GLOBALFOUNDRIES Appoints Michael Hogan as SVP and General Manager to Support New Market Engagement Strategy

New business segments and leadership appointments further position the company for long-term growth and value creation

Santa Clara, Calif., September 24, 2019 – Reinforcing its commitment to deliver specialized solutions to targeted market segments, GLOBALFOUNDRIES (GF) announced today, at its Global Technology Conference (GTC) in Silicon Valley, the appointment of Michael Hogan as senior vice president and general manager of the company’s newly established automotive, industrial and multi-market (AIM) strategic business unit (SBU). Hogan will be responsible for driving market strategy, defining GF’s roadmap for differentiated features and the resultant global expansion for the AIM SBU.

Hogan is a 30-year semiconductor technology veteran and has successfully led premiere companies, including Cypress Semiconductor and Broadcom, in executive-level general manager and senior vice president roles. Most recently, Hogan was the senior vice president and general manager of the IoT, Compute & Wireless business unit at Cypress Semiconductor, where he shaped strategy for the company’s largest and fastest-growing business.

“By aligning our leadership structure around the client experience, our team’s diverse talents and market insights will be leveraged to transform our go-to-market strategy and deliver specialized application solutions that provide real value to clients,” said Thomas Caulfield, CEO of GF. “The addition of Mike Hogan comes at a time when GF is positioned for strong growth and requires seasoned leaders to further enhance and scale our capabilities. Mike’s vast knowledge in the semiconductor space especially in automotive and wireless connectivity, as well as his proven track record of success, will bring great value that will drive growth today and into the future.”

In conjunction with Hogan’s appointment, GF has established dedicated strategic business units around three core market groups, automotive, industrial and multi-market (AIM); mobile and wireless infrastructure (MWI); and computing and wired infrastructure (CWI) to grow market share in the large and growing $47 billion addressable foundry market for 12nm technologies and above.  Hogan will work in close collaboration with Bami Bastani, who has been appointed senior vice president and general manager of the MWI SBU; and with Mike Mendicino, who has been appointed interim vice president for the company’s CWI SBU.

These new strategic business units and leadership appointments position the company for greater scalability and growth, building upon the strategy that began in 2018 with the company’s pivot and continuing into this year with its transformational transactions.

About GF

GLOBALFOUNDRIES (GF) is the world’s leading specialty foundry. We deliver differentiated feature-rich solutions that enable our clients to develop innovative products for high-growth market segments. GF provides a broad range of platforms and features with a unique mix of design, development and fabrication services. With an at-scale manufacturing footprint spanning the U.S., Europe and Asia, GF has the flexibility and agility to meet the dynamic needs of clients across the globe. GF is owned by Mubadala Investment Company. For more information, visit globalfoundries.com.

Contact:

Erica McGill
GLOBALFOUNDRIES
(518) 795-5240
[email protected]

格芯任命Michael Hogan为高级副总裁兼总经理,以支持新的市场深耕策略

新的业务部门和领导层的任命将进一步推动公司的长期增长和价值创造

加利福尼亚州圣克拉拉,2019年9月24日 — 为加强其向目标市场提供专门解决方案的承诺,格芯今日在硅谷举办的2019格芯全球技术大会(GTC)上宣布,任命Michael Hogan为格芯新成立的汽车、工业和多市场(AIM)战略业务部门的高级副总裁兼总经理。Hogan将负责推动格芯市场战略,规划格芯的差异化功能路线图,以及汽车、工业和多市场(AIM)战略业务部门的全球扩张。

格芯汽车、工业和多重市场战略业务部门高级副总裁兼总经理Michael Hogan

Hogan拥有30年的半导体技术经验,曾成功领导过包括赛普拉斯半导体(Cypress Semiconductor)和博通(Broadcom)在内的业内领先公司。加入格芯之前,Hogan是赛普拉斯半导体物联网、计算和无线业务部门的高级副总裁兼总经理,他在该部门为公司最大、增长最快的业务制定了战略。

“通过围绕客户体验来调整领导结构,格芯的多样化人才和市场洞察力将被用来转变我们的市场战略,并为客户提供差异化、功能丰富的解决方案。”格芯首席执行官汤姆·嘉菲尔德(Thomas Coulfield)表示,“Michael Hogan的加入正值格芯位于强劲增长之际,我们需要经验丰富的领导者进一步提升和扩展我们的能力。Michael在半导体领域,尤其是在汽车和无线连接领域的丰富知识以及成功经验,将带来巨大的价值,推动格芯今天和未来的发展。”

在Hogan就任的同时,格芯围绕三大核心市场(汽车、工业及多市场(AIM)、移动与无线基础设施(MWI)以及计算与有线基础设施(CWI))成立了专门的战略业务部门,为在规模庞大且不断增长的12nm及以上芯片的代工市场中扩大市场份额。Hogan将与被任命为格芯移动与无线基础设施战略业务部门高级副总裁兼总经理的Bami Bastani以及被任命为格芯计算与有线基础设施战略业务部门代理副总裁的Mike Mendicino进行密切合作。

格芯新成立的战略业务部门及新领导层的任命,是基于自2018年开始的战略转型,并延续至今以实现更大的可扩展性和增长能力。

关于格芯:

格芯是全球领先的特殊工艺半导体代工厂,提供差异化、功能丰富的解决方案,赋能我们的客户为高增长的市场领域开发创新产品。格芯拥有广泛的工艺平台及特性,并提供独特的融合设计、开发和生产为一体的服务。格芯拥有遍布美洲、亚洲和欧洲的规模生产足迹,以其灵活性与应变力满足全球客户的动态需求。格芯为阿布扎比穆巴达拉投资公司(Mubadala Investment Company)所有。欲了解更多信息,请访问 https://www.globalfoundries.com/cn

媒体垂询:

杨颖(Jessie Yang)
(021) 8029 6826
[email protected]

邢芳洁(Jay Xing)
86 18801624170
[email protected]

战略转变增强IP合作伙伴关系

  • 作者: Dave Lammers

当记者对半导体公司进行比较时,我们通常会深入探究栅极长度、掩膜层、SRAM单元尺寸以及其他一些面向硬件的指标。只有在经历过一段职业生涯后,我才开始认识到,IP和其他形式的设计支持对晶圆厂和集成器件制造商(IDM)取得成功也同样重要。

当格芯在2018年8月底宣布实施“战略转型”时,公众的很多注意力再次转向晶体管的角色,以及资源如何重新部署到7nm逻辑芯片之外的技术领域。在摩尔定律预测的高增长速度逐渐减缓的时代,格芯将这些资源分散到格芯提供的18种不同技术(及其衍生品)上,此举得到了很多人的理解。

