A look at Apple's R&D expenditures from 1995-2013 | Hacker News Reader

A look at Apple's R&D expenditures from 1995-2013

tuaw.com

75 pointsanderzole69 comments

Not surprising, since most of the core R&D needed happens in the supply chain -- the fabs, the screen manufacturers, the battery makers, sensors, etc. If the iWatch for example, uses new curved battery tech or curved screen, that will have arisen mostly from the R&D done by their partners.

With the acquisition of PA Semi, Apple does seem to be doing basic chip R&D now. But the industry as a whole would be no where without most of the traditional semiconductor manufacturers doing R&D to keep Moore's law going.

I do think it is sad that Apple, with $100+ billion in cash sitting in the bank, isn't spending more on basic research. IBM, HP, Bell Labs, Xerox, et al delivered a lot of serendipitous discoveries by spending money on stuff with no immediate product focus or return. Apple is sucking up the lion share of the mobile revolution's profits, but I think a bigger fraction of that could be channeled back into basic R&D as a whole.

I agree generally that there can always be more. The big difference to me is that in putting the product first, Apple is ensuring that they'll remain in a position to continue doing this. The trend is in the right direction though, their revenue for 2013 increased less than 10%[1] but their R&D expenditures have (based on a pretty coarse chart) gone up around 20%.

[1] http://www.wikinvest.com/stock/Apple_(AAPL)/Data/Total_Reven...

..and Microsoft, Atari, Sun, et al sunk billions of dollars into "research" groups with very little to show for it other than a bunch of concept videos and the resumes of the researchers. It's a double-edged sword.

That's the nature of R&D. There are 100 failures for every 1 success, because often, the usefulness of basic research is unknown.

Modern cryptography is based on advances in number theory and abstract algebra. None of the mathematicians over the last 2 centuries who contributed to those advancements foresaw the applications.

Modern GPS satellites are dependent upon general and special relativity. Einstein never foresaw it.

Much of the military industrial spending that funded the creation of Silicon Valley was not specifically targeted at creating the industries that followed.

The whole point of basic R&D is you don't know the economic outcome. It is a quest for knowledge. It may create the next disruptive technology or industry, or, it may not, and just be a another curiosity to be studied by academics.

It's a double edged sword, but why should we want swords with only one edge (short term economic success)?

Microsoft R&D funds for example, like IBM, mathematic researchers. These people are not working on products or applications, they're working on proving theorems.

I don't fault Microsoft for "wasteful" spending. Research and Development is a genetic algorithm for searching the fitness frontier of knowledge space. A whole lot of organisms, or ideas, have to die, before local optima can be broken out of to find the next peak.

"100 failures for every 1 success"

There aren't a lot of examples of big tech companies dumping money into R&D - with no end goal in mind - and then cashing out big time. Even companies that that have gotten useful discoveries from hiring smart people to sit around and invent (Bell Labs, Xerox Park, Microsoft Research etc.) don't ever seem to end up raking in the big bucks.

The only thing that comes to mind is MS and their Kinect.

Unfortunately the ROI is pretty bad when you just throw money and hope it gives you results. It's cheaper to just buy up startups with interesting ideas

All of the startups you're talking about are sitting on the shoulders of much longer term R&D that was funded by corporate labs or government/academic R&D.

When was the last time a small startup produced a huge breakthrough in physics or manufacturing that did not build off of research funded by the public or big consortiums, corporate labs, or universities?

I don't care if Bell Labs didn't cash out, just like I don't care if NASA, DARPA, or Sandia Labs makes a profit. Quantum Theorists need work too, and Y-Combinator isn't going to fund them.

"I don't care if Bell Labs didn't cash out"

Yeah, well that's nice that you don't care - but Apple's shareholders do. The point is that this kind of R&D doesn't pay-off. Companies have tried it in the past and it didn't work.

If you want R&D then I completely sympathize, but you'll have to go get your government to fund it - don't expect Apple do it as a charitable donation to society.

PS: Examples like Leap Motion come to mind.

That's an unsupportable blanket claim. Such research approaches have paid off in the past. Take IBM:

Invented: * DES * Hard Disks * DRAM * RISC * Relational Database * Laser Eye Surgery * Barcodes * PC

Apple shareholders apparently care, because the company is currently being valued as if there are no more disruptive breakthroughs that will produce significant growth in its bottom line.

