CPUThe core is the core chip in the CPUmonocrystalline siliconMade to complete all calculations, receive/store commandsProcessing dataEtc. is the core of digital processing.
The core (Die), also known as the core, is the most important component of the CPU.The raised chip in the center of the CPU is the core, which is made of monocrystalline siliconProduction processAll CPU calculations, receiving/storing commands and processingDataExecuted by the core.variousCPU coreHave fixedLogical structure,L1 cache、L2 Cache、execution unit, instruction level units, andBus interfaceetc.Logical unitThere will be a scientific layout.
mostCPUBoth adopt a reverse corePackaging formThat is to say, the CPU core we usually see is actually the bottom of the silicon chip, which is flipped and packaged on the ceramic circuit board. This advantage is that the CPU core can directly contact the heat sink.This technology is also used on most CPUs today.The other side of the CPU core, which is covered under the ceramic circuit board, should be connected to the external circuit.CPU has tens of millions of transistors, which are connected to the external circuit. The connection method is to solder a wire to each transistorExternal circuitOn.For example, 3000 wires need to be welded on the Duron core, whilePentium 4The quantity of is 5000, which is used for the server64 bit processorItanium7500.There are so many on such a small chipSolder jointThese solder joints must be very small and carefully designed.Since all calculations are carried out on a very small chip, the CPU core will emit a lot of heat, and the internal temperature of the core can reach up to Baidu (unit)surface temperature There will also be dozens of degrees. If the temperature is too high, the CPU will run abnormally or even burn out. Therefore, many computer books or magazines often emphasize the importance of cooling the CPU.The CPU should also have a specific motherboard. For example, the i7 CPU can only use a dedicated motherboard.
Kernel Type
Announce
edit
In order to facilitate the management of CPU design, production and sales, CPU manufacturers willcoreGive the corresponding code, which is also calledCPU core type。
CPU core
Different CPUs (different series or the same series) will have different core types (for examplePentium 4OfNorthwoodWillamette, K6-2 CXT, K6-2+ST-50, etc.), even the same core can have different versions (for example, Northwood core can be divided into B0 and C1 versions). The core version changes are to correct some errors in the previous version and improve certain performance, but ordinary consumers pay little attention to these changes.Each core type has its corresponding manufacturing process (such as 0.25um, 0.18um, 0.13um, 0.09um, 0.065um, and 0.045um), core area (which is the key factor determining the CPU cost, and the cost is basically proportional to the core area), core voltage, current size, number of transistors, and levelscacheSize, main frequency range, pipeline architecture and supportedInstruction set(These two points determine the actual CPU performance andwork efficiencyKey factors), power consumption andCalorific valueSizePackaging method(e.g. S.E.P, PGA, FC-PGA, FC-PGA2, etc.), interface type (e.gLGA775、Socket 370, Socket A, Socket 478, Socket T, Slot 1, Socket 940, etc.)Front end bus frequency(FSB), etc.Therefore, the core type determines the CPU performance to some extent.
Generally speaking, new core types often have better performance than the old core types (for example, the same frequency Northwood core Pentium 4 1.8A GHz has higher performance than the Willamette core Pentium 4 1.8GHz), but this is not absolute. This usually happens when new core types are just launched, due to incomplete technology or immature new architecture and manufacturing process, etc,This may cause the performance of new core types to be inferior to that of old core types.For example, the early Willamette coreSocket 423The actual performance of the Pentium 4 interface is not as good as that of the Socket 370 interfaceTualatinCore Pentium III andceleron The actual performance of the low-frequency Prescott core Pentium 4 is not as good as that of the Northwood core Pentium 4 on the same frequency. However, with the progress of technology and the continuous improvement and perfection of CPU manufacturers on the new core, the performance of the new core products in the middle and later stages will inevitably surpass the old onesCore products。
Core development
Announce
edit
The development direction of CPU core is lower voltage, lower power consumption, more advanced manufacturing technology, more integrated transistors, and smaller core area (which will reduce theproduction costsWhich will eventually reduce the sales price of CPU), more advanced pipeline architecture and moreInstruction set, HigherFront end bus frequencyIntegrate more functions (such as integrationMemory controllerEtc.) and dual core andMulti core(That is, there are two or more cores in one CPU), etc.For ordinary consumers, the most significant thing about the progress of CPU core is that they can buy CPU with better performance at a lower price.