必须更多强调一点,更新的IP在某种程度上是通过战略转型来实现的。

格芯生态系统合作伙伴关系副总裁Mark Ireland指出,12LP (FinFET)工艺就是重新部署IP资源的很好例子。在初始阶段,格芯的12LP工艺主要用于CPU、GPU和类似的高性能产品。现在,12LP正在进入一系列更广泛的市场,包括消费型产品、网络、5G无线、人工智能-机器学习(AI-ML)。这些应用通常需要不同的IP,特别是多协议SERDES、低功耗存储器、高速存储器接口。

“在消费型产品中,数字视频和机顶盒正在向FinFET迁移。消费型产品不是12LP节点的引领者,但现在却在向FinFET迁移。Ireland表示:“我们看到了更加广泛的市场和客户群体,这一点必须在我们注重的IP合作中体现出来。”

他表示,人工智能SoC也需要更多的IP,包括高速SERDES和低功耗存储器。

适用于5G基站的高速SERDES

同样,5G无线标准“扩大了引入一些将用于5G基站和其他用途的SERDES IP的需求。”他指出:“我们的客户需要这种类型的IP,才能在这些市场上取得成功。”他还指出,无线客户可以选择22FDX全耗尽式绝缘体上硅、12LP FinFET或其他工艺,这要取决于他们的应用需求。

格芯和Rambus宣布推出适用于22FDX工艺的28-Gbps和32-Gbps SERDES,就在设计自动化大会之前,格芯和Synopsys表示双方正在准备开发采用12LP工艺的25-Gbps SERDES。Ireland说:“这种IP具有更广泛的市场应用,对于5G基站至关重要。”

另外,格芯与Analog Bits近期达成协议,将Analog Bits的模拟和混合信号IP设计套件引入12LP技术,包括低功耗锁相环(PLL)和扩频时钟生成(SSCG),以及工艺、电压、温度(PVT)传感器IP等。

Ireland表示:“我们正在与更广泛的市场建立更深入的合作伙伴关系,从而满足他们对更多IP的需求。我们正在推动这一进程,这其中不缺少机会。格芯目前面临的挑战是及时获取最高质量的IP。”

每个芯片上的射频模块

格芯的客户解决方案副总裁Subi Kengeri表示,他们有更多IC设计团队正在使用FD-SOI或传统异构集成方法,开发复杂的设计来处理射频和混合信号,而并不简单依赖于粗放的扩展。对于复杂的射频和模拟SoC,Kengeri指出:“IP将成为实现SoC产品技术差异化的载体。设计人员要通过这种方式挖掘技术的差异化价值。因此,IP必须经过完全优化,具备最高的质量,这一点非常重要。”

格芯在射频技术领域有着出色的过往业绩,并且不断在射频技术领域投入巨资,这也是战略转变后策略的一部分。“通信现在变得前所未有的重要,每个芯片上都将有一个射频模块。射频非常复杂,整个行业掌握的这方面技能也比较有限。我们是射频领域的领跑者,并且在射频IP、设计服务和射频参考模块方面进行了投资,因而我们处于非常有利的地位,能够帮助客户加快产品上市、降低成本和减少风险。提到射频,就想到格芯。”

跟踪IP就绪性

晶圆厂IP和客户工程副总裁John Kent表示,一个芯片设计可能需要20个甚至更多的不同IP。Kent说道:“我们跟踪IP就绪情况,这个词的意思是当客户希望进行设计时,我们是否拥有了所有必需的IP。”就绪性指标是“我们是否能够为客户提供服务的一大关键指标。”Kent说,另外一个重要指标是一次性正确率,目的是确保IP的所有DC参数都是准确的。

他说:“在与新客户合作时,我们的亲身实践经验可以告诉我们:我们在哪些方面是世界一流的,在哪里方面尚待改进。我们作为一个团队面临的最大挑战是,在我们放弃7纳米工艺之后,利用团队在7纳米工艺或其他平台上积累的经验,重新平衡我们的资源。”

Kent表示,其他格芯技术平台吸引了更多关注,包括他们长期重视的PDK(产品开发套件)改进。Kent说:“过去十年内,我们在PDK方面的知识不断积累,我们学会了及时执行。通过这个过程,以及我们的这次战略转变,我们将首要PDK开发重点从FDX和FinFET转移到格芯为客户提供的其他18个系列产品上,从而将PDK资源重新部署到这些技术上。

在22FDX基础IP方面,格芯主要但不完全依赖于Invecas,该公司包括以前的IBM存储器IP团队成员。Kent谈到了Invecas:“他们是一个优秀的团队,提供非常出色的产品。”

Ireland说:“我们的22FDX基础IP来自于Invecas,最近我们还扩展了生态系统,包括来自Synopsys的汽车IP。我们的目的是与多家供应商展开合作。”与Synopsys的协议包括基础IP,以及面向各种汽车应用的模拟和接口IP,包括ADAS、动力总成、5G和雷达。

基础IP可能非常复杂

基础IP(即FIP)的复杂度从简单到中等。多个电压的通用型IO可能涉及多个不同的金属堆栈,其设计可能非常复杂。Kent说:“通常在我们发布库时,FIP内部包括数千个单独的库单元。”

存储器,包括静态RAM、ROM、闪存和更新的MRAM,也属于FIP的一部分,因为它们与I/O相似,都是设计的基础。但存储器IP非常复杂,存在复杂的信号传输问题,需要进行纠错。

所谓的复杂IP通常包括大量的模拟和混合信号内容。32-Gbps SERDES可能具备很多数字模式功能,还有复杂的混合信号,以便支持信号和功率参数。

格芯一直在与Everspin携手共同开发新的IP,支持基于22FDX和FinFET工艺的嵌入式MRAM。Kent表示,MRAM相对于闪存具有诸多优势,包括亚纳秒级的写入时间(而闪存的写入时间则长达数毫秒)和非常强大的防故障能力。Kent说:“我们正在开发新的IP来支持MRAM,具备能够与SRAM相媲美的性能。”

汽车应用是MRAM的主要目标。Kent说:“未来的汽车将采用大量传感器,所有部件必须安全运行。由于集成电路必须在汽车中工作更长时间,比如它应该超过计算机的使用寿命,因此我们正在考虑采用MRAM。”

关于作者

Dave Lammers
 

Dave Lammers是固态技术特约撰稿人,也是格芯的Foundry Files的特约博客作者。他于20世界80年代早期在美联社东京分社工作期间开始撰写关于半导体行业的文章,彼时该行业正经历快速发展。他于1985年加入E.E. Times,定居东京,在之后的14年内,足迹遍及日本、韩国和台湾。1998年,Dave与他的妻子Mieko以及4个孩子移居奥斯丁,为E.E Times开设德克萨斯办事处。Dave毕业于美国圣母大学,获得密苏里大学新闻学院新闻学硕士学位。

GF Plays a Role in Quantum Ecosystem

By: Dave Lammers

I first met Sorin Voinigescu in 1995, when – with a newly minted Ph.D. in hand — he was at the International Electron Devices Meeting (IEDM) presenting some of the early work on RF circuits crafted in CMOS technology.

