Radeon RX 6500 XT基于Navi 24核心，采用6nm工艺制造，配置了16MB的无限缓存，拥有16个CU，即1024个流处理器，4GB的GDDR6显存，速率为18 Gbps，显存位宽为64位。该款显卡有几个争议点，分别是显存容量只有4GB、PCIe 4.0接口仅为四通道、不支持4K H.264/H.265编码以及AV1解码。

就在Radeon RX 6500 XT正式解禁前夕，KitGuru发现AMD将社区中一篇名为“Game Beyond 4GB”的文章删除。该文章发表于2020年6月份，内容说明了当时AMD想要向PC游戏玩家的想法，配合宣传拥有6GB或8GB显存的Polaris/RDNA架构显卡。AMD希望引导游戏玩家选择Radeon RX 5500 XT 8GB版，而不是4GB版，表示相比英伟达GeForce GTX 1650 SUPER有着性能的优势。

经过一段时间的消失后，文章最终被AMD放回来了。这表示AMD很可能没有直接删除，应该只是隐藏了，至于为什么又放回来，这操作反映的微妙心态变化就比较有趣了，至少能说明一些问题。

事实上，在Radeon RX 6500 XT以及RX 6400推出前，Radeon RX 6000系列其他产品的配置和定位都相当不错，显存配备上相比竞争对手的产品普遍占有优势。这也间接导致英伟达多次更新GeForce RTX 20/30系列产品线，加大了显卡的显存容量。

Radeon Pro W6500M拥有16个CU，即1024个流处理器，配备了16MB的无限缓存，FP32计算性能为5.3T FLOPS，4GB的GDDR6显存，速率为16 Gbps，显存位宽为64位。Radeon Pro W6300M拥有12个CU，即768个流处理器，配备了8MB的无限缓存，FP32计算性能为3.37T FLOPS，2GB的GDDR6显存，速率为16 Gbps，显存位宽为32位。

AMD表示，Radeon Pro W6400预计会在2022年第一季度开始供货，建议零售价格为229美元。Radeon Pro W6500M和W6300M预计会在今年晚些时候发售，仅在特定国家或地区的OEM系统中提供。

AMD针对更高端的Mac Pro推出了专属的Radeon Pro W6800X/W6900X/W6800X Duo，而针对普通PC的Radeon Pro W6000系列仅有Radeon Pro W6800/W6600/W6400三款产品，分别是对应Navi 21、Navi 23和Navi 24核心，缺少Navi 22核心的型号。让人好奇的是，AMD在基于Navi 24核心的Radeon Pro工作站显卡上，是否还会提供Radeon RX 6500 XT/RX 6400缺失的4K H.264/H.265编码以及AV1解码。

这次AMD更新的重点在于对显存的升级，Radeon RX 6x50 XT系列将配备速率为18 Gbps的GDDR6显存，目前只有Radeon RX 6900XT Liquid Edition和Radeon RX 6500 XT才拥有这样的配置。在上个月，三星已宣布向合作伙伴提供速率为20 Gbps和24 Gbps的16Gb（2GB）GDDR6显存模块，传闻这些速率为18 Gbps的显存模块是通过20 Gbps的芯粒降速而来的。

不过大家更想知道的是，原有的Navi 21和Navi 22是否会改成N6工艺。根据台积电（TSMC）的说法，N6工艺与N7工艺相比，晶体管密度将提高18%，而且可以相同的设计规则、设计基础设施、SPICE模拟程序和IP，可以有效降低生产和开发成本。这意味着在没有投入太多额外成本的情况下，GPU效能会有进一步提升，英特尔明年即将发布的锐炫系列GPU（Alchemist）和AMD最新的Navi 24都采用了N6工艺。不过到目前为止，并没有看到这方面的消息。

三星表示，Xclipse GPU是一种独一无二的混合图形处理器，Xclipse代表了Exynos的“X”和“eclipse”这个词的组合，预示着Xclipse GPU将结束移动游戏的旧时代，并标志着将开启激动人心的新篇章。

