[FoRK] costs of computing

Eugen Leitl eugen at leitl.org
Thu Mar 8 06:29:16 PST 2012


On Thu, Mar 08, 2012 at 09:12:05AM -0500, Bill Kearney wrote:

> The trouble with the marketdroids comparison of computational power  
> alone don't account for throughput.  Sure, once you get something 'on  
> chip' any given new device might be a lot 'faster'.  But that doesn't  
> factor how the older high-performance gear was rigged to handle getting  
> data in and out of it.  I'd daresay current desktops and tablets are  
> stil woefully behind at actual throughput to support such tasks.  That's  
> fine, as they're not intended for that purpose.

Right.

> So, please, spare us spreading the marketing hype.

Nobody is spreading any marketing or hype. It's just a Register guy
engaging in their trademark brand of tech journalism. Anyone
who's treating it more than a couple of amusing data points
should have his Internet license revoked.


More information about the FoRK mailing list