Top Tech Videos Of 2020

What engineers were watching in 2020.

popularity

2020 shaped up to be a year of major upheaval, emerging markets and even increased demand in certain sectors. So it’s not surprising that videos focusing on AI, balancing power and performance, designing and manufacturing at advanced nodes, advanced packaging, and automotive-related subjects were the most popular.

Of the 68 videos published this year, the following were the most viewed in each subject area:

AI/ML

  • 3 Types Of AI Hardware looks at the differences and commonalities of AI chips.
  • High-Performance Memory For AI And HPC examines the current performance bottlenecks in high-performance computing, drilling down into power and performance for different memory options, and what are the best solutions for different applications.
  • Memory Access In AI Systems talks about how memory affects overall power consumption, why partitioning of on-chip and off-chip is so critical to performance and power, and how this changes from the cloud to the edge.
  • Changes In AI SoCs discusses the tradeoffs in AI chips in the data center and in edge devices.
  • AI Inference Acceleration talks about considerations in choosing an AI inference accelerator, how that fits in with other processing elements on a chip, what tradeoffs are involved with reducing latency, and what considerations are the most important.

Power & Performance

  • Thermal Challenges In Advanced Packaging delves into transient and static effects in multi-chip configurations.
  • PCIe 5.0 Drill-Down focuses on the new PCI Express standard, why it’s so important for data centers, how it compares with previous versions of the standard, and how it will fit into existing and non-von Neumann architectures.
  • How Chips Age talks about circuit aging, whether current methods of predicting reliability are accurate for chips developed at advanced process nodes, and where additional research is needed.
  • Visualizing Differences In Analog Design explains the challenges of managing analog versus digital IP, including how to deal with dozens or even hundreds of versions of a schematic, and why visualization is so important for identifying changes and updates to an analog design.
  • Where Timing and Voltage Intersect drills into the limitations for power delivery networks and what processors can handle, why the current solutions to these issues are causing failures, and how voltage reduction can affect timing.

Systems & Design

  • Rising Packaging Complexity explains why advanced packaging is so difficult.
  • Speeding Up Verification Using SystemC shows how HLS plus formal can significantly reduce optimization and debug time.
  • Timing Closure at 7/5nm examines how to determine if assumptions about design are correct, how many cycles are needed for a particular operation, and what happens if signals get out of phase.
  • Analog Simulation at 7/5/3nm talks about analog circuit simulation at advanced nodes, why process variation is an increasing problem, the impact of parasitics and finFET stacking, and what happens when gate-all-around FETs are added into the chip.
  • Context-Aware Debug examines the growing complexity of debug, which now includes software, power intent and integration, multiple clocking and reset domains, and much more, where the limitations are for debug, and how automotive, functional safety and mixed signal affect the overall process.

Manufacturing, Test & Analytics

  • Challenges At 3/2nm discusses issues at upcoming process nodes, the move to EUV lithography and nanosheet transistors, and how process variation can affect yield and device performance.
  • Scan Diagnosis explains the difference between scan test and scan diagnosis, what causes values in a scan test to change, how this can be used to hone in on the actual cause of a failure in a design, and how to utilize test hardware more efficiently.
  • Ensuring HBM Reliability drills into the difficulties that crop up in advanced packaging, what’s redundant and what is not when using HBM, and how continuous in-circuit monitoring can identify potential problems before they happen.
  • Using Big Data For Yield And Reliability talks about the importance of clean data for traceability, yield improvement and device reliability, where and how it gets cleaned, and why that needs to be accompanied by domain expertise.
  • Using ML In Manufacturing shows how to apply machine learning to different use cases to limit early life failures.

Automotive

Architectures, New Standards and More

  • eFPGAs Vs. FPGA Chiplets discusses which approach works best where.
  • CXL Vs. CCIX explains how these two standards differ, which one works best where, and what each was designed for.
  • DDR PHY Training talks about how to train the DRAM physical layer using firmware, why that is so important for flexibility, and what kinds of issues engineers encounter when using this approach.
  • Finding Hardware Trojans shows why locating security threats in hardware is so difficult.

For more Semiconductor Engineering videos, check out the full list here.

Looking for a technology topic not covered in our videos or Knowledge Center? Send a message to: linda@semi-mags.com.



Leave a Reply


(Note: This name will be displayed publicly)