AI芯片扎堆寒武纪还能成中国版“英伟达”？

作者：李白文章来源：陆玖商业评论

2026-06-09 10:17

作者：李白文章来源：陆玖商业评论

2026-06-09 10:17

邦小白快读

全文速览

本文围绕国产AI芯片企业寒武纪的发展和股价波动，梳理了国产AI芯片行业的整体现状，核心干货信息如下：

1. 寒武纪近两年股价经历大幅波动，2025年8月股价超越贵州茅台成为A股第一高价股，2026年多次暴涨暴跌，市值在8000亿左右震荡，涨跌是时代机遇、自身实力和市场规律共同作用的结果。

2. 涨跌背后的逻辑清晰：上涨源于国产替代趋势下AI芯片自主可控需求爆发，寒武纪业绩实现爆发式增长，同时产品力提升，全栈自研路线获得市场认可；下跌源于估值远高于行业平均，业绩增速有放缓迹象，市场竞争格局加剧，引发市场对泡沫的担忧。

3. 行业未来判断：未来3-5年国产AI芯片将经历惨烈洗牌，最终仅剩下2-3家有实力的企业，普通投资者需警惕概念炒作带来的泡沫风险，避免盲目追高。

本文对AI芯片行业的分析，能给国产AI芯片及相关硬科技品牌提供诸多参考，核心干货如下：

1. 需求趋势：当前国产AI芯片受益于自主可控的国产替代趋势，下游云厂商、智算中心采购需求旺盛，寒武纪订单已经排至2027年，预付款持续走高，需求确定性较强。

2. 产品研发与竞争方向：全栈自研、打造云-边-端全场景产品矩阵是可行路线，同时必须重视软硬件生态构建，仅靠芯片性能无法搭建核心护城河。

3. 竞争现状与风险：当前赛道有数十家玩家厮杀，价格战已经打响，利润空间被压缩；华为昇腾已经占据国内一半市场份额，英伟达仍把持高端市场，头部品牌需要应对份额被挤压的风险。

4. 运营风险提示：高度依赖单一业务、客户集中度过高会大幅降低抗风险能力，品牌需要尽量布局多元业务，分散客户结构。

本文梳理了国产AI芯片赛道的现状，给相关赛道卖家提供了机会、风险和策略参考，核心干货如下：

1. 市场机会：当前外部技术封锁背景下，国产替代需求爆发，下游云厂商、智算中心大幅增加国产AI芯片采购量，整体市场需求旺盛，赛道仍有明确增长空间。

2. 风险提示：当前赛道玩家扎堆内卷，价格战已经打响，未来3-5年行业将迎来惨烈洗牌，大量没有核心技术、靠概念炒作的企业会被淘汰，卖家需要警惕合作方暴雷的风险。

3. 合作与布局策略：优先绑定有核心技术积累、全栈布局、产品落地能力强的头部企业，避开没有实际产品的概念项目；当前市场格局已经初步形成，华为昇腾占一半份额，寒武纪位列第三，卖家需要及时调整合作策略适配新的市场格局，同时可以关注高端训练芯片等细分赛道的机会。

本文对AI芯片赛道的分析，给芯片相关生产制造工厂提供了商业机会和转型参考，核心干货如下：

1. 产品生产需求：当前国产AI芯片对制程工艺要求不断提升，头部厂商的主力产品已经采用7nm工艺，能效和算力向国际先进水平看齐，上游代工厂需要加快先进制程工艺的研发升级，满足国内芯片企业的生产需求。

2. 商业机会：当前AI芯片行业需求旺盛，寒武纪等头部主流厂商的订单已经排至2027年，下游客户备货意愿强烈，给上游生产工厂带来了充足且稳定的订单机会，工厂可以优先绑定有实力的头部核心客户，保障产能利用率。

3. 转型与风险提示：未来行业价格战会进一步压缩利润空间，工厂需要加快推进数字化升级，提升生产效率、控制生产成本；同时行业即将迎来洗牌，工厂要避免过度依赖单一芯片企业客户，提前布局多元客户结构，警惕行业退潮后订单下滑的风险。

本文梳理了国产AI芯片行业的痛点和发展趋势，给AI芯片相关服务商提供了行业参考和方向指引，核心干货如下：

1. 行业发展趋势：国产AI芯片行业正从前期的概念炒作逐步走向价值回归，未来3-5年将会经历惨烈的行业洗牌，行业集中度会大幅提升，服务商可以将资源向头部有真实技术、产品的核心企业倾斜，获取更稳定的收益。

