(phi) |
|||
(5 intermediate revisions by 2 users not shown) | |||
Line 1: | Line 1: | ||
{{intel title|Xeon Phi}} | {{intel title|Xeon Phi}} | ||
− | '''Xeon Phi''' is a family of [[x86]] [[many-core microprocessors]] designed by [[Intel]] since [[2012]] for supercomputers and workstations primarily aimed at scientists and researchers. Xeon Phi differs from other [[GPGPU]] and many-core processors in its unique quality that it retains almost full [[compatibility]] with existing [[x86]] software and libraries allowing those processors to tap into the extensive x86 ecosystem. | + | {{ic family |
+ | | title = Xeon Phi | ||
+ | | image = xeon phi (2015).png | ||
+ | | caption = Xeon Phi Logo since 2015 | ||
+ | | developer = Intel | ||
+ | | manufacturer = Intel | ||
+ | | type = Microprocessors | ||
+ | | first announced = 2012 | ||
+ | | first launched = 2012 | ||
+ | | arch = Many-core x86 microprocessors | ||
+ | | isa = x86-64 | ||
+ | | microarch = Knights Ferry | ||
+ | | microarch 2 = Knights Corner | ||
+ | | microarch 3 = Knights Landing | ||
+ | | microarch 4 = Knights Mill | ||
+ | | word = 64 bit | ||
+ | | proc = 22 nm | ||
+ | | proc 2 = 14 nm | ||
+ | | tech = CMOS | ||
+ | | clock min = | ||
+ | | clock max = | ||
+ | | package = | ||
+ | | package 2 = | ||
+ | | package 3 = | ||
+ | | socket = | ||
+ | | socket 2 = | ||
+ | |||
+ | | succession = | ||
+ | | predecessor = | ||
+ | | predecessor link = | ||
+ | | successor = | ||
+ | | successor link = | ||
+ | }} | ||
+ | '''Xeon Phi''' is a family of [[x86]] [[many-core microprocessors]] designed by [[Intel]] since [[2012]] for supercomputers and workstations primarily aimed at scientists and researchers. Those processors are based on Intel's {{intel|MIC Architecture}}. Xeon Phi differs from other [[GPGPU]] and many-core processors in its unique quality that it retains almost full [[compatibility]] with existing [[x86]] software and libraries allowing those processors to tap into the extensive x86 ecosystem. | ||
+ | |||
+ | == Overview == | ||
+ | {{empty section}} | ||
+ | |||
+ | == Member == | ||
+ | === Knights Ferry === | ||
+ | {{main|intel/microarchitectures/knights_ferry|l1=Knights Ferry}} | ||
+ | {{empty section}} | ||
+ | === Knights Corner === | ||
+ | {{main|intel/microarchitectures/knights_corner|l1=Knights Corner}} | ||
+ | {{empty section}} | ||
+ | === Knights Landing === | ||
+ | {{main|intel/microarchitectures/knights_landing|l1=Knights Landing}} | ||
+ | {{empty section}} | ||
+ | === Knights Mill === | ||
+ | {{main|intel/microarchitectures/knights_mill|l1=Knights Mill}} | ||
+ | Launched in December 2017, Knights Mill processors are a special variant of Knights Landing specifically designed to accelerate deep learning type of workloads. | ||
+ | <!-- NOTE: | ||
+ | This table is generated automatically from the data in the actual articles. | ||
+ | If a microprocessor is missing from the list, an appropriate article for it needs to be | ||
+ | created and tagged accordingly. | ||
+ | |||
+ | Missing a chip? please dump its name here: https://en.wikichip.org/wiki/WikiChip:wanted_chips | ||
+ | --> | ||
+ | {{comp table start}} | ||
+ | <table class="comptable sortable tc5 tc6 tc7 tc8 tc9 tc10"> | ||
+ | {{comp table header|main|9:List of Knights Mill-based Xeon Phi Processors}} | ||
+ | {{comp table header|main|9:Main processor}} | ||
+ | {{comp table header|cols|Launched|Price|Family|Cores|Threads|L2$|TDP|Base|Turboo}} | ||
+ | {{#ask: [[Category:microprocessor models by intel]] [[microarchitecture::Knights Mill]] | ||
+ | |?full page name | ||
+ | |?model number | ||
+ | |?first launched | ||
+ | |?release price | ||
+ | |?microprocessor family | ||
+ | |?core count | ||
+ | |?thread count | ||
+ | |?l2$ size | ||
+ | |?tdp | ||
+ | |?base frequency#GHz | ||
+ | |?turbo frequency (1 core)#GHz | ||
+ | |format=template | ||
+ | |template=proc table 3 | ||
+ | |userparam=11 | ||
+ | |mainlabel=- | ||
+ | }} | ||
+ | {{comp table count|ask=[[Category:microprocessor models by intel]] [[microarchitecture::Knights Mill]]}} | ||
+ | </table> | ||