There is a method to expand the CPU core. This method can add 1000 cores to the computerSpeed ratioThe average computer is 20 times faster.
It's mainstreamPentium 4andceleron The core adopted is the same asWillametteThe biggest improvement of the core is that 0.13um manufacturing process is adopted, andSocket 478Interface, core voltage is about 1.5V,L2 Cache128KB (Celeron) and 512KB (Pentium 4) respectively, and the front-end bus frequency is 400/533/800MHz (Celeron only has 400MHz),Dominant frequencyThe range is 2.0GHz to 2.8GHz (Celeron), 1.6GHz to 2.6GHz (400MHz FSB Pentium 4), 2.26GHz to 3.06GHz (533MHz FSB Pentium 4) and 2.4GHz to 3.4GHz (800MHz FSB Pentium 4) respectively, and 3.06GHz Pentium 4 and all 800MHz Pentium 4 areSupport Hyper Threading Technology(Hyper Threading Technology), packaged inPPGAFC-PGA2 and PPGA.According to Intel's plan,Northwood CoreWill soon bePrescottCore replaced.
Prescott
This is the CPU core of Intel. Only Pentium 4 is used instead of low-end Celeron. Its biggest difference from Northwood is that it uses 0.09um manufacturing process and morePipeline structureSocket 478 interface is used at the initial stage and will be switched to LGA 775 interface later. The core voltage is 1.25-1.525V. The front-end bus frequency is 533MHz (Hyper Threading Technology is not supported) and 800MHz (Hyper Threading Technology is supported). The main frequencies are 2.4GHz and 2.8GHz of 533MHz FSB and 2.8GHz, 3.0GHz, 3.2GHz and 3.4GHz of 800MHz FSB respectively. Compared with Northwood, L1Data cacheFrom 8KB to 16KB, whileL2 CacheThen it will increase from 512KB to 1MB, and the packaging mode will be PPGA.According to Intel's plan, the Prescott core will soon replace the Northwood core and will soon introduce the 533MHz FSB Celeron Prescott core.
Smithfield
CPU core
This is the core type of Intel's first dual core processor, which was released in April 2005. Basically, it can be considered that the Smithfield core is the product of loosely coupling two Prescott cores together. This is a loose coupling scheme based on independent cache. Its advantages are simple technology, but its disadvantages are not ideal performance.Pentium D8XX series andPentium EE8XX series adopts this core.Smithfield core adopts 90nm manufacturing process, all of which adoptSocket 775Interface, core voltage is about 1.3V, and the packaging mode is PLGA, both of which supportHardware antivirustechnologyEDBand64 bit technologyEM64TAnd supports energy-saving technology except for Pentium D 8X5 and Pentium D 820EIST。The front-end bus frequency is 533MHz (Pentium D 8X5) and 800MHz (Pentium D 8X0 and Pentium EE 8XX), and the main frequency range is 2.66GHz to 3.2GHz (Pentium D) and 3.2GHz (Pentium EE).The biggest difference between Pentium EE and Pentium D is that Pentium EE supports hyper threading technology while Pentium D does not.The two Smithfield cores have 1MBL2 CacheThe two cores inside the CPU are isolated from each other, and the synchronization of its cache data is realized by the arbitration unit located on the North Bridge chip of the motherboard, which transmits data between the two cores through the front-end bus, so the data delay problem is serious and the performance is not satisfactory.According to Intel's plan, the Smithfield core will soon be replaced by the Presler core.
