AMD联手IBM与Zyphra发布ZAYA1：14T tokens预训练，架构创新性能对标Qwen3

ITBEAR科技资讯11-28

全球人工智能领域迎来一项重要突破——AMD联合IBM与AI初创企业Zyphra共同推出名为ZAYA1的混合专家（MoE）基础模型。该模型成为首个完全基于AMD硬件架构完成训练的同类产品，在数学、科学推理等核心能力上展现出与头部模型竞争的实力，标志着AMD在AI训练生态中的技术突破。训练集群采用IBM Cloud部署的128个节点，每个节点配备8张AMD Instinct MI300X加速器，总计...

网页链接

免责声明：本文观点仅代表作者个人观点，不构成本平台的投资建议，本平台不对文章信息准确性、完整性和及时性做出任何保证，亦不对因使用或信赖文章信息引发的任何损失承担责任。

精彩评论

我们需要你的真知灼见来填补这片空白

发表看法

{"i18n":{"language":"zh_CN"},"isChannel":false,"data":{"share":"https://www.laohu8.com/m/news/2586622473?lang=zh_CN&edition=full","thumbnail":"","is_english":false,"pubTime":"2025-11-28 01:29","share_image_url":"https://static.laohu8.com/9a95c1376e76363c1401fee7d3717173","id":"2586622473","market":"us","top_or_hot":-1,"title":"AMD联手IBM与Zyphra发布ZAYA1：14T tokens预训练，架构创新性能对标Qwen3","media":"ITBEAR科技资讯","content":"<div>\n<p>全球人工智能领域迎来一项重要突破——AMD联合IBM与AI初创企业Zyphra共同推出名为ZAYA1的混合专家（MoE）基础模型。该模型成为首个完全基于AMD硬件架构完成训练的同类产品，在数学、科学推理等核心能力上展现出与头部模型竞争的实力，标志着AMD在AI训练生态中的技术突破。训练集群采用IBM Cloud部署的128个节点，每个节点配备8张AMD Instinct MI300X加速器，总计...</p>\n\n<a href=\"http://gu.qq.com/resources/shy/news/detail-v2/index.html#/?id=nesSN20251128015448951e1edf&s=b\">网页链接</a>\n\n</div>\n","source":"tencent","html":"<!DOCTYPE html>\n<html>\n<head>\n<meta http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\" />\n<meta name=\"viewport\" content=\"width=device-width,initial-scale=1.0,minimum-scale=1.0,maximum-scale=1.0,user-scalable=no\"/>\n<meta name=\"format-detection\" content=\"telephone=no,email=no,address=no\" />\n<title>AMD联手IBM与Zyphra发布ZAYA1：14T tokens预训练，架构创新性能对标Qwen3</title>\n<style type=\"text/css\">\na,abbr,acronym,address,applet,article,aside,audio,b,big,blockquote,body,canvas,caption,center,cite,code,dd,del,details,dfn,div,dl,dt,\nem,embed,fieldset,figcaption,figure,footer,form,h1,h2,h3,h4,h5,h6,header,hgroup,html,i,iframe,img,ins,kbd,label,legend,li,mark,menu,nav,\nobject,ol,output,p,pre,q,ruby,s,samp,section,small,span,strike,strong,sub,summary,sup,table,tbody,td,tfoot,th,thead,time,tr,tt,u,ul,var,video{ font:inherit;margin:0;padding:0;vertical-align:baseline;border:0 }\nbody{ font-size:16px; line-height:1.5; color:#999; background:transparent; }\n.wrapper{ overflow:hidden;word-break:break-all;padding:10px; }\nh1,h2{ font-weight:normal; line-height:1.35; margin-bottom:.6em; }\nh3,h4,h5,h6{ line-height:1.35; margin-bottom:1em; }\nh1{ font-size:24px; }\nh2{ font-size:20px; }\nh3{ font-size:18px; }\nh4{ font-size:16px; }\nh5{ font-size:14px; }\nh6{ font-size:12px; }\np,ul,ol,blockquote,dl,table{ margin:1.2em 0; }\nul,ol{ margin-left:2em; }\nul{ list-style:disc; }\nol{ list-style:decimal; }\nli,li p{ margin:10px 0;}\nimg{ max-width:100%;display:block;margin:0 auto 1em; }\nblockquote{ color:#B5B2B1; border-left:3px solid #aaa; padding:1em; }\nstrong,b{font-weight:bold;}\nem,i{font-style:italic;}\ntable{ width:100%;border-collapse:collapse;border-spacing:1px;margin:1em 0;font-size:.9em; }\nth,td{ padding:5px;text-align:left;border:1px solid #aaa; }\nth{ font-weight:bold;background:#5d5d5d; }\n.symbol-link{font-weight:bold;}\n/* header{ border-bottom:1px solid #494756; } */\n.title{ margin:0 0 8px;line-height:1.3;color:#ddd; }\n.meta {color:#5e5c6d;font-size:13px;margin:0 0 .5em; }\na{text-decoration:none; color:#2a4b87;}\n.meta .head { display: inline-block; overflow: hidden}\n.head .h-thumb { width: 30px; height: 30px; margin: 0; padding: 0; border-radius: 50%; float: left;}\n.head .h-content { margin: 0; padding: 0 0 0 9px; float: left;}\n.head .h-name {font-size: 13px; color: #eee; margin: 0;}\n.head .h-time {font-size: 11px; color: #7E829C; margin: 0;line-height: 11px;}\n.small {font-size: 12.5px; display: inline-block; transform: scale(0.9); -webkit-transform: scale(0.9); transform-origin: left; -webkit-transform-origin: left;}\n.smaller {font-size: 12.5px; display: inline-block; transform: scale(0.8); -webkit-transform: scale(0.8); transform-origin: left; -webkit-transform-origin: left;}\n.bt-text {font-size: 12px;margin: 1.5em 0 0 0}\n.bt-text p {margin: 0}\n</style>\n</head>\n<body>\n<div class=\"wrapper\">\n<header>\n<h2 class=\"title\">\nAMD联手IBM与Zyphra发布ZAYA1：14T tokens预训练，架构创新性能对标Qwen3\n</h2>\n\n<h4 class=\"meta\">\n\n\n2025-11-28 01:29 北京时间&nbsp;&nbsp;&nbsp;<a href=http://gu.qq.com/resources/shy/news/detail-v2/index.html#/?id=nesSN20251128015448951e1edf&s=b><strong>ITBEAR科技资讯</strong></a>\n\n\n</h4>\n\n</header>\n<article>\n<div>\n<p>全球人工智能领域迎来一项重要突破——AMD联合IBM与AI初创企业Zyphra共同推出名为ZAYA1的混合专家（MoE）基础模型。该模型成为首个完全基于AMD硬件架构完成训练的同类产品，在数学、科学推理等核心能力上展现出与头部模型竞争的实力，标志着AMD在AI训练生态中的技术突破。训练集群采用IBM Cloud部署的128个节点，每个节点配备8张AMD Instinct MI300X加速器，总计...</p>\n\n<a href=\"http://gu.qq.com/resources/shy/news/detail-v2/index.html#/?id=nesSN20251128015448951e1edf&s=b\">网页链接</a>\n\n</div>\n\n\n</article>\n</div>\n</body>\n</html>\n","isBrief":false,"type":0,"news_type":1,"symbol":"LU0823421333.USD","symbol_name":"BNP PARIBAS DISRUPTIVE TECHNOLOGY \"C\" (USD) ACC","start_time":0,"source_url":"http://gu.qq.com/resources/shy/news/detail-v2/index.html#/?id=nesSN20251128015448951e1edf&s=b","article_id":"2586622473","we_media_id":null,"thumbnails":[],"rights":{"source":"tencent","url":"http://gu.qq.com/resources/shy/news/detail-v2/index.html#/?id=nesSN20251128015448951e1edf&s=b","rn_cache_url":null,"customStyle":"body{padding-top:10px;}#news_title{font-weight:bold;#titleStyle#;}#news_description span{font-size:12px;#descriptionStyle#;}.footer-note{#statement#}","selectors":".mod-LoadTzbdNews, body","filters":".relate-stock, .hot-list, .recom-box, .wx-sou","directOrigin":true},"url":"https://stock-news.laohu8.com/highlight/detail?id=2586622473","pubTimestamp":1764264562,"columns":[],"sourceInfo":{"source_id":"tencent","name":"腾讯"},"weMediaInfo":null,"summary":"全球人工智能领域迎来一项重要突破——AMD联合IBM与AI初创企业Zyphra共同推出名为ZAYA1的混合专家基础模型。