10/28/2016

高通收購恩智浦

高通 (QCOM-US) 斥 470 億美元併購恩智浦 (NXP),研究機構集邦 TrendForce 旗下拓墣產業研究所表示,近年全球半導體產業掀起併購潮,高通將使全球半導體產業競爭加劇,不過整併恩智浦時,包含重疊的產品,以及供應鏈管理,還有恩智浦的產品周期較長,有別於高通擅長的手機晶片 1、2 年就進行汰換,內部管理仍有許多挑戰。 
拓墣指出,併購案就機會面而言,高通可以快速搶進車用、工業市場,提升與晶圓代工業者的議價能力,強化系統層級方案的完整性與中國大陸智慧型手機市場的影響力。在挑戰面,高通將遭遇供應鏈管理難度大幅提升與部份產品線市場區隔難以界定的課題。 
拓墣表示,高通身為晶片設計業者,與身為 IDM (整合元件製造)業者的恩智浦,其間最大差別在供應鏈管理;再者由於車用電子與安全 (Secure) 應用產品佔恩智浦一半以上營收,這些應用領域的產品生命週期極長,與高通擅長的智慧型手機產業 1、2 年就汰換的特性有極大不同,合併後供應鏈管理將是首要挑戰。 
此外,恩智浦與高通同時擁有音訊處理與藍牙的產品線,在車用處理器方面也出現重疊,加上高通在日前也推出嵌入式領域專用的處理器,未來如何作出產品區隔,避免影響其客戶關係,是另一重要課題。 
拓墣進一步指出,車用、工業或基礎建設等產業特性相對封閉,需花費更多的時間與資源打進其供應鏈,併購成立後,高通能在最短時間內取得這些市場的主要客群,在智慧型手機市場成長動能疲乏,ASP (平均銷售利潤) 也逐漸下滑,有助分散市場風險。再者,恩智浦擁有豐富的 MCU 與 MPU 產品線,與代工廠的伙伴關係也相當密切,高通買下恩智浦後也能提升與代工廠的議價能力。 
高通與恩智浦在部分產品線上也有很大的互補,恩智浦在行動支付與安全領域是領導廠商,解決方案可搭配高通處理器提升完整度,尤其是中國大陸行動支付市場方興未艾,安全性更是近期重要課題,併購成立有助於強化高通在中國大陸智慧型手機市場的影響力。 
拓墣表示,即使是重複的產品研發領域,如電動車,併購後仍可帶來綜效,近年高通在電動車無線充電領域投入相當多心力,恩智浦在電源管理與電動車領域也深耕已久,因此從系統層級角度,極為完整的解決方案與充裕的研發資源將加速高通在車用領域的發展。
蔡宗憲高通併恩智浦 搶奪規格制定權 聯發科壓力加劇鉅亨網2016/10/27
高通 (QCOM-US) 宣布,以每股 110 美元,總金額 470 億美元金額,收購恩智浦,溢價幅度約 11.5%,高通將藉此把領域觸角延伸到車用與更多新的領域,雙方將具很高的互補效益,對台灣 IC 設計公司來說,聯發科 (2454-TW) 在手機晶片上站穩位置,面對未來高通整合恩智浦後,可能在包含車聯網、物聯網相關晶片領域上,具有更多規格主導性,對聯發科布局將是一個很大的壓力,而對台灣代工、封測領域,則是訂單版圖的變化。 
高通宣布斥 470 億美元,收購恩智浦半導體,收購後後年營業額可望超越 300 億美元,而恩智浦過去專注在安全與近距離晶片產品,之前也收購飛思卡爾,藉此進入車用領域,高通併入恩智浦後,可能在車用、物聯網相關領域上有更多的規格制定權。 
由此角度來看,聯發科積極布局的物聯網、車聯網市場,前方恐又面臨高通壓陣,對未來布局將成為很大的壓力。 
而對台灣晶圓代工、封測廠而言,台積電 (2330-TW) 董事長張忠謀日前指出,若高通合併恩智浦,對半導體產業是好事情,且兩家都是台積電重要客戶,台積電樂觀其成,且台積電與這兩家客戶都有很好的關係,希望合併後對產業是產生好的影響。 
不過,若以合併的角度去看,高通與恩智浦兩家公司文化相當不同,一家是美國公司,一家是歐洲公司,因此合併後文化上的衝突,及雙方整併是否可以產生綜效,都需要時間磨合,如同張忠謀所說,至少需要 1 年時間。 
同樣對台灣封測廠而言,兩家都下單在台灣廠商居多,只是未來高通併了恩智浦,若後段封測也由高通主導,則原本吃恩智浦封測訂單的廠商,就有訂單流失的風險。

DON CLARK and  TIM HIGGINS, Qualcomm to Buy NXP Semiconductors for $39 Billion, Wall Street Journal, Oct. 27, 2016
Qualcomm derives most of its revenue from designing and selling chips, but it earns more than half of its profits from licensing its wireless patents to nearly all makers of mobile phones. It had reasons to diversify beyond its processors and wireless chips for smartphones, a business that has reached a plateau. 
The hurdles are several: NXP has about 44,000 employees, according to public filings, many more than the 30,000 Qualcomm disclosed last year in its most recent annual report. NXP also owns seven factories in five countries that turn silicon wafers into chips. 
In addition, analysts say, NXP has older factories that wouldn’t be easily adapted to make Qualcomm’s chips. The company inherited operations that began more than 60 years ago as part of the Dutch giant Philips NV and Motorola Inc., which spun off its chip business to create Freescale. 
Mr. Mollenkopf said he expects veteran NXP executives to help run the manufacturing operations and points to its successful integration of other operations. “We get a lot of expertise coming in,” he said. 
Qualcomm said it expects the purchase to add significantly to its adjusted earnings. The company projects generating $500 million of annualized run-rate cost synergies within two years after the transaction closes.... 
Qualcomm pioneered what the semiconductor industry calls the “fabless” model. That avoids the cost of operating and building advanced fabs, which can cost more than $10 billion each for facilities that can create the most advanced chips. 
Industry executives say running factories requires a different set of management skills than designing chips. Those include tracking the age and performance of manufacturing equipment, overseeing a supply chain of materials and managing production workers, who are represented by labor unions in some NXP locations.
Aaron Pressman, Here's Why the Qualcomm-NXP Deal Makes Sense, Fortune, OCTOBER 27, 2016.
The potential of the automotive and smart device markets “looks to me like the cellphone looked in 2000,” Mollenkopf told Fortune after the deal was announced. “These are two opportunities that are at the beginning and we think we can drive that forward.” 
But despite its big lead, NXP  NXPI 0.43%  was facing several challenges that combining with Qualcomm may help solve. Its smart devices chips were suitable for many current applications, but lacked strong wireless connectivity features that will become increasingly important in the future, especially with the growth of faster 5G networks. That’s a core strength of Qualcomm. NXP’s self-driving car effort also lagged in machine learning capabilities, another Qualcomm strength. 

沒有留言:

張貼留言