This is the core of the Pentium 4 6X1 series and Celeron D 3X2/3X6 series, which has appeared since the end of 2005.The biggest difference between it and Prescott core is that it uses the 65nm manufacturing process, while other aspects have little change. It can basically be regarded as the 65nm manufacturing process version of Prescott core.The Cedar Mill core adopts Socket 775 interface, core voltage is about 1.3V, and the packaging mode is PLGA.Among them, all Pentium 4 are 800MHz FSB, 2MBL2 Cache, all support hyper threading technology, hardware anti-virus technology EDB, energy-saving technology EIST and 64 bit technology EM64T;Celeron D is a 533MHz FSB, 512KB second level cache, which supports hardware anti-virus technology EDB and 64 bit technology EM64T, but does not support hyper threading technology and energy saving technology EIST.Cedar Mill core is also the last single core processor core type of Intel processors on NetBurst architecture. According to Intel's plan, Cedar Mill core will gradually be replaced by Conroe core of Core architecture.
Presler
New technology increases the number of CPU cores rather than the frequency
This is the core adopted by Pentium D 9XX and Pentium EE 9XX, which was launched by Intel at the end of 2005.Basically, it can be considered that the Presler core is the product of loosely coupling two Cedar Mill cores together. It is a cache based loosely coupled scheme. Its advantages are simple technology, but its disadvantages are that its performance is not ideal and independent.The Presler core adopts 65nm manufacturing process, socket 775 interface, core voltage of about 1.3V, and PLGA packaging mode. It supports hardware anti-virus technology EDB, energy saving and power saving technology EIST, and 64 bit technology EM64T. Besides Pentium D 9X5, it supports virtualization technologyIntel VT。The front-end bus frequency is 800MHz (Pentium D) and 1066MHz (Pentium EE).Similar to the Smithfield core, Pentium EE supports hyper threading technology while Pentium D does not, and the two cores have 2MBL2 Cache。The two cores inside the CPU are isolated from each other, and the synchronization of its cache data is also achieved by the arbitration unit located on the North Bridge chip of the motherboard, which transmits data between the two cores through the front-end bus, so the data delay problem is also serious, and the performance is also not satisfactory.Compared with the Smithfield core, the Presler core has almost no innovation in technology, except that it uses the 65nm process, increases the secondary cache of each core to 2MB, and adds support for virtualization technology. It can basically be regarded as the 65nm process version of the Smithfield core.Presler core is also the core type of the last dual core processor on the NetBurst architecture of Intel processors. It can be said that it was the last swan song before NetBurst was abandoned. Later, all Intel desktop processors will be transferred to the Core architecture.According to Intel's plan, the Presler core will gradually be replaced by the Conroe core of the core architecture from the third quarter of 2006.
Yonah
There are dual core Core Duo and single core Core Solo using Yonah core CPU, andCeleron MIt also adopts this core. Yonah was launched by Intel in early 2006.This is a core type of single/dual core processor. Its application features great flexibility, which can be used for both desktop platforms andMobile platform;It can be used for dual core or single core.Yonah core comes from famous processor on mobile platformPentium MIt has the advantages of fewer pipeline stages, high execution efficiency, powerful performance and low power consumption.Yonah core adopts 65nm manufacturing process, the core voltage varies from 1.1V to 1.3V depending on the version, and the packaging mode adopts PPGA,Interface typeIt is an improved new version of Socket 478 interface (incompatible with the previous desktop Socket 478).In terms of front-end bus frequency, both Core Duo and Core Solo are 667MHz, while Yonah core Celeron M is 533MHz.stayL2 CacheIn terms of core Duo and core Solo, both are 2MB, while Yonah core Celeron M is 1MB.Yonah core supports hardware anti-virus technology EDB and energy-saving technology EIST, and most models support virtualization technology Intel VT.But its biggest regret is that it does not support 64 bit technology, only 32-bit processors.It is worth noting that for the dual core Core Duo, its 2MB L2 cache architecture is different from that of all X86 processors. All other X86 processors have their own L2 cache for each core, while the Yonah core of Core Duo usesIBMOfMulti coreProcessor similar cache scheme - two cores share 2MB L2 cache!The shared two-level cache, together with Intel's "Smart cache" shared cache technology, realizes the real cache data synchronization, significantly reduces data latency, and reduces the occupation of front-end bus.This is the true dual core processor in the strict sense!Yonah core is a tight coupling scheme of shared cache. Its advantages are ideal performance, but its disadvantages are complex technology.However, according to Intel's plan, processors on all Intel platforms will be transferred to the Core architecture in the future. Yonah core is actually just a transitional core type. From the third quarter of 2006, it will be replaced by Conroe core on the desktop platform, and will be replaced by Conroe core on the mobile platformMeromCore replaced.