Also, to say Bell Labs didn't cash out is to be charitable. Ma Bell dominated for decades before the government broke them up. Did they fail because of failure to cash out on inventions, or because Bell Labs was split off from the parent company that was funding them.

Also, if you suggest the government fund it, then maybe the government tax Apple's cash, I wonder how their shareholders will like that?

Exactly right. There's a big difference between "go think up a way to figure your position on the globe to within 10 cm" and "here's a pile of cash and some beanbags, think up clever stuff and we'll commercialize anything that looks interesting."

The only thing that comes to mind is MS and their Kinect.

Um, didn't Primesense develop the imaging sensor used in the Kinect? This seems more of the Apple model of "let's figure out a cool use for this interesting device".

Actually, it was "go think up a way to deliver an ICBM between the US and Russia with meter accuracy" and "here's several hundred billion dollars to figure it out and build it."

The GPS system would never have been built by any startup, period. It took decades to build up the space based infrastructure and engineering know-how to make it, with the backing of a very deep pocketed government that was looking for results, not for profits.

We need diversity in research approaches. No one is saying that profit-focused R&D is wrong, just that monoculture is. If you've got $100 billion in the bank, do you need to focus all of it on short term projects, or can you try several strategies.

Good insights. Apple didn't spend too much R&D money on its own innovation. It integrated all the great technologies together and produced a few best selling products.

Can you please reveal what the following technologies are coming from?

- Tablet (Palm Pilot?)

- Touch screen (HTC?)

- Battery

- Nano PCB technology

It's hard to believe the touch screen technology is suddenly improved and Apple used it for its success almost for free. In 2009, I was busy bidding touch screens on eBay, which were so dumb and expensive. Suddenly, everybody has one in hand. Magic!

The question is why those companies that created the technologies didn't profit that much? At least Apple's R&D expense should be sky-rocketing compared with the previous years when iPod touch and iPhone being released even if they didn't created the technologies.

WAT?

Pie chart: The world's worst way to display time-series data.

http://o.aolcdn.com/hss/storage/adam/2c59e6862a532820a9854bd...

For the statisticians and data scientists in the room : do pie charts ever have a valid use case (as defined by being superior to another form of visualization)?

Of course. Pie charts look natural and easy to comprehend when visualizing things like market share of certain products, for example browser market share (http://www.netmarketshare.com/browser-market-share.aspx?qpri...). In fact I'd argue that that's a much better representation than a bar chart, like http://www.w3counter.com/globalstats.php, because the former makes it easier to comprehend roughly what proportion of the total market each browser represents, which is not as straightforward to see in a bar chart (in bar chart it's only easy to see how different browser market shares compare to each other). So yes, pie charts may be appropriate at times, although I agree that it's a poor choice in this particular case, as the root comment explains.

How about a stacked single-bar chart? I'd suggest that it has all of the benefits of a pie chart without introducing perception issues of angles and areas.

Well, I guess you could make a fat bar and still run into problems... Ah, visualisation!

This pie chart -- http://www.youthblog.org/pie%20chart2.jpg -- is accurate and presents the data better than any alternative.

(so no one has to click blindly, it's a photo of a pie showing how much of it was eaten or not)

Whenever the data to be displayed is most easily conceived of as a proportion of a whole. If all of your data points added together equals the population you're sampling, then a pie chart might be appropriate.

It's natural to ask: "Of all R&D expenditures, what percentage came from each year?"

A time-series series (cumulative sum) would also answer the question, and could actually be plotted on the same chart as the original series.

On a side note, time series is not a great phrase for the original data set ("time sequence" would have made more sense, since the numbers shown are not cumulative)

Yes, a cumulative sum would have done a much better job than that pie chart at illustrating their point (from the text: "Apple over the last four years spent more on R&D than it did during the preceding 15 years combined")

Also note that 'time series' is a well-defined and widely-used term that exactly describes the dataset presented here (http://en.wikipedia.org/wiki/Time_series).

But if that is displayed as a line chart, that clearly displays the year-over-year differences. So if someone wants to ask "what is the overall trend of spending over the past five years?" that is nearly impossible with the pie chart.

Pie charts are by far one of the worst forms of graphs ever devised and should be killed off as a concept for good, as in, please remove this from Excel.

