Why is RISC-V successful?

2022-06-19 0 By

It is indisputable that the introduction of risC-V processor architecture is exciting.However, while many call it a harbinger of a broader open source hardware movement, the reasons behind its success are not obvious, and the impact on extending more open source kernels is far from certain.”The adoption of RISC-V as the architecture of choice for many chip developers has created a wave of innovation in the hardware development community,” said Stephano Cetola, Director of technology programs at RISC-V International.”Designers are now taking their RISC-V based designs and implementing them in various industries.”This is not the first time a processor implementation or instruction set architecture (ISA) has been placed in the public domain.The industry is full of them, including OpenPOWER, OpenSPARC, OpenRISC, and more.While each gained a degree of traction, they all paled in comparison to the success risC-V achieved in a short period of time.In talking to people in the RISC-V community, two words are often repeated — free and free.Some people want a free kernel, while others want the freedom to do whatever they want with the kernel.For these people, free is almost irrelevant because they will spend a lot of money to get what they want.The rise of risC-V in the changing market coincides with other events in the industry.The first is the slowing down of Moore’s Law, which means that the increase in total processing power no longer comes with each new manufacturing node.The second is the rapid development of machine learning, which requires a significant increase in processing power.So is it just a matter of being in the right place at the right time?In fact, during this time, the perception of processors has changed.”Risc-v is driven by the community’s demand for freedom in the hardware design process,” said Simon Davidmann, founder and CEO of Imperas Software.Today, everything requires some form of machine learning.Whether we’re talking about your phone, taking better photos, whatever it is, takes a lot of computing.People realized they needed a lot of processors.They need their own processor architecture.You need to configure them the way you want.Off-the-shelf technology won’t help you.As a result, the electronics market has changed, and “we need the freedom to build chips, and the freedom to build processors and the processor architecture in those chips.”All of this needs to be viewed in the context of a new generation of systems companies entering the market, each with unique economic reasons.But one thing these systems companies do have in common is that they don’t try to sell the chips they develop.Instead, they sell services that are driven in part by those products.They cannot buy the right product from the available market, so they are going to develop the chip themselves, while also contributing and collaborating to drive some of the necessary innovation.In this case, RISC-V plays an important role.Risc-v’s differences RisC-V creates breakthroughs in multiple areas, and they succeed in each area for different reasons.To understand this, it is necessary to separate the various aspects of RISC-V’s success.The first is the architecture itself.The second is the large number of open source implementations of the architecture available.The third area is the support kernel around the processor kernel.Finally, there are the necessary tools to help implement and validate the RISC-V processor.It was originally created to meet a specific need.”It now has a lot of resources,” said Davidmann of Imperas.”In the beginning, smart people from universities, academics, universities created a good thing.After leaving Berkeley in the middle of Silicon Valley, it gained some momentum from former Berkeley graduates.The interest is much higher than OpenRISC.The universities needed it, and they pushed it.”Risc-v is now an open standard ISA, contributed by the University of California, Berkeley, and managed by risC-V International, an industry nonprofit.Many universities have created open cores, such as Berkeley’s Rocket Core, ETH Zurich and its Pulp platform.Today, there are many industry collaborative groups that bring industry and academia together to build open source cores and make them available to the entire community.Examples include the CHIPS Alliance and the Open Enhw Working Group.Many countries have developed initiatives to meet local needs.India had its Shakti program and was eventually expelled by IIT Madras.In Israel, the GenPro Alliance brings together industry and academia.There are other similar projects in Japan and China that are building risC-V cores as open source in order to make them available to their communities and meet their specific interests.Risc-v is the first open and customizable ISA.”Right now, the main industry interest associated with RISC-V is not open source implementations, but open source instruction sets,” said Andy Heinig, Head of the Advanced Systems Integration Group for IIS and Head of High-performance Electronics at Fraunhofer..”With this, the environment is standardized, but the actual implementation is company-specific and company-owned.We see similar activity in the area of chip-to-chip interfaces, where different standards are being prepared and discussed.Here, too, the standard allows interoperability between chips from different vendors.Risc-v allows interoperability on the software side.”The ability to make changes is important.”The emergence of open source ISA (such as RISC-V) that supports custom extensions offers processor designers incredible freedom,” said Shubhodeep Roy Choudhury, CEO and co-founder of Valtrix Systems.”At the same time, it presents a