![]() |
|
|||||||||||||||||||||||||
| Ravi Pandya software | nanotechnology | economics |
ARCHIVES
ABOUT ME
DISCLAIMER
|
|||||||||||||||||||||||||
|
Sun 15 Mar 2009 Road Kill on the Information Highway I happened across this 1993 piece by Nathan Myhrvold: Personal Computers I've saved the best for last. Our own industry is also doomed, and will be one of the more significant carcasses by the side of the information highway. The basic tasks that PCs are used for today will continue for a long as it makes sense to predict, so it isn't a question of the category disappearing. The question is one of who will continue to satisfy these needs and how? .... The technical needs of computers on the information highway, or IHCs [i.e. netbooks - Ravi] are quite different than for PCs. .... Most IHCs will certainly need to be cheaper than PCs by an order of magnitude and this will inevitably cause them to be less capable in many ways, but some of their requirements are far more advanced. Another way to say this is that the rich environment of software for PCs is largely irrelevant for IHCs. Windows, NT, System 7 and Cairo do not solve the really important technical problems required for IHC applications, and it is equally likely that the early generations of IHC software won't be great platforms for PC style apps. ... PCs will remain paramount within their domain for many years (we'll still have a computer on every desk) but IHCs will start to penetrate a larger and larger customer base on the strength of its new and unique applications. The power of having the worlds information - and people - on line at any time is too compelling to resist. ... One day however people will realize that their little IHCs are more powerful and cheaper than PCs - just as we have finally done with mainframes. There will be a challenge for the IHC software folks to write the new systems and applications software necessary to obviate PCs, just as we had to work pretty hard to come up with NT, but this battle will clearly go to the companies who own the software standards on IHCs. The PC world won't have any more say about how this is done than the companies who created MVS or VMS did about our world. Of course, some of the VMS people were involved, but as discussed above it is very hard for organizations to make the transition. This may sound like a rather dire prediction, but I think that for the most part it is inevitable. The challenge for Microsoft is to be sufficiently involved with the software for the IHC world that we can be a strong player in that market. If we do this then we will be able to exploit a certain degree of synergy between IHCs and PCs - there are some natural areas where there is benefit in having the two in sync. The point made above is that those benefits are not sufficiently strong that they alone will give us a position in the new world. We'll live or die on the strength of the technology and role that we carve out for ourselves in the brave new world of the information highway. 22:08 # Mon 14 Apr 2008
Supple:
automatically generating user interfaces, by Krzysztof Gajos and
Dan Weld 05:48 # Sun 20 Jan 2008 Economics of Software as a Service Delivering Software as a Service - McKinsey Quarterly Two interesting data points: While SaaS companies have lower operating margins than packaged
software companies overall, if you restrict the comparison to
comparably-sized packaged software companies, the numbers look
essentially identical: However, if you look at the customer cost side, there's a huge
advantage to SaaS: 16:05 #
Multistage Collaboration in CACHE: The Bayes Community Model 15:41 #
A
Logic of Filesystems
Computation Spreading: Employing Hardware Migration to Specialize CMP
Cores On-the-fly A
Predictive Model for Transcriptional Control of Physiology in a Free
Living Cell Forever
Minus a Day? Some Theory and Empirics of Optimal Copyright 08:33 # Fri 23 Nov 2007 My colleague Eric Northup has mentioned these a few times, and I'm glad I looked them up. The Tornado OS (from my alma mater, U of T) and its successor K42 (at IBM Research) use a fine-grained object-oriented approach to all operating system structures (processes, memory regions, etc.), with built-in clustering for replicated instances across processors. This reduces lock contention and increases cache locality by operating on the per-processor instance as much as possible. Since objects are generally expected to be local, it can optimize for this case, and track cross-processor operations as a special case. There are some policy choices (e.g. maintaining replica tables for all processors) that would probably need to be adapted for manycore. The scalability architecture is best described in this paper. The memory manager was key, e.g. for locality-aware allocation, padding to cache line size to avoid false sharing, deferring deletion until quiescence to avoid existence locks, etc. An insight as the basic Tornado model was applied to real workloads was that creation-time object specialization isn't sufficient, instead it is better to for example start with an unshared implementation and then upgrade to shared implementation when multiple processes share an object. They were able to improve their 24-proc scalability from "terrible" to pretty good in 2 weeks of work because of good OO discipline and tracing infrastructure. Overall, I found it striking how the scalability architecture mirrored that for distributed systems - state partitioning, replication, dynamic upgrade, etc. I had expected this from general principles, but it was valuable to see it confirmed in practice with significant workloads. The scalability graphs are impressively linear. Security isn't mentioned, but I expect that the same OO design that gives the OS good modularity and scalability could be applied to give it good capability discipline as well. 12:04 # |
||||||||||||||||||||||||||
© 2002-2004 Ravi Pandya | All Rights Reserved |