2. 核心客户痛点：当前国内AI芯片企业普遍存在开发者生态薄弱的问题，头部企业寒武纪的软件栈和英伟达积累十几年的CUDA生态还有很大差距，企业需要大量的生态建设支持；此外赛道内卷导致利润压缩，大量中小芯片企业面临资金、成本、人才的多重压力，存在较多未被满足的服务需求。

3. 业务机会：服务商可以围绕芯片企业的开发者生态构建、人才招聘、数字化生产、供应链管理、融资对接等需求开发针对性解决方案，抓住国产替代浪潮下的行业增长机会。

本文对AI芯片行业的分析，给布局AI芯片相关产业的平台商提供了战略和运营参考，核心干货如下：

1. 行业企业的核心需求：当前AI芯片赛道玩家众多，大量中小芯片企业普遍面临资金不足、客户资源少、生态配套不完善的问题，需要平台提供资金对接、供需匹配、生态资源整合等多方面的支持。

2. 招商布局方向：平台招商时优先引入有核心技术积累、全栈自研能力、已经实现产品落地的AI芯片企业，严格筛选掉只靠概念炒作没有实际产品的项目，降低平台整体风险。

3. 运营与风险规避：平台需要向入驻企业提示行业泡沫和洗牌风险，引导企业注重核心技术研发和生态构建，同时提前做好企业退出的风险预案，应对行业洗牌带来的波动；平台可以搭建国产AI芯片供需对接专区，对接下游云厂商、智算中心和国内芯片企业，抓住国产替代的行业红利。

本文梳理了国产AI芯片产业的最新动向、存在的问题，对产业研究有较高的参考价值，核心干货如下：

1. 产业最新动向：当前AI芯片赛道热钱涌入，国内已有数十家企业布局赛道，国产替代需求驱动行业快速发展，头部企业寒武纪已经完成从烧钱研发到盈利造血的转变，华为昇腾已经占据国内AI芯片市场50%的份额，市场格局初步形成。

2. 产业存在的新问题：当前产业存在明显的估值泡沫，头部企业寒武纪的动态市盈率达到260多倍，远高于半导体行业60倍的平均水平；赛道内卷严重，价格战已经打响，多数企业存在业务结构单一、客户集中度高、开发者生态薄弱的问题，高端芯片领域和国际龙头英伟达仍有较大差距。

3. 产业趋势与研究方向：目前行业预判未来3-5年产业将经历惨烈洗牌，最终仅会剩下2-3家头部企业，产业从概念炒作走向价值投资，当前头部企业的商业模式仍在探索阶段，先行者寒武纪的发展路径、生态构建模式都值得深入研究。

声明：快读内容全程由AI生成，请注意甄别信息。如您发现问题，请发送邮件至 run@ebrun.com 。

我是品牌商卖家工厂服务商平台商研究者帮我再读一遍。

Quick Summary

This article reviews the development and stock price volatility of domestic Chinese AI chip firm Cambricon Technologies, and outlines the current state of China's domestic AI chip industry. Key takeaways are as follows:

1. Cambricon's share price has seen extreme volatility over the past two years. In August 2025, its share price surpassed Kweichow Moutai to become the highest-priced stock on the A-share market. In 2026, it experienced multiple rounds of sharp rallies and crashes, with its market capitalization fluctuating around RMB 800 billion. These price movements reflect the combined impact of industry-wide opportunities, company-specific strengths and broader market forces.

2. The drivers behind the price swings are clear: Gains were fueled by surging demand for domestically developed, controllable AI chips amid the domestic substitution trend, alongside Cambricon's explosive revenue growth, improved product performance and growing market recognition of its full-stack in-house development strategy. Pullbacks stemmed from its valuation far exceeding industry averages, slowing growth momentum, intensifying competition and growing investor concerns about market froth.

3. Industry outlook: Over the next 3 to 5 years, China's domestic AI chip sector will undergo a brutal consolidation wave, leaving only 2 to 3 capable players standing. Retail investors are advised to watch for bubble risks from concept-driven speculation and avoid chasing overinflated valuations blindly.

