For quite a while, since the rise of smartphones in the late 2000s, the computer processors market has been dominated by ARM central processing units (CPUs) based on the reduced instruction set computer (RISC) architecture. Recently, however, a strong competitor has emerged with a considerably different approach towards the CPU architecture in microprocessors, mobile systems and microcontrollers. The name of this potential ARM killer is RISC-V (pronounced as “risk-five”). 

Over the last couple years, the debate regarding the competition between ARM and RISC-V has been getting more and more vibrant. 

Will RISC-V ultimately replace ARM as the top CPU specification or will both technologies coexist? Let’s take a closer look at these two computer processor architectures, their technical specifications and how they are different from each other. 

What is ARM? 

ARM (originally known as Acorn RISC Machine, ARM stands for Advanced RISC Machines) is a family of RISC instruction set architectures for computer processors, available for various computing devices and environments. 

The ARM CPU architecture is developed by the Arm Ltd company, which licenses the architectures to other companies, allowing them to design their own products that incorporate different components, including interfaces and memory. 

There have been a number of generations of ARM architecture. The original version, ARM1, was introduced in 1985, almost 40 years ago. First application for ARM processors was as an additional second processor for the BBC Micro, providing support to speed up the simulation software. ARM1 used 32-bit internal structure but also had 26-bit address space, limiting it to 64 MB of main memory. This limitation was removed in ARM 3.

ARM 8-A, released in 2011, received the support for 64-bit address space and 64-bit arithmetic. 

ARM processors quickly gained popularity due to their low power consumption, lower costs compared to available alternatives, and minimal heat generation. 

Even though ARM CPUs were widely used since the initial release of this architecture, they really came to power in the late 2000s, upon the release of the first smartphones. Being the best CPU choice for portable devices due to light weight and low power consumption, ARM processors are preferred by the manufacturers of smartphones, tablets and laptops. For the same reasons, ARMs are also widely used in embedded systems. 

According to the official data, more than 200 bln ARM chips have been produced around the world as of 2021. 

What is RISC? 

Since we already mentioned the RISC a number of times, a few words about it need to be said as well. 

RISC is a technology designed to simplify the individual instructions provided to the computer to perform certain tasks. The difference between RISC and CISC (a complex instruction set computer) is that RISC architecture typically requires more instructions provided to a computer in order to complete tasks as individual instructions in RISC are written in simpler code. 

One of the key concepts of RISC computers is that every instruction performs only one function during single CPU cycle. 

What is RISC-V?

RISC-V is basically the fifth generation of the RISC architecture, provided as an open standard instruction set architecture (ISA) based on the RISC standard principles. Unlike the majority of other ISA designs, it is provided under the open source license, so it’s free to use for all the computer chip producers.

The RISC-V specification defines both 32-bit and 64-bit address space options, and additionally includes a description of a 128-bit flat address space variant. 

The RISC-V is a load–store architecture, using IEEE 754 floating-point instructions. RISC-V ISA also includes instruction bit field locations as a way to simplify the use of multiplexers in CPUs. 

Started with a goal to create a practical open source ISA that will be easily deployable in various hardware and software designs, including embedded systems, the RISC-V ISA is a continuation of a long history of CPUs architecture design projects developed at the University of California, Berkeley, since the late 1980s.

History of the RISC-V specification development

The project to develop RISC-V specification was originally started in 2010 by the University of California experts with an intent to create a practicable instruction set that will be available for practical use in various CPUs manufacturing. 

Dr. Krste Asanović, a professor of computer science at UC Berkeley, was an author of the project to develop RISC-V. Eventually, Dr David Patterson, another UC Berkeley professor and one of the creators of the original RISC chips back in the early 1990s, joined the project.

As any ISA needs to be stable for commercial use, the RISC-V Foundation was formed in 2015 with a goal to develop, maintain and publish the intellectual property related to the RISC-V specification. The original authors of the project at UC Berkeley have transferred all the rights to this non-profit corporation controlled by its members.

Currently, the RISC-V Foundation comprises over 325 members, including representatives from companies such as Google, NVIDIA, Microsemi, Western Digital. The RISC-V Foundation members participate in the development of the RISC-V ISA specification and related projects. 

In 2019, due to the U.S. trade regulations concerns as the main reason, the RISC-V Foundation relocated to Switzerland. In 2020, the organization was renamed as RISC-V International, becoming a Switzerland-registered nonprofit business association.