Conroe
This is the core type of the updated Intel desktop platform dual core processor. Its name comes fromTexas, USAOfsmall city“Conroe”。The Conroe core was officially released on July 27, 2006. It is a new core(CoRE )Microarchitecture(Core Micro Architecture) is the first CPU core applied on desktop platform.Core 2 Duo E6x00 series and Core 2 Extreme X6x00 series adopt this core.Compared with the previous generation of Pentium D and Pentium EE with NetBurst microarchitecture, the Conroe core has the advantages of fewer pipeline levels, high execution efficiency, strong performance and low power consumption.Conroe core adopts 65nm manufacturing process, core voltage is about 1.3V, packaging mode is PLGA, and the interface type is still the traditional Socket 775.In terms of front-end bus frequency, both Core 2 Duo and Core 2 Extreme are 1066MHz, while the top Core 2 Extreme will be upgraded to 1333MHz;In the first level cache, each core has a 32KB data cache and a 32KBInstruction Cache, and the first level data cache of the two cores can exchange data directly;stayL2 CacheIn terms of the Conroe core, both cores share 4MB.Conroe core supports hardware anti-virus technology EDB, energy-saving technology EIST, 64 bit technology EM64T and virtualization technology Intel VT.Similar to the caching mechanism of Yonah core, the second level cache of Conroe core is still shared by the two cores, and the improved Intel Advanced Smart Cache(IntelAdvanced intelligenceCache)Shared cache technology to achieve cache data synchronization.Conroe core is the most advanced desktop platform at presentProcessor core, in high performance andlow power consumptionHas found a good balance point on the, comprehensively overwhelming all desktop platform dual core processors, and has a very goodOverclocking capability, is indeed the most powerful desktop CPU core at present.
The number of CPU cores and GPU cores can be flexibly adjusted
This is the core type of Intel desktop platform dual core processor released at the same time with Conroe. Its name comes from the United StatesCaliforniaAllendale, a small city in the south.The Allendale core was officially released on July 27, 2006. It is still based on the new Core (Core) microarchitecture. The 1066MHz FSB Core 2 Duo E6x00 series uses this core, and the 800MHz FSB Core 2 Duo E4x00 series will be released soon.Allendale coreL2 CacheThe mechanism is the same as the Conroe core, but the shared L2 cache is reduced to 2MB.The Allendale core still adopts the 65nm manufacturing process, the core voltage is about 1.3V, the packaging mode is PLGA, the interface type is the traditional Socket 775, and it still supports the hardware anti-virus technology EDB, energy saving and power saving technology EIST, 64 bit technology EM64T, and virtualization technology Intel VT.Except that the shared L2 cache is reduced to 2MB and the L2 cache is 8-way 64Bytes instead of 16 way 64Bytes of the Conroe core, the Allendale core is almost the same as the Conroe core, which can be said to be a simplified version of the Conroe core.Of course becauseL2 CacheIn the case of the same frequency, the Allendale core performance is slightly inferior to the Conroe core.
