Earlier this year, Iyer replaced longtime ip-industry executive Adam
Kablanian as CEO, a situation he described as part of the company's
"It was great having Adam," he said. "We brought Adam on to ...
launch the company, to brand the company and to set up a structure."
The Memoir team in 2012 with former CEO Adam Kablanian
(third from left) and cofounder Sundar Iyer (far left, standing).
But as the company has put more of an emphasis on the system level,
it "was felt that I was the right choice as CEO to move forward."
Iyar noted that the good part about being an ip company is that
product development can be quick and iterative.
The bad news?
"We're kind of like pond scum. We are at the bottom of the food
chain," he said.
Iyer said the way to improve Memoir's position on the food chain is
to build ip "intelligently."
"There are tons of ip companies that fail and ips
that fail because it's a whole purge of code that eventually
breaks, has a corner case that takes eons to verify. We've got
to make sure that our platform is built in a way that each new
ip or feature that we add can be built and verified very fast."
In addition, the platform approach needs "to start with an ethos of
automation bottom up, both in writing code and in verifying code."
Lastly the company needs to be mindful of how to scale effectively,
"If we start building ips that are one offs, we'll fail. We're got
to figure out the right level in the stack, this is the clean break
point, these are clean APIs that architects can innovate on," he
"If you do not id right way to build ip, we can never scale. It will
either be too simple that no one will want to buy it from us or it
will be too application specific," he added.
He's careful to position Memoir -- 28 people strong here and in
India and Armenia -- far from the physical memory space where
companies like Synopsys and an ARM dominate. Instead, Memoir
complements physical memories by "adding intelligence on top of the
them," Iyer said.
Memoir, capitalized for $5 million by Lightspeed Ventures, is
nearing break-even, Iyer said.
Roadman, I haven't been able to track down Adam, but I'm still working on that for a follow-up. Clearly Sundar positions his company in as positive a light as possible. Can't fault him for that, and execs aren't going to answer questions they don't want to.
Anything you can share? These guys are worth circling back with in the second half of the year to see how product-development is proceeding and where they are on break-even goals.
Did you get comments from Adam on this article? on why he left. I am seeing lot of FUD by Sundar and team here. Let me put things in perspective bit they have IBM Power socket for caching IP, may be a Hauwei, Cisco probably told them to go pound sand, as they bought their prior stuff Nemo. So they have 22 people out in foreign countries, at low cost for great window dressing. This makes it great Life Style Company, not a future breakthrough corp that someone will want for 10X bucks, like there VC Friends would want as well. Go ask lightspeed why they have not funded this company another $20M bucks, or any other VC's in the valley. May be you will some clarity, unless you are part of this FUD campaign.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.