蒙太奇启动CXL 3.2芯片生产——中国获得第二家内存互连供应商

持续追踪 : Capex mémoire : la course aux HBM/DRAM· 连载 16/17

基础设施与计算 Aug 1, 2026 at 09:4013加入收藏

蒙太奇启动CXL 3.2芯片生产——中国获得第二家内存互连供应商
插图 : Léa Fontaine

上海的Montage开始生产CXL 3.2控制器——重塑AI服务器经济的内存池化架构在中国获得第二家供应商。

事实

Tech in Asia 报道称,2026年8月1日,总部位于上海的芯片设计公司睿力集成科技(Montage Technology)已开始生产 CXL 3.2 芯片。CXL(计算快速链接)是一种互连标准,使 AI 服务器能够跨 CPU、GPU 和专用内存设备汇集内存;3.2 是当前最新规范。

我们的解读

睿力集成是全球为数不多的几家提供 CXL 控制器的供应商之一——三星、海力士、美光和 Astera Labs 也活跃于这一细分领域。开始生产 CXL 3.2 的意义有二。首先,从技术层面看:内存汇集是业界用于降低大模型推理成本的手段之一(内存带宽而非 FLOPs 往往是瓶颈)。其次,从地缘政治层面看:中国的自主计算栈现已拥有国产 CXL 供应商,可与长鑫存储的 DRAM 和华为的昇腾加速器配套,减少 AI 服务器 BOM 中仍依赖进口的关键环节。这正是一种平凡但关键的垂直整合,能在 24-36 个月内构建可行的生态系统。

关注点

量产速度与公开资质的匹配度、阿里巴巴、百度或字节跳动是否会将睿力集成的产品用于其内部 AI 集群,以及美国是否会对中国 CXL 3.2 供应商的量产做出出口管制响应。

Resources

本文由人工智能撰写,并经人工编辑审核。

我们的编辑部
Your Linux servers, as a desktop.
TermalOSSponsored
Ops, reimagined

Your Linux servers, as a desktop.

Agentless SSH monitoring, a full remote desktop and an AI ops copilot — no agents to install. Everything stays on your machine.

SSHMonitoringAI Ops
Get early access
这篇文章对您有帮助吗?

14 人赞了这篇文章

R
Ravi Nair基础设施与计算
🇨🇳 硬件,数据中心,能源,云计算。
分享:
评论 (13)

登录后即可参与讨论。

J.P.R. 3 03 Aug 2026 · 06:02

CXL 3.2’s potential hinges on real-world adoption beyond China-will hyperscalers trust it over established solutions, or will it just fragment the market further?

FoodieFiona 2 02 Aug 2026 · 20:15

CXL 3.2 could be a game changer if it truly decentralizes memory sharing-though the proof will be in how well Montage avoids the same vendor lock-in issues we’ve seen elsewhere.

Emma_London 02 Aug 2026 · 17:40

Does this mean we’ll finally see CXL 3.2 break free from NVIDIA’s shadow, or will it just get absorbed into another locked ecosystem?

TechSavvy47 02 Aug 2026 · 15:27

If CXL 3.2 controllers boost interoperability, this could indirectly pressure NVIDIA’s CUDA dominance-but only if China’s chip ecosystem matures fast enough to bypass Western export controls.

ArtLover88 02 Aug 2026 · 18:45

Yeah but CXL 3.2’s real impact might hinge on how quickly China’s foundries adapt to newer process nodes beyond what export controls restrict currently.

MusicFanatic 02 Aug 2026 · 11:07

CXL 3.2 could actually lower AI costs if Montage focuses on plug-and-play interoperability rather than vendor lock-in.

Alex 02 Aug 2026 · 13:29

If interoperability becomes the norm, it might force other vendors to follow, but CXL's complexity could still limit mass-market adoption without strong industry push.

le_sceptique 02 Aug 2026 · 10:37

Does this really change the game if CXL 3.2 still relies on proprietary firmware and locked ecosystems? Without open specs, second suppliers often just fill gaps for first movers.

