近日，推特用户@OneRaichu表示，手上也有酷睿i9-12900K的QS版，在Cinebench R20里，单线程基准测试成绩超过了810，而多线程基准测试成绩超过了11600，这成绩比目前任何LGA 1200和AM4平台的处理器都要好。其中最强的AMD Ryzen 9 5950X，单线程基准测试和多线程基准测试成绩分别为640和10400左右，酷睿i9-12900K QS版分别高出了25%和11%。

鉴于英特尔的新平台还处于早期阶段，未来的成绩有可能会继续提高。

预览var tableid = settableid(); document.write(''); Table().init({ title: "英特尔酷睿桌面处理器参数表", header: ["规格","Coffee Lake-S","Comet Lake-S","Rocket Lake-S","Alder Lake-S","Raptor Lake-S","Meteor Lake-S","Lunar Lake-S"], width: '92%', columnWidth: 'MAX', //表格单元宽度，"MAX"最大内容显示，""AVG"平均，"AUTO"自动，指定[15,16,20] id: tableid, data: [ ["发布年份","2018","2020","2021","2021Q4","2022","2023?","2024?"], ["工艺节点","14nm","14nm","14nm","10nm Enhanced SuperFin","10nm Enhanced SuperFin?","7nm Enhanced SuperFin?","TBC"], ["核心架构","SkyLake","SkyLake","Cypress Cove","Golden Cove+Gracemont","Golden Cove+Gracemont?","Redwood Cove+Gracemont?","TBC"], ["核显","Gen9","Gen9.5","Gen12","Gen12?","Gen12?","Gen 12?","Gen 13"], ["最大核心数","8","10","8","16 (8+8)","16 (8+8)","TBC","TBC"], ["插座","LGA1151","LGA1200","LGA1200","LGA1700","LGA1700","LGA1700","TBC"], ["内存支持","DDR4","DDR4","DDR4","DDR4/5","DDR5","DDR5","DDR5"], ["PCIe版本","PCIe 3.0","PCIe 3.0","PCIe 4.0","PCIe 4.0/5.0","PCIe 5.0","PCIe 5.0","PCIe 5.0"], ["主板芯片组","Intel 300系列","Intel 400系列","Intel 500系列","Intel 600系列?","TBC","TBC","TBC"] ] });由于仍采用14nm工艺，酷睿i9-11900K的功耗就非常高，甚至360规格的水冷想压制都不是轻松的事情，不少人希望英特尔在第12代酷睿系列处理器上能有所改善，毕竟是采用10nm工艺的产品。据推特用户@OneRaichu透露，功耗仍将是英特尔新一代产品将面临的严峻问题，酷睿i9-12900K QS版全速状态下轻松地超过了200W。

按照前几天网上流传的讯息，酷睿i9-12900K QS版的PL1为125W，PL2为228W，这种情况似乎并不让人感到太意外，不过暂时没办法确认该信息。

预览var tableid = settableid(); document.write(''); Table().init({ title: "Alder Lake-S系列处理器规格参数对比", header: ["型号", "Core i9-12900K", "Core i7-12700K", "Core i5-12600K"], width: '92%', columnWidth: 'MAX', //表格单元宽度，"MAX"最大内容显示，""AVG"平均，"AUTO"自动，指定[15,16,20] id: tableid, data: [ ["核心和线程数量", "8C16T+8c", "8C16T+4c", "6C12T+4c"], ["P-Core睿频(1-2个核心)", "5.3 GHz", "5 GHz", "4.9 GHz"], ["P-Core睿频(全核)", "5 GHz", "4.7 GHz", "4.5 GHz"], ["E-Core睿频(1-2个核心)", "3.9 GHz", "3.8 GHz", "3.6 GHz"], ["E-Core睿频(全核)", "3.7 GHz", "3.6 GHz", "3.4 GHz"], ["L3缓存容量", "30MB", "25MB", "20MB"], ["PL1", "125W", "125W", "125W"], ["PL2", "228W", "228W", "228W"], ] });