This analysis of the AI chip industry offers key insights for domestic Chinese AI chip and related hard technology brands. Key takeaways are as follows:

1. Demand trends: Domestic AI chips are benefiting from the domestic substitution trend that prioritizes technological self-reliance. Downstream cloud providers and AI computing centers are seeing robust procurement demand, with Cambricon's order backlog extending into 2027 and advance payments continuing to rise, pointing to strong demand certainty.

2. Product R&D and competitive strategy: A full-stack in-house development approach, paired with a product portfolio covering cloud, edge and endpoint devices, is a viable path to market leadership. Brands must also prioritize building software and hardware ecosystems, as chip performance alone is not enough to establish a sustainable competitive moat.

3. Competitive landscape and risks: Dozens of players are currently competing in the sector, where a price war has already broken out, squeezing profit margins. Huawei Ascend already holds half of China's domestic market share, while Nvidia continues to dominate the high-end segment, leaving leading brands at risk of share erosion.

4. Operational risk warning: Overreliance on a single business line and excessive customer concentration drastically reduces a brand's resilience. Brands should diversify their business lines and customer base to mitigate risks.

This article maps the current state of China's domestic AI chip sector, and outlines opportunities, risks and strategic insights for sellers active in related fields. Key takeaways are as follows:

1. Market opportunities: Amid external tech restrictions, domestic substitution demand has surged. Downstream cloud providers and AI computing centers are sharply increasing purchases of domestic AI chips, driving strong overall market demand and leaving clear room for continued sector growth.

2. Risk warning: The sector is currently crowded with competitors and facing cutthroat internal competition, with a price war already underway. A brutal consolidation wave will hit the industry over the next 3-5 years, wiping out a large number of firms that lack core technology and rely purely on concept hype. Sellers should be alert to the risk of partner defaults or business failures.

3. Partnership and positioning strategy: Prioritize partnerships with leading companies that have core technology expertise, full-stack布局 and strong product commercialization capabilities, and avoid concept-driven projects with no actual products. The market structure has now taken initial shape: Huawei Ascend holds half the market, and Cambricon ranks third. Sellers should adjust their partnership strategies to align with the new market landscape, while exploring opportunities in niche segments such as high-end training chips.

This analysis of the AI chip sector outlines business opportunities and transformation insights for chip-related manufacturing facilities. Key takeaways are as follows:

1. Production requirements: Domestic AI chip makers are constantly raising the bar for process technology. Leading firms' flagship products already use 7nm process technology, with energy efficiency and computing power approaching international advanced levels. Upstream foundries need to accelerate R&D and upgrades of advanced process technologies to meet the production demand of domestic chip firms.

2. Business opportunities: The AI chip industry is currently seeing booming demand, with order backlogs for leading mainstream players such as Cambricon extending to 2027. Downstream customers have strong willingness to build inventory, bringing abundant and stable order opportunities for upstream manufacturing facilities. Factories can prioritize partnerships with capable leading core clients to maintain high capacity utilization.

3. Transformation and risk warning: Ongoing industry price wars will further compress profit margins. Factories need to speed up digital transformation to improve production efficiency and control production costs. As the sector is set for consolidation, factories should avoid overreliance on a single chip-making client, build a diversified customer base in advance, and guard against order declines after the sector's boom cools.

This article outlines key pain points and development trends in China's domestic AI chip industry, offering industry insights and strategic direction for AI chip-related service providers. Key takeaways are as follows:

1. Industry development trends: China's domestic AI chip industry is gradually moving away from early-stage concept hype and back to fundamentals-driven value. Over the next 3 to 5 years, it will undergo a brutal consolidation wave that will sharply increase industry concentration. Service providers can reallocate resources toward leading core firms with proven technology and products to secure more stable revenue.

2. Core client pain points: Domestic AI chip companies generally suffer from weak developer ecosystems. Even leading firm Cambricon's software stack still lags far behind Nvidia's CUDA ecosystem, which has been built over more than a decade, leaving firms with strong demand for ecosystem-building support. In addition, cutthroat competition has squeezed profits, leaving a large number of small and mid-sized chip firms facing simultaneous pressure from funding shortages, high costs and talent gaps, with many unmet service needs remaining.

3. Business opportunities: Service providers can develop tailored solutions to address chip firms' needs around developer ecosystem building, talent recruitment, digital production, supply chain management and financing matching, to capture growth opportunities amid the domestic substitution wave.