Today, the RISC-V International publishes all the documentation and specifications related to RISC-V designs, which remains open source and available for everyone to use free of charge. However, only the members of RISC-V International can vote to approve any changes to RISC-V specifications. 

ARM vs RISC-V Comparison 

Here’s a table comparing technical specifications of ARM and RISC-V. 

Final thoughts. ARM vs RISC-V: Which one to choose? 

As you can probably tell from the comparison chart above, there is no simple answer to this question. 

In many ways, right now, ARM-based CPUs are still a better option, mainly due to much longer lifecycle and the fact that ARM Ltd has invested billions of dollars into this specification over the years. ARM processors have a huge market share, being used in the majority of smartphones, as well as laptops and even PCs that are choosing ARM instead of x86 architecture-based designs. 

We could say, however, that RISC-Vs are the future and a very strong contender to the throne of the most used computer processors architecture. RISC-V can provide better performance using a minimum amount of power. The fact that RISC-V is open source and free to use by any processor manufacturers is also a huge advantage.

Some manufacturers, such as Western Digital, for example, have already started implementing the RISC-Vs in their microcontrollers attached to RAMs and SSDs. 

RISC-V is also getting increasingly popular in IoT devices and embedded systems of various kinds, due to its highly scalable nature. But it will undoubtedly take several years for industry players to transition to using RISC-V instead of ARM-based designs. 

The Tauro Technologies’ team of electronic engineers and designers has a proven track record of successfully designing custom hardware for various kinds of products in multiple technology fields. Drawing on the specific needs of our clients, we select and apply various engineering methods to electronic product development and manufacturing in order to achieve the desired result. Utilizing our in-house PCB assembly and debug expertise, we are able to build and evaluate your prototypes before high-volume manufacturing, rapidly and cost-efficiently. 

Interested to know more? Get in touch with us for details.

The post RISC-V vs ARM. Which One To Choose? appeared first on Tauro Technologies.

One of the main challenges designers face when designing embedded systems is the need to meet rigorous and precise requirements for performance, efficiency, and reliability of the system. In order to deliver an embedded system serving the specific purposes of an application, experts in this technology field typically have to come up with complex combinations of hardware and software that are unique to the system requirements. 

The high level of reliability and performance of an embedded system is achieved by multiple components and modules working together and the operating system naturally plays a fundamentally important role in these processes. Resource management, the ability to multitask, divide computing resources among different tasks and prioritise them is extremely important for an embedded OS. 

This is the reason why the developers of modern embedded systems use real-time operating systems (RTOS) as a foundational software layer, powering these solutions. 

What is a real-time operating system (RTOS)?

RTOS is a type of operating system specifically designed for real-time applications: to be able to process various events and data with predictable latency. Real-time operating systems are designed for mission critical applications where missing certain deadlines may result in major consequences. 

Real-Time Operating Systems deployed in embedded products are typically built to be extremely small and include only the most essential components and features needed (often no UI at all). The goal of a typical RTOS solution is to enable timely execution of tasks, which requires the system to have flexible control and resource management capabilities. Specifically, an RTOS needs to be able to rapidly switch between tasks, handle multiple events concurrently, schedule and maintain the response to events within predictable time limits, send data across the network, record data, etc.

RTOS vs. GPOS. What’s the difference? 

As you must have noticed, real-time operating systems are very different from general purpose operating systems like Windows or Linux that are commonly used in consumer electronics and commercial work stations.

Standard operating systems typically need a lot more disk space and computing resources, serving as the core that controls all the processes, from execution of applications and memory storage to user interface, feedback from hardware, and other system operations, as well as all communication between user and the device. 

But the most crucial difference between RTOS and GPOS is the time it takes for the OS to react to certain events, which essentially goes down to the general approach to computing and hardware resource management. RTOS, as we just learnt, are designed to be event-driven and preemptive, providing highly deterministic response, which requires RTOS to use a sophisticated scheduling algorithm. 

Standard operations systems, on the other hand, typically utilize a non-deterministic approach to process execution, with no guarantees of task completion and no precise scheduling of process completion. 