Xclipse GPU是以AMD RDNA 2架构为主体，继承了PC平台上硬件加速的光线追踪和可变速率着色（VRS）等高级图形功能。光线追踪作为一项革命性的技术，通过模拟光在现实世界中的物理行为，计算光线从表面反弹时的运动和颜色特性，可以为图形渲染的场景产生逼真的照明效果。近两年来，光线追踪技术受到了不少游戏迷的关注。三星这次与AMD合作，在移动GPU实现了业界首个硬件加速的光线追踪功能，可以为移动设备提供身临其境的用户体验。

Exynos 2200在CPU部分采用了三丛集群结构，配备了单个Arm Cortex-X2内核、三个Cortex-A710大核、以及四个Cortex-A510小核。三星也对NPU进行了升级，性能翻倍，并支持FP16运算。Exynos 2200集成了3GPP Release 16 5G 调制解调器，支持sub-6GHz和mmWave（毫米波）频段，速度最高达10Gbps。此外，Exynos 2200 配备了集成安全单元（iSE）来存储私有加密密钥。

这次三星对Exynos 2200的图像信号处理器（ISP）重新进行了设计，可支持最新的图像传感器，实现最高200 MP的超高分辨率。在30 fps情况下，单摄像头最高108 MP、双摄像头最高64+36 MP。新的ISP可以连接多达7个独立的图像传感器，并同时驱动4个。视频录制上，最高支持4K HDR（或8K）的分辨率。通过NPU，ISP利用先进的内容感知AI相机，基于机器学习和场景识别，对颜色、白平衡、曝光、动态范围等方面进行最佳设置，以获得精致和真实的拍照效果。

Exynos 2200配备了先进的多格式编解码器（MFC），可以实现240 fps的4K视频解码，或者60 fps的8K视频解码，并以120 fps和4K分辨率或30 fps和8K分辨率编码。MFC还集成了AV1解码器，具有HDR10+功能，提供高达144Hz的刷新率。

三星将会在2月8日将举办“Galaxy Unpacked 2022”，发布Galaxy S22、Galaxy S22 Plus和Galaxy S22 Ultra三款机型。按照原计划，Galaxy S22系列将分为三星Exynos 2200，以及高通Snapdragon 8 Gen 1两个平台。传言三星会在欧洲和南美地区使用Exynos 2200，北美、中国大陆、中国香港、东南亚、印度和韩国使用骁龙平台。

考虑到AMD已经在Radeon RX 6500 XT上采用速率为18 Gbps的GDDR6显存，接下来其他型号有类似的升级并不会让人感到奇怪。在上个月，三星已宣布向合作伙伴提供速率为20 Gbps和24 Gbps的16Gb（2GB）GDDR6显存模块，速率比现有普遍的16 Gbps模块提高了不少，有可能出现在AMD和英伟达下一轮的显卡更新中。

AMD Radeon RX 6000系列目前阵容基本完整，不过仍有添加新型号的空间，比如Navi 22核心仅有Radeon RX 6700 XT一款产品，所谓的Radeon RX 6700一直没有出现，有可能要等到Radeon RX 7000系列中采用6nm工艺和RDNA 2架构的发布，采用对应的产品。此外，目前Radeon RX 6600和Radeon RX 6500 XT之间的规格差异较大，不知道AMD将如何弥补产品线的缝隙。

近日，PowerColor推出了一款Mini-ITX规格的AXRX 6500 XT 4GBD6-DH，厚度为双槽，整体尺寸为165 x 125 x 40 mm。其采用了一个双滚珠轴承技术的90mm风扇进行散热，并配有静音风扇技术，以Radeon RX 6500XT的功耗情况，散热方面应该没什么问题。

