Editing Desk-size computers / Supercomputers / Mainframes
Jump to navigation
Jump to search
Warning: You are not logged in. Your IP address will be publicly visible if you make any edits. If you log in or create an account, your edits will be attributed to your username, along with other benefits.
The edit can be undone.
Please check the comparison below to verify that this is what you want to do, and then save the changes below to finish undoing the edit.
| Latest revision | Your text | ||
| Line 1: | Line 1: | ||
| + | {{WIP}} | ||
==Terminology== | ==Terminology== | ||
'''Desk-size Computers''', as a slang of '''MiniComputers''' due to their size and usual appearance, are a type of smaller general-purpose computer with size, price, and extensibility less than a mainframe computer, but still higher than microcomputers, later commonly known as personal computers (PCs). Usage and demands of minicomputers have greatly shrunk and are being taken over by PC-based server clusters over the years, but some of them are still serving irreplaceable roles for their high reliability and extensibility. | '''Desk-size Computers''', as a slang of '''MiniComputers''' due to their size and usual appearance, are a type of smaller general-purpose computer with size, price, and extensibility less than a mainframe computer, but still higher than microcomputers, later commonly known as personal computers (PCs). Usage and demands of minicomputers have greatly shrunk and are being taken over by PC-based server clusters over the years, but some of them are still serving irreplaceable roles for their high reliability and extensibility. | ||
| − | |||
'''Supercomputers''' are computers specifically built for high-performance mathematical/scientific calculation with a high level of performance as compared to a general-purpose computer. | '''Supercomputers''' are computers specifically built for high-performance mathematical/scientific calculation with a high level of performance as compared to a general-purpose computer. | ||
| Line 7: | Line 7: | ||
Notice that the identification and distinction between PC-based clusters/servers, minicomputers, mainframes, and supercomputers aren't based on size, performance, or price, but system architecture, extensibility, and reliability; As integrated circuit technology evolves, a mainframe could be the size of a minicomputer (e.g. IBM System 390 Integrated Server 3006), a minicomputer could be a size of a relatively big personal computer/PC based server (e.g. IBM i5 515), while a regular laptop PC could achieve the performance of multi-million dollar mainframe in 1980s just by emulating it. | Notice that the identification and distinction between PC-based clusters/servers, minicomputers, mainframes, and supercomputers aren't based on size, performance, or price, but system architecture, extensibility, and reliability; As integrated circuit technology evolves, a mainframe could be the size of a minicomputer (e.g. IBM System 390 Integrated Server 3006), a minicomputer could be a size of a relatively big personal computer/PC based server (e.g. IBM i5 515), while a regular laptop PC could achieve the performance of multi-million dollar mainframe in 1980s just by emulating it. | ||
| − | + | {|class="wikitable" | |
| − | {|class="wikitable | ||
|+ Comparison | |+ Comparison | ||
! Category | ! Category | ||
| Line 20: | Line 19: | ||
|Large scaled, sophisticated, self-contained devices interconnected with a proprietary bus | |Large scaled, sophisticated, self-contained devices interconnected with a proprietary bus | ||
|Between PC and mainframe, some devices are self-contained and directly interconnected while others rely on the central processor and system bus | |Between PC and mainframe, some devices are self-contained and directly interconnected while others rely on the central processor and system bus | ||
| − | |Most to all components rely on the | + | |Most to all components rely on the central processor and system bus |
|Large amounts of self-contained PCs interconnected with Ethernet or other networking standards e.g. InfiniBand | |Large amounts of self-contained PCs interconnected with Ethernet or other networking standards e.g. InfiniBand | ||
|Large amounts of computation nodes interconnected with a proprietary bus, while each node may or may not be self-contained | |Large amounts of computation nodes interconnected with a proprietary bus, while each node may or may not be self-contained | ||
|- | |- | ||
|Statistic Reliability | |Statistic Reliability | ||
| − | |Exceeding 99.99999% uptime per year | + | |Exceeding 99.99999% uptime per year |
| − | |99.999% to 99.9999% uptime per year | + | |99.999% to 99.9999% uptime per year |