据VideoCardz报道，英特尔已经在过去的几周里，将Alder Lake-S处理器分发给了合作伙伴，这些QS版的规格很可能就是最终零售版的规格，英特尔在进入批量生产钱应该不会再改变了。同时，有人再次泄露了Alder Lake-S处理器的部分信息，而且有三款，分别是酷睿i9-12900K、酷睿i7-12700K和酷睿i5-12600K。由于采用了big.LITTLE混合架构，大核心称为P-Core，小核心称为E-Core。这三款K系列处理器的PL1均为125W，PL2均为228W。

预览var tableid = settableid(); document.write(''); Table().init({ title: "Alder Lake-S系列处理器规格参数对比", header: ["型号", "Core i9-12900K", "Core i7-12700K", "Core i5-12600K"], width: '92%', columnWidth: 'MAX', //表格单元宽度，"MAX"最大内容显示，""AVG"平均，"AUTO"自动，指定[15,16,20] id: tableid, data: [ ["核心和线程数量", "8C16T+8c", "8C16T+4c", "6C12T+4c"], ["P-Core睿频(1-2个核心)", "5.3 GHz", "5 GHz", "4.9 GHz"], ["P-Core睿频(全核)", "5 GHz", "4.7 GHz", "4.5 GHz"], ["E-Core睿频(1-2个核心)", "3.9 GHz", "3.8 GHz", "3.6 GHz"], ["E-Core睿频(全核)", "3.7 GHz", "3.6 GHz", "3.4 GHz"], ["L3缓存容量", "30MB", "25MB", "20MB"], ["PL1", "125W", "125W", "125W"], ["PL2", "228W", "228W", "228W"], ] });

据HotHardware报道，桌面版的Alder Lake-S已经完成了ES阶段，进入了QS阶段，已经接近于零售版了。有人在网上声称手上有酷睿i9-12900K的QS版，配置了8个Golden Cove性能（大）核心和8个Gracemont效率（小）核心（没有超线程），其基础频率为3.9 GHz，睿频为5.3 GHz。当然这不意味着最终的零售版也是这个频率，仍然有可能会调整。

据了解，这块酷睿i9-12900K QS版在Cinebench R20里的成绩约为11300，相比此前的ES版有较大幅度提高。作为参考，目前目前i9-11900K处理器Cinebench R20里的成绩大概为6200-6300左右，而基于Zen 3架构、拥有16核心32线程的AMD Ryzen 9 5950X处理器的成绩约为10400左右，相比之下酷睿i9-12900K QS版高出了大概8.6%。

还有其他用户表示，如果想发挥这款处理器的全部性能，必须要在Windows 11操作系统上运行，尽管此前某些ES版的Alder Lake-S处理器没办法用在Windows 11上。一直有传言指，Windows 11在调度机制上有重大的调整，所以这样的消息并不让人感到意外，此前Lakefield系列处理器的测试一定程度上已证实了这点。

预览var tableid = settableid(); document.write(''); Table().init({ title: "英特尔酷睿桌面处理器参数表", header: ["规格","Coffee Lake-S","Comet Lake-S","Rocket Lake-S","Alder Lake-S","Raptor Lake-S","Meteor Lake-S","Lunar Lake-S"], width: '92%', columnWidth: 'MAX', //表格单元宽度，"MAX"最大内容显示，""AVG"平均，"AUTO"自动，指定[15,16,20] id: tableid, data: [ ["发布年份","2018","2020","2021","2021Q4","2022","2023?","2024?"], ["工艺节点","14nm","14nm","14nm","10nm Enhanced SuperFin","10nm Enhanced SuperFin?","7nm Enhanced SuperFin?","TBC"], ["核心架构","SkyLake","SkyLake","Cypress Cove","Golden Cove+Gracemont","Golden Cove+Gracemont?","Redwood Cove+Gracemont?","TBC"], ["核显","Gen9","Gen9.5","Gen12","Gen12?","Gen12?","Gen 12?","Gen 13"], ["最大核心数","8","10","8","16 (8+8)","16 (8+8)","TBC","TBC"], ["插座","LGA1151","LGA1200","LGA1200","LGA1700","LGA1700","LGA1700","TBC"], ["内存支持","DDR4","DDR4","DDR4","DDR4/5","DDR5","DDR5","DDR5"], ["PCIe版本","PCIe 3.0","PCIe 3.0","PCIe 4.0","PCIe 4.0/5.0","PCIe 5.0","PCIe 5.0","PCIe 5.0"], ["主板芯片组","Intel 300系列","Intel 400系列","Intel 500系列","Intel 600系列?","TBC","TBC","TBC"] ] });