PowerColor官方介绍里该款显卡配置了一个8Pin外接供电接口，不过提供的产品图片显示的是6Pin外接供电接口。此外还配备了一个HDMI 2.1端口，以及一个DisplayPort 1.4端口。

由于Radeon RX 6500XT仅有4GB显存，PCIe 4.0接口仅为四通道，加上不支持4K H.264/H.265编码以及AV1解码，很可能会影响部分用户的选购。虽然官方建议零售价为199美元，但实际价格可能会高出不少，不过PowerColor并未提供定价和上市时间。

预览var tableid = settableid(); document.write(''); Table().init({ title: "AMD Radeon RX 6000系列 GPU 参数表", header: ["型号","Radeon RX 6900 XT","Radeon RX 6800 XT","Radeon RX 6800","Radeon RX 6700 XT","Radeon RX 6700","Radeon RX 6600 XT","Radeon RX 6600","Radeon RX 6500 XT","Radeon RX 6400"], 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: [ ["GPU","Navi 21 XTX","Navi 21 XT","Navi 21 XL","Navi 22 XT","Navi 22 XL","Navi 23 XT","Navi 23 XL","Navi 24 XT","Navi 24 XL"], ["CU数量","80","72","60","40","TBC","32","28","16","12"], ["流处理器数量","5120","4608","3840","2560","TBC","2048","1792","1024","768"], ["Infinity Cache","128MB","128MB","128MB","96MB","96MB","32MB","32MB","16MB","16MB"], ["基础频率","1825MHz","1825MHz","1815MHz","1604MHz","TBC","2064MHz","1626MHz","TBC","TBC"], ["游戏频率","2015MHz","2015MHz","1815MHz","2424MHz","TBC","2359MHz","2044MHz","2610MHz","2039MHz"], ["Boost频率","2250MHz","2250MHz","2105MHz","2581MHz","TBC","2589MHz","2491MHz","2815MHz","2321MHz"], ["显存容量","16GB G6","16GB G6","16GB G6","12GB G6","12GB/6GB G6?","8GB G6","8GB G6","4GB G6","4GB G6"], ["显存位宽","256-bit","256-bit","256-bit","192-bit","192-bit","128-bit","128-bit","64-bit","64-bit"], ["显存速度","16 Gbps","16 Gbps","16 Gbps","16 Gbps","14 Gbps?","16 Gbps","14 Gbps","18 Gbps","14 Gbps"], ["TGP","300W","300W","250W","230W","TBC","160W","132W","107W","53W"], ["发布时间","Dec 2020","Nov 2020","Nov 2020","Mar 2021","TBC","Aug 2021","Oct 2021","Jan 2022","Jan 2022"], ] });

目前Ryzen 6000系列APU仅限于移动平台，许多消费者也期待新一代产品出现在桌面平台上。据TomsHardware报道，AMD副总裁兼客户渠道业务总经理David McAfee表示，Ryzen 6000系列APU什么时候引入到下一代AM5平台是AMD考虑的问题之一，这将以DDR5内存的供应和价格稳定为前提。McAfee认为，DDR5内存和RDNA 2架构核显是最佳搭档，新一代内存更高的带宽将有利于核显的发挥。

AMD的态度也说明了，在桌面平台上，Ryzen 6000系列APU仅支持DDR5内存。这与英特尔的做法有所区别，Alder Lake-S同时支持DDR4和DDR5内存，在DDR5内存尚未成熟的时期，这种做法看起来更为合适一些。由于全球半导体供应链上的短缺，以及新冠疫情影响，DDR5内存持续供应紧张，价格居高不下。

AMD将会在2022年下半年发布Ryzen 7000系列CPU，这款代号Raphael的Zen 4架构处理器将是首款采用AM5平台的产品。根据AMD首席执行官苏姿丰博士的说法，新的AM5平台与AM4平台一样，会成为同样长期使用的平台。

