At this year's cadence annual event CDN LIVE INDIA 2017, the attention and most discussed subjects were artificial intelligence and self driving cars.
The next level of deep node chip design is no more driven by computing market, the latest 7 nm designs are now driven by mobile phone SoC chip design requirements. QUALCOMM's Venugopal Puvvada shared to the VLSI chip design engineers, the latest design trends in 7 nm. The role of artificial intelligence, machine learning and deep learning in designing advanced chips was briefly explained by him by giving some examples of image processing. He also highlighted the role of 5G in breaking the barriers which are holding further market growth in areas such as IIOT, mission-critical systems such as vehicle to vehicle communications, safety and security systems. While designing advanced low power chips, matching the cost and performance is the challenge. Machine learning looks to be a opportunity there.
To address the growing complexity of increased memory cells in the die area, Vinod Kariat of Cadence suggested to use Cadence latest tool called Legato, which was officially announced at the CDN live India 2017.
Since more and more semiconductors are finding its way into mission-critical systems such as automotives, healthcare, robots, and smart factories, the semiconductor chips need to be lot more reliable, rugged, secure and safe, suggests Vinod Kariat.
Jaswinder Ahuja of Cadence was very positive on the latest technologies been explored to increase the density of transistors down the 7 nm. He was very hopeful to see 1 nm chips someday in future. He shared the top massive growing markets; IOT, Deep learning, augmented reality/virtual reality, connected car, and cloud/datacenter to the audience.
On the 2nd day:
Cadence CEO Lip Bu Tan supposed to give a keynote but he could not make it this time. However there was interesting talk on AI by Dinakar Munagala, CEO of ThinCi. ThinCi is a VLSI startup focused on AI chips. He touched upon complete AI covering little basics to emerging application areas. He calls AI a paradigm shift. He also presented in a slide how ThinCi VCb-16 AI processor could offer 31x performance/Watt improvement over Nvidia Tesla P4.
Picture: Jaswinder presenting a momento to Dinakar