The SEMI/Gartner Market Symposium was held Semicon West 2014 at San Francisco, on July 7. Am grateful to Ms. Becky Tonnesen, Gartner, and Ms Agnes Cobar, SEMI, for providing me the presentations. Thanks are also due to Ms Deborah Geiger, SEMI.
Dean Freeman, research VP, Gartner, outlined the speakers:
• Sunit Rikhi, VP, Technology and Manufacturing Group, GM, Intel Custom Foundry Intel, presented on Competing in today’s Fabless Ecosystem.
• Bob Johnson, VP Research, Gartner, presented the Semiconductor Capital Spending Outlook.
• Christian Gregor Dieseldorff, director Market Research, SEMI, presented the SEMI World Fab Forecast: Analysis and Forecast for Fab Spending, Capacity and Technology.
• Sam Wang, VP Research Analyst, Gartner, presented on How Foundries will Compete in a 3D World.
• Jim Walker, VP Research, Gartner, presented on Foundry versus SATS: The Battle for 3D and Wafer Level Supremacy.
• Dr. Dan Tracy, senior director, Industry Research & Statistics, SEMI, presented on Semiconductor Materials Market Outlook.
Let’s start with Sunit Rikhi at Intel.
As a new player in the fabless eco-system, Intel focuses on:
* The value it brings to the table.
* How it delivers on platforms of capability and services.
* How it leverage the advantages of being inside the world’s leading Integrated Device Manufacturer (IDM)
* How it face the challenges of being inside the world’s leading IDM.
Intel has leadership in silicon technologies. Transistor performance per watt is the critical enabler for all. Density improvements offset wafer cost trends. Intel currently has ~3.5-year lead in introducing revolutionary transistor technologies.
In foundry capabilities and services platforms, Intel brings differentiated value on industry standard platforms. 22nm was started in 2011, while 14nm was started in 2013. 10nm will be starting in 2015. To date, 125 prototype designs have been processed.
Intel offers broad capability and services on industry standard platforms. It also has fuller array of co-optimized end-to-end services. As for the packaging technology, Intel has been building better products through
multi-component integration. Intel has also been starting high on the yield learning curve.
Regarding IDM challenges, such as high-mix-low-volume configuration, Intel has been doing configuration optimization in tooling and set-up. It has also been separating priority and planning process for customers. Intel has been providing an effective response for every challenge.
Some of Intel Custom Foundry announced customers include Achronix, Altera, Microsemi, Netronome, Panasonic and Tabula.
Engineers designing FPGA applications face many challenges. Using Plunify’s automation and analysis platform, engineers can run 100 times more builds, analyze a larger set of builds and quickly zoom in on better quality results. Using data analytics and the cloud, Plunify created new capabilities for FPGA design, with InTime being an example.
Kirvy Teo said: What happens when you need to close timing in FPGA design and still can’t get it to work? Here is a new way to solve that problem – machine learning and analytics. InTime is an expert software that helps FPGA design engineers meet timing and area goals by recommending “strategies”. Strategies are combination of settings found in the existing FPGA software. With more than 70 settings available in the FPGA software, no sane FPGA design engineer have the time or capacity to understand how these affect the design outcomes.
One of the common methods now is to try random bruteforce using seeds. This is a one-way street. If you get to your desired result, great! If not, you would have wasted a bunch of time running builds with you none the wiser. Another aspect of running seeds is that the variance of the results is usually not very big, meaning you can’t run seeds on a design with bad timing scores.
However, using InTime, all builds become part of the data that we used to recommend strategies that can give you better results, using machine learning and predictive analytics. This means you will definitely get a better answer at the end of the day, and we have seen 40 percent performance improvements on designs!
How has Plunify been doing this year so far? According to Teo, Plunify did a controlled release to selected customers in first quarter of 2014, who are mainly based in China. It is easier to guess who as we nicknamed them “BCC” – Big Chinese Corporations.
Unsurprisingly, they have different methodologies to solving timing problems and design guidelines, many of which were done to pre-empt timing problems at the later stage of the design. InTime was a great way to help them to achieve their performance targets without disrupting their tool flows.
Plunify is announcing the launch of InTime during DAC and will be looking to partner with sale organizations in US.
What’s the future path likely to be? Teo added: “Machine learning and predictive analytics are one of the hottest topics and we have yet seen it being used much in chip design. We see a lot of potential in this sector. Beyond what InTime is doing now, there are still many chip design problems that can be solved with similar techniques.