Merom
This is the core type of Intel mobile platform dual core processor released at the same time as Conroe. Its name comes fromIsraelDomesticJordan RiverNext to it is a lake called Merom.Merom core was officially released on July 27, 2006, which is still based on the new Core (Core) microarchitecture. This is the first time that Intel's full platform (desktop, notebook and server) processors adopt the same microarchitecture design. Core 2 Duo T7x00 series and Core 2 Duo T5x00 series with 667MHz FSB adopt this core.Similar to the desktop version of Conroe core, the Merom core still uses the 65nm manufacturing process, the core voltage is about 1.3V, the packaging method is PPGA, and the interface type is still an improved version of Socket 478 interface compatible with Yonah core Duo and Core Solo (not compatible with the previous desktop socket 478) orSocket 479The socket 479 slot is still used as the interface.Merom core also supports hardware anti-virus technology EDB, energy-saving technology EIST, 64 bit technology EM64T and virtualization technology Intel VT.Merom CoreL2 CacheThe mechanism is also the same as that of Conroe core. The shared L2 cache of Core 2 Duo T7x00 series is 4MB, while the shared L2 cache of Core 2 Duo T5x00 series is 2MB.The main technical characteristics of Merom core are almost the same as that of Conroe core, except that on the basis of Conroe core, various means are used to strengthen power control, making its TDP power consumption almost half of Conroe core, so as to meet the power saving requirements of mobile platforms.
Penryn
Penryn uses 45nm high k manufacturing technology (usingChromium alloyHigh K and metal gate transistor design), andMicroarchitectureEnhanced.Compared with the 65nm process, the 45nm high k process technology can nearly double the number of transistors. The next generation of Intel Core 2Quad core processor820 million transistors will be used.With the newly invented high-k metal gate transistor technology, these 820 million transistors can be switched more efficiently at the speed of light. The switching speed of the transistors has been increased by more than 20%, achieving higher core speed, and increasing theClock cycleThe number of instructions for.Dual Core Processor The size of silicon core in is 107 mm2, which is 65Nanoproducts25% smaller, only aboutOrdinary stampA quarter size of, providing more for adding new features and achieving higher performancefree space。At the same time, because the leakage current is reduced, the power consumption can be reduced. Compared with Intel's existing dual core processors, the new generation processors can run at the same or even lower power consumption. For example, the cooling design power consumption of the Penryn processor is 40 watts/65 watts/80 watts for dual cores and 50 watts/80 watts/120 watts for quad cores.
New feature: fast Raidix-16Divider, EnhancedVirtualization technology, larger cache, separate load cache enhancement, higherBus speed、IntelSSE4Command, super Shuffle engine, deep shutdown technology, enhancedDynamic acceleration technology, slot compatibility, etc.These new features enable Penryn to improve its performance, power consumption, digital media applications, virtualization applications, etc. For example, compared with current products, Penryn processors with 1600MHz front-end bus and 3GHz can improve their performance by about 45%.
No longer use lead as raw material
Intel said that its new generation processors no longer use lead as raw material, and it is expected to stop using halogen by 2008.Through these measures, the harm of Intel processors to the environment will be greatly reduced.One of the biggest features of Intel's new processors is the use of hafnium, which can effectively solve the problem of electric leakage, and improve the power efficiency of the processor by 30%.With the shrinking size of transistors, electrical leakage is more serious, which leads to the problem of processor heating and excessive power consumption.To some extent, electric leakage has become an obstacleProcessor performanceBottleneck for further improvement.
Minimum power consumption: 25W
IntelStephen Smith, director of digital enterprise group, said that Penryn processorMaximum power consumptionNot more than 120 watts.Penryn to be launched in the first quarter of 2008Notebook processorIts power consumption is 25 watts, while the current 65 nanometer notebook processor consumes 35 watts.According to Smith, the Penryn processor has been added to speed upimage processingAnd high definitionvideo coding New instructions for.Compared with the previous generation of products, the video and graphics performance of the Penryn processor has improved by 40% to 60%.Thanks to hardware enhancements,virtual machineThe performance of is also improved by 75%.