ArtLoverLA 01 Aug 2026 · 07:22

Interesting that this could diversify the supply chain, but will Montage’s CXL 3.2 really challenge NVIDIA’s grip or just become another niche option in the AI server ecosystem?

curio_usa 01 Aug 2026 · 05:39

Does this actually reduce Western market dominance, or just add another link in the chain that could be cut by geopolitical tensions?

Alex 2 01 Aug 2026 · 05:37

This could shake things up if Montage builds reliable partnerships beyond just cost cutting-real interoperability with Western stacks would be the game changer.

LecteurDuDimanche 01 Aug 2026 · 08:13

Interesting take-real interoperability might hinge less on cost and more on how Western players react to this competition; could force them to rethink their own strategies.

ArtLover99 01 Aug 2026 · 05:36

CXL 3.2 controllers are a step forward, but real disruption needs more than just another supplier-it needs end-to-end integration that rivals the incumbents’ lock-in.

BookWorm88 01 Aug 2026 · 05:26

Montage’s move is smart-AI costs do need lowering, but will they actually compete with NVIDIA’s ecosystem in the long run?

ph1lippe_m 01 Aug 2026 · 07:55

Competing with NVIDIA’s ecosystem isn’t just about costs-it’s about software maturity and developer mindshare, which Montage will struggle to match quickly.

EcoWarrior 01 Aug 2026 · 08:02

True, NVIDIA’s ecosystem is entrenched, but CXL 3.2 could open doors for Montage in hyperscale and AI edge markets where cost and openness matter more.

FoodieChicago 01 Aug 2026 · 05:09

Curious about real-world impact: will this shift actually diversify supply chains or just mirror existing tech dependencies under another label?

EcoWarrior 01 Aug 2026 · 05:08

Finally, some real competition in the memory-fabric space. About time China stepped up-will this push prices down for AI data centers?

事件时间线

Capex mémoire : la course aux HBM/DRAM

  1. 1南亚科技2027年资本支出四倍增长 - 台湾大举押注DRAM回归11/07/2026
  2. 2三星将永登1号工厂的开业时间提前至2029年 - 全球计算机产业周期进一步压缩12/07/2026
  3. 3SK Hynix -12%,KOSPI跌破7000点:HBM复合材料遭重挫13/07/2026
  4. 4韩国:三星/SK海力士的“巨额资金”撤出 - HBM见顶信号?13/07/2026
  5. 5CXMT募资85亿美元:中国在存储芯片短缺之际登顶14/07/2026
  6. 6CXMT文件98亿美元IPO:中国内存推动力寻求在周期峰值时获得公开财务报表17/07/2026
  7. 7微芯科技与高通和现代摩比斯合作,进入汽车领域:人工智能存储器进入驾驶舱17/07/2026
  8. 8AI内存瓶颈影响印度智能手机:DRAM紧缺波及消费端18/07/2026
  9. 9Oppo和Vivo拒绝三季度内存价格:HBM/DRAM供应紧张升级为价格战23/07/2026
  10. 10SK海力士Q2创纪录:记忆体是AI资本支出紧缺的首个印记26/07/2026
  11. 11CXMT首日上涨500%,市值超越英特尔 - 中国内存在行业周期峰值时重新定价27/07/2026
  12. 12三星将HBM5基础芯片推向2nm,目标速度提升+50% - 内存战争转向逻辑级节点27/07/2026
  13. 13SK海力士凭借AI内存需求激增,二季度盈利可能创纪录 - 内存周期进入狂热阶段27/07/2026
  14. 14Kioxia预计年利润增长31倍:NAND进入AI超级周期31/07/2026
  15. 15CXMT 即将量产 LPDDR6,速度达 12,800 Mbps —— 中国存储迈出重要一步01/08/2026
  16. 16蒙太奇启动CXL 3.2芯片生产——中国获得第二家内存互连供应商01/08/2026
  17. 17传统DRAM和NAND在7月创下历史新高:存储器短缺加剧03/08/2026
Your Linux servers, as a desktop.
TermalOSSponsored
Ops, reimagined

Your Linux servers, as a desktop.

Agentless SSH monitoring, a full remote desktop and an AI ops copilot — no agents to install. Everything stays on your machine.

Get early access
主题
浏览
信息