What the Large Hadron Collider scientists teach us about avoiding collisions

-October 14, 2013

Engineering teams at these five institutions knew from past experience, having worked on distributed projects that it was critical to have some kind of a managed handshake of data exchange between the sites. The LBNL team had done a thorough evaluation of commercial tools and selected SOS from ClioSoft. A master repository of all design data was hosted at LBNL and each site mirrored the data locally and synchronized as necessary.

The fact that SOS had seamless integration for Cadence, Synopsys, Mentor, and SpringSoft allowed each site to remain independent, although they all chose to stay on a common capture, place, and route system (namely the Virtuoso analog mixed-signal design environment from Cadence Design Systems). However, each site did use different tools for verification, especially for the digital side of the chip.

This team of 14 scientists (who, like us, also wear SoC and ASIC designers hats) across five different institutions and two continents had to coordinate chip development. Since each block was being developed on a different timeline, availability of each block layout was time-delayed to the availability of the schematic. The teams needed timed synchronization points to prevent them from stepping and tripping over each other. Sound Familiar? SOS’s core data repository and its inherent check-in/check-out features, overlaid with a basic methodology of notifications when blocks were available for integration, provided a simple, clean and elegant working environment and prevented engineers from colliding into each other and stepping on each other’s data. Presto, collision avoided!

The ATLAS detector was instrumental in one of the crowning achievements of this LHC project, namely the confirmation of the existence of the Higgs boson particle in 2012. When the LHC runs again in 2015 with still more ambitious plans, ATLAS will include a brand new measurement layer closer to the collisions than ever before using the FEI4 chip.

In their summary, the designers found the following features of ClioSoft’s SOS to be the most valuable, and I quote:
  • World-wide access to design data in real time
  • Revision management (backup, snapshots, versions)
  • Graphic diff tool to view changes in schematics and layouts
  • Simple and flexible administration, mostly via GUI
  • Very robust (We never lost data.) 
  • Very well supported (Help desk responds within the hour.) 
  • Fast (using caching, data access about the same as local access) 
  • Flexible: All data types can be shared in repository (design databases for both analog and digital parts, documents, etc.) 
  • Only our own design work is shared. Third-party IP, such as design kits and standard cell libraries, were obtained directly by each site. Repositories can link to local libraries/kits in a seamless way. 
  • Low-cost educational licenses2

Next steps


As these teams look ahead, they clearly see that their projects are getting larger and more complicated. They see the value of collaboration and will continue to leverage this even further. According to Vladimir Zivkovic, a design automation, verification, and test engineer at the NIKHEF institute, SOS was instrumental in improved team productivity by efficient data and deliverable exchange with the other sites. While in the last go-around the development teams for the different detectors worked in isolation, and possibly in competition, Garcia-Sciveres feels that they will need to collaborate more in the future and leverage each other’s development to achieve their end goals. As several of these research groups are already on common platforms like SOS and Virtuoso, they should be able to easily merge repositories and share their development across all sites.

Conclusion

The LHC project set out to uncover a key missing piece of the puzzle of how all matter has mass, which sheds light on how the universe was formed. It did this by pumping up particles to high energies and smashing them into one another in order to transform that energy into mass in the form of new particles. With the identification of the Higgs boson particle, one might even say that it was a smashing success!

But this is just the beginning. Some of the data is simply mind-boggling. The ATLAS detector contains about 150 million sensor elements collecting data 40 million times per second. The data flow from all the LHC experiments was about 700 MBps, which works out to be about 15 petabytes a year. This enormous amount of data will be accessed and analyzed by thousands of scientists from every single continent except Antarctica. The LHC produces about 600 million collisions per second. But after filtering, they expect only about 100 collisions of interest per second. Physicists predicted that they would have enough statistics only after two to three years of data collection. Talk about a needle in a haystack.1

Starting in 2013, the LHC will be shut down to give the engineers and scientists the opportunity to upgrade the machine. When the LHC turns back on in 2015, using the chip in the new detectors we read about earlier, scientists expect it to be run close to its design energy and expect to study about 200 quadrillion-particle collisions.3

Scientists expect to observe the Higgs particle’s properties, its spin, parity, and detailed interactions, which might either confirm or deny that the particle conforms to the predicted standard model. Alternative models like supersymmetry theorize other new particles might exist, including the presence of dark matter or even that the Higgs particle exists in more than our three dimensions of space.

And to think that these cutting-edge discoveries are made possible only with your neighborhood EDA tools. Maybe they are worth a lot more than you think!

About the author
Kumar Venkatramani is president and CEO of Silicon Ideas, a consulting company. His 25 years of experience blend both hardware and software design disciplines. In senior technical and management roles at Cadence Design Systems, he architected and nurtured electronic system level design tools and technologies, and subsequently developed platform-based design methodologies for which fifteen patents were awarded. He has published several papers, presented technical seminars, and co-authored standards in the areas of IP Reuse. He received his BSEE from National Institute of Technology (Tiruchirapalli, India) and his MSEE from Columbia University.

References

  1. Lefevre, Christiane, "LHC: The Guide (English Version)," CERN, 2009, Web Aug 23 2013.
  2. Garcia-Sciveres, Maurice, "The New ATLAS Pixel Chip: FEI4," CERN N.p., March 8, 2011. Web Aug 23 2013.
  3. WennersHerron, Ashley, and Kathryn Jepsen. "What's next for the Large Hadron Collider," Symmetry Magazine, N.p., Feb 4 2013, Web Aug 23 2013.

Loading comments...

Write a Comment

To comment please Log In

FEATURED RESOURCES