In 2009 (TICK time), Intel's processor manufacturing process entered the 32nm era. In 2010, in the TOCK time, Intel launched the processor code named Sandy Bridge, which uses the 32nm process.Sandy Bridge (formerly called Gesher) isNehalemThe successor of, and also its process upgrade version, from45nmIt evolved to 32nm.Sandy Bridge will have eight core versions, and the L2 cache is still 512KB, butL3 cacheIt will be expanded to 16MB.The main feature of Sandy Bridge is the addition of game instrument AVX (Advanced Vectors Extensions) technology, which is the former VSSE.Intel claims that the matrix calculation using AVX technology will be 90% faster than SSE technology.Its importance is comparable to that of 1999Pentium IIIIntroduce SSE.
On April 24, 2012,IntelThe Third Generation of Intelligence Held in BeijingCoRE processorIvy Bridge launch.The first processors will include a mobile versionCore i7Premium Edition, six new intelligent Core i7 processors, sixCore i5Processor.Compared with the previous generation Sandy Bridge, Ivy Bridge combines 22nm and 3D transistor technology to significantly improve transistor density,Core graphics cardSome of the performance even has more than doubled.According to information, Ivy Bridge processorapplication program20% improvement in 3D performance, and supports three independent displaysUSB 3.0Etc
0.13um manufacturing process is adopted, core voltage is about 1.65V, secondary buffer is 512KB, packaging mode is OPGA, front-end bus frequency is 333MHz and 400MHz.
The main difference between it and Clawhammer is that the second level cache is reduced to 512KB (which is also why AMD promotes 64 bits for market needs and speed upCPUThe results of the relatively low price policy are basically the same.
Wincheste is a relatively new AMDAthlonThe 64 CPU core is a 64 bit CPU with 939 interfaces and 0.09 micron manufacturing process.This core uses 200MHzfsb , support 1GHyperTransprot bus, 512KL2 CacheGood cost performance.Wincheste integrationDual channelMemory controller, supporting dual channelsDDR memory, Wincheste'sfeverVolume ratioThe old Athlon is small and its performance has been improved.
Troy
Troy is the first AMD to use the 90nm manufacturing processOpteronCore.Troy CoreIs onSledgehammerIt is based on a number of new technologies, usually 940Pin, 128KL1 cacheAnd 1MB (1024KB)L2 cache.Also use 200MHzfsb , support 1GHyperTransprot bus, integrated memory controller, support dual channelsDDR400Memory, and can supportECCMemory.In addition, the Troy core also providesSSE-3 is the same as Intel Xeon. In general, Troy is a good CPU core.
Venice
Venice core is evolved from Wincheste core, and its technical parameters are basically the same as Wincheste: the same based onX86-64Architecture, integrated dual channel memory controller, 512KBL2 Cache, 90nm manufacturing process, 200MHz external frequency, support 1GHyperTransprot bus.There are three main changes in Venice. First, it uses Dual Stress Liner (DSL for short) technology to integrate semiconductorstransistorOfresponse speedIncrease by 24%, so that the CPU has more frequency space and is easier to useOverfrequency;Second, it provides support for SSE-3, which is the same as Intel's CPU;Third, further improvementMemory controllerTo some extent, it increases the performance of the processor, and more importantly, it increases the memory controller for differentDIMMModules and different configurationscompatibility。In addition, the Venice core also uses dynamic voltage. Different CPUs may have different voltages.
The SanDiego core, like Venice, evolved from Wincheste core, and its technical parameters are very close to Venice's. The SanDiego core has the same new technologies and functions as Venice.howeverAMDThe company positions SanDiego core on the top Athlon 64 processor, even forServer CPU。SanDiego can be regarded as an advanced version of the Venice core, butCache capacityFrom 512KB to 1MB.Of course, due to the increase of L2 cache, the size of SanDiego core has also increased, from 84square millimetreIt is increased to 115 square millimeters, and the price is also higher.