Nearly 24 years later, Voinigescu is doing equally innovative pathway research in quantum computing, using the 22FDX® process from GLOBALFOUNDRIES (GF) to investigate how to integrate qubits with the RF control and readout circuits. And Voinigescu sees a type of Moore’s Law for quantum devices, in which scaled-down qubits and support circuits are able to operate at higher temperatures, perhaps obviating the need for the scarce helium consumed in today’s cryogenics.

Quantum devices today are largely Josephson Junction superconducting devices operating at milliKelvin temperatures, with wires connecting the qubits to the control and measurement electronics. Voinigescu’s lab at the University of Toronto is studying how to create semiconductor-type qubits that could be controlled with millimeter wave signals. The present-day superconducting qubits have quantum energy separation levels in the 5-10 GHz range. In order to operate the quantum gate, the microwave control signals need to be at that frequency, the 5-10 GHz range.

“All qubits, regardless of implementation, mimic a spin, and the control is performed with a signal that must resonate with the electron spin resonance frequency of that qubit,” Prof. Voinigescu explained. One way of thinking of it, he said, is that each quantum gate could require the equivalent of a 5G cellular signal, perhaps in the 60 GHz range. In fact, he was attracted to the field of quantum computing a few years ago, when he attended a session on quantum computing at IEDM and realized that his two decades of research in high-frequency circuits could play a role in the quantum computing field.

22FDX at 3.3 degrees Kelvin

There is what he calls a “trinity” in the search for higher-temperature quantum computing, where devices must be isolated from any heat or disturbances. The smaller the gate width of the transistor, the higher the frequency required to excite the qubit gate, and the higher temperature it can be operated at. Now, the 22FDX-based devices the Voinigescu lab is investigating have a gate width of 50 nm (the gate length is 18nm, and the channel thickness is 6-7nm). As the gate width is reduced, a somewhat higher temperature environment can be used for the qubits, control, and measurement circuits.

And the cool (excuse the pun) thing about the 22FDX process is something the Toronto lab and its partners recently discovered: at the extremely low temperatures required of quantum systems, the performance of the active and passive high-frequency devices actually improves.

The University of Toronto team, working with GF and industrial partners Lake Shore Cryotronics and Keysight Technologies, among others, reported at the 2019 RFIC conference, held in Boston in June,  how the 22FDX process was used to create monolithically integrated double quantum dots with readout transimpedance amplifiers (TIAs), with the output matched to 50 Ω.

More importantly for circuit design, the researchers found that the high-frequency performance of all the active and passive devices created in a production-type 22nm 22FDX technology improved at 3.3 degrees K, with no variance of the polysilicon resistors and improved quality factor of the MOM capacitors.

“What is unique to FD-SOI is that at low and high frequencies, the circuits are not affected by de-ionization, as bulk MOSFETs are known to be affected. Because of that we essentially get significantly better performance at low temperatures, as measured down to 2 degrees Kelvin. In fact, we see significant improvements down to 60-70 degrees K, and below that performance essentially remains flat,” he said. Transconductance, mobility, and fmax all improved, and that has important implications for space, satellites, and other low-temperature environments as well.

At low temperatures, threshold voltages increase for n-MOSFETS and decrease for p-MOSFETs, regardless of the technology. With FD-SOI, the back gate can be used to adjust the Vts to the optimal operation point. Circuits can be designed at room temperature, and then at low temperatures can be “validated,” tuning the Vt’s with back-gate biasing. Circuits that find a “sweet spot” at room temperature, Voinigescu said, can maintain that current density down to 2 degrees Kelvin.

Source: International Workshop on Cryogenic Electronics for Quantum Systems, Professor Sorin Voinigescu, University of Toronto, June 2019

Smaller Dimensions Help Raise Temps

Jamie Schaeffer, the product offering manager for the 22FDX and 12FDX FD-SOI platforms at GF, said the qubits are created in the six or seven-nanometer active layer, providing confinement for Coulomb and spin blockade devices that are, in a sense, boxed in by the buried oxide. “We have to get the spin layers to interact, and with more advanced dimensions we can get those closer. As we are going from 22 to 12 FDX, the smaller dimensions serve the goal of higher temperature quantum computing,” Schaeffer said.

Nigel Cave, a technologist who works in the CTO office at GF, said as semiconductor-type qubits are scaled to smaller dimensions, it may be possible to bring the quantum system’s operating temperature above 4 degrees Kelvin, from the 10-100 milliKelvin in today’s systems. This would enable the use of a standard helium cryostat versus a dilution cryostat, thus reducing costs and also allowing 1-2 watts of total power to be removed from the system. “The ability to remove more power potentially paves the way to co-integrate the Qubits and their control circuitry in the same FDX based device” Cave said.

Schaeffer said IBM, Google, Intel, Microsoft, and others have large quantum research programs underway. “In our case, we believe we can contribute something that is enabling for our partners who are doing meaningful work in the quantum sciences. We have a toolset that is manufacturable and leveraging our process integration capabilities is a way to get to lower costs.”


Source: International Workshop on Cryogenic Electronics for Quantum Systems, equal1.labs https://equal1.us/technology

Two Camps in Quantum Ecosystem

Ted Letavic, a vice president and senior fellow at GF, said from a ten-thousand-foot level, the quantum compute community can be divided into two camps: those who are pushing for ways to create thousands of qubits in order to increase the quantum compute power; and a camp that argues more attention needs to be paid to how to use the roughly 50-100 qubit systems that now exist in order to solve real-world problems.

“One faction says we need thousands of qubits, the other faction says we have 50-100 qubit systems now and don’t know what to do with them. One answer is to provide free access in consortia, and together we can best figure out how to use them, how to create economic value and advance our economy,” he said.

GF has “some key technologies that can help,” acting as a foundry for startups, universities, and others as they investigate different approaches. Letavic, along with Cave and John Pellerin, deputy CTO and vice president of worldwide R&D, provided input to the Department of Energy, which earlier this year put out a Request for Information regarding how best to organize the Quantum Information Science Centers (QISCs).

They argued that while current exploratory R&D is largely being done in non-standard university labs, GF could provide a process integration and early manufacturing effort for researchers, startups, and others participating in the QISCs. Working with foundries would ensure that “devices intended to unlock the promise of quantum systems can be fabricated in volume within existing manufacturing assets.”

Letavic pointed to the work being done with Prof. Voinigescu as one real-life example, where the FD-SOI devices proved advantageous for I/O at 4 degrees Kelvin, and hold promise as a source of qubit transistors confined in the very thin FD-SOI layer. The Toronto effort used wafer shuttles that were processed at GF’s Dresden, Germany fab.

GF also has a silicon germanium platform, as well as a silicon photonics capability, that could play a role in “unlocking the promise of quantum.”

“I do believe in quantum, but it is going to be additive to classical compute,” Letavic said. “The society that gets to a quantum compute infrastructure first will have a very large economic advantage over the rest. And whether you are in the camp of ‘let’s chase the maximum number of qubits,’ or the camp that says ‘let’s figure out how to best use quantum systems to the best of our ability,’ GF is playing in both.”

