Integrated forecast: better hardware, suspect software

0

This article is part of Electronics Design Forecast 2022 publish

What you will learn:

  • The influence of PCI Express, both with PCIe Gen 4 and the emergence of PCIe Gen 5.
  • HBM and its support for ray tracing.
  • Why RISC-V and Small FPGAs Will Become More Popular.
  • The fate of free software.

To me, some articles seem to write themselves. Forecasting items do not fall into this category, probably because I see the consolidation and dependencies occurring within our industry that have made the design of the system more complex. I remember when size, weight and horsepower (SWaP) plus price were the main concerns for developers. Nowadays, it also includes details like security, artificial intelligence, availability, communications, compatibility, etc.

The challenge we have is the breadth of technology we cover, from microcontrollers to cloud servers. The differences along the way have never been so marked. Integrating many different components into a single chip is more common at the lower end of the scale as disaggregation becomes the norm in the cloud. In the latter case, storage and communication are distributed from the server box to their own farms connected via high-speed networks.

The growth of PCI Express (PCIe) 5 and the emerging PCIe 6 standard support the CXL and CCIX standards which allow direct access to peripherals and storage. CXL appears to be gaining momentum after absorbing Gen-Z.

PCIe Gen 4 will remain the mainstay this year, but PCIe Gen 5 platforms are state of the art. The challenge will be availability rather than technology. This is more of a concern for high-end applications, as many on-board applications do well, even with 4 Gtransfers / s / pin of PCIe Gen 3 compared to the rate of 64 GT / s / pin for PCIe. Gen 6.

We are seeing similar support in external memory space, with DDR4 flavors remaining the technology of choice even though DDR5 is raising its head this year. Price and availability will limit adoption of DDR5, especially in the in-vehicle space. DDR6 remains on the drawing board.

High Bandwidth Memory (HBM) is also becoming the norm, taking advantage of stacked memory arrays that provide higher on-chip memory capacity and higher throughput. This is due to the large memory bus, which cannot be replicated with off-chip memory like DDR or GDDR DIMMs. HBM3 is the latest generation. HBM storage started on GPUs, but is used on CPUs and accelerators (Fig. 1).

Graphics processing can take advantage of HBM, and greater ray tracing support results in higher demand for on-chip memory while providing significantly improved graphics presentation. NVIDIA has pushed ray tracing – the company has moved this support from its high-end solution to the mid-range, while Imagination Technologies has provided support for mobile applications.

Ray tracing is useful but not limited to gaming apps and multimedia creation. For example, it will prove effective in virtual reality (VR) applications. Virtual reality and augmented reality (AR) are also becoming mainstream due to improvements in hardware.

The Metaverse is more than hype, but it will be some time before mainstream audiences use it en masse. It has as much to do with software as it does with hardware. In the meantime, industrial AR / VR applications will be where the money lies, especially when tied to digital twin support, which is growing significantly in the industry. (Fig. 2). The reason for this success has less to do with lower costs but rather with software improvements and the fact that training and support costs can be drastically reduced using this technology.

But back to computers and processors

While incremental improvements in single cores continue, the ROI in x86 and Arm platforms remains with multicore solutions. Mixing different performance cores in an SoC is already common, but will become more and more popular. Likewise, security processors will be more prevalent and easier to use, especially when moving to deployment. Security has finally gone from being a possibility to a requirement when connecting to the cloud.

The elephant growing in the room is RISC-V. Yes, it’s just an instruction set definition, but it basically provides an architecture for building hardware and there are many sources out there these days. Stand-alone standard chips are available as well as integrated RISC-V FPGAs and so on. Outfits like Imagination Technologies have added RISC-V to their portfolio. This doesn’t mean that the x86 or Arm architectures are going to go away, but they now have a strong, distributed competitor.

Small FPGAs are another item to watch out for. Although they’ve been around for a long time, the focus often seems to be on high-end, better-performing solutions. What low-count logic devices can do is provide hardware customization that offers a power and price advantage over software alternatives. The development tools are easier to use, so all you need is a logic diagram. Given the limited availability of chips in general due to the implications of the COVID-19 outbreak (Fig. 3)it makes sense to use a collection of more readily available FPGAs that can be customized.

Free software challenges

There is a lot of proprietary software and libraries available, but the trend is to use open source software for everything from operating systems to applications. Microsoft Windows Flavors work on most PCs. However, versions of Linux, like Android, work on most in-vehicle systems.

Linux is one of the many open source success stories, but it also enjoys significant industry support. This is not the case for the majority of open source projects. Of course, most open source projects aren’t used by a large audience.

On the other hand, a large audience of unsupported developers are still using many of these projects. This has led to some challenges for developers who use the fruits of these projects when problems arise. These can range from poor design or use that opens up security holes like Apache Log4j or bugs such as OpenSSL’s Heartbleed.

Software inventory and software composition software is already an essential tool for some developers who need to track and verify their use of free software. Using these Software Composition Analysis (SCA) tools will be important for all developers using open source libraries and tools. It will also increase interest in companies providing open source software that they support, relieving the customer of this chore. Things get a bit more difficult when the end product has to be certified.

Read more articles in the Electronics Design Forecast 2022 publish

Share.

Comments are closed.