高通选择代工伙伴的纠结之心

手机芯片大厂高通(Qualcomm)CEO Paul Jacobs 近日表示，他们不排除拥有一家芯片制造厂的可能性。尽管高通是排名全球第五大芯片供货商，手头还有60亿美元左右的现金，但笔者并不认为Jacobs的意思是高通将在短时间内脱离无晶圆厂芯片公司的行列；取得所有权的手段有很多种，而且几乎都与策略联盟或是寻求共同点有关。 果然，近日据台湾中经社旗下《经济产业新闻》的报道称，高通公司已经和三星和联电（UMC）签订合作协议，后两者将会为高通代工供应28nm制程的工艺芯片，以解决高通公司骁龙Snapdragon S4系列处理器供应短缺的状况。 业界普遍认为，高通此举将缓解台积电28nm制程工艺产能不足而带来的相应芯片缺货现象。目前高通绝大多数产品均由台积电代工生产。 报道中称，预计总部位于台湾新竹的联电将在2012年第四季度开始供应28nm制程工艺的Snapdragon S4处理器以及3G/4G基带芯片。 根据估算，联电供应给高通的产能在3000-5000片晶圆/每月，约为台积电产能的20-33%。据称高通骁龙S4以及28nm制程基带产品已经使用联电28nm CMOS工艺流片成功并通过验证。而关于高通和三星的合作，国外Slashgear网站报道，高通把28nm的部分订单交给了三星。三星并未透露这笔订单的具体情况。。 高通在想些什么？ 作为业界有名的高富帅，高通要建立自己的晶圆厂应该不难，为什么要跑到竞争对手的工厂中送钱呢？在选择工厂时，他们又看重哪几点呢？我们来揣测一下高通的想法。 要建立晶圆厂不但前置时间很长、费用也很高──目前最先进制程晶圆厂兴建成本约在50~100亿美元──高通会跨足芯片制造几乎是不可能的事情。高通也可以 去收购或是取得某家现有晶圆厂的部分股权，来生产其处理器；部分获利状况不佳的晶圆厂也许会对这样的交易有兴趣，但是几乎大部分具备28nm、20nm制程能力的芯片制造厂都很赚钱。 但虽然先进制程晶圆厂的资本支出负担庞大，高通确实有很多可用现金，也有意改变芯片缺货、无法满足客户需求的现状。因此还有一种可能性较高的方法是，高通会被要求先投入一笔现金，以保留某家现有先进制程晶圆厂的部分产能；这种一方面是投资、一方面 是预付款项的交易在过去就已经存在，不过通常都无法长久维持。 Jacobs表示，28nm制程芯片供应情况已经逐渐改善，高通预期在2012年底就能满足客户的需求；所以，如果他的意思是高通正在寻找新的晶圆代工伙伴，谁会是最有可能的对象？首先，Jacobs真正会有兴趣合作的，应该只有那些具备优良先进制程技术能力的厂商，就像高通长期伙伴台积电。 另一家也来自台湾的晶圆代工厂联电(UMC) 几乎不太需要考虑，虽然联电会很需要高通的金援，但是并不清楚高通的投资值不值得。英特尔目前也拥有先进制程技术，但对Jacobs来说很困扰的一点是， 英特尔会是想趁高通28nm应用处理器供应不足时见缝插针的一个竞争对手；剩下的还有同是竞争对手的三星，还有技术成熟的纯晶圆代工业者 GlobalFoundries。 三星正积极扩张其晶圆代工业务，并已经在为苹果(Apple)的iPhone与iPad用 芯片代工；对三星来说，何不花点钱扩充生产线、并为高通保留部分产能？高通也许会是三星在苹果之外的另一个助益良多的替代/补充客户，因为据说苹果可能会考虑将部分芯片代工订单转往台积电，而且业界消息指出，Jacobs近日曾拜访三星讨论半导体供应问题。 再来还有 GlobalFoundries，该公司在美国纽约州有一座全新的晶圆厂；对GlobalFoundries来说，越快进入量产越好，而且该公司在晶圆代工产业界算是新秀，应该会偏好全心服务少数客户。笔者可以预见，高通等几家少数业者会愿意花钱取得GlobalFoundries纽约州新晶圆厂的第一批保证产能。 除非Globalfoundries 的幕后金主──ATIC──政策出现大转弯，笔者不认为该公司会把纽约州晶圆厂卖给高通，或是由高通所主导的无晶圆厂芯片供货商联盟之类；不过 Globalfoundries有可能会向这类客户取得一笔预付款──这是延续其在NAND闪存领域的习惯──再将该笔款项投资在纽约州晶圆厂的产 能扩充，甚至在阿布达比兴建另一座晶圆厂。 高通最后没有选择执行力有点问题，但不存在竞争关系的GlobalFoundries，而是带点竞争色彩，但执行力优良、技术能力紧追英特尔之后的三星，会不会养虎为患呢？ 编译：Judith Cheng 本文授权编译自EE Times，版权所有，谢绝转载 本文下一页：参考英文原文：On Qualcomm's manufacturing options，by Peter Clarke