“First, there is a need to determine the type of problems that can be solved with these techniques. Second, we are re-looking at existing design problems and wondering, if I can throw 100 or 1000 machines to this problem, can I get a better result? Third, how to get that better result without even running it!
“As you know, we do offer a FPGA cloud platform on Amazon. One of the most surprising observations is that people do not know how to use all those cheap power in the cloud! FPGA design is still confined to a single machine for daily work, like email. Even if I give you 100 machines, you don’t know how to check your emails faster! We see the same thing, the only method they know is to run seeds. InTime is what they need to make use of all these resources intelligently.
Why would FPGA providers take up the solution?
The InTime software works as a desktop software which can be installed in internal data centers or desktops. It is on longer just a cloud play. It works with the current in-house FPGA software that the customer already own. We are helping FPGA providers like Xilinx or Altera, by helping their customers with the designs. They will feel: How about “Getting better results without touching your RTL code!”
Xilinx Inc. has announced of its 20nm All Programmable UltraScale portfolio with product documentation and Vivado Design Suite support.
Neeraj Varma, director-Sales, India, Xilinx, said: “We are enabling All Programmable and smarter systems. We are using smart IP. We are aligning to produce smarter systems. We are helping customers to differentiate their products faster.
“In future, we will go with concurrent nodes with FPGAs, SoCs and 3D ICs. As per our estimates, 28nm will have a very long life. We shipped the 20nm device in early Nov. 2013. It complements 28nm or new high-performance architectures. 16nm complements 20nm with FinFET, multiprocessing, memory.”
Strategy execution has kept Xilinx a generation ahead. As of Dec. 2013, its 20nm portfolio is available to customers. There are two major announcements from Xilinx.
* Xilinx 20nm All Programmable UltraScale portfolio now available with ASIC-class architecture and ASIC-strength design solution.
* Xiilinx doubles industry’s highest capacity device to 4.4 mn logic, delivering density adantage, a full generation ahead.
KINTEX UltraSCALE – XCKU035, 040, 060, 075, 100, 115.
VIRTEX UltraSCALE – XCVU065, 080, 095, 125, 145, 160.
There is a family migration path. There is scalability for derivative applications. You can leverage PCB investment across platforms. It is future-proof with migration path to 16nm. For making these happen, Xilinx is using the TSMC 20SoC.
Varma added, “We have increased the logic cells in Kintex and Virtex, and added 100G Ethenet blocks and 150G Interlaken blocks.”
The second announcement – highest density in FPGAs in industry. The XCVU440 is the largest in the industry by 4X, a full generation ahead, and uses 50M equivalent ASIC gates. Xilinx is delivering an ASIC-class advantage through silicon, tools and methodology.
There is UltraSCALE ASIC-class architecture, and ASIC-class capabilities. There is also the Vivado ASIC-strength design suite.
UltraFAST is the design methodology. UltraSCALE will support networking, digital video and wireless.
Interconnect bottlenecks impede next generation performance.
* Routing delay dominates overall delay.
* Clock skew consumes more timing margin.
* Sub-optimal CLB packing reduces performance and utilization.
Varma added: “We have solved these issues – as UltraSCALE re-architects the core. There is 90 percent utilization now with maximum performance. We added next-generation routing, ASIC-like clocking – have clocks by segment, and logic cell packing.
“Block-level innovations optimize critical paths for massive bandwidth and processing. We are going to support DDR4, and there will be a lot more security features.”
The Vivado design suite accelerates productivity. Analytical placer solves the interconnect issue.
UltraSCALE apps include:
VIRTEX: 400G OTN switching, 400G transponder, 400G MAC-to-Interlaken bridge, 2x100G Muxponder, ASIC prototyping.
KINTEX: 4×4 mixed mode radio, 100G traffic manager NIC, super high-vision processing, 256-channel ultrasound, 48-channel T/R radar processing.
Agnisys Inc. was established in 2007 in Massachusetts, USA, with a mission to deliver innovative automation to the semiconductor industry. The company offers affordable VLSI design and verification tools for SoCs, FPGAs and IPs that makes the design verification process extremely efficient.
Agnisys’ IDesignSpec is an award winning engineering tool that allows an IP, chip or system designer to create the register map specification once and automatically generate all possible views from it. Various outputs are possible, such as UVM, OVM, RALF, SystemRDL, IP-XACT etc. User defined outputs can be created using Tcl or XSLT scripts. IDesignSpec’s patented technology improves engineer’s productivity and design quality.
The IDesignSpec automates the creation of registers and sequences guaranteeing higher quality and consistent results across hardware and software teams. As your ASIC or FPGA design specification changes, IDesignSpec automatically adjusts your design and verification code, keeping the critical integration milestones of your design engineering projects synchronized.