About Author

Dave Lammers is a contributing writer for Solid State Technology and a contributing blogger for GF’s Foundry Files. Dave started writing about the semiconductor industry while working at the Associated Press Tokyo bureau in the early 1980s, a time of rapid growth for the industry. He joined E.E. Times in 1985, covering Japan, Korea, and Taiwan for the next 14 years while based in Tokyo. In 1998 Dave, his wife Mieko, and their four children moved to Austin to set up a Texas bureau for E.E. Times. A graduate of the University of Notre Dame, Dave received a master’s in journalism at the University of Missouri School of Journalism.

 

GLOBALFOUNDRIES: Then and Now

By: Gary Dagastine

The first of a three-part series looking back at GF’s first 10 years, and ahead at the next decade and beyond.

As GLOBALFOUNDRIES celebrates its 10-year anniversary, the company finds itself at a key inflection point. Driven by financial imperatives and changing business opportunities, CEO Tom Caulfield has initiated a sweeping strategic transformation. The move is designed to better utilize the company’s resources and grow the return on investment by focusing on applications where GF’s diverse, differentiated technologies offer significant advantages.

To accomplish this it’s essential that GF employees share common goals, and a company-wide effort to facilitate that, called ONEGF, is taking place. But that isn’t an easy process for any company, and it’s harder still when you consider GF’s origins: It began as an spinoff of AMD’s in-house manufacturing operations in Dresden, Germany; then, GF acquired the Chartered Semiconductor foundry in Singapore; built a greenfield foundry in Malta, NY; and, as if that wasn’t enough, acquired IBM’s former in-house technology development group and chip manufacturing operations in New York and Vermont.

Given all of the change that has taken place, Foundry Files wanted to get the perspectives of long-time employees from various parts of the company on how GF got where it is today, to learn if there are lessons in that experience which can help with the challenges that lie ahead.

Read on to learn why the following GF employees say that a sense of shared purpose, a laser-like focus on the customer, and finding satisfaction in different types of technological innovations may well be the keys to success in the company’s next decade.

Learning by Doing in Dresden: From an IDM to a Foundry

“We can show plenty of scars,” said Jens Drews, Director of Communications & Government Relations for Fab 1 in Dresden, referring to the fab’s challenging transition from a dedicated manufacturing resource for AMD microprocessors to a foundry satisfying the wide-ranging demands of new customers, while at the same time ramping multiple new technologies.

Asked to describe the site’s journey over the last ten years, Jens thought for a moment and then came back with the opening line of Charles Dickens’ A Tale of Two Cities: “It was the best of times, it was the worst of times, … it was the season of light, it was the season of darkness; it was the spring of hope, it was the winter of despair,” adding that the fab’s momentum and growth potential clearly pointed towards the “best of times.”

Jens, the voice of GF’s Dresden site to employees and to the German and European media and governments, is in a position to know because he has been with the site for more 23 years and has witnessed the changes firsthand.

“We started out in Dresden with a straightforward goal – to compete with Intel in CPUs – and we were tremendously proud when we were able to do that,” he said. “But during the last few years when we were still a part of AMD, the chip industry went through a big paradigm shift away from the original IDM model towards a model that saw the rise of fabless and fab-light companies and foundries. For a while, AMD Dresden continued to be an island of stability in a sea of change, and we were going about our work as always, only concerned with one customer, one technology and one product family at a time. But it had become very clear by the early 2000s that changes in our industry would eventually catch up with Dresden – and probably sooner rather than later.”

“Then, when we became part of GF, we found ourselves literally overnight in a completely different ball game. It was like going from playing volleyball to playing rugby. You can imagine that it took time and quite a few hard knocks before we adjusted to the rules of the new game,” he recalls.

“Welcome to GLOBALFOUNDRIES Fab 1” was projected on the façade of the Dresden office building in March 2009.

For example, Jens mentioned the steep hiring ramp in 2009/10 that brought plenty of “new blood” from companies like Qimonda, Chartered Semiconductor and from the solar industry. They joined a fairly homogeneous team deeply steeped in the AMD culture. “Although this brought excitement and fresh outlooks, it still was a shock to the system because until then we had been a merry band of AMD brothers and sisters, so to speak, and we now had to cope with change everywhere you looked. For sure, we went through a period of growing pains,” he said.

But that was then, and now GF Dresden is a force to be reckoned with in the worldwide foundry industry. For example, GF’s innovative 22FDX® FD-SOI technology is ramping and is drawing more and more interest for some of the fastest-growing applications in the industry, and so are the 28, 40 and 55/65 nm platforms.

“Today, we have a long-term strategy in place that positions Dresden as a high-mix ‘More than Moore’ foundry fab with a focus on demanding markets such as automotive, security, 5G and AI. With that, we are now becoming part of key European industrial value chains such as automotive,” he said.  Jens noted that the company’s pivot in 2018 dovetailed with the Dresden site’s strategic move away from straightforward scaling towards feature-rich platforms for new markets besides computing and communications.

“We have never seen better alignment between corporate and site strategies, which allows us to stay fully focused on serving the diverse needs of new and old customers and their exciting markets.”

His conclusion: “The future of Dresden looks bright, a ‘season of light’ as Charles Dickens would say. We have successfully redefined the value we bring to our customers and their markets. Our growth potential is real and we have learned, sometimes the hard way, what is expected of us: Strong platforms that make the difference for our customers’ customers, continued innovation, and of course solid execution with focus on quality, cost and the bottom line.”

A Focus on the Customer

If building a sense of shared purpose was the imperative at Fab 1, a world away at GF’s Fab 7 in Singapore keeping customers happy and generating solid financial returns have been the main goals. That’s according to Peter Benyon, Vice President & General Manager of Fab 7. Peter, a 20-year employee split equally between GF and Chartered Semiconductor, was recently named Vice President and General Manager of Fab 8 in Malta, NY, effective July 1.

GF’s Singapore fabs offer a number of mature 200mm and 300mm processes, and soon also will offer GF’s 8SW technology for RF applications, which will be moving there from Fab 10 in East Fishkill, NY. Fab 10 is being sold to ON Semiconductor. (GF’s East Fishkill-based 45RFSOI and silicon photonics technologies will be moved to Malta.)

From the outset, employees in Singapore have had a strong customer focus and a “first-time-right” mentality in the fab that has led to both customer satisfaction and good margins for GF.

GF’s “Opening Day” in Singapore

“We’ve been a foundry from the beginning, and we’ve always had an intense focus on the customer because when it comes to mature technologies, the pricing and other deliverables offered by competing foundries are generally in the same ballpark. Therefore the question on the customer’s mind is always, “Why should I come to you?” Peter said.

“Our answer to that is to focus on their needs by giving priority to their work as required, offering flexibility our competitors usually don’t, and doing what we say we’ll do. As a result, we have built many strong, longstanding and mutually beneficial relationships,” he said.

