In recent years, China's progress in the chip field has been remarkable, but the gap compared to international advanced levels remains significant and cannot be ignored. Technologically, the mainstream chip manufacturing process in China is currently at 28 nanometers, while the most advanced international technology has reached 10 nanometers and even 7 nanometers, indicating a substantial technological gap. For instance, Yangtze Memory Technologies Co., Ltd., with a total investment of approximately 160 billion yuan, primarily produces the currently popular 3D NAND flash memory. It is expected to achieve a monthly production capacity of 300,000 wafers by 2020 and 1 million wafers by 2030. However, it is crucial to recognize that chip technology evolves rapidly. By the time our production capacity investments are in place, the technology may have already become outdated, potentially rendering the heavily invested production lines obsolete. On the other hand, the global wave of chip industry consolidation poses significant challenges for China. In the smartphone sector, Chinese companies like Huawei, Xiaomi, OPPO, and Vivo have all entered the global top ten. However, except for Huawei, other manufacturers heavily rely on foreign companies like Qualcomm for their chip supply chains. Any disruptions in the supply chain, such as mergers, acquisitions, or price hikes, would immediately impact Chinese manufacturers. While China can impose certain conditions through merger reviews to protect domestic industries, it is challenging to prevent the mergers and acquisitions themselves.

The chip industry requires high-intensity investment, persistence, and the ability to endure solitude. Samsung's rise in the chip industry is a testament to its perseverance. Today, Samsung is reaping massive profits from its chip business, but few remember that Samsung's chip division suffered losses for many years. Samsung's chip technology was also acquired through patents, and when it spent $500 million to purchase Sandisk's technology patents, many considered it a foolish move. However, Samsung has proven the wisdom of its decision. For China's chip industry to achieve a late-mover advantage, it will still take time. With the 5G era approaching, it is unrealistic for China's chip industry to completely eliminate its dependence on foreign technology by 2020. However, with financial and policy support, and the collective efforts of companies like Huawei, ZTE, and Tsinghua Unigroup, through a combination of independent innovation and capital operations, collaboration between industry, academia, and research, and a transformation from quantity to quality, China has the potential to rise strongly in the global chip industry within the next 10 to 15 years, completely shedding its reliance on foreign technology and becoming a global chip industry powerhouse.

The multi-party game triggered by Toshiba's sale of its chip business
Toshiba incurred massive bad debt losses due to its acquisition of Westinghouse Electric, resulting in losses of up to $10 billion. To avoid being delisted from the Tokyo Stock Exchange, Toshiba had to quickly sell its most profitable flash memory chip business to secure much-needed cash flow and improve its financial situation. Flash memory chips are essential components for smartphones and other mobile devices, and since 2016, there has been a supply shortage, leading to continuous price increases. Toshiba is the world's second-largest flash memory chip manufacturer, with a market share of about 17%, second only to Samsung Electronics of South Korea. If not for its dire financial situation, Toshiba would not have resorted to selling its chip business. Clearly, this business is a "golden goose," attracting many major companies to bid, including Foxconn, Western Digital, SK Hynix, and Apple.

Considering factors such as technology retention and industrial security, despite Foxconn's offer of $27 billion, Toshiba decided to prioritize a consortium led by Bain Capital (including several major Japanese financial institutions, SK Hynix, Apple, Dell, etc.). However, Western Digital, also interested in the deal, persistently obstructed Toshiba's transaction and demanded that Toshiba sell the chip business to them. The situation escalated to legal battles between the two parties. Eventually, after Toshiba offered favorable terms, the relationship between the two parties eased, and Western Digital agreed to let Toshiba sell the chip business to the Bain Capital consortium.

The transaction is now largely finalized, awaiting approval from regulatory bodies, including China. This incident illustrates that technology transactions are not always about the highest bidder. Foxconn's $27 billion offer was not as attractive as Bain Capital's $18 billion, not because Toshiba did not want the money, but because the Japanese government did not approve. There were comprehensive considerations regarding technology and industrial security. In fact, Japanese institutions contributed over 50% of the funding in the Bain Capital-led consortium, meaning that after the sale, control of Toshiba's chip business would remain in Japanese hands. This was the fundamental reason for Bain Capital's success. It also reflects that chip business is not just about companies but is closely tied to national high-tech industries and political factors. In the chip industry, money is not everything.

