The Xilinx All Programmable product portfolio based on 28nm and 20nm planar and 16Fin FET+ technologies keeps customers a generation ahead of their competition

Mr. Giles Peckham, Regional Marketing Director for Xilinx Inc shares Xilinx FPGA vision and provide insights on the megatrends we are seeing .

• Describe the role of FPGAs in reprogrammable system?

Over 50 Billion devices and machines will be connected by 2020. Once connected, they must be secure to deter intruders, right down to the hardware level. As these devices, machines, systems, and networks become more context aware, they must adapt to their environments and demands, being more programmable and software defined. They also must be scalable, with ever more functions virtualized and efficiently mapped onto shared compute resources. As data and video is captured from sensors and cameras everywhere, analytics must enable these machines to recognize, interpret, decide, and act.

These systems and networks must also meet the growing demands of impatient end users and real time scenarios that require immediate, low latency response. Yet, behind the scenes, they must process an exponentially growing amount of data, packets, and pixels with ever more sophisticated algorithms while consuming the lowest possible power. And they must be highly differentiated, or they will fail in the increasingly competitive and cost sensitive worldwide market. This can only be accomplished by combining software intelligence with hardware optimization and any-to-any connectivity.

Xilinx All Programmable Solutions enable Smarter, Connected, and Differentiated Systems, integrating the highest levels of software-based intelligence with hardware optimization and any-to-any connectivity.

• How FPGA powers high performance computing?

Xilinx powers high performance computing with the SDAccel™ development environment for OpenCL™, C, and C++. This development environment enables up to 25X better performance/watt for data center application acceleration leveraging FPGAs. SDAccel, member of the SDx™ family, combines the industry’s first architecturally optimizing compiler supporting any combination of OpenCL, C, and C++ kernels, along with libraries, development boards and the first complete CPU/GPU like development and run-time experience for FPGAs.

First Architecturally Optimizing Compiler for OpenCL, C, and C++

• Architecturally optimizing compiler delivers up to 25X better performance/watt compared to CPU/GPU
• Delivers 3X the performance and resource efficiency of other FPGA solutions
• Enables new or existing OpenCL, C and C++ code for creating high performance accelerators

First Complete CPU/GPU-Like Development Experience on FPGAs

• First complete software development environment targeting FPGAs
• Optimize applications on FPGA platforms with little to no FPGA experience
• Easily migrate applications to FPGAs while maintaining and reusing OpenCL, C and C++ code

First complete CPU/GPU-Like Run-time Experience on FPGAs

• Supports large applications with multiple programs and CPU/GPU-like on-demand loadable compute units
• Maintains system functionality during program transitions and keeps critical system interfaces and functions live during application execution
• Allows FPGA accelerators to be shared across multiple applications using on-the-fly compute unit reconfiguration

• Define megatrends in system design and what are the key challenges?

Xilinx has defined 6 megatrends: 5G Wireless, NFV/SDV, Cloud Computing, Video/Vision, Industrial IoT, ADAS

5G: 5G networks must be more scalable, intelligent, and heterogeneous. Technologies such as distributed small cells, massive-MIMO with hundreds of antennas, and centralized base-band processing via CloudRAN, will dramatically increase coverage and data throughput. Networks will need to connect securely through backhaul and optical fronthaul for processing.

SDN/NFV: SDN creates centralized control of programmable and intelligent networks, leveraging standard interfaces to all network functions. NFV typically works within an SDN network with functions that are virtualized, readily accessible, configurable, and highly scalable. In order to meet the throughput, latency and power requirements of the network, many of these virtualized functions require optimized acceleration solutions including workload acceleration, efficient management and routing of data flows, a wide range of communication protocols, and programmable data plane acceleration on demand.

Cloud Computing: Data centers need to be workload optimized so they can rapidly adapt to changing throughput, latency, and power requirements from a wide range of virtualized on-demand software applications. These applications include machine learning, video transcoding, image and speech recognition, CloudRAN, and Big Data analytics, along with storage and networking acceleration and flexible, high performance connectivity.