Foundry Files asked how that mindset might be made to percolate throughout the entire company. “It would be wrong to try to accomplish that by force-fitting a culture onto other GF locations. That has been tried before and it didn’t work. But ensuring customer needs are a priority, always and everywhere, needs to be part of our DNA,” he said.

Peter will take up that challenge in his new role at Fab 8, where his customer focus and proven operational excellence will enhance GF’s ability to serve a rapidly expanding customer base, and will provide a competitive edge over and above the fab’s world-class technologies.

Taking Delight in New Ways to Innovate

“Innovation is what gets our people excited, but scaling isn’t the only kind of innovation. For people who want to do new and unique things, the opportunities are endless. Silicon photonics is an example. Today it’s a small market. However, tomorrow it may be a very large one, and we’re helping to define it,” said Neil Peruffo.

Neil is Vice President & General Manager of Fab 10 in East Fishkill, and has been with IBM and GF for more than 30 years in technology development, characterization and deployment roles.

GF acquires IBM Microelectronics fab in East Fishkill, NY in July 2015

IBM has had a long and illustrious history in semiconductors, with a list of notable accomplishments and individuals too long to mention here. But because the major focus of its semiconductor operations was to support IBM’s mainframe and server businesses, there was concern within the company’s chip unit over a “narrowing corridor of applicability,” as Neil puts it, for its most advanced and costly technology.

He said the concern turned into excitement when the announcement was made in 2015 that GF was acquiring IBM’s semiconductor operations, because it created a sense that new horizons were opening up.

“Even as we were ramping advanced 14nm SOI technology for IBM servers we saw that our fab loading was going down, and although we knew we were core to IBM it was clear that core was requiring less silicon,” he said. “We recognized that our survival depended on expanding into areas such as RF and silicon photonics to reverse the loading trend. So, we went through a pivot of our own to do exactly that even before the GF corporate pivot took place.”

Neil said that after the acquisition many team members felt proud and relieved.  They felt that being part of a pure-play foundry enabled them to continue their careers in the semiconductor industry, while providing new opportunities for growth.

What about now, though, with GF’s pivot away from scaling and toward differentiated and derivative technologies? “When you’re scaling, it’s easy to see what’s next because there’s a roadmap and you basically know what the next steps are going to be. However, on this path of differentiation that GF is taking we have to create our own path forward, which means there’s some uncertainty and we have to think differently about what it means to innovate,” he said.

“What does AI become? What about the IoT? 5G? We have to think beyond the scaling roadmap because there simply is no roadmap to create solutions in these different spaces. For our people, who are curious by nature, who are excited by technology, and who want to do new and unique things, the future holds great opportunities.”

In the next installment in this series, Gregg Bartlett, GF’s Sr. VP of Strategy & Asset Management, tells journalist Dave Lammers about the ups, downs and unexpected twists in the evolution of GF’s corporate strategy over the years.

About Author

Gary Dagastine

Gary Dagastine

Gary Dagastine is a writer who has covered the semiconductor industry for EE Times, Electronics Weekly and many specialized media outlets. He is a contributing editor at Nanochip Fab Solutions magazine and also is the Director of Media Relations for the IEEE International Electron Devices Meeting (IEDM), the world’s most influential technology conference for semiconductors. He started in the industry at General Electric Co. where he provided communications support to GE’s power, analog and custom IC businesses. Gary is a graduate of Union College in Schenectady, New York.

 

Making New Memories: 22nm eMRAM is Ready to Displace eFlash

By: Martin Mason

GF’s strategy is to deliver highly differentiated, high-added-value solutions for clients in high-growth markets, and a tangible result of that commitment is our 22nm FD-SOI (22FDX®) embedded MRAM non-volatile memory (NVM) technology, which has entered pilot production for several large IoT clients.

The embedded STT-MRAM (spin-transfer torque magnetoresistive RAM) technology, developed in partnership with Everspin Technologies, Inc., is aimed at IoT, general-purpose microcontrollers, automotive, edge AI and other applications where low-power operation and fast, robust, non-volatile code and data storage is a critical requirement.

GF’s eMRAM technology is unique in that it is a robust MRAM solution, having been designed as a high-volume embedded flash (eFlash) replacement. It has passed rigorous real-world production testing, and enables delivery of persistent data retention and endurance at extended temperatures. That is key for microcontroller applications and wireless connected IoT, where embedded memory must retain code and data at high temperatures, including through solder reflow at 260°C during PCB assembly. It also offers erase and (re)write speeds an order of magnitude faster than eFlash (200 nanoseconds vs. 10’s of microseconds), with comparable read speeds, providing a power advantage over eFlash in many applications.

Although it initially makes use of GF’s power-efficient 22FDX process, MRAM is deployed in ‘back-end-of-line’ metallization, which allows for a robust roadmap with planned derivatives on both FDX™ and GF’s FinFET technology. That’s because STT-MRAM as a memory technology allows for process variations which can be used to tune the memory bit cell. Accordingly, we expect to offer two “flavors” of eMRAM: eMRAM-F on 22FDX for code/data storage right now, and eMRAM-S as a working memory to augment SRAM at 1x nodes in the future.

GF is currently running multi-project wafers (MPWs) with 22FDX-based eMRAM-F designs for several clients, and multiple production tape-outs are scheduled in the next three quarters. Custom design services are available from GF and our design partners, and 22FDX eMRAM process design kits are available with macro densities ranging from 4Mb to 32Mb for individual macros. A 48Mb macro is scheduled for release in 4Q19 as well.

Industry is at a Transition Point

What’s driving great interest in alternative embedded NVM technologies at present is the fact that the industry is at a transition point: The 28nm node is possibly the last cost-effective node for eFlash, and the transition to 22nm geometries is making it imperative to find an alternative that is suitable for new and fast-growing low-power applications.

Many new eNVM memory technologies look interesting but aren’t yet production-ready. For example, RRAM (resistive RAM), which stores data by changing the electrical resistance of a dielectric, is the subject of much research and development but its maturity on 2x nm process nodes is limiting its adoption. Likewise adoption for PCM memory is limited by a lack of foundry support below 28nm.

Source: GF

By contrast, eMRAM from GF is an especially compelling and timely solution. Although the technology is complex and has required a significant investment of time and money to develop and deploy, it offers tremendous performance and versatility. In addition to the power benefits gained by combining the power-efficient base FDX silicon-on-insulator process with eMRAM, GF’s FDX process also has industry-leading RF connectivity capabilities, and extensive IP is available from GF. This all enables delivery of a unique high performance, high integration, low power and small size solution, which brings tremendous value to clients.

In fact, GF is on its third generation of MRAM technology with the 22nm eMRAM product, having also produced Everspin’s 256Mb 40nm and 1Gb 28nm standalone MRAM products as part of a joint development effort.

