Virtual prototypes technology enables the creation of software models that can fully represent the functionality of an embedded system. They combine high-speed processor instruction-set simulators and abstract, transaction-level models (TLM) of the hardware building blocks to create a high-performance software development environment. This promotes continuous, pre-silicon hardware/software integration, instead of waiting to bring hardware and software together near the end of the silicon development.
Virtual prototypes improve software development productivity by allowing unlimited observability and controllability of the target hardware, and through predictable and repeatable execution of debug scenarios. Virtual prototypes allow developers to boot operating systems, create applications and multimedia codecs and develop low-level drivers.
Automotive Embedded Systems Software Benefits from Virtual Prototyping
Automotive design has grown to include the design of electronic control units (ECUs) in addition to the mechanical design. This industry is challenged by an increasing need to reduce design cycle times and cost while satisfying global regulations on passenger safety, emissions and fuel economy. The embedded software component of ECU design continues to grow even faster than the hardware, and automotive electronics companies are experiencing high growth in their ranks of software engineers. Automotive software and hardware are focused on reliability and quality, both for safety-critical functionality where errors can be life threatening and for other subsystems where reliability is important to control warranty costs.
Automotive manufacturers use Synopsys’ virtual prototypes as a specification communication medium and for architectural development and evaluation, while Tier 1 automotive suppliers focus on using Synopsys virtual prototype for embedded software development.
Synopsys VP's facilitate quantitative architectural analysis and evaluation. Increasingly, automotive electronics development starts with formal models defined using MATLAB/Simulink® and universal modeling language (UML). These models are used to evaluate candidate architectures for networked ECUs and to investigate the performance required for each ECU and the network traffic that the ECU partition generates. Using Synopsys technology a virtual prototype can be created for each ECU including specific processor, cache size, clock rate and peripheral set. This virtual prototype becomes the scaffold for the development of the ECU software and provides developers with greater visibility into operation and performance than can be obtained with a bench setup and actual hardware.
The virtual prototype technology enables quantitative analysis of ECU design candidates. MATLAB/Simulink and UML models can be linked into networks to perform a simulation at an abstract level. As refinements are added in the form of architectures for ECUs and protocols for networks, the entire analysis can be repeated at an increasingly detailed level. At the lowest level, software development on a virtual prototype of an ECU (or a subsystem of networked ECUs) provides very high levels of error checking and visibility into the internals of the hardware design.
Embedded System Complexity In Telecom and Wireless
Synopsys’ solutions for embedded systems design are used extensively in the wireless and telecommunications industries. Synopsys’ virtual prototypes are unique in their ultra-high performance and very high timing accuracy. This combination of speed and accuracy has special value to our customers.
Synopsys Virtual Prototypes Cut Development Cycles
Wireless handset development is a system-level problem, entailing development of sophisticated hardware and a significant and growing body of software to perform the many functions demanded by the market. With high volumes and short product lifecycles, bringing a product to market early is the key to profitability.
The growing complexity of embedded designs requires new design methods and automation. A complex multimedia handset typically consists of a digital signal processing subsystem, a subsystem with a general-purpose microprocessor and often a third processor to handle multimedia functions. These subsystems are connected by complex multi-level on-chip buses to which are connected many peripheral models such as timers, the radio interface and a subscriber identity module (SIM) card interface.
Synopsys’ electronic system-level (ESL) design tools, models and virtual prototyping technology enable the design and development of these increasingly complex wireless devices faster and with increased confidence in their functionality and performance.
Evaluate Performance and Energy Usage
Using the Synopsys Metrix tool enables the evaluation of algorithms on the basis not just of performance but also of the amount of energy used. For example, accurate predictions of talk time or the number of digital photographs that can be processed can be made for a given battery capacity.
The gating factor in reaching the marketplace is very often the completion of the embedded software development. Synopsys’ cycle-accurate and high-performance virtual prototyping solutions allow developers to design and optimize their system-level architecture and to develop hardware and software concurrently, improving quality and reducing time to market. One Synopsys customer in the handset industry recently saved nine months out of a two-year development cycle — nearly the equivalent of an entire product generation.
Superior Systems Engineering Tools for Embedded Systems Design
Synopsys provides manufacturers of consumer electronics such as cameras and printers with the systems engineering tools required to get to market sooner with higher quality products that cost less to build. Using Synopsys’ tools and virtual prototyping technologies, consumer electronic designers rapidly explore multiple design alternatives without having to build costly and time-consuming physical prototypes.
Many areas of consumer electronics, such as digital cameras, MP4 players, printers and digital radios, contain increasingly large amounts of embedded software interacting with complex hardware. Much of this software operates in an environment of inflexible real-time constraints. All consumer electronics products are developed in an environment of intense cost sensitivity where design challenges cannot afford to be solved by over-engineering the SoC.
Optimal Embedded Systems Design Increases Profit Margins
Using Synopsys embedded systems design tools and models to create an optimal architecture for a SoC can make a big difference to the profit margin for the chip and the system built around it. For example, a cache memory that is unnecessarily large results in wasted silicon area and thus increased cost, while a cache memory that is too small results in lower performance than optimum for the application. Synopsys’ virtual prototypes enable architectural exploration that produces quantitative results on aspects such as power consumption, performance and cost.
In most consumer products, aspects of the system performance are visible to the end purchaser as capabilities of the product, such as the rate at which a digital camera can take photos or the ability of an MP4 player to play music continuously under all circumstances. The Synopsys suite of systems engineering tools and models allows developers to evaluate architectures and then run the hardware and software together to ensure that the product will have the right performance and features to meet market requirements.
A growing number of consumer products are battery powered. The power measurement capabilities of Synopsys’ virtual prototype solutions means that estimating the actual power usage for features such as the number of photographs per recharge can be minimized before the real chip is available.
Using Synopsys’ virtual prototype, embedded software development can begin even before the detailed hardware design has been finalized. Porting of operating systems and communication stacks, together with the development of both driver and application code, can be overlapped with the hardware design and development. This concurrent design and development means that product cycle times and time to market for new and iterated products are substantially reduced.