相关阅读：
• 无晶圆厂经营模式在14nm的鸡血下仍将坚挺
• 450mm，欧洲芯片制造业的十字路口
• SEMI: 2013年晶圆厂设备支出将增长17%，达 463亿美元新高N9yesmc

{pagination} On Qualcomm's manufacturing options Peter Clarke So Paul Jacobs CEO of Qualcomm Inc. (San Diego, Calif.) doesn't rule out owning the means of chip production? (see Qualcomm wafer fab not ruled out). Although Qualcomm, as fifth largest chip company, has about $6 billion of cash on hand, I don't think Jacob's reported sentiment means that Qualcomm is going to leave the ranks of the fabless any time soon. There are many ways to finesse ownership issues and almost all of them will be about partnership and finding common ground. So for me this disclosure is more about the likes of Globalfoundries, Samsung and Intel, versus Taiwan Semiconductor Manufacturing Co. Ltd. and playing one of against another than it is about the creation of the Qualcomm Semiconductor Manufacturing Company. Such is the long lead-time and high expense of bring up wafer fabs – typically between $5 billion and $10 billion at today's leading-edge – it is almost impossible to think Qualcomm would start from scratch on chip manufacturing. An alternative would be for Qualcomm to buy all or a part-share in an existing wafer fab that can produce its processors. Some chip manufacturers are struggling for profitability and might be willing to cut a deal, although most of the chip companies able to manufacture at 28-nm and 20-nm are in a more profitable situation. Nonetheless at the leading-edge capital expenditure is ferocious and Qualcomm has been piling up the cash and has interest in changing a situation that has left it short of chips and disappointing customers. Another more likely possibility is that Qualcomm would be asked to put a significant amount of cash down to reserve a proportion of the output of an existing leading-edge wafer fab's output. Such part equity, part pre-payment deals have existed in the past although the circumstances that have driven their creation have not always been long-lived. And Jacob's himself has said that the 28-nm chip supply situation is improving and Qualcomm would hope to be meeting customers' requirements by the end of 2012. So if Jacob's musings are about partnership who is he thinking about partnering with? Samsung or Globalfoundries First of all Jacob's can only be really interested in working with companies that are at least as good in terms of manufacturing process technology as its incumbent long-term supplier TSMC and who also have a good manufacturing technology roadmap. That almost but not quite takes United Microelectronics Corp. (Hsinchu, Taiwan) out of contention. A cash injection from Qualcomm might be what UMC needs but it is not clear that is where Qualcomm would be advised to send its "big checks." Intel has an edge in manufacturing process technology right now, which must be worrying for Jacobs, but it is also a competitor seeking to capitalize on Qualcomm's 28-nm application processor supply problems. This leaves Samsung also a competitor, but one with a more mature and distinct foundry operation, and GlobalFoundries. Samsung is eager to expand its foundry operation, which is already doing essential work for Apple's iPhone and iPad products. Why not take a chunk of cash to convert additional manufacturing lines and reserve the output for Qualcomm? For Samsun Qualcomm might also make a useful replacement or complement for Apple which may be thinking of switching at least some of its production to TSMC. That Jacobs has, reportedly, visited Samsung recently for discussions that included semiconductor supply adds weight to this possibility. And then there is GlobalFoundries Inc. with its brand new wafer fab in Malta, Saratoga County, in upstate New York. The faster the ramp the better for GlobalFoundries and as a relative beginner in the foundry business it probably prefers supporting relative few customers wholeheartedly. I could see a deal in which Qualcomm and a few others write "big checks" to get a guaranteed first call on the output of the New York wafer fab. Unless the Abu Dhabi government and the sovereign wealth fund – the Advanced Technology Investment Company – that controls Globalfoundries suffers a complete U-turn in policy I don't see them selling off the New York wafer fab to Qualcomm or to a consortium of fabless chip companies assembled around Qualcomm. But it could take some up-front money from such a set of customers – as has become customary in the NAND flash memory business – and then re-apply that money to expansion in New York, and building a next wafer fab in Abu Dhabi. And whether that wafer fab would have a G or a Q on the outside of the building is a little way down the road. The choice comes down to the non-competitive pure-play foundry, GlobalFoundries, against whom there is a question mark on execution; and the competitive Samsung, a company which usually executes well and which is closer on the heels of Intel.