据Cowcotland报道，目前在法国已经有零售商将Radeon RX 6500 XT挂出，实际价格达到了299欧元。虽然AMD还没有公布Radeon RX 6500 XT在欧洲地区的建议零售价，但以美国地区作为参照，结合过往的定价情况，正常情况下应该会在200欧元左右，这意味着高出了50%左右。

此前已有报道指出，Radeon RX 6500 XT的PCIe 4.0接口仅为四通道，且不支持4K H.264/H.265编码以及AV1解码。若价格较高，会更加影响部分用户的选择。相比之下，英伟达GeForce RTX 3050显存大了一倍，其他功能似乎也没有减少，性能应该更强，官方定价比Radeon RX 6500 XT高了25%。

预览var tableid = settableid(); document.write(''); Table().init({ title: "AMD Radeon RX 6000系列 GPU 参数表", header: ["型号","Radeon RX 6900 XT","Radeon RX 6800 XT","Radeon RX 6800","Radeon RX 6700 XT","Radeon RX 6700","Radeon RX 6600 XT","Radeon RX 6600","Radeon RX 6500 XT","Radeon RX 6400"], 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: [ ["GPU","Navi 21 XTX","Navi 21 XT","Navi 21 XL","Navi 22 XT","Navi 22 XL","Navi 23 XT","Navi 23 XL","Navi 24 XT","Navi 24 XL"], ["CU数量","80","72","60","40","TBC","32","28","16","12"], ["流处理器数量","5120","4608","3840","2560","TBC","2048","1792","1024","768"], ["Infinity Cache","128MB","128MB","128MB","96MB","96MB","32MB","32MB","16MB","16MB"], ["基础频率","1825MHz","1825MHz","1815MHz","1604MHz","TBC","2064MHz","1626MHz","TBC","TBC"], ["游戏频率","2015MHz","2015MHz","1815MHz","2424MHz","TBC","2359MHz","2044MHz","2610MHz","2039MHz"], ["Boost频率","2250MHz","2250MHz","2105MHz","2581MHz","TBC","2589MHz","2491MHz","2815MHz","2321MHz"], ["显存容量","16GB G6","16GB G6","16GB G6","12GB G6","12GB/6GB G6?","8GB G6","8GB G6","4GB G6","4GB G6"], ["显存位宽","256-bit","256-bit","256-bit","192-bit","192-bit","128-bit","128-bit","64-bit","64-bit"], ["显存速度","16 Gbps","16 Gbps","16 Gbps","16 Gbps","14 Gbps?","16 Gbps","14 Gbps","18 Gbps","14 Gbps"], ["TGP","300W","300W","250W","230W","TBC","160W","132W","107W","53W"], ["发布时间","Dec 2020","Nov 2020","Nov 2020","Mar 2021","TBC","Aug 2021","Oct 2021","Jan 2022","Jan 2022"], ] });

根据华擎官方的资料，Radeon RX 6500 XT的PCIe 4.0接口仅为四通道，比基于Navi 23核心的Radeon RX 6600 XT的八通道更少。PCIe 4.0 x4的带宽为8GB/s，等于PCIe 3.0 x8或PCIe 2.0 x16的带宽。除了华擎外，其他显卡厂商暂时还没有提及这一点。

此外，Radeon RX 6500 XT还缺乏其他Radeon RX 6000系列GPU一些重要的功能。根据3DCenter汇总的信息，Navi 24核心不支持4K H.264/H.265编码以及AV1解码。如果用户对于这些功能有刚性需求，那么搭载Navi 24的Radeon RX 6500 XT/RX 6400显卡就不是理想的选择了。