近日，推特用户@KOMACHI_ENSAKA发现，英特尔向SATA-IO组织提交了即将推出的PCH有关SATA的信息。文件涵盖了Alder Lake和Raptor Lake的PCH系列某些ID，即7AB0到7ACB和7A30到7A4B，证实了这两款处理器可以使用相同的芯片组。当然，这不代表一定是完全兼容的，可能还考虑到其他的因素。事实上，英特尔官方至今还没确认Raptor Lake是Alder Lake的继任者。若不是其他文件的泄露，一直以为采用7nm工艺制造的Meteor Lake才是Alder Lake接班人。

此前有消息指出在桌面平台上，Alder Lake仅支持PCIe 4.0标准，同时600系芯片组也只支持PCIe 4.0标准。如果这个传闻是确切的，那么英特尔可能在700系芯片组才支持PCIe 5.0标准了，这有点类似400系与500系芯片组之间的情况。

预览var tableid = settableid(); document.write(''); Table().init({ title: "英特尔酷睿桌面处理器参数表", header: ["规格","Coffee Lake-S","Comet Lake-S","Rocket Lake-S","Alder Lake-S","Raptor Lake-S","Meteor Lake-S","Lunar Lake-S"], width: '92%', columnWidth: 'MAX', //表格单元宽度，"MAX"最大内容显示，""AVG"平均，"AUTO"自动，指定[15,16,20] id: tableid, data: [ ["发布年份","2018","2020","2021","2021Q4","2022","2023?","2024?"], ["工艺节点","14nm","14nm","14nm","10nm Enhanced SuperFin","10nm Enhanced SuperFin?","7nm Enhanced SuperFin?","TBC"], ["核心架构","SkyLake","SkyLake","Cypress Cove","Golden Cove+Gracemont","Golden Cove+Gracemont?","Redwood Cove+Gracemont?","TBC"], ["核显","Gen9","Gen9.5","Gen12","Gen12?","Gen12?","Gen 12?","Gen 13"], ["最大核心数","8","10","8","16 (8+8)","16 (8+8)","TBC","TBC"], ["插座","LGA1151","LGA1200","LGA1200","LGA1700","LGA1700","LGA1700","TBC"], ["内存支持","DDR4","DDR4","DDR4","DDR4/5","DDR5","DDR5","DDR5"], ["PCIe版本","PCIe 3.0","PCIe 3.0","PCIe 4.0","PCIe 4.0/5.0","PCIe 5.0","PCIe 5.0","PCIe 5.0"], ["主板芯片组","Intel 300系列","Intel 400系列","Intel 500系列","Intel 600系列?","TBC","TBC","TBC"] ] });

在5W部分，随着英特尔停产了Lakefield系列处理器，这个市场将会有Alder Lake-M5接替，采用1个大核心和4个小核心，再由GPU划分成48个EU和64个EU两种型号。在9W部分则提供了2个大核心+8个小核心以及2个大核心+4个小核心的配置，前者的GPU有96个EU，后者则是80个EU。

15W部分和9W部分在处理器规格配置上是一致的，英特尔通过增加TDP以提高频率。在28W部分，英特尔在GPU方面统一配置了96个EU，小核心数量也统一为8个，通过大核心数量区分型号，分别配置了6个和4个。35W-45W部分与28W部分类似，只是酷睿i7在组合上稍微有点不同，会多增加两个大核心。45W-55W部分则统一使用8个大核心+8个小核心，但GPU只配置了32个EU。

目前在某购物网站上，已经有商家在出售LGA 1700插座了，其插座盖和前几天曝光带LGA-18XX字样的款式是一致的，目前显示的价格为88元。商家提供的这款产品非常地完整，包括了整个安装扣具，一般来说是用于主板插座坏的时候进行更换。如果不是专业人员，普通用户不会购买这样的产品，平常不会接触到。