Orleans
This is the first one released at the end of May 2006Socket AM2The core type of the interface single core Athlon 64, whose name comes from a French cityOrleans(Orleans)。ManilaThe core is located in the desktop mid tier processor, which adopts 90nm manufacturing process and supportsVirtualization technologyAMD VT1000MHzHyperTransport bus,L2 Cache512KB, the biggest highlight is that it supports dual channelsDDR2667 memory, which is different fromsingle channelThe biggest difference between socket 754 interface Athlon 64 of DDR 400 memory and socket 939 interface Athlon 64 that only supports dual channel DDR 400 memory.The Orleans core Athlon 64 is also divided intoTDP power consumption62W standard version (core voltage is about 1.35V) and 35W TDP ultra-low power version (core voltage is about 1.25V).In addition to supporting dual channel DDR2 memory and supportingVirtualization technologyIn addition, compared with the previous socket 754 interface and socket 940 interface, the Orleans core Athlon 64 has no architectural changes, and its performance is not much better.
Paris Core is the successor of Barton Core, mainly used for AMD'sFlashing dragonThe early 754 interface flash dragon part uses the Paris core.Paris adopts 90nm manufacturing process to support iSSE2Instruction set, generally 256KL2 Cache, 200MHz external frequency.The Paris core is a 32-bit CPU, derived fromK8Core, so it also has memorycontrol unit 。CPU built-inMemory controllerThe main advantage of is that the memory controller canCPU frequencyOperation, compared with the traditional locationNorth BridgeThe memory controller has less latency.Compared with the socket A interface flash dragon CPU, the performance of the flash dragon CPU using the Paris core is significantly improved.
Palermo core is mainly used for AMD'sFlashing dragonCPU,Socket 754 interface, 90nm manufacturing process, 1.4V voltage, 200MHzfsb 128K or 256KL2 Cache。Palermo core originates from Wincheste core of K8, new E6SteppingThe version already supports 64 bits.In addition to having the same internal architecture as AMD high-end processors, it also has EVP, Cool 'n' Quiet; andHyperTransportAnd other unique technologies of AMD to bring more "calm" and higherComputing powerExcellent processor.Due to the birth of ATHLON64 processor, Palermo also has a memory control unit.
Manila
This is the core type of the first Socket AM2 interface Sempron released at the end of May 2006. Its name comes fromthe PhilippinescapitalManila(Manila)。The core of Manila is positioned at the low-end processor of the desktop. It adopts the 90nm manufacturing process, does not support the virtualization technology AMD VT, and still uses the 800MHz HyperTransport bus,L2 CacheIt is 256KB or 128KB, and the biggest highlight is that it supports dual channelsDDR2667 memory is the biggest difference from Sempron, which only supports single channel DDR 400 memory.The Manila core Sempron is divided into the standard version with TDP power consumption of 62W (core voltage of about 1.35V) and the ultra low power version with TDP power consumption of 35W (core voltage of about 1.25V).In addition to supporting dual channel DDR2, there is no architectural change in the Manila core Sempron compared with the previous Socket 754 interface Sempron, and its performance is not much outstanding.