Register verification and sequences consume up to 40 percent of project time or more when errors are the source of re-spins of SoC silicon or an increase in the number of FPGA builds. IDesignSpec family of products is available in various flavors such as IDSWord, IDSExcel, IDSOO and IDSBatch.
IDesignSpec more than a tool for creating register models!
Anupam Bakshi, founder, CEO and chairman, Agnisys, said: “IDesignSpec is more than a tool for creating register models. It is now a complete Executable Design Specification tool. The underlying theme is always to capture the specification in an executable form and generate as much code in the output as possible.”
The latest additions in the IDesignSpec are Constraints, Coverage, Interrupts, Sequences, Assertions, Multiple Bus Domains, Special Registers and Parameterization of outputs.
“IDesignSpec offers a simple and intuitive way to specify constraints. These constraints, specified by the user, are used to capture the design intent. This design intent is transformed into code for design, verification and software. Functional Coverage models can be automatically generated from the spec so that once again the intent is captured and converted into appropriate coverage models,” added Bakshi.
Using an add-on function of capturing Sequences, the user is now able to capture various programming sequences in the spec, which are translated into C++ and UVM sequences, respectively. Further, the interrupt registers can now be identified by the user and appropriate RTL can be generated from the spec. Both edge sensitive and level interrupts can be handled and interrupts from various blocks can be stacked.
Assertions can be automatically generated from the high level constraint specification. These assertions can be created with the RTL or in the external files such that they can be optionally bound to the RTL. Unit level assertions are good for SoC level verification and debug, and help the user in identifying issues deep down in the simulation hierarchy.
The user can now identify one or more bus domains associated with Registers and Blocks, and generate appropriate code from it. Special Registers such as shadow registers and register aliasing is also automatically generated.
Finally all of the outputs such as RTL, UVM, etc., can be parameterized now, so that a single master specification can be used to create outputs that can be parameterized at the elaboration time.
How is IDesignSpec working as chip-level assertion-based verification?
Bakshi said: “It really isn’t an assertion tool! The only assertion that we automatically generate is from the constraints that the user specifies. The user does not need to specify the assertions. We transform the constraints into assertions.”
FPGAs serve highly diverse applications. Tailored devices are serving diverging market needs. According to Vince Hu, VP Product & Corporate Marketing, Altera Corp., next-generation portfolio involves an ideal mix of process technologies.
There is greater diversity and capabilities for the broadest range of applications. Finally, Altera has added the 55nm EmbFlash that extends Altera’s tailored approach. Hu was speaking at the 13th Globalpress Electronics Summit being held in Santa Cruz, USA.
Addressing needs of higher-volume systems is key. Industrial and automotive systems tend to be cost sensitive, low power and limited in broad areas. There is an increased pressure to innovate leading to a strong demand for programmable solutions with enhanced features. Altera is expanding the capabilities of non-volatile programmable logic devices (PLDs).
Altera is also bolstering high-volume system solutions. TSMC leading-edge embedded flash technology is a device tailored for high-volume applications. It adds more functionality to non-volatile PLDs. It also re-inforces Altera’s commitment to high-volume applications.
In addressing power/performance challenges, 20SoC is said to be the quickest path to next-generation process. It is tailored for a range of performance and bandwidth-critical applications. There is up to 60 percent lower power vs. 28nm. One of the latest results with 20SoC process is the first 32Gbps transceivers that are operating in 20nm silicon.
Currently, high-end applications are pushing the envelope. Intel’s 14nm tri-gate is said to be a game changer for FPGAs. Tri-gate
surrounds channel on the three sides. It increases channel performance and reduces power. Tri-gate is a proven, second- generation technology. The 14nm tri-gate maintains the Moore’s Law.
Driving toward 400G OTN systems
Altera has acquired OTN IP provider TPACK. It accelerates the company’s OTN roadmap and builds on the Avalon acquisition in 2010. OTN IP, combined with high-performance silicon, positions Altera for continued growth in the high-end networking market.
Tailored devices are now serving diverging market needs. It is an extension of Altera’s tailored approach. There are even greater diversity and capabilities, serving the broadest range of applications. A mixture of application-specific IP provides even greater tailored solutions. Altera is mixing the advanced FinFET process, traditional HKMG planar process and embedded flash technology.
Singapore based Plunify claims that chip design companies can design faster and better using cloud computing. Stressing on the company’s go-to-market strategy, Plunify’s founder, Harn Hua Ng, said the Plunify partners with tool vendors, their distributors and complementary sales representatives.