| − | |Poorly built ones could run into trouble every few hours under a high burden; unplanned downtime of proper prebuilt servers is around a few minutes to 1 hour per month | + | |Poorly built ones could run into trouble every few hours under a high burden; unplanned downtime of proper prebuilt servers is around a few minutes to 1 hour per month |
| − | |Varies (usually less than 99.9999% per year) depending on SLAs and configuration; each node follows the failure model of regular PC-based servers; nodes failing may cripple performance or capacity | + | |Varies (usually less than 99.9999% per year) depending on SLAs and configuration; each node follows the failure model of regular PC-based servers; nodes failing may cripple performance or capacity |
| − | |Not a priority, some supercomputers even run periodically every day; nodes failing may cripple performance or capacity | + | |Not a priority, some supercomputers even run periodically every day; nodes failing may cripple performance or capacity |
|- | |- | ||
|Servicability | |Servicability | ||
| Line 39: | Line 38: | ||
|- | |- | ||
|Redundancy | |Redundancy | ||
| − | |Multiple redundancy and backup at components and devices level; device-as-a-whole level redundancy is optional | + | |Multiple redundancy and backup at components and devices level; device-as-a-whole level redundancy is optional |
| − | |Redundancy at components and devices level; device-as-a-whole level redundancy is optional | + | |Redundancy at components and devices level; device-as-a-whole level redundancy is optional |
| − | |Data and power redundancy could be achieved by practices and external devices like RAID, ECC RAM, UPS etc., otherwise no. Device-as-a-whole level redundancy is always used | + | |Data and power redundancy could be achieved by practices and external devices like RAID, ECC RAM, UPS etc., otherwise no. Device-as-a-whole level redundancy is always used |
|Redundancy relying on large amounts of nodes, each node itself is not redundant; Software has to be programmed in a way to handle tasking reordering and data consistency in case of node failing | |Redundancy relying on large amounts of nodes, each node itself is not redundant; Software has to be programmed in a way to handle tasking reordering and data consistency in case of node failing | ||
|Not a priority; nodes aren't redundant and software seldom considers node failing as tasks are usually parallel to make use of peak performance | |Not a priority; nodes aren't redundant and software seldom considers node failing as tasks are usually parallel to make use of peak performance | ||
| Line 59: | Line 58: | ||
|Least concerned; programs are usually specifically built for specific supercomputer; sometimes supercomputers are even built for a specific task (e.g. Anton series) | |Least concerned; programs are usually specifically built for specific supercomputer; sometimes supercomputers are even built for a specific task (e.g. Anton series) | ||
|} | |} | ||
| − | |||
| − | |||
<div style="max-width:100%; overflow:auto;"> | <div style="max-width:100%; overflow:auto;"> | ||
{|class="wikitable sortable" | {|class="wikitable sortable" | ||
| Line 243: | Line 240: | ||
| style="background: Gainsboro;"| None | | style="background: Gainsboro;"| None | ||
| | | | ||
| − | | | + | | |
|- | |- | ||
|Cray Research XMS | |Cray Research XMS | ||
| Line 255: | Line 252: | ||
| style="background: Gainsboro;"| None | | style="background: Gainsboro;"| None | ||
| | | | ||
| − | | | + | | |
|- | |- | ||
|Csiro Csirac | |Csiro Csirac | ||
| Line 369: | Line 366: | ||
| style="background: Gainsboro;"| None | | style="background: Gainsboro;"| None | ||
| | | | ||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
|- | |- | ||
|Fujitsu Facom VP series | |Fujitsu Facom VP series | ||
| Line 508: | Line 499: | ||
| | | | ||
|[https://www.old-computers.com/museum/computer.asp?c=980&st=1 Old Computers] [http://simh.trailing-edge.com/ SIMH] | |[https://www.old-computers.com/museum/computer.asp?c=980&st=1 Old Computers] [http://simh.trailing-edge.com/ SIMH] | ||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
|- | |- | ||
|IBM System370 / ESA390 | |IBM System370 / ESA390 | ||
| Line 529: | Line 514: | ||
|IBM AS/400 / System i | |IBM AS/400 / System i | ||
|1988 | |1988 | ||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| style="background: Gainsboro;"| None | | style="background: Gainsboro;"| None | ||
| | | | ||
| Line 873: | Line 852: | ||
|} | |} | ||
</div> | </div> | ||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
[[Category:Computers|*]] | [[Category:Computers|*]] | ||
| − | [[Category: | + | [[Category:Emulated By MAME|*]] |
