(Ŭ·¯½ºÅÍ)
      • HPC¿Í AI °íÀ¯ ¼Ó¼º
        • HPC °íÀ¯ ¼Ó¼º
        • AI °íÀ¯ ¼Ó¼º

°ü·Ã ½ÃÀå

Âü°í ÀÚ·á

  • °ü·Ã Á¶»ç
  • ¿ä¾à
KSA 25.02.26

This IDC study lays the groundwork for sizing and forecasting the performance-intensive computing (PIC) competitive market. PIC combines infrastructure approaches across two major and fast-growing composite workloads: modeling and simulation (M&S aka HPC) and artificial intelligence (AI). It also combines infrastructure on premises and off premises and self-managed and as-a-service deployments. As a compute, storage, and connectivity paradigm, PIC offers the most powerful and efficient way to execute mathematically intensive, very complex instructions or to perform a relatively simple instruction on massive amounts of data."In the past few years, a common infrastructure paradigm has emerged, thanks to the importance of mathematically intensive computations in many use cases found in digital organizations," said Madhumitha Sathish, research manager at IDC's Infrastructure Systems, Platforms and Technologies Group. "IDC believes that performance-intensive computing offers an opportunity for vendors to approach these use cases in a more streamlined and targeted fashion."

IDC's Worldwide PIC (HPC, AI, and Analytics) Infrastructure and Services Taxonomy

PIC (HPC, AI, and Analytics) Infrastructure and Services Taxonomy Changes for 2025

Taxonomy Overview

  • Advice for the Technology Supplier
    • Performance at All Costs
    • Package Performance-Intensive Computing Software Stacks
    • Low-Latency Interconnect and Network Support
    • Hybrid Approaches

Definitions

  • What Is Performance-Intensive Computing?
    • Performance-Intensive Computing Workloads
      • Composite Workloads
    • Alignment with IDC's Enterprise Workloads
  • Performance-Intensive Computing Infrastructure Segments
  • Infrastructure Procurement and Deployment
    • Vendor Type
    • Buyer Type
    • Control Plane
    • Deployment Location
    • Management Type
      • Procurement Type
  • Infrastructure Hardware
    • Infrastructure Hardware: Compute
      • Coprocessors (Accelerated Computing)
    • Infrastructure Hardware: Storage
      • Internal Storage (Server-Based Storage Platforms or Storage-Intensive Servers)
    • Converged Infrastructure
      • Integrated Infrastructure
      • Hyperconverged Infrastructure
      • Composable Infrastructure
    • Infrastructure Software
      • Physical and Virtual Computing Software
      • Storage Software
      • Systems Management Software
      • Other Infrastructure Software
      • PIC Dedicated and Public Cloud Services
      • PIC Infrastructure - Environmental Attributes
      • Workload Abstraction
        • Bare Metal
        • Virtualized
        • Containerized
      • Workload Profile
        • Compute Intensive
        • Memory Intensive
        • Data Intensive
        • Network Intensive
        • Hybrid
      • Parallelization (Cluster)
        • Traditional
        • Cloud Native
      • Scaling (Cluster)
        • Level 1: Intra-Node (Processor)
        • Level 2: Intra-Node (Accelerator)
        • Level 3: Inter-Node (Cluster)
        • Level 4: Inter-Node (Datacenter)
        • Level 5: Inter-Node (Geodispersed)
      • Size (Cluster)
      • Node-to-Node Communications Protocol (Cluster)
      • HPC and AI-Specific Attributes
        • HPC-Specific Attributes
        • AI-Specific Attributes

Related Markets

Learn More

  • Related Research
  • Synopsis
ºñ±³¸®½ºÆ®
0 °ÇÀÇ »óǰÀ» ¼±Åà Áß
»óǰ ºñ±³Çϱâ
Àüü»èÁ¦