Veteran EDA investor Jim Hogan has a practical interest in system-on-chip (SoC) Realization, and he said so at the start of his keynote speech at the North American SystemC User Group (NASCUG) meeting today (Feb. 28, 2011). "My perspective is how I make money off this," he said. "I'm in business to sell companies or do IPOs. I'm looking for opportunities to make money."
The NASCUG meeting is co-located with the DVCon conference, also beginning Feb. 28 in San Jose, Calif. The NASCUG meeting also includes standards updates and user presentations.
In his speech, Hogan said he will focus his investments on SoC Realization as articulated in the EDA360 vision. "I like the Cadence lexicon," he said. (However, as Hogan's own speech shows, EDA360 is no longer just "Cadence lexicon" but a concept that is being discussed and applied across the industry).
SoCs Are Where the Money Is
Hogan began his talk by pointing to semiconductor industry growth of around 30% in 2010, with an expected 10% growth in 2011. But the money, he noted, is increasingly in SoCs rather than discrete ICs. One slide he showed suggested that SoCs have gross margins of 40-60%, compared to 10-20% for discrete ICs.
However, there are challenges. "The value chain is broken and the shift in value is going back to the systems guys," Hogan said. "The guys on the bottom need to figure out how to make money." Intel, he noted, is the best manufacturing company in the world, but Intel is not strong in applications. ARM is dominating the mobile market and moving into servers, and Intel will become an ASIC supplier, Hogan predicted.
While ARM processors have "won the war," Hogan noted that "the ARM market cap is nothing compared to their licensees. They're not really capturing the value. Where the value is being captured, it's at the system or SoC level. IP guys are getting some of that value and EDA is getting almost nothing."
A New Direction for EDA
Hogan briefly reviewed the EDA360 categorization of the marketplace in terms of Silicon Realization, SoC Realization, and System Realization. "Silicon Realization is what I call EDA classic," he said. "It's been around forever and the ecosystem is there, and it's generally not a place I want to go."
SoC Realization, however, "is where I want to focus my EDA investments over the next 5 years." Part of SoC Realization, he said, is validating IP quality and functionality. In this context, "adaptive" IP makes sense, he said; "I shouldn't be confined to a single processor architecture."
Hogan cited other aspects of SoC Realization, including:
- Design assembly and IP integration
- Design data management
- Power, performance, and area feasibility checks
- Debug and analysis tools
- Memory and memory controllers
- Bare metal software development
"You've got to be able to provide the SoC to the system customer with software capability up to the operating system," he said.
Right now, Hogan said, SoC Realization has a "nascent" ecosystem. Dozens of companies are offering IP sourcing, SoC creation, or SoC handoff, but nobody is offering all of these. That's good news for Hogan. "Whenever there's a fracturing of the supply chain, and the value changes or disaggregates, that means I can make money," he said.
"The whole shift in emphasis is going to be SoC Realization. It's an opportunity for us to make money," Hogan concluded.
For a closer look at Hogan's SoC Realization "shopping list," see today's EDA360 Insider blog post by Steve Liebson.