Zyphra团队表示，当前版本为预训练基础模型，后续将发布经过指令微调的增强版本。AMD方面透露，此次合作验证了MI300X加速器与ROCm软件栈在大规模MoE模型训练中的可行性。这项进展为AI训练硬件生态的多元化发展提供了新的技术路径选择。","collect":0,"end_time":0,"defaultTopTitle":"qq.com","property":[],"viewcount":null,"language":"zh","relate_stocks":{"LU0823421333.USD":"BNP PARIBAS DISRUPTIVE TECHNOLOGY \"C\" (USD) ACC","IBM":"IBM","LU0056508442.USD":"贝莱德世界科技基金A2","BK4585":"ETF&股票定投概念","LU1242518857.USD":"FULLERTON LUX FUNDS - ASIA ABSOLUTE ALPHA \"I\" (USD) ACC","LU2264538146.SGD":"Fullerton Lux Funds - Global Absolute Alpha A Acc SGD","LU0823421416.USD":"BNP PARIBAS DISRUPTIVE TECHNOLOGY \"C\" (USD) INC","LU2098885051.SGD":"JPMorgan Funds - Multi-Manager Alternatives A (acc) SGD","BK4538":"云计算","BK4602":"量子计算概念","LU1852331112.SGD":"Blackrock World Technology Fund A2 SGD-H","LU2458330169.SGD":"FRANKLIN SHARIAH TECHNOLOGY \"A\" (SGD) ACC","LU0868494617.USD":"UBS (LUX) EQUITY SICAV - US TOTAL YIELD SUSTAINABLE \"P\" (USD) ACC","LU1880398471.USD":"AMUNDI FUNDS GLOBAL EQUITY \"A2\" (USD) ACC","LU1880398554.USD":"AMUNDI FUNDS GLOBAL EQUITY \"A2\" (USD) INC","LU1064131342.USD":"Fullerton Lux Funds - Global Absolute Alpha A Acc USD","LU0109392836.USD":"富兰克林科技股A","LU0823434583.USD":"BNP PARIBAS US GROWTH \"C\" (USD) ACC","LU1861558580.USD":"日兴方舟颠覆性创新基金B","LU1951198990.SGD":"Natixis Thematics AI & Robotics Fund H-R/A SGD-H","BK4134":"信息科技咨询与其它服务","LU0823434740.USD":"BNP PARIBAS US GROWTH \"C\" (USD) INC","LU1894683264.USD":"AMUNDI FUNDS US EQUITY RESEARCH VALUE \"A2\" (USD) ACC","BK4612":"AI芯片","LU1951200564.SGD":"Natixis Thematics AI & Robotics Fund R/A SGD","LU0082616367.USD":"摩根大通美国科技A（dist）","LU0127658192.USD":"EASTSPRING INVESTMENTS GLOBAL TECHNOLOGY \"A\" (USD) ACC","LU0274383776.USD":"MANULIFE GF US SMALL CAP EQUITY \"AA\" (USD) INC","LU1242518931.SGD":"Fullerton Lux Funds - Asia Absolute Alpha A Acc SGD","AMD":"美国超微公司","BK4614":"Manus概念股","BK4588":"碎股","LU1585245621.USD":"EASTSPRING INV GLOBAL LOW VOLATILITY EQUITY FUND \"A\" (USD) ACC B","LU1803068979.SGD":"FTIF - Franklin Technology A (acc) SGD-H1","LU1894683348.USD":"AMUNDI FUNDS US EQUITY RESEARCH VALUE \"A2\" (USD) INC","LU1430594728.SGD":"Eastspring Investments - Global Low Volatility Equity AS SGD","BK4532":"文艺复兴科技持仓","LU1923623000.USD":"Natixis Thematics AI & Robotics Fund R/A USD","BK4512":"苹果概念","LU1674673691.USD":"HSBC GIF GLOBAL LOWER CARBON EQUITY \"AD\" (USD) INC","IE00BMPRXR70.SGD":"Neuberger Berman 5G Connectivity A Acc SGD-H","IE00BMPRXN33.USD":"NEUBERGER BERMAN 5G CONNECTIVITY \"A\" (USD) ACC","LU0081259029.USD":"UBS (LUX) EQUITY FUND - TECH OPPORTUNITY \"P\" (USD) ACC","LU0198837287.USD":"UBS (LUX) EQUITY SICAV - USA GROWTH \"P\" (USD) ACC","BK4141":"半导体产品","BK4534":"瑞士信贷持仓","GFS":"GLOBALFOUNDRIES Inc.","BK4529":"IDC概念","BK4575":"芯片概念","BK4566":"资本集团"},"translate_title":"AMD teams up with IBM and Zyphra to release ZAYA1: 14T tokens pre-training, architecture innovation performance benchmarking Qwen3","themeId":null,"isJumpTheme":false,"ttsUrl":null,"symbols_score_info":{"AMD":1.5,"GFS":1.5,"IBM":1.5},"content_text":"全球人工智能领域迎来一项重要突破——AMD联合IBM与AI初创企业Zyphra共同推出名为ZAYA1的混合专家（MoE）基础模型。该模型成为首个完全基于AMD硬件架构完成训练的同类产品，在数学、科学推理等核心能力上展现出与头部模型竞争的实力，标志着AMD在AI训练生态中的技术突破。训练集群采用IBM Cloud部署的128个节点，每个节点配备8张AMD Instinct MI300X加速器，总计1024张GPU通过InfinityFabric互联技术构建。该集群依托ROCm软件栈实现高效协同，峰值算力达750PFLOPs。训练过程消耗14万亿tokens数据，采用渐进式课程学习策略，从通用网页数据逐步过渡到数学、代码及推理专项数据集。Zyphra团队表示，当前版本为预训练基础模型，后续将发布经过指令微调的增强版本。模型架构包含两项关键创新：其一为卷积压缩注意力（CCA）机制，通过将卷积操作与注意力头压缩技术结合，在保持性能的同时降低32%显存占用，长序列处理吞吐量提升18%；其二为线性路由混合专家系统，通过细化专家粒度并引入负载均衡正则化，在Top-2路由策略下实现2.3个百分点的精度提升，即使在70%稀疏度条件下仍能维持90%以上的硬件利用率。基准测试显示，ZAYA1-Base版本在MMLU-Redux、GSM-8K、MATH、ScienceQA等综合评估中与Qwen3-Base持平，在CMATH数学推理和OCW-Math专业测试中表现显著优于对比模型。值得注意的是，该模型未经指令微调即在STEM领域展现出接近专业版本的推理能力，验证了架构设计的有效性。Zyphra计划于2026年第一季度推出指令微调版本和强化学习对齐（RLHF）版本，届时将开放API接口及模型权重下载。AMD方面透露，此次合作验证了MI300X加速器与ROCm软件栈在大规模MoE模型训练中的可行性。公司正与多家云服务提供商推进\"纯AMD\"训练集群的标准化部署，目标在2026年训练超千亿参数模型时，使总拥有成本（TCO）达到与NVIDIA方案相当的水平。这项进展为AI训练硬件生态的多元化发展提供了新的技术路径选择。","kind":"news","is_publish_news":true,"is_publish_highlight":false,"is_publish_live":false,"is_publish_wemedia":null,"editions":null,"column":"","sentiment":"1","news_tag":"productRelease","news_rank":0,"symbols":[],"gpt_button":0,"need_auth":false,"code":"91000000","status":"200"},"commentList":[],"isCommentEnd":true,"newsSizeData":{"likeSize":0,"commentSize":0,"repostSize":0,"favoriteSize":0,"likeStatus":false,"favoriteStatus":false},"APP":{"userAgent":"Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; ClaudeBot/1.0; +claudebot@anthropic.com)","isDev":false,"isTTM":false,"deviceId":"web-server-community-laohu8-v3","version":"4.35.11","shortVersion":"4.35.11","platform":"web","vendor":"web","appName":"laohu8","isIOS":false,"isAndroid":false,"isTiger":false,"isTHS":false,"isWeiXin":false,"isWeiXinMini":false,"isWeiBo":false,"isQQ":false,"isBaiduSwan":false,"isBaiduBox":false,"isDingTalk":false,"isToutiao":false,"isOnePlus":false,"isHuaWei":false,"isXiaomi":false,"isXiaomiWebView":false,"isOppo":false,"isVivo":false,"isSamsung":false,"isMobile":false},"href":"/m/news/2586622473","isCrawlerRequest":true}