The fierce battle between Broadcom and Qualcomm over acquisition and counter-acquisition
The acquisition of Qualcomm, the world's fourth-largest chip giant, by Broadcom, the fifth-largest, was somewhat unexpected. The transaction, valued at over $140 billion, would be the largest merger and acquisition in the history of the chip industry if successful. Qualcomm was caught off guard by Broadcom's aggressive takeover bid but rejected the proposal, citing an undervaluation of Qualcomm's worth and regulatory uncertainties. Broadcom then nominated new board members for Qualcomm, further pressuring the company; Qualcomm rejected all of Broadcom's nominations. Subsequently, Broadcom increased its offer to $82 per share, but Qualcomm's board rejected it again, citing an undervaluation of the company. The back-and-forth between the two companies has been a dramatic spectacle of acquisition and counter-acquisition. The outcome remains uncertain, but Broadcom is clearly prepared and will not give up easily, potentially resorting to hostile takeovers with financial backing if necessary. Qualcomm's stance against the acquisition is firm, but how it will fend off the "barbarians at the gate" remains to be seen.

Regardless of whether this merger succeeds, its impact on the global chip industry and the broader technology sector will be profound. If successful, the combined entity of Broadcom and Qualcomm would become the world's third-largest chip company,仅次于三星和英特尔，在移动、汽车芯片、WiFi、物联网领域将成为首屈一指的霸主。在即将到来的5G时代，联合体的地位和话语权还将进一步提升，对产业链的制衡作用将进一步强化。如果不成功，这也是足以载入历史的大事件，与近几年芯片产业大整合的趋势相一致，对整机产业的警示和震慑作用不可低估。

高通与最大客户苹果彻底闹翻
高通与苹果曾经有过良好的合作，苹果大量采购高通芯片，高通返还部分专利授权费给苹果。双方的交恶源于2017年初，苹果在部分产品中采用高通死敌英特尔的芯片，同时在美国国际贸易委员会发起的对高通垄断调查案中，提交了不利于高通的证词。高通遂以苹果违约为由，拒绝返还10亿美元的专利授权费给苹果。苹果一怒之下将高通告上了法庭，要求索赔。高通反诉苹果专利侵权，要求在美国和中国禁售苹果产品，双方互不相让，发誓要拼个鱼死网破。

与苹果闹翻导致高通损失惨重。除每年失去约20亿美元的利润之外，苹果供应商也不再缴纳专利授权费。高通专利授权、芯片排他性采购是绑在一起的一种商业模式，这种商业模式在全世界广受诟病，导致中国、韩国、台湾地区监管机构相继对高通开出天价的罚单。但高通专利授权商业模式没有被否定，苹果质疑的恰恰是高通商业模式，指责高通双重收费。这触碰了高通的底线，有可能颠覆高通长期以来生存发展的基础，因此高通才不惜一切代价也要与苹果一争高下。

高通在3G/4G/5G领域积累了海量的移动技术专利，同时在基带芯片领域占据超过70%的市场份额。高通很清楚，苹果再强势，也绕不过高通，也要用它的产品。因此高通才敢跟苹果叫板。

苹果目前是全球最高市值的科技公司，尽管自身也有芯片业务，但由于手机、电脑等产品出货量太大，在芯片业务上仍然依赖于高通、三星等竞争对手。这也从一个侧面表明了芯片对整机厂商的重要性。

全球芯片业整合风起云涌
东芝出售闪存、博通收购高通，其实是全球范围内兴起的芯片业整合大潮中的两朵大浪花。近几年来，为抢占未来科技制高点，全球芯片业整合风起云涌。根据国际半导体产业协会公布的报告显示，2015年，全球芯片行业的并购交易额超过600亿美元，2016年和2017年可能分别为1160亿美元和930亿美元。该协会表示，2016年似乎是并购狂潮的高峰期。这些兼并和收购主要是在成熟市场上增加规模和竞争力。

