761 Days

Clouds over San Francisco

761 days.

That’s 2 years, 1 month, and 3 days.

761 days ago, I hosted a small group of interested EDA folks, journalists, and bloggers in a small room in the Doubletree hotel after one of the evenings after DVCon.

Most of the discussion that year was around OVM and VMM and which methodology was going to win out and which was really open and which simulator supported more of the System-Verilog language. Well, all that is put to bed. This year at DVCon, 733 days later, we all sang Kumbaya as we sat around and our hearts were warmed by the UVM campfire.

But, back to that small group that I hosted 761 days ago. Those that attended this conclave had shrugged off all the OVM and VMM hoopla and decided to come hear this strange discussion about Cloud Computing and SaaS for EDA tools. Some, no doubt, thought there was going to be free booze served, and they were certainly disappointed. Those that stayed, however, heard a fiery discussion between individuals who were either visionaries or lunatics. For many, this was the first time they had heard the term cloud computing explained, and their heads spun as they tried to imagine what, if anything would come of it for the EDA industry.

Over the 761 days since, the voices speaking of cloud computing for EDA, once very soft, grew slowly in volume. All the reasons that it would not work were thrown about like arrows, and those objections continue. But slowly, over time, the voices in support of this model have grown to the point where the question no longer was “if” but “when”.

761 days, that’s when.

Yesterday, to the shock of many at SNUG San Jose, including many in attendence from Synopsys, Aart DeGeus personally answered the question asked 761 days earlier. Indeed, those individuals gathered in that small room at the Doubletree were visionaries, not lunatics.

There are many reasons why Synopsys should not be offering its tools on the cloud via SaaS:

  • Customers will never let their precious proprietary data off-site
  • It will cannibalize longer term license sales
  • The internet connection is too slow and unreliable
  • There’s too much data to transfer
  • The cloud is not secure
  • It’s more expensive
  • It just won’t work

But, as it turns out, there are better reasons to do it:

  • Customers want it

Sure, there are some other reasons. The opportunity to increase revenue by selling higher priced short-term pay-as-you-go licenses. Taking advantage of the parallelism inherent in the cloud. Serving a new customer base that has very peaky needs.

But in the end, Aart did what he does best. He put on his future vision goggles, gazed into the future, saw that the cloud was inevitable, and decided that Synopsys should lead and not follow.

761 days. Now the race is on.

1 Star2 Stars3 Stars4 Stars5 Stars (2 votes, average: 3 out of 5)
Loading ... Loading ...

Tags: , , ,

3 Responses to “761 Days”

  1. System-Level Design » Blog Archive » Blog Review: April 6 Says:

    […] Harry Gries, aka The ASIC Guy, looks at Synopsys’ push into the cloud and why this is about to change the very fabric of EDA. We’re still waiting for the other big shoes to drop. […]

  2. Brad Pierce Says:

    According to Dan Sullivan, Babs Smith, and Michel Néray, “People in perfection-based corporate cultures are always on the defensive, and they seem to be most interested in rationalizing why initiatives won’t work. People in progress-based corporate cultures are more interested in assisting other to achieve extraordinary goals, which also creates opportunities for themselves.”

    http://bradpierce.wordpress.com/2010/04/11/forget-about-perfection-and-focus-on-progress/

  3. Brad Pierce Says:

    Regarding the “skinny straw” problem, you list the objection “The internet connection is too slow and unreliable”.

    Turnaround time is important, but EDA customers are also worried about the high network traffic fees charged by external cloud providers.

Leave a Reply