å¾æ¥ããããARMã¯x86ããï¼é»åçã«ï¼å¹ççã ãã¨ããè¨èª¬ãããã¾ããããã¯åç´ã«ãARMã¯çé»åãªã¹ããåãã§ãx86ã¯é»åãé£ãPCåããç¨åº¦ã®ã¢ãã¦ããªã¤ã¡ã¼ã¸ã®ãã¨ãããã¾ãããåä¸ç´ã®RISC vs CISCè«äºã®ãããããããARMã¯x86 (x64ãå«ã)ã«æ¯ã¹å½ä»¤ã»ãããã·ã³ãã«ãªã®ã§ãå½ä»¤ãã³ã¼ãã«ãããé»åãå°ãªãã¦æ¸ãã§å¹ççãã¨ããè°è«ã®å½¢ãã¨ããã¨ãããã¾ãã ãã®è°è«ã«ã¤ãã¦ã¯ãåå°ä½ã¨ã³ã¸ãã¢ã®å¤ãã¯ãARMãx86 ããå¹çãè¯ãã¨ããã®ã¯ããã¯ãéå»ã®ç¥è©±ã(in todayâs age it is a very dead argument)ã¨ããèªèãå ±æãã¦ããã¨è¨ã£ã¦ããã§ããããæåãªã¨ããã§ã¯Apple CPU (ARM)ã¨Zen (x86)ã®ä¸¡æ¹ãéçºããã¸ã ã»ã±ã©ã¼æ°ã®ã¤ã³ã¿ãã¥ã¼ã§ãè¨ããã¦ãã¾ãããChips and Cheeseã¨ã
ããã¾ã§ãã³ã³ããé¢é£ã®è¨äºã¯ãããªãã«æ¸ãã¦ãã¾ããããæ¹ãã¦ææ°äºæ ã«åããã¦ç·´ãç´ãããè¦æ¸¡ãã¦ã¿ãã¨ã大ããªã¨ããããå°ããªã¨ããã¾ã§é¸æè¢ãå¤ãããã¨æãã¾ããã ã³ã³ããç³»ã¢ã¼ããã¯ãã£ã丸ã£ã¨ä»æã®æ§æã§çä¼¼ãããã¨ã£ã¦ãããããã»ã¨ãã©ãªãã¦ãåèã«ãã¤ã¤èªåæµã«ç¯ãä¸ãã¦ããã§ããããããä»åã¯ç¯ãã«ããã£ã¦ã©ãããé¸æè¢ãããã®ãã«ãã©ã¼ã«ã¹ããå¤åç³»ã§æ»ãã¦ã¿ããã¨æã£ã次第ã§ã:-) ç®æ¬¡ ä»å¹´ä¸çºç®ã®é·ããã¤ã§ããååã¯å¦ç¿ææç¨ãååã¯é楽ãªãã¤ã¹ãã§ãã ã¯ããã« åºç¤ ã¤ã³ã¹ã¿ã³ã¹ or ã³ã³ãã ECS or EKS on EC2 or FARGATE X86 or ARM64 ãã¼ããã©ã³ãµã¼ ã¡ã³ããã³ã¹ï¼ALB or ECS Service å ±æ or ï¼ç°å¢æ¯ ã¢ã¯ã»ã¹ãã°ï¼ALB or WEBãµã¼ãã¼ ECS / EKS ãããã¤ï¼Blue/Gr
æè¿x86ã¨Armã«ç¶ã第3ã®å¢åã¨ãã¦ãRISC-V(ãªã¹ã¯ãã¡ã¤ã)ã®ååãèããã¨ãå¤ããªã£ããRISC-Vã®å ´åãx86ã¨Armã¨ç°ãªãã®ã¯ãã¾ãã¾ãªãã³ãã¼ããã¾ãã¾ãªã³ã¢ãç¨æãã¦ãããã¨ã§ãã¾ã ç¾ç¶ã¯IPã販売ãã¦ããã¬ãã«ã®ä¼æ¥ã®æ¹ãå¤ããããããã®æä¾ãéå§ãã¦ããã¡ã¼ã«ã¼ãåºå§ãã¦ããã ã¤ã¡ã¼ã¸ã¨ãã¦ã¯ã1980年代æ«ï½90年代ã®x86å¸å ´ãèããã°è¯ããã¤ã³ãã«ã¨AMD以å¤ã«Cyrix/IDT/TI/IBM/NexGenãªã©å¤æ°ã®ã¡ã¼ã«ã¼ããç¬èªã®å®è£ ã«åºã¥ãx86ããã»ããµã¼ãå¸å ´æå ¥ãã¦ããæ代ã«è¿ãã ãã¡ããããããç°ãªãç¹ãããã®ã ãã2010年代ååã¯ããã»ããµã¼ã¨ããã°x86ã¨Armãããã«PowerPC/POWERã¨ãã£ãç¨åº¦ãããããã ã£ãã®ã«ããªã2010年代å¾åããæ¥éã«RISC-Vãçãä¸ãã£ãã®ããã¨ããä¸é£ã®æµããæ°åã«åãã¦èª¬æãããã
ãªãã·ã¼ã®ç®ã被ã£ãæè³ç¨ãã³ã·ã§ã³å±ã®GA technologiesããé¢è«ããã°5ä¸åã®ã®ããå¸ãåºåã§åã£ã¦ããã¦åºãæ¸ãã¾ãã件ããã¬ã¦æ¨å£é¾ç¤¾é·ãæ ã¦ã¦éæ
â ããã¸ã¡ã³ãã®æ¥µæã¯ãèªåã®ãã¨ã¯æ£ã«ããããã㨠ã¿ã¤ãã«ã¯ https://qiita.com/jnchito/items/0a0b46106681f41f2f0e ã®ã¤ã³ã¹ãã¤ã¢ã§ãã æã¨ã³ã¸ãã¢ãªã©ããã£ã¦ããæã«ãããã¼ã¸ã£ãä¸å¸ããä½ãã³ã¡ã³ããåããã¨ãã¨ã¯è¨ã£ã¦ãããã®ã³ã¼ããæ¸ããªãã®ã«ãããã¨ãããããªæ°æã¡ã«ãªã£ãçµé¨ãããèªåãå®éã«ããã¸ã¡ã³ããããç«å ´ã«ãªãã¨ããã¯ããReact ã¨ããã¾ãããããã®ã§æ¹éã¨ãåºãã«ãããªããã¨ãªã£ã¦æ¢ã¾ã£ã¦ãã¾ããã¨ãããã¾ãã æ¨ä»ã®ã½ããã¦ã§ã¢ã¨ã³ã¸ãã¢ãªã³ã°ã¯å¹ ãæ·±ããç°æ¬¡å ã®ã¬ãã«ã¾ã§åºãã£ã¦ããã®ã§ãå ¨ã¦ã®ãã¨ãããã¸ã¡ã³ããå®è·µã§ããã¨ããã®ã¯æ£ç´ç¡çãªè©±ã§ããèªåãã§ãããã¨ããããã¸ã¡ã³ãã§ããªããªããã½ããã¦ã§ã¢éçºã®ä¸çã§ã¯ä½ãã§ããªãã®ã«çããã§ãã ããã§å¿ è¦ãªãã¨ã¯ãèªåã®ãã¨ã¯æ£ã«ããããã§ã
NVIDIAã«ããè²·åã失æããã°Armã¯æ¥ç¸¾ä½è¿·ãï¼ä¸¡ç¤¾ãè±å½ã»ç«¶äºå¸å ´åºã«åè«ï¼1/2 ãã¼ã¸ï¼ è±å½æ¿åºå½å±ãç¾å¨é²ãã¦ãããNVIDIAã®Armè²·åã«é¢ãã調æ»ã®ä¸ç°ã¨ãã¦çºè¡¨ããææ¸ã«ããã¨ãããNVIDIAã«ããè²·åææ¡ã失æã«çµãã£ãå ´åãArmã¯ã¹ã¿ã³ãã¢ãã³ä¼æ¥ã¨ãã¦æé·ãã¦ããä¸ã§ãé大ãªéå£ã«ç´é¢ãããã¨ã«ãªãã¨ããã è±å½æ¿åºå½å±ãç¾å¨é²ãã¦ãããNVIDIAã®Armè²·åã«é¢ãã調æ»ã®ä¸ç°ã¨ãã¦çºè¡¨ããææ¸ã«ããã¨ãããNVIDIAã«ããè²·åææ¡ã失æã«çµãã£ãå ´åãArmã¯ã¹ã¿ã³ãã¢ãã³ä¼æ¥ã¨ãã¦æé·ãã¦ããä¸ã§ãé大ãªéå£ã«ç´é¢ãããã¨ã«ãªãã¨ããã 29ãã¼ã¸ã«åã¶ãã®ææ¸ã¯ãè±å½æ¿åºã2021å¹´11æã«è±å½ã®ç«¶äºå¸å ´åºï¼CMAï¼Competition and Markets Authorityï¼ã«ãããªã調æ»ã®å®æ½ãæ示ãããã¨ã«å¯¾ããArmï¼NVIDIAã