Everyone is this thread is conveniently ignoring that the data is initially presented as a bar graph. The pie chart is secondary for a very narrow purpose, and it's far from terrible in this application.

I disagree, that was a very effective use of a pie chart. By looking at it for a couple of seconds you can tell that the last 4 years take up over half of the total.

I suspect that many people will need to be told that. So the graph is a useful demonstration, but not useful to discover that information.

And the first bar chart which is not zero-based! (groan)

I am not an expert, but it's my understanding that what is R&D vs. "only" engineering is largely an accounting question. I don't think Microsoft is spending all of that $10B on MSR, for example.

As a metric I don't think it provides much insight.

you are spot on with that thought. accounting for r&d is pretty much an accounting/financial call. rather than accounting for costs, you capitalise that cost/capex as an asset, which can later be amortised in x years. i'm guessing that given that apple generates billions in profits, it would rather account much of its r&d as an operating expense, reducing the immediate tax bill.

MSR is only "R" (and that too only a subset of the total research that happens throughout the company - the stuff that can be classified as "basic science"/longer-term), the "D" happens all over the company in the various departments.

http://o.aolcdn.com/hss/storage/adam/3924d8cabc4e8469af6fa91...

One question: Why is VW's R&D spending so high? It's BILLIONS more than other car manufacturers.

They build cars in many segments. Tesla build one model, while VW has 27 models (from their german website) + commercial vehicles. Also, they own several brands like Audi, Bentley and some others. Most of them has a wide model range.

Yeah, take a look: https://en.wikipedia.org/wiki/Volkswagen_Group#Subsidiaries_...

Bugatti and Lamborghini are fun to note, though probably insignificant against VW's entire R&D budget.

I noticed that as well - One wonders what % of that expenditure is on improving R&D processes to lower costs, assuming that most of that cost is on non-renewable hardware modeling (hurling one-off test cars at brick walls and such). Also, is racing considered part of the R&D budget?

This could be interesting to analyse but those charts are not really the best, mainly the pie chart. Maybe the best way to look at research costs is to compare it to the other costs of the company (maybe with the exception of those related to the cost of building the products).

I'm in no way an analyst so I might be way off on this recommendation though.

The article feels a little shallow on the analysis side.

For instance I would expect this oddity to be caused by the fact that they aren't a traditional hardware manufacturer (they don't have any fabs and outsource their production) nor are they a traditional software manufacturer (their software focus pales in comparison to Google or Microsoft).

Since developing silicon and software are two of the biggest drivers in the industry and Apple doesn't focus on either, it makes sense they have a low R&D budget.

Don't Apple's A7 and other chips count as "developing silicon"? And you don't see Apple focusing on "developing software"?

Yes, but it is abstract development that relies on lower level discoveries. It's like designing a medical treatment that relies on anti-biotics, if no new antibiotics are discovered, and the old ones become less valuable, then new treatments hault.

The A7 is a marvelous chip, but the reason these super smartphones exist is silicon manufacturing prowess. If we could not make these chips smaller and less power hungry, the mobile revolution would not have happened. Apple stood on the shoulders of giants who went before them in this regard.

Intel, IBM, TSMC, et al, plow a lot of money into advancing silicon manufacturing further. Lots of other companies plow billions into tools to support this sector.

In some ways, taking in huge margins on silicon upstream (while the downstream players are squeezed to minimize their margins), without plowing money back into the system, could be seen as somewhat parasitical.

I used to work at IBM Research, and one of the things I used to love doing was reading IBM Systems Journal. Research, like on scanning electron microscopes, measuring quantized magnetic flux in superconducting circuits, new kinds of giant magneto resistive effects, it was all very exciting to know that right down the hall, stuff I'd normally read about in Scientific American, Nature, or other places was being done. IBM T.J. Watson Research was given relative autonomy when I was there, they could spend money without have to justify it as being linked to a product, and I think that was very valuable for long term development.

The new players in Silicon Valley don't seem to have the same commitment to basic R&D, long term R&D. Everything has to be linked to something shiny that can be sold in 2-3 years.

Apple is indirectly plowing billions of dollars into that research via their contracts with Intel and the foundries. If the margins on those contracts aren’t sufficient to provide for further research, that’s on the head of the supplier’s pricing teams, not Apple.

