Apple Lost its Focus on Science

When Apple was about the business of launching the new Mac OS X in 2001, the feeling was that there was much to be gained by engaging the science community explicitly. UNIX was and is the lingua franca of science and computing, and Apple’s strategy worked. These days, Apple doesn’t appear so interested in explicit science support because of changes in Apple’s market and in the science and research community. Here’s how it all happened.

Hadron ColliderPart of the Large Hadron Collider

The Early 2000s

Back in 2001-2005, Apple was a new kid on the block when it came to shipping a mass-market UNIX OS. Also, it was an era when powerful UNIX workstations on the desktop made sense for computation and visualization — something that Apple knew all about. Apple made considerable headway with the idea that Macs made better sense as scientific UNIX workstations with beter graphics and a high level of security, especially because Windows was not as secure an OS in those days.

Apple also worked on the fringes of clusters and supercomputing. The company sold a few supercomputers, but focused on primarily workgroup clusters, 4 to 32 nodes, that could be managed by an individual or two.

The culmination of that effort was at apple.com/science, which I’ll turn to below.

Apple, Unix, scienceImage Credit: Apple

The Late 2000s

In time, what changed was the availability of relatively inexpensive Linux-based clusters of all sizes. While supercomputers with thousands of nodes were typical at the U.S. National Laboratories, the affordability and availability of supercomputing assets rose dramatically for many other organizations. It made sense to look at PCs and Macs as front ends to massive cloud computing assests, both financially and technically. The U.S. Dept of Energy (DOE) has spearheaded this effort. The strategy makes sense for all science areas, from physics and astronomy to life sciences. While Apple succeeded with first-class dekstop hardware, it couldn’t compete with “white box” commodity Linux computing power in large clusters.

That technical reality, recognized by Apple, has caused a little bit of consternation by those who saw Apple, historically, as all-in when it came to science. For example, some take the Mac Pro as Apple’s stand, a line in the sand, that they’ll always deliver the strongest possible Mac for personal computation and research. Others have a broader perspective. George Providakes, a very senior scientist at a major research organization told me:

Well, not sure things are a bleak as portrayed [with Apple and science.] Regarding Xserve, no one was buying them so it seems to be more a blip as everyone is moving to Linux blades or the cloud.

Regarding the Pros, I am a bit mixed. I am not sure Intel has provided the chip sets that make a refresh sensible. A 12 core MacPro is pretty powerful, but without newer high performance chips, the changes would be modest. My sense is that a bigger performance gain is being looked at before a new machine comes out. It is certainly possible Apple will drop Pro lines as its business evolves, but I am inclined to say no, since it still invests in the Pro software Final Cut et al.

Therefore, Ivy Bridge roll out is a better timing measure for Apple Pro refresh. If the 3D chips provide the performance gain and Apple fails to act, it’s pretty clear where things are heading.

From a larger business perspective, it looks to me that the scientific market is moving to the cloud not to enormously more powerful workstations. This must color the business decisions of Apple.”

Coincident with the cloud developments has been the phenomenal rise of Apple, mobility, smartphones, and tablets. If one looks at Apple’s revenue from iPhones and iPads, it far exceeds that of desktops and notebooks. The Xserve fell out of the bottom in sales as Apple’s customers of all kinds came to use mobile devices as front ends to larger computing assets, supercomputers, clusters, various computational clouds and the Internet as a whole.

Super ComputerModern Supercomputer

Today

As a result, when Apple bowed out of supercomputing and clusters in 2005, it wittingly or unwittingly bowed out of the formal support for the science community. But not necessarily out of the market for excellence in computing products. To put it simply, one path to success for Apple has been to steer away from the low ROI, pure science market and instead smartly focus on the global market for mobility (including MacBooks) and tablets. In turn, those superior products have found their way into certain science-related markets, for example, medicine. In fact, the medical and aviation industries have embraced iPads wholeheartedly. Many major science and technology related organizations continue to emphasize Apple products — with good reason. In that sense, the products stand on their own.

Apple products, today, are often seen as better designed, very secure, very capable products used to access, rather than host, a wide variety of back end services. It’s probably not an understatement that virtualization has played a major role.  It’s true that for certain personal, creative activities, like video editing, writing, web development and art, a very powerful desktop workstation is desirable. But, in general, the nature of the really big technical problems in science demands both elegant front ends as well as community developed cloud resources. Apple is giving us the great front ends, yet with UNIX under the hood.

Apple and Science

It’s probably an exaggeration to suggest that Apple has completely lost interest in supporting the sciences. Rather, Apple has become so mainstream and so highly accepted that its products are used in many new ways mentioned above.