如果与此前泄露的规格图进行比较，各方面都对得上。这应该是第一次出现英特尔Alder Lake-S处理器使用的LGA 1700插座的实物照片，后继的Raptor Lake-S也将继续使用LGA 1700插座。

预览var tableid = settableid(); document.write(''); Table().init({ title: "英特尔酷睿桌面处理器参数表", header: ["规格","Coffee Lake-S","Comet Lake-S","Rocket Lake-S","Alder Lake-S","Raptor Lake-S","Meteor Lake-S","Lunar Lake-S"], width: '92%', columnWidth: 'MAX', //表格单元宽度，"MAX"最大内容显示，""AVG"平均，"AUTO"自动，指定[15,16,20] id: tableid, data: [ ["发布年份","2018","2020","2021","2021Q4","2022","2023?","2024?"], ["工艺节点","14nm","14nm","14nm","10nm Enhanced SuperFin","10nm Enhanced SuperFin?","7nm Enhanced SuperFin?","TBC"], ["核心架构","SkyLake","SkyLake","Cypress Cove","Golden Cove+Gracemont","Golden Cove+Gracemont?","Redwood Cove+Gracemont?","TBC"], ["核显","Gen9","Gen9.5","Gen12","Gen12?","Gen12?","Gen 12?","Gen 13"], ["最大核心数","8","10","8","16 (8+8)","16 (8+8)","TBC","TBC"], ["插座","LGA1151","LGA1200","LGA1200","LGA1700","LGA1700","LGA1700","TBC"], ["内存支持","DDR4","DDR4","DDR4","DDR4/5","DDR5","DDR5","DDR5"], ["PCIe版本","PCIe 3.0","PCIe 3.0","PCIe 4.0","PCIe 4.0/5.0","PCIe 5.0","PCIe 5.0","PCIe 5.0"], ["主板芯片组","Intel 300系列","Intel 400系列","Intel 500系列","Intel 600系列?","TBC","TBC","TBC"] ] });

至于这些新东西什么时候来嘛，推特用户@PJ_Lab_UH公布了一系列新平台主板的发布时间，他是Uniko's Hardware网站的编辑所以消息来源还是比较靠谱的。

在今年第四季度Intel会发布Z690主板，这代表着桌面版的Alder Lake-S处理器也会一同发布，这个其实Intel自己也说过了，只不过没有明说到底会先发那个版本的Alder Lake而已，其实有消息说主板厂其实早在台北电脑展期间要求展示Z690主板了，只不过被Intel拒绝了而已。而B660与H610这些主流级主板则会推后一点，在2022年第一季度推出，估计那个时候Intel才会拿出全部的第12代酷睿桌面处理器。

AMD会在2022年第二季度推出全新的AM5平台，届时肯定会推出Zen 4架构处理器，代号为Raphael，目前对它所知甚少，可以确定的是它会支持DDR5内存，AM5平台也会采用LGA设计，至少拆散热器的时候不会把CPU一同拔出来，不过呢，根据AMD以往的记录，第二季度发布不等于当季就会发售，大家能买到估计都是第三甚至第四季度的事了。

到了2022年第三季度Intel会推出Z790主板，它是搭配第13代酷睿处理器Raptor Lake所准备的，它是Alder Lake的改良版，但具体改了啥现在还不太清楚，接口同样是LGA 1700，选在这个时间点发布估计是Intel对AMD新平台的回应。

其实此前Alder Lake已经多次出现在Geekbench的数据库里面，而近日又有一个14核的样品出现了，这是颗移动平台的Alder Lake-P，平台显示为AlderLake-P DDR5 RVP，也就是Intel的公版测试平台，配了64GB的DDR5内存，内存频率应该是4800MHz的。

目前软件还是不能完全正确识别Alder Lake处理器的信息，14核20线程没错，Alder Lake-P最高的SKU就是6个Golden Cove加8个Gracemont的，前者支持超线程后者不支持，所以就变成了14核20线程。基础频率1.4GHz没问题，睿频频率明显就识别错了，27.1GHz你说这个谁信？缓存识别还是有些问题，L1与L2只识别到Gracemont的，L1的数量还有问题，在Alder Lake里面，每个Gracemont则是每个核心有64KB的L1指令缓存，32KB的L1数据缓存，每4个Gracemont共享2MB的L2缓存；而每个Golden Cove拥有32KB L1指令缓存，48KB L1数据缓存，1280KB的L2缓存；L3缓存方面，每个Golden Cove配3MB，而每4个Gracemont也配3MB，所以有24MB L3缓存。