预览var tableid = settableid(); document.write(''); Table().init({ title: "AMD Radeon RX 6000系列 GPU 参数表", header: ["型号","Radeon RX 6900 XT","Radeon RX 6800 XT","Radeon RX 6800","Radeon RX 6700 XT","Radeon RX 6700","Radeon RX 6600 XT","Radeon RX 6600","Radeon RX 6500 XT","Radeon RX 6400"], 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: [ ["GPU","Navi 21 XTX","Navi 21 XT","Navi 21 XL","Navi 22 XT","Navi 22 XL","Navi 23 XT","Navi 23 XL","Navi 24 XT","Navi 24 XL"], ["CU数量","80","72","60","40","TBC","32","28","16","12"], ["流处理器数量","5120","4608","3840","2560","TBC","2048","1792","1024","768"], ["Infinity Cache","128MB","128MB","128MB","96MB","96MB","32MB","32MB","16MB","16MB"], ["基础频率","1825MHz","1825MHz","1815MHz","1604MHz","TBC","2064MHz","1626MHz","TBC","TBC"], ["游戏频率","2015MHz","2015MHz","1815MHz","2424MHz","TBC","2359MHz","2044MHz","2610MHz","2039MHz"], ["Boost频率","2250MHz","2250MHz","2105MHz","2581MHz","TBC","2589MHz","2491MHz","2815MHz","2321MHz"], ["显存容量","16GB G6","16GB G6","16GB G6","12GB G6","12GB/6GB G6?","8GB G6","8GB G6","4GB G6","4GB G6"], ["显存位宽","256-bit","256-bit","256-bit","192-bit","192-bit","128-bit","128-bit","64-bit","64-bit"], ["显存速度","16 Gbps","16 Gbps","16 Gbps","16 Gbps","14 Gbps?","16 Gbps","14 Gbps","18 Gbps","14 Gbps"], ["TGP","300W","300W","250W","230W","TBC","160W","132W","107W","53W"], ["发布时间","Dec 2020","Nov 2020","Nov 2020","Mar 2021","TBC","Aug 2021","Oct 2021","Jan 2022","Jan 2022"], ] });

AMD Radeon RX 6500 XT和RX 6400均采用Navi 24核心，采用6nm工艺制造，配置了16MB的无限缓存、1MB的L2缓存、以及128KB的L1缓存。前者将配置16个CU，拥有1024个流处理器，后者只有12个CU，流处理器数量为768个。两款显卡会配备4GB的GDDR6显存，显存位宽为64位。

据VideoCardz报道，Radeon RX 6500 XT的加速频率为2815 MHz，游戏频率为2610 MHz，比搭载Navi 23的Radeon RX 6600 XT高了约9%至10%。此前，Radeon RX 6500 XT的GDDR6显存速率将为18 Gbps，而不是之前传闻的16 Gbps，这使得带宽从128 GB/s提高到144 GB/s。Radeon RX 6400的显存速率可能仍为14 Gbps，对应的显存带宽是112 GB/s。

预览var tableid = settableid(); document.write(''); Table().init({ title: "AMD Radeon RX 6000系列 GPU 参数表", header: ["型号","Radeon RX 6900 XT","Radeon RX 6800 XT","Radeon RX 6800","Radeon RX 6700 XT","Radeon RX 6700","Radeon RX 6600 XT","Radeon RX 6600","Radeon RX 6500 XT","Radeon RX 6400"], 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: [ ["GPU","Navi 21 XTX","Navi 21 XT","Navi 21 XL","Navi 22 XT","Navi 22 XL","Navi 23 XT","Navi 23 XL","Navi 24 XT?","Navi 24 XL?"], ["CU数量","80","72","60","40","TBC","32","28","16?","12?"], ["流处理器数量","5120","4608","3840","2560","TBC","2048","1792","1024?","768?"], ["Infinity Cache","128MB","128MB","128MB","96MB","96MB","32MB","32MB","16MB?","16MB?"], ["基础频率","1825MHz","1825MHz","1815MHz","1604MHz","TBC","2064MHz","1626MHz","TBC","TBC"], ["游戏频率","2015MHz","2015MHz","1815MHz","2424MHz","TBC","2359MHz","2044MHz","2610MHz?","TBC"], ["Boost频率","2250MHz","2250MHz","2105MHz","2581MHz","TBC","2589MHz","2491MHz","2815MHz?","TBC"], ["显存容量","16GB G6","16GB G6","16GB G6","12GB G6","12GB/6GB G6?","8GB G6","8GB G6","4GB G6","4GB G6"], ["显存位宽","256-bit","256-bit","256-bit","192-bit","192-bit","128-bit","128-bit","64-bit?","64-bit?"], ["显存速度","16 Gbps","16 Gbps","16 Gbps","16 Gbps","14 Gbps?","16 Gbps","14 Gbps","16 Gbps?","14 Gbps?"], ["TGP","300W","300W","250W","230W","TBC","160W","132W","107W?","53W?"], ["发布时间","Dec 2020","Nov 2020","Nov 2020","Mar 2021","TBC","Aug 2021","Oct 2021","Jan 2022?","Mar 2022?"], ] });