Industrial IoT: The Industrial Internet of Things (IIoT) is driving the fourth wave of the industrial revolution. It is dramatically altering manufacturing, energy, transportation, cities, medical, and other industrial sectors. Most experts believe that IIoT is happening now and with very tangible, measurable business impact.

The IIoT enables companies to aggregate data from sensors to maximize efficiency of machines and the throughput of an entire operation. Applications include motion control, machine to machine communications, preventative maintenance, smart energy and smart grid, big data analytics, and smart and connected medical systems.

Only Xilinx provides a flexible, standards based solution that combines software programmability, real-time processing, hardware optimization and any-to-any connectivity with the security and safety needed for Industrial IoT systems. Xilinx SDAccel™, SDSoC™, and Vivado^® High-Level Synthesis enable customers to quickly develop their smarter, connected, and differentiated applications.

Video/Vision:The cost of higher resolution cameras and image sensors is falling rapidly. High resolution displays are becoming more prevalent. Video is expected to exceed 80% of Internet traffic by 2020. This creates new applications with added intelligence using high performance image/video processing and tightly integrated analytics. Video/Vision applications are found in numerous markets including consumer, broadcast, automotive, military, communications infrastructure, and cloud computing.

Xilinx All Programmable technologies are enabling a wide range of video/vision applications, including industrial machine vision for inspection systems and robotics, 8k display and transport, ADAS and the road to autonomous vehicles, military visions systems, surveillance systems, and drones. All of these solutions require high performance processing and encoding of images using programmable hardware, often tightly coupled to analytics of identified objects using software, and any-to-any connectivity to sensors, machines and the cloud.

ADAS: Ongoing development and deployment of ADAS and V2X connectivity are paving the way for the next advancement in driver assistance – autonomous driving with predictive intelligence. New HMI trends, like heads-up displays with augmented reality and gesture recognition, are rapidly changing the way drivers interact with their automobiles.

Xilinx All Programmable technologies enable next generation automotive systems including Infotainment, Driver Information, and Advanced Driver Assistance Systems. Applications span from collision avoidance – encompassing image and video processing, object recognition, and analytics – to the heart of the secure connected car with semi-autonomous operations and vehicle to vehicle communications.

For these next-generation automotive and ADAS systems, designers can use SDSoC™ development environments and heterogeneous processing of Zynq^®UltraScale+™ MPSoC devices to address connectivity, video encoding, graphics processing, and real-time control and safety requirements. Xilinx is shipping FPGA and Zynq^®-7000 All Programmable SoC-based devices for ADAS modules that are currently being used by 16 OEMs, and have shipped approximately 100M devices to the automotive and transportation industries in support of most major OEMS in production today. Xilinx provides fully automotive qualified devices and comprehensive solutions that include operating systems, IP, and platforms.

• What are recent Xilinx transformations?

The Xilinx All Programmable product portfolio based on 28nm and 20nm planar and 16Fin FET+ technologies keeps customers a generation ahead of their competition with an expansion of its offerings from three perspectives:

• Portfolio: UltraScale™ architecture-based All Programmable FPGAs, 3D ICs and SoCs
• Product: Co-optimized with the Vivado™ Design Suite for extra performance, power, and integration
• Productivity: Unmatched time to integration and implementation

UltraScale+ Family: Expanding the 20nm UltraScale Architecture

Xilinx is committed to staying a generation ahead with aggressive roadmaps across each of the three elements of its broader portfolio, with each element supporting and reinforcing the previous generation. With the addition of UltraScale+, Xilinx has built upon the UltraScale architecture which allows for simple migration between planar and FinFET nodes. This allows customers to migrate their 20nm designs and benefit from the performance per watt advantages of FinFET technology.

Innovations at 16nm: UltraScale+ Family

Building on the core UltraScale architecture at 20nm, Xilinx’s 16nm UltraScale+™ family of FPGAs, 3D ICs and MPSoCs, combine new memory, 3D-on-3D and multi-processing SoC (MPSoC) technologies, delivering a generation ahead of value. To enable the highest level of performance and integration, the UltraScale+ family also includes a new interconnect optimization technology, SmartConnect. These devices extend Xilinx’s UltraScale portfolio – now spanning 20nm and 16nm FPGA, SoC and 3D IC devices – and leverage a significant boost in performance/watt from TSMC’s 16FF+ FinFET 3D transistors. Optimized at the system level, the UltraScale+ family delivers far more systems integration and intelligence, and the highest level of security and safety than previous generations of technology.