"Locality is efficiency, Efficiency is power, Power is performance, Performance is King", Bill Dally ãã«ãã¹ã¬ããã£ã³ã°ã¨ã¯ï¼ CPUã¨GPUã®ãã«ãã¹ã¬ããã£ã³ã°ã®éããããã°ã«ã¾ã¨ãã¦ãããã©ä¾ã«ãã£ã¦èª°ãèå³ãªãããâ arutema47 (@arutema47) 2021å¹´8æ16æ¥ ã¤ã¶ããããèªã¿ããæ¹ãå¤ããã ã£ãã®ã§å®æããã¾ããã ãã«ãã¹ã¬ããã£ã³ã°ã¨ã¯ã¡ã¢ãªé 延ãé è½ãã¹ã«ã¼ããããä¸ãããã¼ãã¦ã§ã¢ã®ãã¯ããã¯ã§ãã ãã CPUã¨GPUã§ä½¿ããæ¹ãããªãç°ãªãããããã®éãã«ã¤ãã¦èãã¦ã¿ãè¨äºã§ãã ï¼SIMDã«ã¤ãã¦ä¸¦åããã°ã©ãã³ã°ã®è¦³ç¹ãã触ããã¹ãã§ããããæéç¡ããã«ãã¹ã¬ããã£ã³ã°ã«æ³¨ç®ããããåçã§ã¯çãã¾ãããï¼ æ¬è¨äºã«ã¤ã㦠æ¬è¨äºã¯CPUã¨G
AMDã®AthlonãZenãã¤ã¯ãã¢ã¼ããã¯ãã£ãApple A4ãªã©æ°ã ã®ãããéçºã«æºãããã天æã¨ã³ã¸ãã¢ãã¨é«ãè©ä¾¡ãåããã¸ã ã»ã±ã©ã¼æ°ã¯2021å¹´1æã«Intelãçªå¦éè·ããè¨äºä½ææç¹ã§ã¯AIãããã®ã¹ã¿ã¼ãã¢ããã§ããTenstorrentã®ç¤¾é·å ¼æé«æè¡è²¬ä»»è ãåãã¦ãã¾ãããããªã±ã©ã¼æ°ã«ãæè¡ç³»ãã¥ã¼ã¹ãµã¤ãã®AnandTechãã¤ã³ã¿ãã¥ã¼ãè¡ã£ã¦ãã¾ãã An AnandTech Interview with Jim Keller: 'The Laziest Person at Tesla' https://www.anandtech.com/show/16762/an-anandtech-interview-with-jim-keller-laziest-person-at-tesla ãªããã¸ã ã»ã±ã©ã¼æ°ã®çµæ´ã¯ä»¥ä¸ã®éãã å ¥ç¤¾éè·å¹´ä¼æ¥è©æ¸ãé¢ä¸ãã製