+ | {{comp table end}} | ||
+ | |||
+ | === Knights Hill & Knights Peak === | ||
+ | {{main|intel/microarchitectures/knights_hill|intel/microarchitectures/knights_peak|l1=Knights Hill|l2=Knights Peak}} | ||
+ | Announces on November 18 2014, Knights Hill was planned to be fabricated on their [[10 nm process]]. Severe delays with their 10 nm process has caused them to push the architecture all the way back to 2018. In late 2017 Intel finally announce that they had cancelled the architecture altogether in favor of an entirely new microarchitecture aimed at [[exascale]] computing. Consequently, Knights Peak, Hill's successor was also cancelled. |
Latest revision as of 00:38, 19 December 2017
Xeon Phi | |
Xeon Phi Logo since 2015 | |
Developer | Intel |
Manufacturer | Intel |
Type | Microprocessors |
Introduction | 2012 (announced) 2012 (launch) |
Architecture | Many-core x86 microprocessors |
ISA | x86-64 |
µarch | Knights Ferry, Knights Corner, Knights Landing, Knights Mill |
Word size | 64 bit 8 octets
16 nibbles |
Process | 22 nm 0.022 μm , 14 nm2.2e-5 mm 0.014 μm
1.4e-5 mm |
Technology | CMOS |
Xeon Phi is a family of x86 many-core microprocessors designed by Intel since 2012 for supercomputers and workstations primarily aimed at scientists and researchers. Those processors are based on Intel's MIC Architecture. Xeon Phi differs from other GPGPU and many-core processors in its unique quality that it retains almost full compatibility with existing x86 software and libraries allowing those processors to tap into the extensive x86 ecosystem.
Contents
Overview[edit]
This section is empty; you can help add the missing info by editing this page. |
Member[edit]
Knights Ferry[edit]
- Main article: Knights Ferry
This section is empty; you can help add the missing info by editing this page. |
Knights Corner[edit]
- Main article: Knights Corner
This section is empty; you can help add the missing info by editing this page. |
Knights Landing[edit]
- Main article: Knights Landing
This section is empty; you can help add the missing info by editing this page. |
Knights Mill[edit]
- Main article: Knights Mill
Launched in December 2017, Knights Mill processors are a special variant of Knights Landing specifically designed to accelerate deep learning type of workloads.
List of Knights Mill-based Xeon Phi Processors | |||||||||
---|---|---|---|---|---|---|---|---|---|
Main processor | |||||||||
Model | Launched | Price | Family | Cores | Threads | L2$ | TDP | Base | Turboo |
7235 | 18 December 2017 | Xeon Phi | 64 | 256 | 32 MiB 32,768 KiB 33,554,432 B 0.0313 GiB | 250 W 250,000 mW 0.335 hp 0.25 kW | 1.3 GHz 1,300 MHz 1,300,000 kHz | 1.4 GHz 1,400 MHz 1,400,000 kHz | |
7285 | 18 December 2017 | Xeon Phi | 68 | 272 | 34 MiB 34,816 KiB 35,651,584 B 0.0332 GiB | 250 W 250,000 mW 0.335 hp 0.25 kW | 1.3 GHz 1,300 MHz 1,300,000 kHz | 1.4 GHz 1,400 MHz 1,400,000 kHz | |
7295 | 18 December 2017 | Xeon Phi | 72 | 288 | 36 MiB 36,864 KiB 37,748,736 B 0.0352 GiB | 320 W 320,000 mW 0.429 hp 0.32 kW | 1.5 GHz 1,500 MHz 1,500,000 kHz | 1.6 GHz 1,600 MHz 1,600,000 kHz | |
Count: 3 |
Knights Hill & Knights Peak[edit]
- Main articles: Knights Hill and Knights Peak
Announces on November 18 2014, Knights Hill was planned to be fabricated on their 10 nm process. Severe delays with their 10 nm process has caused them to push the architecture all the way back to 2018. In late 2017 Intel finally announce that they had cancelled the architecture altogether in favor of an entirely new microarchitecture aimed at exascale computing. Consequently, Knights Peak, Hill's successor was also cancelled.
designer | Intel + |
first announced | 2012 + |
first launched | 2012 + |
full page name | intel/xeon phi + |
instance of | microprocessor family + |
instruction set architecture | x86-64 + |
main designer | Intel + |
manufacturer | Intel + |
microarchitecture | Knights Ferry +, Knights Corner +, Knights Landing + and Knights Mill + |
name | Xeon Phi + |
process | 22 nm (0.022 μm, 2.2e-5 mm) + and 14 nm (0.014 μm, 1.4e-5 mm) + |
technology | CMOS + |
word size | 64 bit (8 octets, 16 nibbles) + |