Since pay-as-you-go business models are rare in the semiconductor industry, we went through several steps, of which the first was to better understand the market, the available tools and stake-holders:
* How is the market reacting to cloud computing and licensing schemes?
* What are current tool capabilities with regards to multiple CPUs/servers? Which parts of the chip design workflow can best take advantage of scalable, parallel features?
* What tools are more suitable for a cloud environment?
With these in mind, the next step was to build the cloud platform and the application clients to address immediate concerns – security, accessibility and cost.
“Then, we partner with tool vendors, their distributors and sales reps to bring our solutions to end-users. Companies of different sizes
view the advantages of cloud computing differently, so solutions need to be customized accordingly. Some see Plunify as solving longer term IT problems of scaling and provisioning; while others use us as an immediate way to speed up their design workflows. We are still in the process of learning about the market.”
How can the on-demand cloud computing platform dramatically accelerate chip design workflows? According to Harn Hua Ng, one immediate benefit is an almost instantaneous fulfillment of peak demand IT requirements, for example, a urgent request to do 100 synthesis builds to fix a problem due yesterday. Or if the problem cannot be fixed, at least the design team will find out in a day rather than potentially in three months’ worth of runtime without a cloud solution. The longer term acceleration is a gradual parallelization of the design workflow.
Currently, chip designers tend to visualize the design workflow as a chain of mostly serial steps with many dependencies, just because many steps can be time-consuming (both in terms of runtime and time taken to analyze intermediate results).
With an on-demand compute platform, designers can have more room to experiment and to optimize, more readily incorporating agile practices in hardware development.
Xilinx Inc. has announced its 20nm portfolio strategy. The 20nm portfolio will allow Xilinx to offer twice the performance at half the power. It will increase productivity by 4x, and improve integration by 1.5- 2x. Besides, there will be 20-50 percent lower BOM cost.
Xilinx’s 20nm all programmable portfolio builds on 28nm breakthroughs to stay a generation ahead. “At 20nm, we were able to break out to become an all programmable company,” said Neeraj Varma, country manager, sales, India, Australia and New Zealand, Xilinx India.
The next generation FPGAs, second generation SoCs and 3D ICs will be ‘co-optimized’ with Vivado for the most compelling alternative ever to ASICs and ASSPs. From enabling programmable logic, the Xilinx 20nm portfollio will enable programmable systems integration!
The first SoC strength design suite was shipped in Q2-2012. It has been built from ground up for the next decade of all programmable devices. Today, the Xilnix Vivado is used for over 30 percent of 28nm FPGAs and 100 percent for 3D ICs.
Xilinx has been expanding on its next generation competencies. The 3D IC expertise and supply chain has gone from homogenous to heterogenous. The SoC and embedded software has also undergone change, as have XCVRs and analog mixed signal (AMS), communications BU and applications IP, and next generation design automation. Xilinx is now charting an aggressive course forward.
Xilinx’s 20nm portfolio has been co-optimized for performance, power and integration to address the market needs at 20nm. For the next-generation FPGA,, it will provide unmatched system optimized transceivers at highest channel quality w/ second generation auto equalization. There will be higher bandwidth w/over 100 transceivers @ 33Gb/s.
There will be 2X performance optimization, with faster DSP and BRAM, DDR4, transceivers and 2x memory bandwidth. There will be over 90 percent routing architecture enabling high bandwidth bussing and fast design. One half power optimization will provide an optimized performance/watt. There will be next generation block level power management. There will be 1.5x integration/BOM in terms of 1.5x logic, DSP, BRAM, AMS, VCXO, etc. Read more…
Device volume, variety and complexity are only going to increase. Transformative technologies like virtual prototypes give organizations the tools to transcend challenges. Companies like Altera are creating competitive advantage and innovation with these solutions. Virtual prototyping is now ready for the masses.
Industry trends and challenges make virtual prototyping a must-have solution. New realities make prior adoption barriers mere myths. Virtual prototyping has become a key process for early software development and supply chain enablement. Industry trends also alter design requirements. For instance, earlier, it used to be computing and single core, which has since moved on to connectivity and multi-core.
This opens up implications for SoC development, especially, in terms of increased complexity and volume of software. There is a need to get the architecture right. No amount of downstream tools will compensate for the fundamentally wrong architecture. There is also a need to start software development earlier, in parallel with hardware design. Needless to say, hardware-software integration must be accelerated and system validation will minimize waterfall development process.