Rembrandt还将配置RDNA 2架构的核显，以取代长时间使用的Vega架构，同时会支持PCIe Gen4和DDR5内存，将采用6nm工艺制造，更换为FP7插座。新一代APU属于Ryzen 6000系列，Ryzen 6000U和6000H系列会代替目前代号Cezanne的相关产品。其最多配置8个内核和16线程，传闻最高频率将达到5.0 GHz，这应该是第一个实现该目标的Ryzen产品，同时核显将配置12个CU。

由于这一代产品更换了插座，变成了FP7，和前两代的Renoir和Cezanne的FP6插座都不相同，笔记本制造商需要调整相应的设计，以适应Ryzen 6000系列APU的需要。在CES 2022大展上，苏姿丰博士将展示新一代APU，传闻还有移动平台的新款Radeon RX 6000M系列产品。

预览var tableid = settableid(); document.write(''); Table().init({ title: "AMD移动平台Ryzen APU参数", header: ["型号","Ryzen 4000系列","Ryzen 5000系列","Ryzen 6000系列","Ryzen 7000系列"], 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: [ ["代号","Renoir","Cezanne","Rembrandt","Phoenix"], ["插座","FP6","FP6","FP7","FP8"], ["架构","Zen 2","Zen 3","Zen 3+","Zen 4"], ["制造工艺","7nm","7nm","6nm","5nm"], ["内存类型","DDR4","DDR4","DDR5","DDR5"], ["GPU架构","Vega","Vega","RDNA 2","RDNA 2"], ["PCIe版本","3.0","3.0","4.0","5.0"], ["最大核心/线程数","8C/16T","8C/16T","8C/16T","TBC"], ["推出时间","2020年","2021年","2022年","2023年?"], ] });

据VideoCardz报道，收到了Navi 24芯片的渲染图。作为一颗小型GPU，其裸片尺寸估计为141mm²，比基于RDNA架构的Navi 14（158mm² ）更小一些。Navi 24仅有Navi 23一半的规格，配置了16个CU，拥有1024个流处理器，16MB的无限缓存（Infinity Cache），显存位宽也只有64位。

图：左边为Radeon RX 6500 XT，右边为Radeon RX 6400

Navi 24有可能是最后一款基于RDNA 2架构的Navi 2x成员，但却是AMD第一款采用6nm工艺制造的消费级GPU。AMD应该会在接下来的CES 2022大展上，展示基于Navi 24核心的Radeon RX 6500 XT和RX 6400显卡。