very interesting validation challenge.Ensuring that all designs are compliant and functional requires a change in the way test generators are designed.They need to be highly configurable to allow validation of custom functionality as well as legacy/baseline functionality.”This ISA big leap from open ISA to open source processors.”The concept of open source IP is very tempting because it is reminiscent of the concept of free IP,” says Andy Jaros, VP of IP sales and marketing at Flex Logix.”However, open source is not free.Most companies, unless they want to invest significant resources in IP development, will license a pre-implemented RISC-V kernel from a number of IP vendors such as Open5 and Andes.This saves development time, validation, software development, etc., as well as warranties and compensation.”Having multiple companies develop competitive cores can spur innovation in implementation.”The real value of RISC-V is that it offers the possibility to compete with Arm, not because it’s open source,” Jaros added.”There are multiple RISC-V core vendors to provide choice and promote competition.But with Arm, you can only get an Arm kernel from Arm.”Another driver was the rapid increase in the number of cores, which made it less desirable to pay royalties on an instance basis.”People need processors throughout the design,” says Davidmann.”They want lots of small processors, and the existing licensing terms are quite difficult.It’s expensive, of course, but more importantly, it’s limited in its freedom to change it.I don’t believe risC-V is successful because it’s cheaper or less costly.If you only want to do the same things as an Arm core, then you should definitely buy an Arm core because it’s well proven.It’s very well designed.This is exactly what you want.The only reason to use RISC-V is because you want the freedom to change it and add your own things.”Even with all this, RISC-V probably wouldn’t have succeeded without the ecosystem that grew around it.”The open source community has developed key tools that are critical to making RISC-V-based processors ubiquitous, such as the chip technology process design suite, design validation suite, implementation tools, etc.,” said Cetola of RISC-V International.”This also makes it possible to democratize VLSI design, accelerating the development of designs through the development of higher-level design description languages and advanced open source automation tools, taking RISC-V functionality even further.With design tools and tool chains, RISC-V will soon become ubiquitous.”The OpenHW Group is one of the industry partnerships that have made this possible.It is developing the processor kernel and the peripheral IP that supports the kernel.In addition, it is deploying a full suite of tools to design and validate these kernels.”They do things differently,” says Davidmann.”One is that they give you the source so you can change it.More importantly, they also provide you with a validation environment, so if you make a change, you know it’s still valid.If someone just throws a core at you and you change some code, you might have broken something.You need a sophisticated validation environment to know that you’re not breaking it.This is what sets Open Source hardware apart, because they provide a complete validation environment.If you add a new directive, you know you haven’t broken the rest.I don’t think people will just take the Open Core and use it.It doesn’t make much sense.If you want to save money, you can do this.But what it allows you to do is take it and expand it, which is a very good place to start.That’s the key.You don’t start from scratch.”Can this open source momentum extend beyond processor cores?The processor is a small part of a complete SoC.It also requires a memory controller and memory interface, USB, PCI, and so on.These kernels do not provide differentiation for the product, and many people want these kernels to be open source.The problem is that these cores are very complex, and they contain analog parts that are often custom-designed and implemented for each foundry and process technology.While the controller can be built in open source, an argument can be made that not tightly integrating the digital and analog parts leads to poor product quality.LowRISC, an organization founded in the UK, initially wanted to build an open source equivalent to Raspberry Pi.Today, it develops hardware and software in a fully collaborative framework.This includes the RISC-V kernel, along with the software compilation infrastructure that supports it.More recently, Google created a specification and IP for silicon Trust roots.It has open-source the work and commissioned lowRISC to manage it.Part of the point here is that openness and transparency ultimately improve security and credibility, not the ability to modify specifications.Conclusion RISC-V enabled and promoted innovation.While free may be important to one part of the industry, the real key is freedom.This freedom brings together like-minded people, companies and organizations to break new ground.It is less likely to lead to breadth than depth.While the additional hardware modules may be open source, perhaps the most important gain will be the ability to quickly adopt the open specification of the processor and implement it.When engineers stopped developing their own custom processors, tools for processor development and validation disappeared because they offered little differentiation in the 1980s.Now that processors are again becoming highly differentiated, the industry is collaborating to develop the necessary tools.An unanswered question is whether they will be able to create open source tools faster than the EDA industry offers.