Applications of RTOS

Due to the above-described distinctive features, real-time operating systems are typically used in embedded systems that must register and process large numbers of events, mostly initiated externally, in short time intervals within specified deadlines. This is why RTOS is the preferred operating system for flight control, military, industrial control, manufacturing automation, and in many other fields and industries where performance and reliability play a crucial role. 

Here are some examples of RTOS applications:

• Defense application systems
• Command and control systems
• Medical critical care systems
• Air traffic control
• Stock exchange control systems
• Networked multimedia systems
• Automobile safety control
• Online transaction systems
• VoIP communication 
• Real-time simulations

Types of RTOS

Real-time operating systems can be divided into three categories based on how rigorous their approach to response and execution of tasks is.

Hard RTOS

Hard RTOS approach the deadline with highest strictness. Such a system must guarantee the execution of a task on specified time and within the assigned time interval. Hard RTOS are used in systems that execute most important and crucial operations, such as air traffic control or critical care systems where delayed response can have serious consequences. 

Firm RTOS

Firm RTOS are also designed to follow deadlines, but in applications where missing deadlines on several processes is acceptable and won’t lead to critical effects. Missing a large number of deadlines still cannot happen, however. The most typical applications for RTOS are multimedia control systems, video conferencing systems, tracking and monitoring solutions, etc. 

Soft RTOS 

Soft RTOS, as its name entails, utilize a more flexible approach to the time limit. Missing deadlines for tasks is not as critical for soft RTOS and doesn’t lead to system failure like it would in a hard RTOS. It is also typical for real-time operating systems of this type to be more focused on prioritization and multitasking in order to identify and process highest priority tasks rapidly, while less important ones can take more time.

Benefits of using RTOS 

Now you can probably see why RTOS is essential for many embedded systems: small size, preemptive scheduling, multi-tasking, and real-time context-switching are critical features that allow embedded systems to provide the required performance and predictability. 

Let’s take a closer look at the main advantages of real-time operating systems.

Maximized performance

Naturally, the ability to provide high performance, maximizing the utilization of available resources, would be the main strength of RTOS solutions. Additionally, it is also important to mention that embedded systems powered by RTOS are typically designed to enable 24/7 operations with no interruptions caused by errors, updates or other factors that may disrupt the performance of a regular OS.

Advanced task prioritization and scheduling

Another powerful feature of RTOS is advanced priority-based scheduling that allows such systems to identify critical tasks and switch between processes. The response time in RTOS is measured in microseconds, and modern-day RTOS-based solutions are expected to respond within 3 microseconds. 

Small size and ease of maintenance

Due to small size and event-driven architecture, RTOS solutions normally incorporate fewer components and require minimal number of interconnections between them, which makes it much easier to maintain such a system. This is another huge benefit that allows organizations to minimize the technical support costs and use an embedded system for many years. 

Weaknesses of RTOS 

Since we covered the most important strengths of RTOS, it makes sense to touch on the weaknesses and disadvantages of this technology. Naturally, all drawbacks of RTOS stem from the distinctive features of this OS type described above. 

RTOS can handle only a small number of tasks

Real-time systems are designed to focus on key applications. Having limited computing resources, they can effectively handle only a small number of tasks. 

Real-time processing requires a lot of resources

The main reason why RTOS can’t handle many tasks is that real-time response to critical events requires a considerable amount of system resources. The augmentation of an RTOS-based embedded system with additional hardware components, however, is difficult and expensive. 

Difficult to design and implement

Due to all these distinctions from mainstream operating systems, implementing an embedded RTOS-based solution normally requires the field experts with years of hands-on experience and a considerable time. The architecture and algorithms powering such systems can be not only very complex, but also require high degree of customization depending on the application.  

Summary

Real-time operating systems play a fundamentally important role in modern-day embedded solutions design, enabling almost instant response and exceptional fail safety. But choosing an RTOS and integrating it with other system components is not a trivial task. A project to design and implement an embedded system can not only be difficult and time-consuming, but also very expensive and risky, especially for organizations lacking qualified internal talent to deliver such a system. 

The Tauro Technologies’ engineering team has a proven track record of successfully designing RTOS-based embedded systems across multiple industries. Utilizing our in-house expertise in this field, we are able to deliver custom-made solutions designed to match specific needs of each client, risk-free and in a short time period. 

Interested to know more? Get in touch with us for details.

The post Real-Time Operating Systems and Applications appeared first on Tauro Technologies.