This is the core type of the first dual core processor on the desktop platform released by AMD in April 2005, which evolved from the Venice core. It can basically be seen as two Venice cores coupled together, but the degree of collaboration is relatively close, which is based on independencecacheThe advantage of the tight coupling scheme is that the technology is simple, but the disadvantage is that the performance is still not ideal.Manchester core adopts 90nm manufacturing process, integrates dual channel memory controller, supports 1000MHz HyperTransprot bus, and adopts Socket 939 interface.The two cores of the Manchester core have 512KBL2 Cache, but with Intel'sSmithfieldCore andPreslerCorecacheData synchronizationRely onMainboard North BridgeOn chipArbitration unitadoptFront end bustransmission modeThe big difference is that the two cores in the Manchester core cooperate closely. The cache data synchronization is controlled by the SRI (System Request Interface) built in the CPU, and transmission can be realized inside the CPU.In this way, not only CPU resources are used very little, but also memory bus resources are not needed,Data delayIt is also significantly less than Intel's Smithfield core and Presler core, and its collaboration efficiency is obviously better than these two cores.However, because the Manchester core is still a cache of two coresMutual independenceFrom the perspective of architecture, it is obviously inferior toYonahThe core is Intel's shared cache technology, Smart Cache.Of course, sharecacheThe technology needs to redesign the entire CPU architecture, which is much more difficult than simply coupling the two cores together.
This is AMD's new high-end desktop platform in April 2005Dual coreThe core type of the processor is very similar to the Manchester core. The difference isL2 CacheDifferent.Toledo is evolved from the core of San Diego. Basically, it can be seen as two San Diego cores simply coupled together.Toledo core adopts 90nm manufacturing process, integrates dual channel memory controller, supports 1000MHz HyperTransprot bus, and adopts Socket 939 interface.The two cores of the Toledo core have 1MBL2 CacheThe same as the Manchester core, its cache data synchronization is also transmitted inside the CPU through SRI.Compared with the Manchester core, the Toledo core is identical except that the L2 cache of each core is increased to 1MB. It can be regarded as the advanced version of the Manchester core.
This is the first core type of socket AM2 interface dual core Athlon 64 X2 and Athlon 64 FX released at the end of May 2006. Its name comes from the British place nameWindsor(Windsor)。Windsor core is positioned at the desktop high-end processor, adopts 90nm manufacturing process, supports virtualization technology AMD VT, and still uses 1000MHz HyperTransport bus,L2 CacheOn the other hand, the two cores of the Windsor core still use independent L2 cache. Each core of Athlon 64 X2 is 512KB or 1024KB, and each core of Athlon 64 FX is 1024KB.The biggest highlight of the Windsor core is that it supports dual channel DDR2 800 memory, which is the biggest difference from the socket 939 interface Athlon 64 X2 and Athlon 64 FX that only supports dual channel DDR 400 memory.Windsor core Athlon 64 FX onlyFX-62The TDP power consumption of this product is up to 125W;The Athlon 64 X2 is divided into the standard version with 89W TDP power consumption (about 1.35V core voltage), the low power version with 65W TDP power consumption (about 1.25V core voltage), and the ultra low power version with 35W TDP power consumption (about 1.05V core voltage).The cache data synchronization of the Windsor core is still realized inside the CPU by relying on the SRI (System request interface) transmission built in the CPU. In addition to supporting dual channel DDR2 memory andVirtualization technologyIn addition, compared with the previous Socket 939 interface Athlon 64 X2 and dual core Athlon 64 FX, there is no architectural change, and its performance is not much outstanding. Its performance is still inferior to Intel's upcoming Conroe Core 2 Duo andCore 2 Extreme。And AMDcost reductionIn order to improve competitiveness, in addition to Athlon 64 FX, it has decided to stop production with 1024KBx2L2 CacheOnly the Athlon 64 X2 with 512KBx2 L2 cache is reserved.
CPU Introduction
Announce
edit
Notebook computerThe dedicated CPU is called Mobile CPU in English. It pursues not only performance, but also low heat and power consumption. The earliest laptops directly use desktop CPU, but alsoCPUDominant frequencyThe narrow space of notebook computers cannot quickly dissipate the heat generated by the CPU, and the batteries of notebook computers cannot afford the huge power consumption of desktop CPUs. Therefore, Mobile CPUs specially designed for notebooks began to appear. Its manufacturing process is often more advanced than that of desktop CPUs of the same era, because Mobile CPUs will integrate what is not available in desktop CPUsPower managementTechnology, and will first adopt higher micron precision.