预览var tableid = settableid(); document.write(''); Table().init({ title: "AMD Radeon RX 6000系列 GPU 参数表", header: ["型号","Radeon RX 6900 XT","Radeon RX 6800 XT","Radeon RX 6800","Radeon RX 6700 XT","Radeon RX 6700","Radeon RX 6600 XT","Radeon RX 6600","Radeon RX 6500 XT","Radeon RX 6400"], 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: [ ["GPU","Navi 21 XTX","Navi 21 XT","Navi 21 XL","Navi 22 XT","Navi 22 XL","Navi 23 XT","Navi 23 XL","Navi 24 XT?","Navi 24 XL?"], ["CU数量","80","72","60","40","TBC","32","28","16?","12?"], ["流处理器数量","5120","4608","3840","2560","TBC","2048","1792","1024?","768?"], ["Infinity Cache","128MB","128MB","128MB","96MB","96MB","32MB","32MB","16MB?","16MB?"], ["基础频率","1825MHz","1825MHz","1815MHz","1604MHz","TBC","2064MHz","1626MHz","TBC","TBC"], ["游戏频率","2015MHz","2015MHz","1815MHz","2424MHz","TBC","2359MHz","2044MHz","TBC","TBC"], ["Boost频率","2250MHz","2250MHz","2105MHz","2581MHz","TBC","2589MHz","2491MHz","TBC","TBC"], ["显存容量","16GB G6","16GB G6","16GB G6","12GB G6","12GB/6GB G6?","8GB G6","8GB G6","4GB G6","4GB G6"], ["显存位宽","256-bit","256-bit","256-bit","192-bit","192-bit","128-bit","128-bit","64-bit?","64-bit?"], ["显存速度","16 Gbps","16 Gbps","16 Gbps","16 Gbps","14 Gbps?","16 Gbps","14 Gbps","16 Gbps?","14 Gbps?"], ["TGP","300W","300W","250W","230W","TBC","160W","132W","107W?","53W?"], ["发布时间","Dec 2020","Nov 2020","Nov 2020","Mar 2021","TBC","Aug 2021","Oct 2021","Jan 2022?","Mar 2022?"], ] });

据Wccftech报道，近期得到了有关代号Rembrandt的APU更多的细节，其中AMD Ryzen 9 6900HX将是新系列中最顶尖的产品之一。Ryzen 9 6900HX将是8核心16线程的规格，拥有4MB的L2缓存和16MB的L3缓存。其加速频率为4.6 GHz，与目前的Ryzen 9 5900HX相同，但功耗会更低。

目前仍不清楚新APU将配置多少个CU，但AMD会从这一代产品开始，改变集成显卡的命名方法。Rembrandt使用的RDNA 2架构GPU将称为Radeon 6xxM，对应Ryzen 9 6900HX的核显名为Radeon 680M。由于使用DDR5-4800内存会比以往DDR4内存的带宽有较大幅度增加，这可能更有利于RDNA 2架构核显的发挥。不过Ryzen 9 6900HX针对的是高性能笔记本电脑，大多数厂商会搭配英伟达的高端GeForce移动显卡使用。

此外，AMD可能会在2022年第三季度发布其对应的桌面版，在Zen 4架构的Raphael处理器之后。

预览var tableid = settableid(); document.write(''); Table().init({ title: "AMD移动平台Ryzen APU参数", header: ["型号","Ryzen 4000系列","Ryzen 5000系列","Ryzen 6000系列","Ryzen 7000系列"], 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: [ ["代号","Renoir","Cezanne","Rembrandt","Phoenix"], ["插座","FP6","FP6","FP7","FP8"], ["架构","Zen 2","Zen 3","Zen 3+","Zen 4"], ["制造工艺","7nm","7nm","6nm","5nm"], ["内存类型","DDR4","DDR4","DDR5","DDR5"], ["GPU架构","Vega","Vega","RDNA 2","RDNA 2"], ["PCIe版本","3.0","3.0","4.0","5.0"], ["最大核心/线程数","8C/16T","8C/16T","8C/16T","TBC"], ["推出时间","2020年","2021年","2022年?","2023年?"], ] });

