feature article
Subscribe Now

EDA Taking Its Toll

Cadence’s Dan Holden pointed me to his blog entry that pointed to a YouTube video that Cadence’s CMO John Bruggeman made regarding the iPad’s new role as gatekeeper to content. He discusses this in terms of what he calls “tolltakers” and relates it to the need for the EDA industry to find new ways to get paid for the value they provide. This got me thinking about what it means to be a tolltaker. These thoughts don’t reflect specific support or repudiation of John’s comments, but rather spring from those comments and go on their own from there.

It’s no secret that the EDA industry has struggled to find ways to be paid fair value for the capabilities it feels it brings to semiconductor companies. Here’s how I could summarize an abstracted annual EDA negotiation:

Semi guy: OK, thanks for coming in today. We’ve set our EDA budget for this year, and, just like the last 20 years, it’s pretty much the same as last year, although we did have to cut a bit. And, because there are so many more tools that we need now, your share of this overall budget unfortunately has to come down. So… what can you do for us?

EDA guy: But the technology we’re delivering has so much more content than it did last year or five years ago, and you’re making so much more money based on the capabilities we give you. We feel that the value of our deal should increase commensurately with that value.

Semi guy: Yeah, well, cry me a river. The budget is set, it’s out of my hands. You want in or not?

[EDA guy grumbles and caves]

By John’s picture, however, the iPad shows a way to insert oneself into the value chain and collect a toll based on traffic. Very different model, not immediately obvious how to apply to the EDA world, but interesting to explore.

Here’s a caution, however: people (Americans, anyway) have a very mixed view of tolltakers in general. In fact, I can identify three different categories of activities that could, in the abstract, be considered tolltaking; not all are good, and few are seen as good.

The obvious kind of tolltaker is exemplified by the literal taker of tolls along roads and bridges. These are common in some parts of the world – the East Coast of the US, Japan – and less common elsewhere. Speaking from US experience, where they’re common, people grumble about highway robbery. When visitors from non-toll areas encounter tolls, they react with outright outrage. Never mind that, in the absence of tolls, other taxes would be higher.

This somewhat reflects what would appear to be an American sense of entitlement to all kinds of things like roads for free. (OK, we’ll make allowances if someone else pays for them.) So, at least judging from those with access to microphones, we whine about having to pay tolls and we whine about having to pay taxes. And we whine if the road isn’t there or isn’t in good shape. (This may reflect, to some extent, the American heritage of having a Bill of Rights but not a Bill of Responsibilities.)

But, looking at the big picture, whether you like it or not, roads cost money to build and maintain, and, by taxes or by tolls, the revenue needs to be raised. So here the tolls reflect value that would not otherwise exist. You’re getting something for your money.

(For those of you not in the US, rest assured that last statement has assuredly resulted in a few shouts of disgust, disbelief, and contempt from the whiney contingent. That’s how they roll. And I’m not going to divert this into a no-win discussion about whether or not monies collected are well spent. Totally different matter.)

A second type of tolltaker is more subtle; it reflects more of a gray zone. This can be shown by the example of ATM charges. If you just take a snapshot today, you can easily argue that ATMs provide convenience, and that convenience is worth something. Someone paid to put that ATM in your local Quickie-Mart and deserves to be rewarded for the initiative. Or, you dared to bank with someone else, and so you deserve to be punished for having the temerity to use the ATM of a bank you rejected. After all, it costs them money to maintain the ATM.

So, at first blush, it looks like, here again, the ATM fee is simply a toll for value added. But let’s take a trip in the WayBack Machine to a time long ago, far longer than our memories last (much to the glee of banks) to the dawn of the 1980s when ATMs were first rolled out. Then they were a boon to banks, because now people could use a machine rather than an expensive bank teller to get their cash. Banks started installing ATMs as quickly as possible, running TV ads extolling their convenience and practically begging people to use them so that they could save tons of money by laying off tons of tellers.

It was only after this became the norm and our memories of this time went blank that the positioning went from, “please save us money by using our ATMs” to “hey, these ATMs cost money to support, so you need to pay us.” So, in this case, are we paying tolls for value received, or are we paying for the privilege of saving the bank money while we access our own money? Less clear.

The third category brings us back to black-and-white. Many of us grew up with the story of the Billy Goats Gruff, where they can’t cross the bridge without paying off the troll under the bridge. It wasn’t the troll’s bridge (except that he decided to live there); he didn’t build the bridge, so it’s not like he’s getting a return on an investment; he wasn’t maintaining the bridge. No, he just managed to put himself there in between where the goats were and where they wanted to be, and they had no choice but to knuckle under in order to get across. Why were they paying a toll? Because, in accordance with the classic parental chestnut, “I said so.” (That violence, in the end, provided justice is not the point.)

It could be debated how much of Wall Street belongs in this category. When some individuals take home more in personal income than the total annual profits of small innovative tech companies even in a year when they contributed to the destruction of the economy, you can reasonably ask, what value are we getting for that? “Liquidity?” “Markets?” Is that real or hyperbabble? Are we simply taking a chunk of our economy and handing it to Wall Street for little discernable value in return?

Looking at another example, if broadband providers are allowed to charge you more (or provide slower access) to view internet content that they don’t own, they won’t be providing value, they won’t be recouping cost. They’ll simply be controlling your behavior with punitive economic “incentives.” Contrary to the yips and squeals of some of those that oppose net neutrality, this type of tolltaking squelches innovation because anyone new or small or creative has a higher hurdle than the incumbents with the sweet deals.

A more sinister version of this kind of tolltaking is protection money. This is even worse; not only is it money for no value, but it’s money to protect against intentional damage.