The newly extended Xilinx UltraScale+ FPGA portfolio is comprised of Xilinx’s market leading Kintex^®UltraScale+ FPGA and Virtex^®UltraScale+ FPGA and 3D IC families, while the Zynq^®UltraScale+ family includes the industry’s first all programmable MPSoCs.

ZynqUltraScale+ MPSoC – The 2nd Generation All Programmable SoC

The UltraScale+™ MPSoC Architecture, built on TSMC’s 16nm FinFET process technology, enables next generation ZynqUltraScaleMPSoCs. This new architecture provides processor scalability from 32 to 64 bits with support for virtualization, the combination of soft and hard engines for real time control, and graphics/video processing, waveform and packet processing, next generation interconnect and memory, advanced power management, and technology enhancements that deliver multi-level security, safety and reliability. These new architectural elements are coupled with the Vivado^® Design Suite and abstract design environments to greatly simplify programming and increase productivity.

Industry’s First 3D on 3D Technology

The high end of the UltraScale+ portfolio leverages the combined power of 3D transistors and 3rd generation of Xilinx 3D ICs. Just as FinFETs enable a non-linear improvement in performance/watt over planar transistors, 3D ICs enable a non-linear improvement in systems integration and bandwidth/watt over monolithic devices.

Next Generation Design Suite & Methodology

Built from the ground up for Xilinx’s 28nm portfolio, the Vivado Design Suite has been co-optimized with the UltraScale architecture to deliver significant quality of results, routability, utilization, and productivity advantages. When combined with UltraFast™, a potent methodology that covers all aspects of board planning, design creation, design implementation and closure, programming and hardware debug, design teams will be able to accelerate their time to predictable success.

Productivity for the front end design process is multiplied by more than 4X with high level synthesis and IP integration tools. Productivity in design implementation improves by more than 4X due to faster hierarchical planning and analytic place and route engines as well as support for fast incremental ECOs.

• How Xilinx FPGA differentiates with other?

Xilinx is the world’s leading provider of All Programmable FPGAs, SoCs, MPSoCs and 3D ICs, enabling the next generation of smarter, connected, and differentiated systems and networks. Driven by the industry-wide shifts towards Cloud Computing, SDN/NFV, Video Everywhere, Embedded Vision, Industrial IoT, and 5G Wireless, Xilinx innovations enable these applications that are both, software defined, yet hardware optimized.

Xilinx’s portfolio of software defined and hardware optimized solutions include proven C and IP based design tools that support the development of ‘software defined hardware’, and a new family of software development environments that supports the development of ‘software defined systems’. With this unique combination, Xilinx is addressing the rapidly growing demands for programmability and intelligence with software, while enabling >10X the bandwidth, 1/10th the latency and power, and flexible any-to-any connectivity with optimized hardware.

• Can you highlight software defined data center and how it is advancing with Xilinx FPFA?

SDAccel™ development environment for OpenCL™, C, and C++, enabling up to 25X better performance/watt for data center application acceleration leveraging FPGAs. SDAccel, a member of the SDx™ family, combines the industry’s first architecturally optimizing compiler supporting any combination of OpenCL, C, and C++ kernels, along with libraries, development boards, and the first complete CPU/GPU-like development and run-time experience for FPGAs.

SDAccel’s architecturally optimizing compiler delivers up to 25X better performance/watt compared to CPUs or GPUs and 3X the performance and resource efficiency of other FPGA solutions. SDAccel leverages foundational compiler technology that is utilized by more than 1,000 programmers. SDAccel harnesses the power of this complier and enables software developers to leverage new or existing OpenCL, C, and C++ code for creating high performance accelerators, optimized for memory, dataflow, and loop pipelining in a wide range of data center applications such as compute search, image recognition, machine learning, transcoding, storage compression and encryption.