这颗Alder Lake-P的样品Geekbench 5的得分为单线程1298，多线程9040，单线程性能就比Ryzen 9 4900H高一些，低于Ryzen 9 5980HS，也低于Core i9-11950H，但多线程性能由于核心线程数多，所以是要比他们要好不少的，当然我们不清楚这颗处理器的TDP，也不清楚它正确的睿频频率是多少，所以Alder Lake的性能怎么样现在还不好判断。

近日，Alder Lake-S出现在CapFrameX测试数据库里，搭配的是8GB×4的DDR5内存，频率为4800MHz，同时使用了英伟达GeForce RTX 3080显卡进行了游戏测试。CapFrameX不会记录所有测试硬件的详细规格，但是会记录性能数据和相关的传感器数据，所以只知道测试内容但不知道处理器的详细信息，只能了解到其频率为2.2 GHz，是一个工程样品。

测试的结果显示，在游戏DOTA2中，其成绩为119.98 帧/秒。不过由于不清楚其游戏中的相关设置，所以知道这个成绩也作用不大。这个发现有意义的地方在于，这是迄今为止Alder Lake-S的第一个游戏基准测试，而且说明使用Alder Lake-S已经可以进行游戏了。

预览var tableid = settableid(); document.write(''); Table().init({ title: "英特尔酷睿桌面处理器参数表", header: ["规格","Coffee Lake-S","Comet Lake-S","Rocket Lake-S","Alder Lake-S","Raptor Lake-S","Meteor Lake-S","Lunar Lake-S"], width: '92%', columnWidth: 'MAX', //表格单元宽度，"MAX"最大内容显示，""AVG"平均，"AUTO"自动，指定[15,16,20] id: tableid, data: [ ["发布年份","2018","2020","2021","2021Q4","2022","2023?","2024?"], ["工艺节点","14nm","14nm","14nm","10nm Enhanced SuperFin","10nm Enhanced SuperFin?","7nm Enhanced SuperFin?","TBC"], ["核心架构","SkyLake","SkyLake","Cypress Cove","Golden Cove+Gracemont","Golden Cove+Gracemont?","Redwood Cove+Gracemont?","TBC"], ["核显","Gen9","Gen9.5","Gen12","Gen12?","Gen12?","Gen 12?","Gen 13"], ["最大核心数","8","10","8","16 (8+8)","16 (8+8)","TBC","TBC"], ["插座","LGA1151","LGA1200","LGA1200","LGA1700","LGA1700","LGA1700","TBC"], ["内存支持","DDR4","DDR4","DDR4","DDR4/5","DDR5","DDR5","DDR5"], ["PCIe版本","PCIe 3.0","PCIe 3.0","PCIe 4.0","PCIe 4.0/5.0","PCIe 5.0","PCIe 5.0","PCIe 5.0"], ["主板芯片组","Intel 300系列","Intel 400系列","Intel 500系列","Intel 600系列?","TBC","TBC","TBC"] ] });

大概是吸取了PCIe Gen4上的教训，英特尔在Alder Lake-S上率先支持PCIe Gen5，以免再遇到PCIe标准长时间落后于竞争对手的窘况。Alder Lake-S还同时支持DDR5和DDR4内存，板卡厂商可以选择其中一种进行搭配。另外Alder Lake-S很重要的一个变化是采用新的插座，LGA1700和目前使用的LGA1200的大小也不太一样。对大部分用户来说，如果选择Alder Lake-S，就需要购买新的散热器支持了。

据Wccftech报道，英特尔已告知合作伙伴，计划于2021年11月发布第12代酷睿系列处理器，时间表方面可能会根据一些变化前移或推迟，但基本上是确定的。有消息指，英特尔可能会采取AMD相同的策略，未来几代产品都使用相同插座的平台（类似AM4/AM5）。

