IDCÀÇ ¿ùµå¿ÍÀÌµå ÆÛÆ÷¸Õ½º ÀÎÅٽúê ÄÄÇ»ÆÃ(HPC, AI, ¾Ö³Î¸®Æ½½º) ÀÎÇÁ¶ó¿Í ¼ºñ½º ºÐ·ù¹ý(2025³â)
»ùÇà ¿äû ¸ñ·Ï¿¡ Ãß°¡
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
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
ÀÎÇÁ¶ó Çϵå¿þ¾î
ÀÎÇÁ¶ó Çϵå¿þ¾î : 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)
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
KSA
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
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
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)
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