And if Walmart's employees and Chinese suppliers have their payments squeezed, it's not Walmart's fault? That ignores the pricing power that a large player has. Apple and Tim Cook, as has been noted in the media, are very good at negotiating down their suppliers.

But no matter how you slice it, these downstream suppliers are doing more with less. They have much smaller margins, fund more important R&D which benefits the entire ecosystem, which takes up a much larger percentage of their overall revenues.

The marginal utility of an extra dollar in Apple's cash reserves seems less useful or effective than the marginal utility of a dollar in say, TSMC's coffers.

Intel obviously has vastly more negotiating power than any of Walmart’s suppliers. This may or may not be true of other fabs, but even in that space there are only a few suppliers and the costs of switching fabs are enormous; that isn’t true when Walmart is searching for a supplier of plastic adirondack chairs.

Apple doesn't buy mobile chips from Intel, they buy them from Samsung.

You're basically arguing (repeatedly) that outsourcing (and thus the whole modern economy) won't work. It's provocative, but seriously citation needed.

I don't I've arguing outsourcing won't work, just that outsourcing can lead to malinvestment.

True, and that r&d $ is hidden under capex and investment in associates. Real r&d expenditure can be tough to measure from the outside.

In detail, that's not quite right about the A7.

Apple designed the A7 themselves. It happens to implement the ARM instructions, but is not an ARM design. The only other player I can think of currently with this in the wild is Qualcomm, and this is one reason Apple and Qualcomm chips maintain a performance edge over other ARM CPUs.

I believe your parent is referring specifically to basic physics/chemistry research toward future process technology, rather than processor design.

They're ignoring that Apple (and others) are competing in design at all and claiming the only progress is in fab process. Fab process is very important, but it is most important in x86 land where you don't have many options but to improve fabrication, hence Intel's massive spend there. Beyond that you have the competing groups of TSMC, GlobalFoundries etc. that will spend way more than any non-specialist is going to.

The rest of the processor industry, including the ARM ecosystem where some people get to compete with ARM themselves, is making a lot of progress in working out how to improve the layout of the transistors, such that two designs on the same process can perform quite differently. In the case of the Ax devices and the Krait this is a surprisingly significant margin, and they tend to be at least a whole generation ahead in performance of the designs coming out of ARM. This is not something to be dismissed.

I'm not dismissing it, but to ignore the fact that the iPhone is wholly dependent on tons of innovation in their supply chain is a problem. No matter how fancy your transistor layout gets, if, starting tomorrow, there were no more process node shrinks, and no more innovation in other ares of silicon (e.g. low-k/high-k), the mobile revolution would be effectively frozen in place.

Phones would have no choice but to get bigger to get more powerful or to get longer battery life and perf/watt would hit a wall.

Point being, Apple spends a disproportionately small sum on R&D relative to the size of their earnings and they don't seem to be funding basic research at all like large corporations of the past, e.g. IBM, Bell Labs, Xerox PARC, etc.

With the government cutting funds for basic research, we need corporations who are sitting on $4 trillion in cash, to pick up the slack.

You can apply the 'They didn't build that' argument to just about anything. It almost seems like a strawman here.

http://krugman.blogs.nytimes.com/2012/07/24/they-didnt-build...

Great analogy actually, because the President and Krugman are exactly right in this regard. We ignore the ecosystem that was built, and which we benefit from, by those before us at our peril.

You should see technology as a forest. You can mine it for medical cures, for wood, it's an enormous externality that you leverage. And you have a duty to continue planting more back into it to keep it going for yourself and everyone who comes after you.

The Apple narrative is too bound up in heroic origination stories, without due credit and acknowledge to the huge role played by the rest of the industry. I'm only saying that if one profits immensely by using knowledge produced by one's forefathers, one has a duty to reinvest and keep driving it forward, not hoarding piles of cash. (oh, and not going insanely litigious and secretive on discoveries either :) )

I should point out A7 was manufactured by Samsung...

It is troubling that the new players in Silicon Valley are not really committed to basic R&D.

"they don't have any fabs"

Isn't this not true anymore? I thought Apple brought chip production "in house" as well as other specific hardware related to their iOS products.