In addition to eMRAM technology, GF also offers clients eFlash and system-in-package flash (SIP Flash) embedded memories, built in a range of technologies from 130nm to 28nm to meet a broad swath of application requirements.

A Successful Strategy

The rollout of 22FDX eMRAM technology truly shows the success of GF’s efforts to intensify investment in areas where we have clear differentiation and where we can add great value for clients.

About Author

Martin Mason

Martin Mason

Martin Mason is Sr. Director of Leading-Edge eNVM at GLOBALFOUNDRIES. Before joining the company, he was at Maxim Integrated Products in Executive Director roles for Core Products and Precision Conversion Solutions, and prior to that, he served in product marketing and design/application engineering positions at Atmel, Actel, Concurrent Logic and GEC Plessey Semiconductors. He is a graduate of University of Newcastle upon Tyne in England.

 

格芯在美国和德国向台积电提起专利侵权诉讼

并申请禁制令以阻止侵权台湾产半导体产品的非法进口

加利福利亚州圣克拉拉,2019年8月26日——格芯(GLOBALFOUNDRIES),总部位于美国的全球领先的特殊工艺半导体代工厂,今日在美国和德国提起了多个法律诉讼,指控台湾积体电路制造股份有限公司(台积电)所使用的半导体生产技术侵犯了16项格芯专利。这些诉讼分别于今天向美国国际贸易委员会(ITC)、美国特拉华联邦地区法院、美国德克萨斯西区联邦地区法院,以及德国杜塞尔多夫地区法院和曼海姆地区法院提出。

在提起法律诉讼的同时,格芯还申请了法院禁制令,以阻止总部位于台湾的、在半导体生产领域处于垄断地位的台积电使用侵权技术生产的产品被进口至美国和德国。这些法律诉讼要求格芯指明台积电的主要客户以及下游电子公司,后者在大多数情况下才是包含了台积电侵权技术产品的实际进口人。格芯还基于台积电使用格芯专有技术而产生的数百亿美元的销售额而向台积电提出了巨额的损害赔偿请求。

 “尽管半导体生产在持续地向亚洲转移,格芯却反其道而行之,在美国和欧洲的半导体行业进行了大量投资。在过去的十年中,格芯共在美国投资超过150亿美元,并在欧洲最大的半导体生产基地投资超过60亿美元。我们提起法律诉讼的目的在于保护这些投资,以及在背后驱动着这些投资的基于美国和欧洲的技术创新”,格芯的工程及技术副总裁Gregg Bartlett如是评价道。“多年来,在我们投入数十亿美元进行本土的技术研发的同时,台积电却在非法从我们的投资中获利。此次采取行动,对于叫停台积电对我们关键资产的非法使用,并保护美国和欧洲的生产基地十分重要。”

格芯希望通过提起法律诉讼来保护其投资、资产和知识产权,并藉此确保半导体行业始终是充满竞争的行业,以保护行业客户的利益。

Media Fact Sheet

关于格芯

格芯是全球领先的特殊工艺半导体代工厂,提供差异化、功能丰富的解决方案,赋能我们的客户为高增长的市场领域开发创新产品。格芯拥有广泛的工艺平台及特性,并提供独特的融合设计、开发和生产为一体的服务。格芯拥有遍布美洲、亚洲和欧洲的规模生产足迹,以其灵活性与应变力满足全球客户的动态需求。格芯为阿布扎比穆巴达拉投资公司(Mubadala Investment Company)所有。欲了解更多信息,请访问 https://www.globalfoundries.com/cn。

 

GLOBALFOUNDRIES Files Patent Infringement Lawsuits Against TSMC In the U.S. and Germany

Injunctions seek to prevent unlawful importation of infringing Taiwanese semiconductors

Santa Clara, Calif. August 26, 2019 – GLOBALFOUNDRIES (GF), the world’s leading specialty foundry based in the United States, today filed multiple lawsuits in the U.S. and Germany alleging that semiconductor manufacturing technologies used by Taiwan Semiconductor Manufacturing Company Ltd. (TSMC) infringe 16 GF patents.  The lawsuits were filed today in the U.S. International Trade Commission (ITC), the U.S. Federal District Courts in the Districts of Delaware and the Western District of Texas, and the Regional Courts of Dusseldorf and Mannheim in Germany.

In filing the lawsuits, GF seeks orders that will prevent semiconductors produced with the infringing technology by Taiwan-based TSMC, the dominant semiconductor manufacturer, from being imported into the U.S. and Germany. These lawsuits require GF to name certain major customers of TSMC and downstream electronics companies, who, in most cases, are the actual importers of the products that incorporate the infringing TSMC technology. GF also seeks significant damages from TSMC based on TSMC’s unlawful use of GF’s proprietary technology in its tens of billions of dollars of sales.

“While semiconductor manufacturing has continued to shift to Asia, GF has bucked the trend by investing heavily in the American and European semiconductor industries, spending more than $15 billion dollars in the last decade in the U.S. and more than $6 billion in Europe’s largest semiconductor manufacturing fabrication facility. These lawsuits are aimed at protecting those investments and the US and European-based innovation that powers them,” said Gregg Bartlett, senior vice president, engineering and technology at GF. “For years, while we have been devoting billions of dollars to domestic research and development, TSMC has been unlawfully reaping the benefits of our investments. This action is critical to halt Taiwan Semiconductor’s unlawful use of our vital assets and to safeguard the American and European manufacturing base.”

GF is filing these lawsuits to protect its investments, assets and intellectual property, which will help to ensure that semiconductor manufacturing remains a competitive industry for the benefit of its clients.

Media Fact Sheet

About GLOBALFOUNDRIES

GLOBALFOUNDRIES (GF) is the world’s leading specialty foundry. We deliver differentiated feature-rich solutions that enable our clients to develop innovative products for high-growth market segments. GF provides a broad range of platforms and features with a unique mix of design, development and fabrication services. With an at-scale manufacturing footprint spanning the U.S., Europe and Asia, GF has the flexibility and agility to meet the dynamic needs of clients across the globe. GF is owned by Mubadala Investment Company. For more information, visit globalfoundries.com.

Contact:

Laurie Kelly, GLOBALFOUNDRIES
(518) 265-4580
[email protected]

 

AI at the Edge Optimizes 5G mmWave Networks

By: Peter A. Rabbeni

AI touches our lives in many different ways, and while some AI-enabled applications are highly visible, like the increasingly popular Amazon Echo and Google Home voice-controlled intelligent digital assistants, others are less obvious. But by no means are they less important.

For example, AI techniques are essential to the successful rollout of 5G wireless communications. 5G is the developing standard for ultra-fast, ultra-high-bandwidth, low-latency wireless communications systems and networks whose capabilities and performance will leapfrog that of existing technologies.

5G-level performance isn’t a luxury; it’s a capability the world critically needs because of the exploding deployment of wirelessly connected devices. A crushing amount of data is poised to overwhelm existing systems, and the amount of data that must be accessed, transmitted, stored and processed is growing fast.

