The Influence and Challenge of 5G Era on FPGA

FPGA is a semi-custom circuit, mainly used in application-specific integrated circuits, and has an irreplaceable position in aerospace/defense, consumer electronics, electronic communications and other fields. In the downstream application of FPGA, communication occupies the largest market segment, which can reach about 60%. Among them, FPGA chips are of indispensable importance for the development of 5G.

The Influence and Challenge of 5G Era on FPGA

FPGAs are an antique in today’s semiconductor world. Although FPGAs are 35 years old, the next 10 years represent growth opportunities not seen since the early 1990s. Why is this happening now?

The global data volume continues to surge, and IDC predicts that by 2025, the world will generate more than 175zb of data per year. With so much data, there is a huge opportunity to analyze it to gain insights that can change and impact the world. Artificial intelligence will play a major role in this data mining operation, and companies are adding employees with deep skills in machine learning and data analysis to meet future challenges…

In addition, due to the substantial increase in the number of 5G channels, the single-site FPGA usage has increased accordingly. China’s 5G commercialization progress is leading in the world, and the large-scale capital expenditure of each generation of mobile communication technology in my country is generally concentrated in the first few years of commercialization. Therefore, the current FPGA will occupy a more important position than the 4G era. Due to the high frequency band used by 5G, the number of 5G base stations may reach 1.5 times that of 4G.

In addition, with the maturity of millimeter wave technology in the “second half of 5G” after 2022, the number of small base stations is expected to reach tens of millions. The business scenarios that 5G needs to meet will far exceed 1G~4G. 5G devices will face more complex physical protocols and algorithms, and require higher logic control and interface rates.

Therefore, experts estimate that in the 5G market, the market value of a single base station side FPGA will reach several times that of 4G. The FPGA threshold is high and the localization rate is low. In the future, it will become the field of localization substitution in the Chinese market. The demand for FPGAs in the Chinese market is the largest in the world (more than 30%), but the localization rate of my country’s FPGA market is very low. At present, in the civilian field, domestic FPGAs account for less than 5% of the Chinese market, and domestic FPGAs are expected to flourish in the future.

1. New block floating point unit for AI/ML workloads

Block floating point (BFP) is a mix of floating point and fixed point arithmetic, where blocks of data are assigned a common exponent. We describe a new arithmetic unit that performs block floating-point operations and creates floating-point results for common matrix arithmetic operations. The BFP arithmetic unit supports a variety of data formats with different precisions and ranges.

Compared to traditional floating point units, BFP saves a lot of power and area by sacrificing some precision. This new arithmetic unit has been implemented in the new 7nm FPGA family from Achronix. In one demonstration, AI and machine learning workloads were benchmarked, and BFP improved both performance and power consumption compared to half-precision (FP16) operation.

2. Can FPGAs Lead the Future?

As a major FPGA (Field Programmable Gate Array) manufacturer, Xilinx believes that Moore’s Law is nearing its end, and the future cannot simply rely on periodic updates of chip design to achieve what it needs, which leads to the need to rely more on chip architecture innovation to break through in the future . Gilles Garcia, marketing director of Xilinx’s communications department, said that the common challenges facing these current challenges are that higher performance, more bandwidth and stronger computing power are required, and the 5G era means greater development opportunities for Xilinx.

Different from the previous communication period, 5G means more fragmented processing nodes, and end-edge-pipe-cloud constitutes a huge computing network. Garcia further pointed out that 5G will bring three main challenges to FPGA manufacturers: how to manage the complexity of the radio with large antenna arrays, how to manage the bandwidth requirements of the fronthaul, and how to meet the larger data backhaul requirements.

Specifically, in terms of fronthaul, the processing of 5G protocol devices is added while satisfying the use of traditional 4G radio devices, which means that it is necessary to support converged access of different protocols; multiple interfaces between metro areas, each interface The maximum rate is 100Gbps, which means that the core trunk transmission interface needs to have at least a transmission capacity of 100-400Gbps. At the same time, the data backhaul throughput of the terminal will increase by 10 times. Coupled with the huge user group, the data volume will show an exponential upward trend.

From the perspective of network characteristics, 5G’s requirements for higher bandwidth and lower latency have further increased. Gilles Garcia introduced that Xilinx’s 16-nanometer process technology can save 60% of power consumption; while RF SoC technology can combine digital and analog on a single-chip SoC (system-on-chip), Package size is reduced by 70%.

3.Where is the next competitive point of FPGA?

An important change in the 5G era is that telecom operators perform edge computing in data centers, which involves network virtualization functions. In the core network part, manufacturers’ requirements for bandwidth continue to increase, so Xilinx believes that data centers will be an important market. Large manufacturers, including Xilinx, have invested heavily in research and development to provide artificial intelligence capabilities, machine learning capabilities, and heterogeneous computing solutions.

In the FPGA market, Intel has previously acquired Xilinx’s competitor Altera, and Intel’s major focus is also in the data center. Previously, Intel acquired eASIC, intending to use eASIC to transition FPGA to ASIC (application-specific integrated circuit) level. There are also opinions in the industry, whether ASIC will succeed FPGA as the next-generation processing solution.

Compared with the past, there are obvious differences in the network in the Internet of Things era. The deployment requirements of edge computing services are diversified. Only companies that accurately grasp the needs of customers can gradually become bigger and stronger. In the field of artificial intelligence learning and reasoning and image intelligence requirements, the popularity of GPUs has driven companies such as NVIDIA to usher in rapid growth, and in the era of 5G edge computing, FPGAs will become the new focus of the new industry.

Compared with other heterogeneous processors, FPGA is more suitable for edge computing scenarios. It can be deeply customized according to its algorithm structure for each specific application to achieve higher computing efficiency and energy efficiency. The architecture can greatly optimize bandwidth and improve computing efficiency; and it has natural advantages in low latency and stability. It can be seen that the diversification of 5G edge scenarios will drive the computing power industry into the FPGA era.

Haoxinshengic is a pprofessional FPGA and IC chip supplier in China. We have more than 15 years in this field。 If you need chips or other electronic components and other products, please contact us in time. We have an ultra-high cost performance spot chip supply and look forward to cooperating with you.

If you want to know more about FPGA or want to purchase related chip products, please contact our senior technical experts, we will answer relevant questions for you as soon as possible