预览var tableid = settableid(); document.write(''); Table().init({ title: "英特尔酷睿桌面处理器参数表", header: ["规格","Coffee Lake-S","Comet Lake-S","Rocket Lake-S","Alder Lake-S","Raptor Lake-S","Meteor Lake-S","Lunar Lake-S"], width: '92%', columnWidth: 'MAX', //表格单元宽度，"MAX"最大内容显示，""AVG"平均，"AUTO"自动，指定[15,16,20] id: tableid, data: [ ["发布年份","2018","2020","2021","2021Q4","2022","2023?","2024?"], ["工艺节点","14nm","14nm","14nm","10nm Enhanced SuperFin","10nm Enhanced SuperFin?","7nm Enhanced SuperFin?","TBC"], ["核心架构","SkyLake","SkyLake","Cypress Cove","Golden Cove+Gracemont","Golden Cove+Gracemont?","Redwood Cove+Gracemont?","TBC"], ["核显","Gen9","Gen9.5","Gen12","Gen12?","Gen12?","Gen 12?","Gen 13"], ["最大核心数","8","10","8","16 (8+8)","16 (8+8)","TBC","TBC"], ["插座","LGA1151","LGA1200","LGA1200","LGA1700","LGA1700","LGA1700","TBC"], ["内存支持","DDR4","DDR4","DDR4","DDR4/5","DDR5","DDR5","DDR5"], ["PCIe版本","PCIe 3.0","PCIe 3.0","PCIe 4.0","PCIe 4.0/5.0","PCIe 5.0","PCIe 5.0","PCIe 5.0"], ["主板芯片组","Intel 300系列","Intel 400系列","Intel 500系列","Intel 600系列?","TBC","TBC","TBC"] ] });

这颗标记为Intel Core-1800的芯片拥有16核24线程，也就是Alder Lake-S最顶级的规格，将会采用big.LITTLE混合架构，配备了8个高性能的Golden Cove和8个高效率的Gracemont核心，采用10nm Enhanced SuperFin工艺制造，基本频率为1.8 GHz。使用的是LGA 1700插座，搭配600系芯片组的主板。

其TDP为125W，可运行56秒，PL2为228W，可运行2.44ms，温度限制为100摄氏度。Golden Cove核心最大频率为4.6 GHz（最多2个核心），频率降低到4.4 GHz（3-4个核心），4.2 GHz（5-6个核心）和4 GHz（7-8个核心）。Gracemont核心则是3.4 GHz（1-4个核心）和3 GHz（5-8个核心）。在4.2 GHz的时候，电压为1.3147V。

预览var tableid = settableid(); document.write(''); Table().init({ title: "英特尔酷睿桌面处理器参数表", header: ["规格","Coffee Lake-S","Comet Lake-S","Rocket Lake-S","Alder Lake-S","Raptor Lake-S","Meteor Lake-S","Lunar Lake-S"], width: '92%', columnWidth: 'MAX', //表格单元宽度，"MAX"最大内容显示，""AVG"平均，"AUTO"自动，指定[15,16,20] id: tableid, data: [ ["发布年份","2018","2020","2021","2021Q4","2022","2023?","2024?"], ["工艺节点","14nm","14nm","14nm","10nm Enhanced SuperFin","10nm Enhanced SuperFin?","7nm Enhanced SuperFin?","TBC"], ["核心架构","SkyLake","SkyLake","Cypress Cove","Golden Cove+Gracemont","Golden Cove+Gracemont?","Redwood Cove+Gracemont?","TBC"], ["核显","Gen9","Gen9.5","Gen12","Gen12?","Gen12?","Gen 12?","Gen 13"], ["最大核心数","8","10","8","16 (8+8)","16 (8+8)","TBC","TBC"], ["插座","LGA1151","LGA1200","LGA1200","LGA1700","LGA1700","LGA1700","TBC"], ["内存支持","DDR4","DDR4","DDR4","DDR4/5","DDR5","DDR5","DDR5"], ["PCIe版本","PCIe 3.0","PCIe 3.0","PCIe 4.0","PCIe 4.0/5.0","PCIe 5.0","PCIe 5.0","PCIe 5.0"], ["主板芯片组","Intel 300系列","Intel 400系列","Intel 500系列","Intel 600系列?","TBC","TBC","TBC"] ] });