Tuesday, December 02, 2008

You've got to be kidding.

Don't take this as a comment one way or the other on D-wave's actual product claims about building adiabatic quantum computers, but check out the image at right from their presentation at a conference on supercomputing about "disruptive technologies". This may be great for raising [edited in response to comment] enthusiasm, but it's almost self-parody. Remember, true disruptive technologies reshape the world. Examples include fire, agriculture, the wheel, the transistor, the digital computer, the laser, and arguably the internet (if you consider it separately from the computer). Must be nice to claim, on the basis of one real data point and one projected data point, that a product is on a disruptive trajectory (in productivity (arbitrary units) vs. time (arbitrary units)). Wow.

11 comments:

CarlBrannen said...

Having spent more time than I want to think about interacting with rich people on investments, I've come to the conclusion that the majority of the rich are no more intelligent than the general public. So sales presentations direct to them should be produced at that level.

This is part of how they all managed to get sold the same bad investment ideas that resulted in the recent financial debacle. Not only are they stupid, but they run in a herd. When the whole herd of them gets moving in the same direction towards the same cliff they do it world-wide and are utterly unstoppable.

If some politician in some obscure country managed to avoid the bank crashes by careful policy, he would have been blamed for stopping the party and voted out of office.

Dave Bacon said...

I agree the plot is kind of silly. On the other hand the point that "disruptive" technologies often take time to "disrupt" probably holds a grain of truth, don't you think?

Douglas Natelson said...

Dave - sure, but we all know the follies of trying to thread an exponential take-off through two points. I know that this was a fluff graph for fundraising types.

Anonymous said...

what comes below low quality use? Useless.

Anonymous said...

I saw that plot and had exactly the same reaction. For us, anyway, it backfired.

Minus the two "data" points, the plot appears in the Wikipedia article on disruptive technology.

However, I find it confusing to apply that terminology to quantum computing. The essence of the Christensen idea is that existing technology, while constantly improving to meet the needs of its mainstream customers, can overshoot the needs of other customers. These customers, who may be neglected by the suppliers or even nonexistent, don't need the full capability of the technology as traditionally measured. Instead, they value other attributes, such as portability or low power consumption. The "disruptive" technology can get a toehold by providing these attributes, even though it falls short on the traditional metrics.

On the plot, then, the disruptive technology is supposed to get in the door by siphoning off the low end, steadily improving until it eats the lunch of the mainstream suppliers who weren't paying attention to the low end.

It seems to me that quantum computing's primary goal is the high end, in terms of the size of problem it can handle, so I have trouble understanding how it lines up with this description at all.

Anonymous said...

The least that can be said is that it is a very, very optimistic plot. I wish them luck, but wouldn't bet my lunch on their claims.

Anonymous said...

Doug: the chart is from wikipedia. The context was to show where QC technology would be if it followed the wikipedia trajectory in a Disruptive Technologies session at SC08. It has absolutely nothing to do with raising money. Try getting your facts straight before acting like a retard in public.

Douglas Natelson said...

Geordie - I stand corrected. I also point out that "retard" is considered by many to be pretty damned offensive.

Anonymous said...

Doug Geordie is a hothead, it is eaasy to get under his skin but he`s ok. He once punched me in the head twice. Then I kicked him in the nuts and now we`re all good.

Anonymous said...

if you extrapolate the use curves on the graph backward about 50 years, they intersect the time axis. does this mean that nobody used anything before the 50's?

Anonymous said...

sorry, i guess it means that nothing worked more than 50 years ago, because the performance would have been zero.