5G Needed for the Upcoming Data Explosion

Every minute, by some estimates, users around the world send 18 million text messages and 187 million emails, watch 4.3 million YouTube videos and make 3.7 million Google search queries. In manufacturing, analysts predict the number of connected devices will double between 2017 and 2020. Overall, by 2021 internet traffic will amount to 3.3 zettabytes per year, with Wi-Fi and mobile devices accounting for 63% of that traffic (a zettabyte is 12 orders of magnitude larger than a gigabyte, or 1021 bytes).

The new 5G networks are needed to handle all of this data. The new networks will roll out in phases, with initial implementations leveraging the existing 4G LTE and unlicensed access infrastructure already in place. However, while these initial Phase 1 systems will support sub-6GHz applications and peak data rates >10GBps, things really begin to get interesting in Phase 2.

In Phase 2, millimeter-wave (mmWave) systems will be deployed enabling applications requiring ultra-low latency, high security, and very high cell edge data rates. (The “edge” refers to the point where a device connects to a network. If a device can do more data processing and storage at the edge – that is, without having to send data back and forth across a network to the cloud or to a data center – then it can respond more quickly and space on the network will be freed up.)

AI and 5G are Perfect Partners

AI functionality is key to edge computing because it provides for more effective control of networks, cells and devices. Without it, many 5G applications that rely on edge computing simply couldn’t be implemented, wouldn’t work well, or would cost too much to deploy.

Take the case of adaptive beamforming, where signals from phased array antennas are combined in ways that increase signal strength in a given direction. It’s important for 5G applications because while spectrum availability in the mmWave frequency range (30GHz – 300GHz) is nearly infinite, signals in these wavelengths are attenuated by atmospheric absorption which limits their usable range to about 300 meters. They also have difficulty penetrating buildings and foliage.

In the past, systems leveraging mmWave frequencies were built accepting these limitations, but that also limited their application. The control of adaptive beam-forming antenna arrays used for mmWave 5G communications is critical to optimizing their operation and performance, and so with the advancement of semiconductors and faster digital signal processing, sensing systems combined with AI can be used to control them. This will lead to dynamically optimized base stations and computing resources which better accommodate changing user needs and environmental conditions. Without AI, this would be much harder to achieve.

Smart Surveillance Cameras

Another way in which AI conserves network resources is its role in the growing use of “smart” surveillance cameras, which make use of diverse semiconductor technologies. More than 120 million IP (internet protocol) cameras were connected to networks globally in 2016, for use in a wide range of applications.

Many of these are so-called “smart” surveillance cameras. (In one notable instance recently, smart surveillance technology enabled police to pick out a wanted man among a crowd of 60,000 concert-goers.)

Without AI to enable edge processing of most of the data generated by a smart camera, though, networks would be overloaded. A single high-definition IP smart camera generates a video stream of 10Mb of data (or 30 frames) per second. Multiply that by the millions of such cameras added in recent years, and the network bandwidth required just for this application would be over a petabyte per second (1015) ─ clearly impractical.

AI to the Rescue at the Edge

Moreover, processing this data in the cloud would be hugely expensive with current technologies. The only real answer is to compute at the edge, using AI techniques for object recognition, gesture detection and classification, and only send minimal metadata over the network.

One might think that the most advanced, leading-edge semiconductor technologies are required to do this, but in fact a number of processes come into play. GF offers the industry’s broadest set of technology solutions for a range of 5G and edge-connected applications, including mmWave front end modules (FEMs), standalone or integrated mmWave transceivers and baseband chips, and high-performance application processors for mobile and networking.

For example, GF’s RF SOI, SiGe and FDX™ FD-SOI offerings are designed to serve applications ranging from sub-6GHz to mmWave frequency bands. RF SOI and SiGe solutions deliver an optimal combination of performance, integration and power efficiency for FEMs with integrated switches, low noise amplifiers and power amplifier applications. FDX offerings are well-suited for the next generation of connected devices such as smart cameras which require ultra-low power technology with intelligence and wireless connectivity built in.

Clients can take advantage of the back gate body-biasing capability of FDX that can be used to dynamically increase performance when needed for image processing, AI/Machine learning, or controlling leakage when a system is in standby. The FDX ecosystem of IP partners includes optimized IP for on-chip  power management, radio sub-systems, low voltage SRAM, instant-on MRAM, eNVM and FPGA blocks for the highly integrated flexible systems-on-chips (SoCs) needed for AI-enabled edge computing.

FDX SoC for future commercial IP cameras. (Source: GF)

Connected Intelligence

Traditionally, the industry has viewed networks non-holistically, on a transactional basis and from the separate and distinct viewpoints of computation, storage and data transport.

But if we now think about adding many edge-connected devices with sensor collection capabilities to the network, we begin to see the value of creating networks with smart sensing capabilities, whose data can be collected and become the basis for intelligent and time-based decisions that improve and optimize the services provided by the network.

This is what we mean by Connected Intelligence – the ability to sense, decide and act upon information collected from devices/sensors connected to the network to create the ultimate user experience.

The addition of AI engines to augment this “sense, decide and act” approach to network optimization can create a very powerful framework to best leverage available network assets.

Underneath it all is the realization that AI-enabled 5G mmWave networks will depend on the advancements and innovation in semiconductor technologies. No single technology solution will serve all potential applications. It’s going to require a range of technologies to make these next-generation applications work together seamlessly and maximize their potential.

GF is working closely with clients to understand their needs in this space, and is leading the industry with our portfolio optimized to help meet the demands of AI at the edge.

About Author

 

Peter A. Rabbeni

Peter is Vice President of Segment Offering Management, Business Development and Marketing. He has more than 25 years’ experience in the wireless industry, in system/circuit engineering, sales, marketing and business development positions at both the OEM and silicon levels. He has held senior positions at Raytheon, Ericsson and IBM, and while at IBM he built a vertically integrated semiconductor portfolio strategy that led to more than $3 billion worth of silicon design wins worldwide. He is responsible for cultivating IBM’s RF foundry presence in Asia, turning it into one of the most successful design-win engines within IBM Microelectronics. He was a key player in helping to lead GF’s successful acquisition and integration of IBM’s RF foundry business.