You are correct. Apple just spent billions of dollars buying up a fab lab. However, it will be years before this investment sees any practical use at Apple.

http://semiaccurate.com/2013/07/12/apple-has-their-own-fab/

Is there a link to substantiate that claim that doesn't put the information behind a paywall?

Not one that I know of.

Charlie isn't the typical blogger though, and I prefer his paywall style over the ad-based models that other sites have done. When Charlie wrote a few honest (but blasting) reports on various companies a while back... the ad-companies held his revenue hostage. They didn't like him writing the way he did. As such, he stopped doing ad-hosted content and was forced into the paywall model.

Now to be fair, Charlie is a bit rambunctious and exaggerates some claims, but his fact-gathering tends to be months or years ahead of others (Do a double-take at the date of this article: http://semiaccurate.com/2012/03/02/sony-playstation-4-will-b.... I personally can verify the date on that article, as I read the article when it came out and didn't believe it.) . His writing style is abrasive though, and ad companies didn't want much to do with him.

He gets things wrong sometimes, because he's really trying to predict the future. But since he gets things right more than 50% of the time, I'm pretty confident in his claims. Overall, he does good work. It is unlikely that other sites will have his information for another few months at least.

He also wrote for The Inquirer, and before that The Register, and before that ... some other print trade mag I can't recall. This goes back to the early 2000s when those papers actually broke stories and not press release. IIRC Semiaccurate is "his" project after leaving The Inquirer a couple years back. And yes, a lot of what Charlie Demerjian reports is rumors with a little corroboration. But that's why he scoops other sources by months.

There are a whole bunch of other sites linking to this piece calling it a rumour, and that there is little detail beyond the paywall and no corroborating evidence since then.

I think it's false.

Fair enough. As I've said before, Semiaccurate is only "mostly" correct. Charlie has been wrong before, mostly because he pushes (likely) hypotheticals.

I have however, read the contents behind the paywall, and Charlie is quite specific. He has the company name, some vague details on the deal, and the location of the facility.

He is stretching the hypothetical a bit to say that "Apple bought a fabrication lab". That may have gone to far... but those details are discussed in the paywalled article.

Chip fabs are extremely expensive. You basically need to invest $5B+ every two years just to stay up-to-date on technology, and then the same amount or more to do the necessary research and build the thing in the first place.

(Reference in point: history of Globalfoundries)

They brought some chip design in house. Chip production is quite expensive that only a few companies do it in house.

I know this is a hearsay but at the height of the apple/samsung law suit I was following a story and one of the commenters claimed Apple's chip R&D was being done in a borrowed facility at a Samsung owned chip factory in Austin, TX. Intrinsity based in Austin had been doing work with samsung and since Intrinsity was bought out by Apple, I think there's some chance it's true. No way I can verify if it's true it'd be pretty funny...

Intrinsity did design, not process. They did some design work for Samsung around the A8/Snapdragon days. Shortly after that Apple acquired Intrinsity and brought them in house ala PA Semi.

Perhaps they spend slightly less than others, but what interests me is that there is an exponential growth of R&D spending over the last five years with no real innovation (in my opinion, since the iPhone). And Microsoft has been spending consistently around 15% of their revenue on R&D. For what? Higher pixel density and an OS with tiles that can be swiped on touch screens? Is R&D really that low ROI? We're talking about tens of billions of dollars here.

In a lot of cases, you won't notice the result of every dollar spent on R&D (until much later on, when the technology becomes mainstream). For e.g. siri: might just seem like voice-to-text to you, but there's probably a ton of stuff happening in the backend in the fields of ai, scalability, personalization of the results, etc. Some of this stuff might have been known problems with known solutions, other things might have been applied research.

Remember that Apple is releasing a new product category in 2014; the R&D for that has been paid over the last few years.

Normalize to show percent of revenue and then we'll talk.

A time series of $$$ allocated to R&D is basically going to show when the company's growing and when it's shrinking, the same way a heat map of who clicks on furry porn is basically going to show population density.

I realize there's a chart of $$$ allocated to R&D as a percentage of sales but that's not quite right either.

The best chart shows percent of sales. Shows how lazy they got the last couple of years ;-)

Agree. They are waiting for other companies to provide more inventive technologies, ;-)

For an investor, that's the best chart. More sales for each dollar of R&D.

Reply on news.ycombinator.com