This analysis of the AI chip industry offers strategic and operational insights for platform operators that are positioning in AI chip-related industries. Key takeaways are as follows:

1. Core needs of industry players: The AI chip sector currently has a large number of participants, and most small and mid-sized chip firms face challenges including insufficient funding, limited customer access and incomplete ecosystem support, requiring platforms to provide support across multiple areas such as financing connection, supply-demand matching and ecosystem resource integration.

2. Investment and sourcing direction: When sourcing and onboarding new companies, platforms should prioritize AI chip firms with core technology expertise, full-stack in-house development capabilities and commercially launched products, and strictly screen out concept-driven projects with no actual products to reduce platform-wide risk.

3. Operations and risk mitigation: Platforms should communicate the risks of industry froth and consolidation to tenant companies, guide firms to prioritize core technology R&D and ecosystem building, and prepare contingency plans for company exits in advance to weather consolidation-driven volatility. Platforms can also build dedicated domestic AI chip supply-demand matching zones to connect downstream cloud providers, AI computing centers and domestic chip firms, and capture growth opportunities from the domestic substitution trend.

This article reviews the latest developments and existing challenges of China's domestic AI chip industry, offering high-value reference for industrial research. Key takeaways are as follows:

1. Latest industry developments: Hot capital is flooding into the AI chip sector, with dozens of domestic firms already positioning in the space. Domestic substitution demand is driving rapid industry growth: leading firm Cambricon has completed its transition from R&D-focused unprofitable growth to profitable self-sustainability, while Huawei Ascend holds 50% of China's domestic AI chip market, marking the initial formation of a clear market structure.

2. Emerging industry challenges: The sector currently has a clear valuation bubble: leading firm Cambricon has a trailing twelve-month P/E ratio of over 260, far exceeding the semiconductor industry average of 60. The sector also faces severe cutthroat competition, with a price war already underway. Most firms suffer from single business lines, high customer concentration and weak developer ecosystems, and China still lags far behind global leader Nvidia in the high-end chip segment.

3. Industry trends and research directions: Industry projections indicate that the sector will undergo brutal consolidation over the next 3 to 5 years, leaving only 2 to 3 leading players standing. The industry is shifting from concept hype to value-driven investment. The business models of leading domestic players are still in the exploration stage, so the development path and ecosystem building model of first-mover Cambricon merit in-depth research.

Disclaimer: The "Quick Summary" content is entirely generated by AI. Please exercise discretion when interpreting the information. For issues or corrections, please email run@ebrun.com .

I am a Brand Seller Factory Service Provider Marketplace Seller Researcher Read it again.

寒武纪的上涨，可以说是时代赋予的机遇，亦有自身的实力；而它的下跌，则是市场规律在内的多方面因素的体现。

除了华为昇腾和寒武纪，沐曦、壁仞、燧原、平头哥、海光、昆仑芯等数十家企业都在AI芯片赛道上厮杀。

即便是当下热钱涌入的AI以及芯片赛道，总有潮水日渐退去，红利渐渐见顶的一天。海外AI算力一片红火，国内人人都想成为下一个英伟达，至少是下一个英特尔或者AMD。但是，产业竞争最终是要拿到商业化结果，前瞻性核心护城河搭建同样至关重要，这对尚处竞赛中的玩家是最大的挑战。

这几年，寒武纪的股价一直在坐过山车，市场对其也是褒贬不一，争论很多。从市场表象上看，这家公司经历了从亏损到单季盈利10亿的华丽转身，又上演了股价超越茅台登顶A股的疯狂戏码。有人说，寒武纪如今正迎来一场迟来的估值回归。事实真的如此吗？

堪比芯片版“茅台”，

但在A股之王边缘荡过秋千

如果只聚焦多年来寒武纪的股价走势，某些时刻确实像极了一场惊心动魄的“过山车”游戏。2024年2月，寒武纪股价还曾一度低至100元下方，到2025年8月28日以1587.91元的收盘价正式超越贵州茅台，成为A股第一高价股，寒武纪只用了不到2年时间。

而且这场“股王”争夺战也充满了戏剧性：2025年8月27日盘中首次短暂超越茅台，尾盘回落后，寒武纪总会时不时高开，时不时就低开低走，股价也是起起伏伏，一直处于漫长的震荡之路。如果打开新闻搜索寒武纪股价，股价上涨和股价跌幅的消息、报道几乎是同时存在的。

这种剧烈波动并非多年前的孤例，即使到了2026年，今年寒武纪的股价也同样坐了一趟过山车。比如，今年2月3日，寒武纪盘中大幅跳水，截至当天中午休市股价跌近13%，总市值回落到4500亿元左右，蒸发超700亿元。