Now that Apple is a mass market, consumer electronics giant, there seems to be less need to promote a desktop UNIX workstation for scientists, scientific app development, and fan the flames of a small, underfunded, eccentric community — merely for the benefit of respectability. And yet, there is something important to be gained there, which I’ll mention below. Even so, Apple’s more general approach combined with a withdrawal from scientific conferences has likely contributed to some isolation from the science mainstream.

The shift described here has left some scientists a bit uncomfortable, and occasionally Apple’s message isn’t clear. The cancelation of the Xserve/RAID, then the Xserve, certain changes to OS X and the lingering concerns about the next Mac Pro, if there is to be one, have left the customer base tense.  The stagnation of apple.com/science (with references to discontinued products) and the withdrawal of support for macresearch.org have added fuel to the concerns.

XserveThe Apple Xserve, high quality, low volume. And gone.

The last two items could be focus and housekeeping issues rather than an explicit statement by Apple. However, confusingly, a more active site for showcasing the Mac has been apple.com/business/mac, which was formerly apple.com/itpro. One almost needs a scorecard to keep up.

Because of that legacy, always delivering the best and fastest is in flux, some scientists who are both Apple fans, legacy customers and adept at extracting maximum performance from a desktop at a high technical level have turned to Linux.  Dr. Gaurav Khanna, Physics Dept., University of Massachusetts and a specialist in Apple technologies, astrophysics and high performance computing told me:

This [shift] has certainly impacted my adoption decisions. This is the first year wherein I moved from my dated 2008 Mac Pro to a combination of a Mac Mini (as my primary desktop) plus a high-end Linux workstation (with 16-core Intel “Sandy-Bridge” Xeon processors and an AMD Radeon HD 7970 + Nvidia “Fermi” CUDA GPU for OpenCL development). I had little choice … given the present trend, its just highly unlikely that I would have been able to get all this technology in a Mac.”

Another astronomer, a project lead at a major observatory out west, summed it up rather nicely when it comes to mixing and matching solutions with his MacBook:

From my perspective Apple products are essentially irrelevant or orthogonal to state of the art in astrophysics, computational astronomy, and informatics. I use a MacBook, but just because it is an easy way to have a Linux shell for access to my workstation, and at the same time run Windows programs I hate, but for compatibility with the rest of world I have to use, i.e., Word, Excel, PowerPoint. In short I don’t use Apple applications, just the platform for access to other platforms.

For research we run a 3 Petabyte, 1,600 core Linux cluster for image processing and reduction and a 200 TB Microsoft SQL Server cluster as the only product that can provide a large database on a cluster. Most of our electro-optical hardware is run on a Windows interface because that is what the manufacturers provide. Our graphic artists use photoshop on a Mac to make cool graphics but mostly for outreach, not for research.”

The Return of a Return

So much for a historical perspective. Now for an opinion based on my experience.

Eventually, Apple will be compared to an incredibly wealthy person who spends a lot of time trying to get wealthier. While the products are great and serve the customers, the bigger question is, what is it that’s worthwhile that both Apple and its customers are doing?

In other words, to make a point, is the iPad simply a mechanism to extract more money from the customer? Or is there a celebration of the things that are worthy of our efforts as human beings? For example, just a few of Apple’s TV commercials, specifically for the iPad, celebrate that, especially with children, but that’s one of the few places where Apple openly discusses the best that humans can achieve with what Apple has given us.

I am not overlooking Apple’s other current efforts. When Apple SVP Phil Schiller introduced the iBooks Author app, he mentioned that education is in Apple’s DNA. And yet, one might have imagined that Apple wouldn’t take such a heavy financial hand in that initiative to the point where analysts are alarmed rather than celebratory.

Despite Apple’s financial success, we know the difference between a conduct that’s all marketing, has strings attached and is all about making money as its only goal and, conversely, efforts that are intended to truly serve the causes of humankind.

Today, the tide has turned in Apple’s favor, and it’s natural to ask what Apple can do to make its community of customers stronger and more productive. It’s no longer a question of ROI; it’s a question of values that are not related to money. It would certainly be a positive outward sign if Apple were to support, more openly, the things that can be achieved with its products, even if they aren’t always the primary computational platform. Perhaps it’s that apparent loss, from a marketing and product standpoint, that’s making some scientists nervous about Apple. It’s understandable in a community that has lofty goals.

Apple, science & moneyAvalanche! (Image Credit: Shuterstock)

Another Kind of ROI

Some wealthy companies lose their way because they get caught up in survival, litigation, profits, the extinguishing of competition and the ability to force change. Those become ends in themselves. Injecting certain connections back to society can help a company maintain its grounding and balance. Apple has always had one foot in the world of science, and so a more contemporary emphasis on how scientists are succeeding with Apple products, at apple.com/science, would accomplish a very laudable goal of a natural, healthy corporate balance. After all, we want the best for Apple too.