图：左边为Radeon RX 6500 XT，右边为Radeon RX 6400

相比发布时间，更让人感兴趣的是Navi 24核心的新传言，这款GPU将采用6nm工艺制造。虽然没有具体说明哪间晶圆代工厂，但很可能仍是台积电，这意味着Navi 24将与英特尔第一代锐炫（Intel Arc）Alchemist显卡采用相同的工艺制造。有趣的是，过去有消息指AMD的下一代的Radeon RX 7000系列将由RDNA 2和RDNA 3架构的GPU组成，其中RDNA 2架构GPU会采用6nm工艺制造。基于RDNA 3架构的Navi 31/Navi 32/Navi 33面向中高端市场，将采用5nm工艺制造，而中低端市场则有6nm工艺的产品负责。

AMD Radeon RX 6500 XT和RX 6400均采用Navi 24核心，配置了16MB的无限缓存、1MB的L2缓存、以及128KB的L1缓存。前者将配置16个CU，拥有1024个流处理器，后者只有12个CU，流处理器数量为768个。两款显卡会配备4GB的GDDR6显存，显存位宽为64位。传闻Radeon RX 6500 XT和RX 6400的GDDR6显存速率分别为16 Gbps和14 Gbps，对应的显存带宽是128 GB/s和112 GB/s。据称AMD将使用新的宣传方式，即具有无限缓存（Infinity Cache）的显存带宽，Radeon RX 6500 XT和RX 6400将标称为232 GB/s和216 GB/s。

功耗方面，Radeon RX 6500 XT和RX 6400分别为107W和53W。从泄露的渲染图看到，基于前者（Navi 24 XT）的显卡是双槽规格，会有一个6Pin外接供电接口，基于后者（Navi 24 XL）的显卡将是单槽半高规格，不需要外接供电。两款显卡都会配备两种显示输出端口，分别是DisplayPort和HDMI。有点遗憾的是，目前仍然不清楚Radeon RX 6500 XT和RX 6400的核心频率是多少，此前传闻Navi 24核心频率可能会比较高，甚至达到2.8 GHz或以上。

预览var tableid = settableid(); document.write(''); Table().init({ title: "AMD Radeon RX 6000系列 GPU 参数表", header: ["型号","Radeon RX 6900 XT","Radeon RX 6800 XT","Radeon RX 6800","Radeon RX 6700 XT","Radeon RX 6700","Radeon RX 6600 XT","Radeon RX 6600","Radeon RX 6500 XT","Radeon RX 6400"], 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: [ ["GPU","Navi 21 XTX","Navi 21 XT","Navi 21 XL","Navi 22 XT","Navi 22 XL","Navi 23 XT","Navi 23 XL","Navi 24 XT?","Navi 24 XL?"], ["CU数量","80","72","60","40","TBC","32","28","16?","12?"], ["流处理器数量","5120","4608","3840","2560","TBC","2048","1792","1024?","768?"], ["Infinity Cache","128MB","128MB","128MB","96MB","96MB","32MB","32MB","16MB?","16MB?"], ["基础频率","1825MHz","1825MHz","1815MHz","1604MHz","TBC","2064MHz","1626MHz","TBC","TBC"], ["游戏频率","2015MHz","2015MHz","1815MHz","2424MHz","TBC","2359MHz","2044MHz","TBC","TBC"], ["Boost频率","2250MHz","2250MHz","2105MHz","2581MHz","TBC","2589MHz","2491MHz","TBC","TBC"], ["显存容量","16GB G6","16GB G6","16GB G6","12GB G6","12GB/6GB G6?","8GB G6","8GB G6","4GB G6","4GB G6"], ["显存位宽","256-bit","256-bit","256-bit","192-bit","192-bit","128-bit","128-bit","64-bit?","64-bit?"], ["显存速度","16 Gbps","16 Gbps","16 Gbps","16 Gbps","14 Gbps?","16 Gbps","14 Gbps","16 Gbps?","14 Gbps?"], ["TGP","300W","300W","250W","230W","TBC","160W","132W","107W?","53W?"], ["发布时间","Dec 2020","Nov 2020","Nov 2020","Mar 2021","TBC","Aug 2021","Oct 2021","Jan 2022?","Mar 2022?"], ] });