而后来几个月又在跌宕中前行，直至最近一两个月就开始狂飙，其中4月30日，寒武纪以7168亿元市值涨停，股价报1699.96元；5月6日市值突破8000亿元；到5月20日，仅用不到三周时间便完成了从7168亿到8600亿的跨越。到6.1日，似乎又处于8000亿市值以下了。

回过头来看，这种大起大落带来的影响是双向的。一方面，芯片股整体上扬和人工智能产业蓬勃发展的大背景，为寒武纪提供了一定的估值支撑。也正是这些实打实的业绩数据，让市场相信寒武纪已经从烧钱研发阶段迈入了盈利+造血的良性循环。

另一方面，外界对寒武纪稳定性的疑虑从未消失。每一次股价暴涨都伴随着“泡沫”的疑问，很多网友在每一次大跌后又都在各种声音中讨论。这种反复的情绪拉扯意味着，投资者既害怕错过下一个国产版英伟达，又担心成为接盘侠，矛盾心理或多或少是存在的。

通常情况下，资本市场都是预期的放大器，一旦某个行业处于风口之上，资金会把未来十年的增长都提前兑现到当下的股价里。

但这种提前透支的狂欢，终究会被现实的冷水浇醒，企业荡秋千行情，本质上是市场在狂热与理性之间的反复摇摆，而话说回来，其实也是国产AI芯片行业从概念炒作走向价值投资的必经之路。

涨有涨的逻辑，跌有跌的难堪

往深处分析，寒武纪市值震荡的核心，是高估值与现实增速、竞争格局之间的博弈，短期看，AI算力需求仍在，但高增长未来不排除难持续、份额被挤压、估值回归理性是大趋势；中长期取决于技术突破、生态构建、客户拓展能否兑现。

具体来看，之前的上涨确实有其内在逻辑。

其中最直接的支撑自然来自业绩的爆发式增长。2026年一季度，寒武纪实现营收28.85亿元，同比增长159%；净利润10.13亿元，同比增长185%；经营现金流首次转正至8.34亿元。客观而言，这份亮眼成绩单的背后，本质属于国产替代趋势下AI芯片自主可控需求的集中释放。

目前全球AI资本开支高达七八千亿美元，在外部技术封锁的大背景下，国内云厂商和智算中心纷纷加大对国产芯片的采购力度，寒武纪作为国内最早布局AI芯片的企业之一，自然成为最大的受益者之一。目前寒武纪的订单已经排至2027年，预付款项和合同负债持续走高，负债单季跳增近4亿元，这也侧面印证了下游客户备货意愿强烈，订单能见度较高，需求确定性极强。

另一个支撑因素是产品力的提升。据悉，寒武纪坚持全栈自研路线，已经形成了“云-边-端”全场景产品矩阵。其主力产品思元590采用7nm工艺，INT8峰值算力达到512TOPS，功效堪比英伟达A10。2026年2月，寒武纪发布了新一代HNLPU架构，进一步让市场看到了寒武纪追赶国际先进水平的希望。

然而，当情绪退潮，一些问题都会被无限放大。首当其冲的是估值是否真的过高？

去年8月经济学家马光远就曾说，彼时寒武纪的市盈率超过了4000倍，英伟达的市盈率只有57倍。即使是当下，寒武纪当前动态市盈率达260多倍，远高于半导体行业约60倍的平均水平。以2025年实际净利润20.59亿元计算，其市值与净利润比率超过400倍。

看高看低本身是市场行为，倒也无可厚非，外界真正担心的是，企业是否有足够的能力支撑这些高估值。

按业绩增速看，寒武纪的增速有放缓的情况。去年第四季度寒武纪营收18.9亿元，环比增长9.4%，净利润4.55亿元，环比下滑19.8%。这是寒武纪连续第二个季度出现净利润环比下滑，2025年第三季度，寒武纪净利润环比下降了17%。某些意义上可以表明，在经历了快速的业绩增长后，寒武纪也得经历波动调整。

再把视线拉回到市场层面，华为昇腾的强势崛起，正在彻底改变中国AI芯片市场的格局。根据IDC及多家投行的数据，华为2026年将占据中国AI芯片市场50%份额，美国芯片公司AMD预计以12%的份额排名第二，寒武纪（Cambricon）可能位列第三。