So how does this relate to the business of trying to get more value for EDA? Every time a business or industry tries to figure out how to change their economic model, they take a look at the revenue stack and ask the question, “How can I get a [bigger] piece of that?” And too often there’s less thought about how they can impact the value stack.

Again, we can look to banks for an example. Sorry to pick on them so much, but, hey, they have pretty much been the bad boys in the economy, first wrecking it through reckless lending, then sitting back like petulant six-year-olds and refusing to lend after getting yelled at.

There was an interesting point made recently on a radio show about how banking has changed. Banks used to add value by taking in savings and then loaning that money back out; they charged more interest than they paid, and that was the old-fashioned way they made their money. At some point they figured out it was just easier to start charging fees for everything. They inserted themselves into as much of the revenue stack as they could, extracting as much of it as they could without causing extreme outrage. (Moderate outrage could be tolerated.)

So where’s the problem? What difference does it make how they earn their money? Simply this: the lending model creates a money multiplier. A dollar saved becomes many dollars spent. This contributes to the economy. Not so with fees: they’re simply a drain on the economy, not a contributor.

The benefit of being a tolltaker is that people pretty much have to pay up. No pay, no play. But, as we’ve seen, there are a number of ways to be a tolltaker, and, while they all might make money, not all contribute overall to the economy.

Which raises another structural problem. In the US, anyway, there is one and only one responsibility, nay, duty that corporate managers must adhere to: make money for the shareholders. Nothing else matters. You don’t need to provide a good (or any) product, you don’t have to contribute to the economy, you don’t have to serve your customers well or protect the neighborhood you live in or provide good jobs or good pay or benefits, you don’t have to do anything unless it makes the shareholders money.

Yeah, there are laws regulating some kinds of behaviors, but businesses chafe at those, and, even in the face of economy-wrecking behavior, attempts at further restructuring and regulating outrageousness is met by an outcry of “you’ll stifle innovation!” whining. As if credit default swaps were a useful innovation.

So when deciding whether or how to be tolltakers, I suppose that, legally, you must do whatever makes money for the shareholders, whether or not you actually add value. But, in a world where we focus three months out and no more, I adhere stubbornly to the old-fashioned sense that, over the long term, those that provide true value will prevail (as long as the shortsighted aren’t allowed to nuke the entire economic environment in the meantime). People are less likely to grumble and more likely to continue paying if you can say, “See? We provided you with this economic value and are asking only so much in return – even though you’re paying us more this year than last, we allowed you, net net, to be further ahead than you would have been without us.”

This Pollyanna view neglects one other important factor: competition. You can do everything you want to win the deal at a fair valuation, but the loser may still tank things. Someone is always at risk of losing, and it’s always better to make less money than to make no money.

Therefore you could argue that the industry needs to restructure together. Which would, of course, constitute collusion and be illegal. So that can’t happen.

So where is this… thing that has turned into an extended rant… leading? As long as this [EDA guy grumbles and caves] scenario is likely, nothing is going to change because, so far, the tools users have held the cards. And with fewer companies designing SoCs, those few companies become more powerful. This means that, even as tool cost and complexity mushroom, it becomes harder to get paid full value. And at some point something has to give.

It feels like, only when big tools providers are faced with actual insolvency rather than just lackluster results, only when things get so bad that someone has to say “This makes no sense anymore,” only when the semiconductor makers start to feel like they may in fact no longer be able to keep making chips because the tools will go away, only then might they open up the purse strings a bit more and begrudge the EDA industry some of their profits.

Is there a better ending to this movie? Here are the constraints: the structure needs to change, tolltaking might be a way to do that, people pay tolls begrudgingly, but less so if there’s value, competitors are always likely to low-ball, and companies can’t collude in bringing change.

That pretty much says to me that change can only happen insidiously. Pick a place to try a new model. The model has to feel fair – if it simply feels like paying more for the same thing, it’s not going to fly. If it feels like it will be more money for more value and less money for less value, then the risk is more balanced and you’re likely to get more takers. Tolls have much more of that characteristic: smaller amounts more closely correlated with actual activity and actual value.

Let it prove itself. If it starts to work, you can get past the toll-haters. Done surgically, it’s harder for competitors to quash. Done strategically, you can make progress without requiring the rest of the industry to come with you. Success can then be met with further migration of the model. And when others see that you’re successful, they’ll come along. And if you’re lucky, you’ll end up getting more of what you deserve, and your customers won’t even hate you for it.

Leave a Reply

featured blogs
Aug 19, 2018
Consumer demand for advanced driver assistance and infotainment features are on the rise, opening up a new market for advanced Automotive systems. Automotive Ethernet allows to support more complex computing needs with the use of an Ethernet-based network for connections betw...
Aug 18, 2018
Once upon a time, the Santa Clara Valley was called the Valley of Heart'€™s Delight; the main industry was growing prunes; and there were orchards filled with apricot and cherry trees all over the place. Then in 1955, a future Nobel Prize winner named William Shockley moved...
Aug 17, 2018
Samtec’s growing portfolio of high-performance Silicon-to-Silicon'„¢ Applications Solutions answer the design challenges of routing 56 Gbps signals through a system. However, finding the ideal solution in a single-click probably is an obstacle. Samtec last updated the...
Aug 16, 2018
All of the little details were squared up when the check-plots came out for "final" review. Those same preliminary files were shared with the fab and assembly units and, of course, the vendors have c...
Jul 30, 2018
As discussed in part 1 of this blog post, each instance of an Achronix Speedcore eFPGA in your ASIC or SoC design must be configured after the system powers up because Speedcore eFPGAs employ nonvolatile SRAM technology to store its configuration bits. The time required to pr...