New realities of prototyping render prior barriers as mere myths. For instance, earlier, it was believed that creating a prototype is hard. IP models, TLMCentral and model creation software have come a long way, in reality. Earlier, there was a need to wait for complete prototype. Now, software can be developed incrementally and VDKs are jumpstarting the software development. Earlier, one felt the need to change software environment. In reality, the very same tools, debuggers and environment used for hardware can be used here.
Also, today, there are multiple use cases, verticals and customers of virtual prototyping. There is industry support for system-level models. The TLMCentral is an open, web-based portal that provides consolidated access to transaction-level models available across the industry, helping virtual prototype developers accelerate the creation and deployment of their prototypes for early software design.
Open and free, TLMCentral is the first industry-wide portal to aggregate available transaction-level models. It has over 1,000 models of most common IP blocks and interfaces for wireless, consumer and automotive applications. TLMCentral is supported by leading IP vendors, tool providers, service companies and universities. It also offers model developers, architects and software engineers an infrastructure for news, forums and blogs.
Integrated into the software development environment, there are popular debuggers, powerful controls and debugging information. VDK is a great starting point and for ongoing use. One can install and start using. There is no need to wait for months for a prototype. Templates, sample software and reference prototypes are available in one place. Post-silicon support and validation is provided, besides early availability for software development and testing.
Key process for earlier software development includes hardware-software integration and system validation. Semis are engaging customers earlier. The VDKs are driving tangible time-to-volume reduction. Tangible benefits of virtual prototyping include faster time to revenue, faster customer success, and faster field and ecosystem readiness.
Moshe Gavrielov, president and CEO, Xilinx Inc. was in Hyderabad, India to inaugurate Xilinx India’s new, expanded India site at Hi-Tech city. The new Hyderabad site is a 131,000 square-foot office building — more than 2X the size of the previous site. It is equipped with engineering labs for end-to-end development and has a larger, energy-efficient data center. It also has facilities for customer and employee events.
This site is Xilinx’s largest R&D centre outside of US headquarters. Pivotal to record-fast delivery of 28nm technologies, it has a stellar worldwide technical support team. Xilinx has been committed to be the first to process nodes. It is pioneering 3-D IC technology and leading edge processing sub-systems. It is also offering programmable analog/mixed signal solutions as well as system to IC tools and IP to enable silicon.
Gavrielov said that Xilinx’s business drivers have been the programmable imperative, relentless system integration and an insatiable intelligent bandwidth. The programmable imperative has further accelerated. For instance, 28nm = 2X 45nm cost. Insatiable bandwidth has now expanded to Latin America. It is said to be almost 2x the size of the US market. There has been 20 percent growth in mobile subscribers and 318 percent growth in Facebook users.
There has been insatiable bandwidth across India as well, while bandwidth is also said to be growing in Africa. Some of the trends driving insatiable intelligent bandwidth include extreme bandwidth – which has seen 5X growth in five years, smart vision, ubiquitous computing and embedded security. Read more…
Xilinx Inc. has announced its first Zynq-7000 Extensible Processing Platform (EPP) shipments to customers. It showcased the first public demonstration of a Zynq-7000 EPP at the ARM European Technical Conference, in Paris, France. where attendees saw the device running a Linux-based application. Xilinx has recently started shipping Zynq, to at least three customers.
The Zynq-7000 family is the world’s first EPP. It combines an industry-standard ARM dual-core Cortex-A9 MPCore processing system with Xilinx 28nm unified programmable logic architecture. This processor-centric architecture delivers a complete embedded processing platform that offers developers ASIC levels of performance and power consumption, the flexibility of an FPGA and the ease of programmability of a microprocessor.
Dave Tokic, senior director, partner Ecosystems and Alliances, said the company had made a number of investments. It has adopted a two-pronged approach: focusing on how it developed the ecosystem, and what it could do by itself. “We need a tool flow applicable across all customers. Our technologies are enabling much more complex designs. We are also raising the bar for the EDA providers. We do provide early access to our tools, etc.”
Tokic added that the company has also invested a lot in training and certification in India. “Our partners are some very good companies. We have 24 members in our program. Eight of those are certified members.” Some of the partner companies include Wipro, TCS, Corel, Mistral, CMC, GDA Technologies (L&T), Mechatronics, etc.
Lawrence Getman, VP of Processing Platforms, added that Xilinx has been seeing how to potentially leverage a cloud. “We are continuing to develop the IP ecosystem. We are also looking to engage expert service needs.”
Commenting on developments, Getman said that Xilinx’s Virtex-7 series FPGAs are based on the high performance low power (HPL) process by TSMC. Xilinx wants to foster more collaborative approach in future for acquiring and working with customers.