更重要的是，华为拥有完整的软硬件生态和强大的客户资源，头部企业可能已追加大额订单。

与此同时，英伟达也没有坐以待毙。

虽然受到出口管制的限制，但英伟达仍然占据着全球高端AI芯片市场大部分份额，英伟达还通过降价和推出特供版芯片等方式，进一步巩固自己的市场地位。在高端训练芯片领域，寒武纪与英伟达之间的差距仍然不小。

通俗点来说，任何行业的发展都遵循着同样的规律：前期是从概念爆发到泡沫膨胀，再到泡沫破裂后的价值回归。寒武纪的上涨，可以说是时代赋予的机遇，亦有自身的实力；而它的下跌，则是市场规律在内的多方面因素的体现。

国产芯片，

会走向移动互联网时代

大鱼吃小鱼吗？

目前，中国AI芯片行业正站在一个关键的十字路口。随着市场逐渐成熟，竞争将变得更加残酷，行业洗牌不可避免。未来的国产芯片市场，很可能会重演移动互联网时代大鱼吃小鱼的故事。

第一个风险是国内厂商的内卷加剧。除了华为昇腾和寒武纪，沐曦、壁仞、燧原、平头哥、海光、昆仑芯等数十家企业都在AI芯片赛道上厮杀。

这些企业各有优势：有人在通用计算领域实力雄厚，平头哥背靠阿里云拥有稳定的客户资源，细分领域玩家则在高端训练芯片上发力。为了争夺市场份额，价格战已经悄然打响，这将进一步压缩整个行业的利润空间。

第二个风险是业务结构单一。

以寒武纪为例，营收几乎全部来自云端AI芯片，2024年云端产品线收入占比高达99.3%，2025年，云端产品线营收占比高达99.7%。这种高度依赖单一业务的模式，抗风险能力有待加强。不仅如此，近三年，公司前五大客户的销售金额合计占营业收入比例分别为92.36%、94.63%和88.66%，客户集中度较高。

一旦后续算力市场降温、行业需求回落，寒武纪将面临巨大的业绩压力。相比之下，华为不仅有昇腾AI芯片，还有鲲鹏服务器芯片、麒麟手机芯片等多条产品线，业务结构更加均衡。

生态构建的进展程度也是潜在风险。众所周知，英伟达的成功，不仅仅是因为芯片性能强大，更因为它建立了一个覆盖数百万开发者的CUDA生态系统。这个生态系统经过十几年的积累，已经形成了强大的网络效应，新进入者很难撼动。

寒武纪虽然也在努力构建自己的NeuWare软件栈，但与CUDA相比还有很大差距。开发者生态的薄弱，将直接影响寒武纪产品的市场接受度和客户黏性。

寒武纪面前的两条路：要么学英伟达，加快构建自己的软件生态，形成软硬件协同的壁垒；要么像华为一样，拥有全栈技术能力和完整的产业链布局，甚至凭借自身实力，去重新定义半导体行业的游戏规则。

但这两条路都不好走。构建生态需要投入巨额的资金和时间，而全栈技术能力则需要长期的技术积累和人才储备。

回望移动互联网时代，曾经涌现出上百家手机厂商，但最终活下来的只有华为、小米、OPPO、vivo等少数几家。AI芯片行业也将遵循同样的规律。

所以，陆玖商业评论提前判断，未来3-5年，中国AI芯片市场将经历一场惨烈的洗牌，不排除有的企业都会被淘汰出局。先倒下的会是那些没有核心技术、只靠概念炒作的公司，然后是那些技术路线错误、资金链断裂的公司。最终，市场可能只会剩下2-3家真正有实力的企业。

当然，资本市场的理性回归，对整个行业来说未必是坏事。它会挤掉泡沫，让资金流向那些真正有技术、有产品、有盈利能力的企业。寒武纪作为国产AI芯片的先行者，拥有深厚的技术积累和先发优势，可它面临的挑战与受到的疑虑，似乎一点也不小。

注：文/李白，文章来源：陆玖商业评论(公众号ID：liujiucaijing69)，本文为作者独立观点，不代表亿邦动力立场。

【版权提示】亿邦动力倡导尊重与保护知识产权。未经许可，任何人不得复制、转载、或以其他方式使用本网站的内容。如发现本站文章存在版权问题，烦请提供版权疑问、身份证明、版权证明、联系方式等发邮件至run@ebrun.com，我们将及时沟通与处理。

已点赞