DAC 51 Morning Meanderings 06_04
The DAC floor only looked a little bit like ‘Shaun of the Dead’ this morning. Folks were stumbling in after the traditional late Tuesday night parties. By sheer random luck, I scheduled my first meeting this morning at 10:00, and got a much needed snooze. I’m faring much better than the gentleman we encountered last night, who after staggering up, informed us that he’d walked down from Edmonton, (yes, Canada) , and inquired as to the location of Folsom St. where he’d been assured that the houses of ill-repute were beyond compare. I hope you’re still alive little guy! OK, that’s enough DAC color commentary for today. Hang in everybody, we’re almost there. It’s not over yet though, for example :
If you have an all-inclusive DAC registration, don’t go home yet! Training Thursday is well… tomorrow. There’s a full list of training classes available. There are tracks on SystemVerilog, ARM Engineering and Embedded Systems among others. If you don’t have an all-inclusive registration, you can still get in on the action. Here’s how much it will cost.
Wow! I wrote about it, but they’ve actually done it! If you’re interested in managing, browsing, and taking action using the metrics from your backend design tools, go visit Dassault Systemes. Expect to see more on this from me soon, but for now, head over and ask to see the PinPoint tool. Among other features, the tool overlays output from timing, I-R drop, and layout tools onto a gmaps style diagram of the chip. Users can view historical reports of tool outputs and track progress towards physical design convergence graphically and, at least as importantly, quantitatively. The tool facilitates a coordinated work flow between disciplines, teams, disparate locations. It also gives managers an objective means of tracking progress and data with which they can constructively influence the flow.
Space Codesign’s tools enable system architects to evaluate software vs. hardware, power vs. performance, and other high level chip design scenarios with a graphical drag and drop interface. The company aims to ease the adoption of high level ESL techniques by insulating the user from the scarier aspects of SystemC TLM development.
Synopsys, 62626, and Automotive Engineering
- The automotive design problem can be divided into three areas: car control, infotainment, and autonomous driving.
- While Ethernet doesn’t funnel any of the vehicle’s mission critical control data yet, more intra-car communications are winding up there prompting a change in the industry terminology from ‘entertainment bus’ to ‘infotainment bus’.
- Fully automated driving systems are expected in the 2025 to 2030 time range. Expectations are that most of the remaining hurdles will be regulatory rather than technical.
- Expect vehicle to vehicle and vehicle to infrastructure communications in the future. These will participate in the autonomous driving systems.
- One of the keys to success in the automotive segment is going to be starting hardware and software co-verification more quickly. In this case, the hardware includes an entire car, as a peripheral. High level system simulation that can include mechanical and electrical models is crucial.
Of Interest from the IoT Panel:
Gary Smith: IoT isn’t a market; it’s a Wall Street buzz word.
Bernard Murphy: The number of edges in the IoT will be about the number of cells that are in a new born baby. There are interesting analogies to be drawn between biological security and IoT security. You need local defenses such as signaling, or perhaps self-destruct capabilities to protect the rest of the network.
Randy Smith: Time to market is much more critical now. There are companies in the mobile phones space putting in sensors that don’t have a use yet, but that are expected to in future phones.