2016年公布的并购交易共60多宗，49宗在当年已并购结束。其中三宗交易占年度并购交易总额75%以上，包括安华高以370亿美元并购博通（目前的博通为安华高并购之后沿用的原名）；软银320亿美元收购半导体知识产权提供商ARM公司；西部数据以190亿美元收购Sandisk。

2017年全球半导体行业有12项交易将会完成，价值超过930亿美元。2017年的最大并购交易预计为高通和恩智浦半导体之间的交易，价值470亿美元，也是高通公司历史上最大的并购交易；价值第二高的交易是亚德诺和凌力尔特之间148亿美元的交易。仅这两笔交易就占了2017年全球交易总量的66%。不过，如果博通和高通的交易达成，2017年的并购交易规模将远远超过该协会的预测。

国际半导体产业协会还预计，接下来十年，半导体产业很有可能从水平整合进入到上下游垂直整合阶段。横向变纵向，厂商综合实力越来越强，产业集中度越来越高，寡头垄断的格局可能得到进一步强化。

5G时代芯片价值更加凸显
与整合大潮相伴相生的，是最近两年芯片价格的不断上涨，最典型的是内存芯片、闪存芯片。以内存条为例，从2016年第二季度开始，连续上涨超过一年，价格几乎提高了两倍。内存和闪存芯片价格的上涨，推动了智能手机平均价格在2017年上升了30%，台式机价格也在被动上涨。这无疑提高了整机厂商的成本，增大了产品销售风险。在整机厂商叫苦不迭的同时，三星、海力士、西部数据、东芝等芯片厂商却赚得盆满钵满。三星依靠闪存芯片涨价创造利润新高，一举超越英特尔跃居全球第一大芯片厂商。

芯片价格上涨，究竟是供需矛盾引发，还是厂商集体主观推动？有没有人为哄抬价格的因素？国家发改委相关官员近日已表态，将对芯片价格异常上涨进行调查。过去几个月，不断有整机厂商向国家发改委反映内存行业情况。发改委近期已经开始关注产业的相关动态，不排除未来对内存芯片厂商进行调查，以确定是否存在合谋涨价的垄断行为的可能。

应该看到，即将到来的5G时代是万物互联的时代，人与人之间的连接创造了万亿级大市场，而物与物的连接，有厂商预测将达到1000亿个，比人与人之间的连接规模大十几倍。芯片作为移动设备的心脏，地位将更加突出，产业规模将成倍扩大。并购大潮的出现，就是巨头间为了抢占未来制高点而采取的行动。而芯片涨价潮的出现，更凸显了其作为产业链上游的主导权和话语权。

中国芯片产业进步巨大但差距仍存
全球范围内掀起的整合并购大潮，必将对中国芯片产业发展产生深远影响。

早在几年前，我国通信网、互联网、电子信息制造业总体规模已位居世界前列，但产业大而不强的矛盾很突出，最大的软肋就是“缺芯少魂”。其中的“芯”指的就是芯片。连续多年，我国芯片的进口额超过石油，成为第一大宗进口商品，每年花费的总金额超过2000亿美元，折合人民币超过万亿元。据国家制造强国建设战略咨询委员会的估算，2015年中国芯片市场规模占全球的三分之一，但95%以上的产品供给都来自外资企业。

芯片虽小但它是战略性、基础性和先导性产业，是发展数字经济的重要支撑，在信息技术领域的核心地位十分突出，可以说是产业领域的“国之重器”。为加快振兴我国芯片产业，2014年6月，国务院发布新的纲领性文件《国家集成电路产业发展推进纲要》，提出了在较短时间内实现我国芯片产业跨越式发展的战略目标。紧接着由国开金融、中国烟草、中国移动等15家企业共同投资的“大基金”成立，主要为芯片产业链中的设计、封测和晶圆制造等关键环节项目提供资金支持。

“大基金”初期计划规模1200亿元，实际募集资金接近1400亿元。同时，各级地方政府成立的集成电路发展基金总规模超过3000亿元。近期有报道称，“大基金”二期募集资金规模将超过2000亿元。统计数据显示，未来10年，预计我国在集成电路领域新增投资总规模将超过10000亿元。