ç±³IBMã¯x86ãµã¼ãä¸ã®Linuxç°å¢ã§COBOLè¨èªã®ã³ã³ãã¤ã«ãå®è¡ãå®ç¾ãããIBM COBOL for Linux on x86 1.1ããçºè¡¨ãã¾ããã IBM COBOL for Linux on x86 1.1ã¯ããIBM COBOLã³ã³ãã¤ã©ãã¡ããªã¼ãã®ææ°è£½åã¨ãã¦ãIBM Enterprise COBOL for z/OSããã³IBM COBOL for AIXã¨ã®äºææ§ãåãã¦ãã¾ãã ã¾ããEnterprise COBOL for z/OSã¨åãæé©åæè¡ãIBM COBOL for Linux on x86 1.1ã«ç¨ããããã³ã³ãã¤ã©ã®åºæ¬çãªããã¯ã¨ã³ããå ±æããã¦ãããããé«éãªå¦çãå®ç¾ã§ããã¨ã®ãã¨ã TXSeriesã®ãã«ããã©ãããã©ã¼ã ã«å¯¾å¿ãDb2 for Linux/UNIX/Windowsã«å¯¾å¿ãDb2ãã¡ã¤ã«ã·ã¹ãã ã¯ã·ã¼ã±ã³ã·ã£ã«ãã¤
Armå ¥éåå¼·ä¼ã¨ã¯ãmacOSãArmã«ç§»è¡ãããã®æ©ã«Armã¢ã¼ããã¯ãã£ã§ã®ããã°ã©ãã³ã°ã«ã¤ãã¦å ¥éããã½ããã¦ã§ã¢ã¨ã³ã¸ãã¢ã®ããã®ä¼ã§ããä»å主å¬ã®@nullpo_head æ°ããArmã®ä»®æ³åæ¯æ´æ©æ§ã«ã¤ãã¦ããã®ä»çµã¿ããæ·±ã説æãã¾ããååã¯ãç¾ä»£ã®ãã¼ãã¦ã§ã¢ä»®æ³åæ¯æ´æ©æ§ãã«ã¤ãã¦ãå ¨ï¼åã ãã¼ãã¦ã§ã¢ä»®æ³åæ¯æ´ã¨ã¯ä½ã ä½ä¼¯ å¦åæ°ï¼å ¥éã»ãã·ã§ã³3ã¤ç®ã¯ãArmã®ä»®æ³åæ¯æ´æ©æ§ãã«ã¤ãã¦ã®å ¥éã»ãã·ã§ã³ã§ããã©ãããããããé¡ããã¾ãã æ¬çºè¡¨ã®ã¹ã¿ã¼ãã¨ã´ã¼ã«ã§ããVMwareã¨ãQemuã¨ã使ã£ããã¨ãããã©ä»®æ³ãã·ã³ã®ä»çµã¿ãªããç¥ããã¨ããã¨ãããã¾ãã¹ã¿ã¼ãã«ãªã£ã¦ãã¾ãã 1åç®ã®ã´ã¼ã«ã¯ãæè¿ã®VMã®ãã£ããããä»çµã¿ã¨ãã¼ãã¦ã§ã¢ä»®æ³åæ¯æ´ã¨ã¯ä½ãã¨ãããã¨ãããããã¨ãããã¦ãã®è©±ã®ãã¨ã«å®éã«Armã®ä»®æ³åæ¯æ´æ©æ§ã®æ¦è¦ã説æããArmã®ä»®æ³
Apple M1ã«é¢ãã¦ä»¥ä¸ã®ãããªãã¤ã¼ãã話é¡ã«ãªã£ãã 18/ Another "magic" trick is how their "Swift" programming language uses "reference counting" instead of the "garbage collection" in Android. They did something in their CPU to double the speed of reference counting.â RobáµÊ³áµ Grahamð·, provocateur (@ErrataRob) 2020å¹´11æ26æ¥ fun fact: retaining and releasing an NSObject takes ~30 nanoseconds on current gen Intel, and ~6.