 

边缘人工智能助力优化5G毫米波网络

作者:Peter A. Rabbeni

人工智能在日常生活中的应用十分广泛,只不过一些支持人工智能的应用比较引人注目,比如越来越受欢迎的Amazon Echo和Google Home语音控制智能数字助手,而另一些应用则不那么明显,但这绝不表示这些不明显的应用不重要。

例如,5G无线通信能否成功推出,在一定程度上取决于人工智能技术的发展。5G是一项正在制定中的标准,主要针对超快、超高带宽、低延迟的无线通信系统和网络,这些系统和网络的容量和性能都将远超现有技术。

5G级性能并非奢侈之物。随着无线互联设备部署数量的爆发式增长,它已成为全球范围内亟需实现的一种性能。海量数据将会超出现有系统的承载能力,而需要访问、传输、存储和处理的数据量也在快速增长。

即将到来的数据大爆炸亟需5G技术

据估计,全世界的用户每分钟会发送1,800万条短信和1.87亿封电子邮件,观看430万个YouTube视频,进行370万次Google搜索查询。有分析师预计,在2017年到2020年期间,制造业采用的互联设备的数量将会翻一番。总体而言,到2021年,互联网流量将达到每年3.3泽字节,Wi-Fi和移动设备的用量将占到总流量的63%(1泽字节比千兆字节高出12个数量级,也就是1021字节)。

而要想处理这类数据就离不开全新的5G网络。新的网络将分阶段推出,初期会利用4G LTE和无授权访问这类现有的基础设施。虽然初期第1阶段采用的这些系统可支持各类6GHz以下的应用,而且支持的峰值数据速率能超过10GBps,但真正的重点在于第2阶段。

第2阶段将会部署毫米波(mmWave)系统,为需要超低延迟、高安全性和极高蜂窝边缘数据速率的应用提供支持。(“边缘”指的是设备与网络相连接的点。如果一台设备能够在边缘位置处理和存储更多数据(也就是无需通过网络在云端或数据中心之间来回传输数据),那么其响应速度会更快,占用的网络空间也就更少。)

人工智能和5G堪称是完美组合

人工智能的功能是决定边缘计算性能的关键,因为它能提供更有效的网络、蜂窝和设备控制。如果没有人工智能的支持,许多依赖于边缘计算的5G应用将根本无法实现,或无法正常运作,抑或是需要花费极高的成本来进行部署。

例如,在自适应波束形成中,相控阵天线发出的信号会通过一定的方式组合在一起,以增加特定方向上的信号强度。这一特点对5G应用来说非常重要,因为虽然毫米波频率范围(30GHz–300GHz)内可使用的频谱几乎没有限制,但这些波长的信号会因大气吸收而衰减,导致其可用范围会限制在约300米以内。此外,这类信号也难以穿透建筑物和植物。

这些限制在过去构建采用毫米波频率的系统时就已存在,系统的应用范围因而也受到了影响。如何控制自适应波束形成天线阵列(用于进行毫米波5G通信)对于优化其操作和性能而言至关重要,因此,随着半导体技术的进步和数字信号处理速度的提高,可以利用具有人工智能功能的感应系统来进行控制。此举将有助于基站和计算资源实现动态优化,更好地满足不断变化的用户需求,适应各种环境条件。而如果没有人工智能,这将很难实现。

智能监控摄像头

如今,“智能”监控摄像头的使用日益广泛,它们基本采用了多种半导体技术,而人工智能在这类摄像头中的应用正是节省网络资源的另一种方式。2016年,全球范围内联网的IP(互联网协议)摄像头的数量已超过1.2亿,应用类型更是多种多样。

其中很多就是我们所说的“智能”监控摄像头。(就在最近,智能监控技术就帮助警察从6万名音乐会观众中找出了通缉犯。)

如果没有人工智能对智能摄像头生成的大部分数据进行边缘处理,那么网络就会出现过载。一个高清IP智能摄像头每秒会产生10Mb(或30帧)大小的视频流。如果再乘以近年来增加的数以百万计的摄像头,单单这一类应用每秒所需的网络带宽就会超过拍字节(1015),这显然是不切实际的。

人工智能利用边缘技术解决难题

此外,在当前的技术条件下,在云端中处理这些数据的成本非常高昂。唯一有效的解决方案就是在边缘进行计算,使用人工智能技术来识别对象、检测手势和进行分类,并只将最少量的元数据通过网络发送出去。

有人可能会认为,要做到这一点,就需要采用最先进、最前沿的半导体技术,但其实,目前市场上的多种产品和技术就已足够解决问题。格芯拥有业界最广泛的技术解决方案,适合各种5G和边缘连接应用,包括毫米波前端模块(FEM)、独立或集成毫米波收发器与基带芯片,以及用于移动和联网的高性能应用处理器。

例如,格芯RF SOI、SiGe和FDX™ FD-SOI产品就旨在为6GHz以下到毫米波频段的应用提供支持。RF SOI和SiGe解决方案为集成开关、低噪声放大器和功率放大器的FEM提供出色的性能、集成度与功效组合。FDX产品则非常适合下一代互联设备,如需要超低功耗技术、内置智能功能和无线连接的智能摄像头。

在系统处于待机状态,需要进行图像处理、人工智能/机器学习或控制漏电时,客户可以利用FDX的背栅极体偏置功能来动态地提高性能。IP合作伙伴的FDX生态系统包括:经过优化的IP(用于片上电源管理)、无线电子系统、低压SRAM、即时开启型MRAM、eNVM和FPGA块,以此打造支持人工智能的边缘计算所需的高度集成且灵活的片上系统(SoC)。

适用于未来商用IP摄像头的FDX SoC。(资料来源:格芯)

互联智能

一直以来,不管是在事务处理层面,还是在计算、存储和数据传输这些相互独立且明显不同的方面,业界都没有全面地看待网络。

但是,如果我们现在考虑向网络中添加许多具备传感器采集功能的边缘互联设备,我们将能了解到打造具备智能感应功能的网络究竟具有何种价值——可以采集数据、以这些数据为依据做出明智、及时的决策,进而改善和优化网络提供的服务。

这就是我们所说的互联智能,即能够感测到联网设备/传感器收集的信息,并以此为依据做出决策、采取行动,提供出色的用户体验。

通过增加人工智能引擎的数量,可以增强这种“感测、决策并采取行动”的网络优化方法的效力,进而形成功能强大的框架,以便更充分地利用可用的网络资产。

究其根本,能否实现支持人工智能的5G毫米波网络,关键在于半导体技术领域的进步和创新。没有哪一种技术解决方案能够服务于所有潜在的应用。要想让这些下一代应用无缝协作,同时充分发挥它们的潜力,我们需要采用一系列的技术。

格芯与客户紧密合作,努力了解他们的相关需求,并利用经过优化的产品组合引领行业发展,不断满足客户对边缘人工智能的需求。

关于作者

Peter A. Rabbeni

Peter目前担任产品管理、业务开发与营销部门副总裁。他拥有超过25年的无线行业从业经验,涉足OEM和芯片两大领域,从事过系统/电路工程、销售、营销和业务开发工作。他曾在雷神公司、爱立信和IBM担任高级职位。在IBM工作期间,他开发出了垂直整合型半导体产品组合战略,帮助公司在全球范围内实现了30多亿美元的芯片设计中标收入。当时,他还负责拓展IBM在亚洲地区的RF代工业务,最终将其打造成了IBM微电子部门最成功的设计中标业务来源之一。在推动格芯成功收购和整合IBM的RF代工业务方面,他发挥了关键的作用。

更多详情