“大基金”成立之后，先后大手笔投资了一批国内芯片领域的龙头企业，包括紫光、中芯国际、中兴通讯、长电科技等。截至2017年年底，国家集成电路产业投资基金已投资额超过700亿元，其中约60％的资金投向半导体制造领域。

在政策和资金双重驱动下，我国芯片产业发展步伐明显加快。根据中国半导体行业协会的统计，2016年中国集成电路产业销售额达到4335.5亿元，同比增长20.1%。其中，位居产业链高端的芯片设计业继续保持高速增长，销售额为1644.3亿元，同比增长24.1%。

2017年中国在芯片产业领域的标志性成就包括华为海思发布了全球首款10纳米技术的AI芯片；国产第三代北斗芯片实现亚米级的定位精度和芯片级安全加密；装备了国产芯片的超级计算机“神威·太湖之光”荣获世界超算领域的三连冠；紫光和海思跻身全球前十大芯片设计企业行列，在全球芯片设计前50强中，中国企业占据了11席；华为也顺利地在高端机型中使用大量海思麒麟芯片，不再受制于人。这些成就，彰显了我国在芯片领域奋起直追的态势。

如果说华为中兴靠自主创新实现产业突破，那么以紫光为代表的企业成功靠的是资本运作。作为中国集成电路产业的“种子选手”，紫光集团2015年2月获得了国家集成电路产业投资基金和国家开发银行总计300亿元投资，2016年3月再获1500亿元投融资支持，助力紫光这几年充分运用资本杆杠，在资产并购上频频出击，引发全球科技界高度关注。

但是我们也注意到，“不差钱”的紫光，似乎总是“雷声大雨点小”，到目前为止，除了对展讯和锐迪科的收购获得成功之外（两家都是中国公司），包括对美光、西部数据、台湾力成等大陆之外芯片公司的入股、并购，几乎无一例外遭遇监管部门的阻拦，最后归于失败。发达国家对中国资本在芯片领域的并购高度警惕，认为会威胁到他们的高科技产业安全，因此一律采取封杀政策。正如习总书记所指出的，真正的核心技术是花钱买不来的。

过去几年中国在芯片领域的进步有目共睹，但与国际先进水平相比差距依然不容忽视。从技术上看，目前国内芯片主流制程技术是28纳米，而国际最先进的技术是10纳米乃至7纳米，中国在技术上差距还是很大。例如，总投资高达约1600亿元的“长江存储”集团的主要产品为当前最热门的3D闪存，预计到2020年形成月产能30万片的生产规模，到2030年形成每月100万片的产能。必须看到，芯片技术迭代发展迅速，等到我们的产能投资到位，技术上可能又大幅落伍，到那时巨额投资的生产线很可能成为落后的产能。

另一方面，全球芯片产业整合大潮，对中国相当不利。从智能手机看，目前中国的华为、小米、OPPO、vivo均跻身全球前十，但除华为外，其他整机企业的芯片供应链高度依赖高通等国外公司。一旦供应链企业出现意外（例如并购、产品涨价），首先受到冲击的就是中国厂商。当然中国可以通过并购审查提出一些有利于国内产业发展的条件，但很难阻止并购行为本身。

芯片产业需要高强度投资，需要持之以恒，需要耐得住寂寞。三星在芯片领域后来居上，靠的就是坚忍不拔。今日三星凭借芯片赚翻了，但有谁注意到三星芯片业务过去多年曾长期亏损？三星芯片技术也是花钱买来的专利，当初5亿美元购买Sandisk技术专利时，多数人认为三星买贵了，是傻瓜行为。但三星用事实证明其决策是正确的。

中国的芯片产业要实现后发超越，仍然需要一定的时间。5G时代即将到来，到2020年中国芯片产业要完全摆脱对国外的依赖还不现实。但在资金、政策支持下，在华为、中兴、紫光等企业的共同努力下，自主创新与资本运作多措并举，产学研用携手，从量变到质变，再经过10到15年时间，中国完全有可能在全球芯片领域强势崛起，彻底摆脱对外依赖，真正成为全球芯片产业强国。