ã¢ã¦ãã»ãªãã»ãªã¼ãã¼å®è¡ã«ã¤ãã¦è£è¶³ ååã®è¨äºã§ãã¢ã¦ãã»ãªãã»ãªã¼ãã¼å®è¡ãã«ã¤ãã¦ç¹ã«èª¬æããã«è©±ãé²ãã¦ãã¾ã£ããã¨ã«æ°ã¥ããã®ã§ãã¾ãã¯ããã«ã¤ãã¦ç°¡åã«è£è¶³ãã¦ãããã ã³ã³ãã¥ã¼ã¿ã¼ã®æ§è½åä¸ã®æ´å²ã¯ã¬ã¤ãã³ã·ã¼ã¨ã®æ¦ãã®æ´å²ã§ãããã colin-scott.github.io ä¸ã®ãµã¤ãã¯å¹´ä»£æ¯ã«ã³ã³ãã¥ã¼ã¿ã¼ã·ã¹ãã ã§ã®å種ã¬ã¤ãã³ã·ã¼ãã©ã®ããã«å¤åãã¦ãã£ãããç´¹ä»ãã¦ããã1990年代ååã¯ãã£ãã·ã¥ã¡ã¢ãªã¨ã¡ã¤ã³ã¡ã¢ãªã¨ã®éã®ã¬ã¤ãã³ã·ã¼å·®ã¯ããã»ã©å¤§ãããªãã£ããããã®å¾ã®æè¡é©æ°ã«ãã£ã¦ç¾å¨ã¯L1ãã£ãã·ã¥ã¨ã¡ã¤ã³ã¡ã¢ãªã¨ã®éã«100åãããã®ã¬ã¤ãã³ã·ã¼å·®ãããããã«ãªã£ã¦ãã¾ã£ããããã¯ã¤ã¾ããããã°ã©ã å®è¡ä¸ã«ã¡ã¤ã³ã¡ã¢ãªã¸ã®ã¢ã¯ã»ã¹ãçºçãã¦ãã¾ãã¨ãããã ãé·ãã¬ã¤ãã³ã·ã¼ã®éCPUã®å¦çãé²ãããã¨ãã§ããªããªã£ã¦ãã¾ããã¨ãæå³ããããã®ãã
Appleã®ã«ã¹ã¿ã ããããM1ãã¨ããããæè¼ããæ°ãããMacBook AirããMacBook ProããMac miniãã¯ãIntelã«ã¨ã£ã¦ã¯æ©ã¿ã®ç¨®ã ãAppleãèªç¤¾è£½ãã¼ã½ãã«ã³ã³ãã¥ã¼ã¿ã¼ããIntelã®ããã»ããµã¼ãæé¤ãããé¢å©æç¶ããã¯ã2å¹´ã»ã©ãããè¦è¾¼ã¿ã ã ã¨ã¯ãããIntelã®å½éãå°½ããããã§ã¯ãªãã Intelã«ã¯ãAppleã®è å¨ããPCå¸å ´ãå®ãå¹¾ã¤ãã®å¼·ã¿ããããApple以å¤ã®PCã¡ã¼ã«ã¼ã¯ãããç°¡åã«ã¯Intelããé¢ããããªããIntelã¯ãã¾ã ã«ãM1ãããã¯ãã«ãªãã¤ã¨ã³ããããã®ãªã¼ãã¼ã ãããã«ãç¶æ³æ¹åã®ããã«å¿ è¦ååãªè³éï¼182å5000ä¸ãã«ç¸å½ã®è³ç£ï¼ãæã£ã¦ããã Linley Groupã®ã¢ããªã¹ããLinley Gwennapæ°ã¯ãçæçã«ã¯ãéè¦ãªé¡§å®¢ã1社失ã以å¤ã«ã¯Intelã®PCäºæ¥ã«ã¨ã£ã¦ã®è å¨ã¯ããã»ã©å¤§
ãªãªã¼ã¹ãé害æ å ±ãªã©ã®ãµã¼ãã¹ã®ãç¥ãã
ææ°ã®äººæ°ã¨ã³ããªã¼ã®é ä¿¡
å¦çãå®è¡ä¸ã§ã
j次ã®ããã¯ãã¼ã¯
kåã®ããã¯ãã¼ã¯
lãã¨ã§èªã
eã³ã¡ã³ãä¸è¦§ãéã
oãã¼ã¸ãéã
{{#tags}}- {{label}}
{{/tags}}