The Xilinx product portfolio based on 28 and 20nm planes and 16nm Fin FET+ technology has been expanded from three aspects, enabling customers to maintain a generation-leading advantage. These three aspects include:
・ Product series: FPGA, 3D IC and SoC based on UltraScale™ architecture
・ Product: Vivado® Design Suite’Co-optimization can enhance performance, improve power consumption and increase integration.
・ Productivity: Unparalleled integration and implementation speed
UltraScale+ series: The ever-expanding 20nm UltraScale architecture
Xilinx has formulated an active development roadmap in its rich product series, which runs through its three product categories, and each category can support and strengthen the development of the previous generation of products, and is committed to continuing to maintain the position of the leading generation. Xilinx launched the latest product of UltraScale+, further expanding the application of UltraScale architecture, which can simplify the migration between the plane node and the FinFET node. This helps customers migrate their 20nm designs and give full play to the performance-to-power advantages of FinFET technology.
Figure 1: Xilinx continues to expand its leadership in these three areas.
16nm innovation: UltraScale+ series
Xilinx 16nm UltraScale+™ series FPGAs, 3D ICs, and MPSoCs are based on the 20nm core UltraScale architecture, which perfectly combines the latest UltraRAM and high-speed memory (HBM), 3D-on-3D and multi-processing SoC (MPSoC) technologies to achieve a leading generation value. In order to achieve the highest level of performance and integration, the UltraScale+ series also includes the latest interconnect optimization technology SmartConnect. These devices not only further enrich the Xilinx UltraScale product series (now including 20nm and 16nm FPGA, SoC and 3D IC devices), but also give full play to the advantages of the significantly improved performance-to-power ratio of TSMC 16FF+ FinFET 3D transistors. The UltraScale+ series has been optimized at the system level. Compared with the previous generation technology, it can achieve higher system integration and intelligence, as well as the highest level of confidentiality and reliability.
The latest expansion of the Xilinx UltraScale+ FPGA series includes Xilinx’s industry-leading Kintex® UltraScale+ FPGA and Virtex® UltraScale+ FPGA and 3D IC series, while Zynq® The UltraScale+ series includes the industry’s first batch of MPSoCs.
Zynq UltraScale+ MPSoC C second-generation SoC
The UltraScale+™ MPSoC architecture is based on TSMC’s 16nm FinFET process technology, enabling a new generation of Zynq® UltraScale MPSoC. This new architecture can provide the following advantages for 32 to 64 bit processors: virtualization support; a combination of software and hardware engines for real-time control; graphics/video processing; waveform and data packet processing; next-generation interconnection and storage; advanced power management ; And can achieve multi-level security and reliability enhancement technology. These latest architectural elements and Vivado® The combination of Design Suite and abstract design environment can significantly simplify programming and increase productivity.
Figure 2: Xilinx UltraScale MPSoC architecture selects the right engine for the right task.
The industry’s first 3D on 3D technology
The high-end UltraScale+ series can take full advantage of the mixed power advantages of 3D transistors and the third-generation Xilinx 3D ICs. Just as FinFETs can achieve nonlinear improvements in the performance-to-power ratio of planar transistors, 3D ICs can achieve system integration and nonlinear improvements in the bandwidth-to-power ratio of monolithic devices.
Figure 3: Xilinx’s third-generation 3D IC will provide homogeneous and heterogeneous configurations.
Next-generation design kits & methods
The Vivado design suite is based on the Xilinx 28nm series and is co-optimized with the UltraScale architecture to achieve significant quality effects, routing, utilization, and productivity advantages. When combined with UltraFast™-an effective design method that covers development board planning, design creation, design implementation and closure, programming and hardware debugging-design teams can accelerate the achievement of expected success.
Through the use of high-level synthesis and IP integration tools, the productivity of the front-end design process has been increased by more than 4 times. Through faster hierarchical planning and multi-dimensional analysis of the placement and routing engine, and support for fast incremental ECO, design productivity is increased by more than 4 times.
Figure 4: Vivado Design Suite combined with UltraFast design method to achieve unparalleled integration and implementation time.