[Beowulf] MS HPC... Oh dear...
landman at scalableinformatics.com
Wed Jun 14 08:18:42 PDT 2006
Douglas Eadline wrote:
> Of course there is no absolute rule on how to maintain a market, but
> stagecoach companies who did not recognize locomotives kind of missed
> the technology curve.
> Joking aside, there are things Linux can do that Windows (and any closed
> source OS for that matter) cannot do. There are market segments where
> these features are invaluable. I hate to keep dropping this link but,
> take a look at "Why Linux On Clusters"
No apologies, it makes good points.
> There are, of course, market segments where some of these features
> are not that important and things like TCO, cost, integration,
> and ease of use, may take precedence.
I have to admit that I see all of these as Linux cluster advantages.
TCO is lower due to upfront acquisition cost and lower long term maint
costs as well as lower long term licensing costs, and no patch-tuesday
fiascos, no anti-virus per node, etc. Integration is IMO simpler as
Linux talks to everything. Ease of use again is in the eye of the
beholder. Most windows admins don't do CLI very well (most that I met),
though there may be some cli masters. Most end users don't care about
CLI or not, and just want to run. As long as it just runs and the
admins can manage it without a huge learning curve, this is a good thing.
> Microsoft recognizes this and I believe they are looking at more
> "turn-key" verticals that are driven by ISV applications.
My belief is that their approach will be to leverage the ISVs, and
convince the C-letter folk in corporations that their product will save
> The one exciting thing about clusters is the technology has
> redefined how supercomputing is done (now called HPC by the way). The
> financial barrier to entry has been lowered by a factor of at
> least 10 and therefore more market growth can be expected.
> Ten years ago a high school or college could not even think
> of doing protein folding, but now with a handful of reasonably
> powerful PCs and Gromacs they can. There are other similar cases as well.
> One technical point, measuring cluster usage is difficult because very few
> vendors know where their hardware ends up. Of course one can count
> the clusters IBM, Dell, HP, push out the door, but the number of
> "dark clusters"* that come from white box companies is basically unknown.
> I just got an idea for new Cluster Monkey poll.
> * dark clusters - I first heard this used by IDC, like the
> proposed dark matter, it is out there but we cannot see it.
Heh... and it may be the most massive thing in the universe ...
>> We are all sooooo technical!
>> It does not work that way....
>> Microsoft HPC will work IF the market wants the technology that they can
>> supply. The answer is in market segments. And the Supercomputer market
>> segment has been the science and engineering users of the last 30 years.
>> It has not changed its market much. As technology got complex, the
>> simulation market has grown some, but not to the extent of the word
>> processing or spreadsheet market place or the home web access or gaming
>> markets have grown over the same time period.
>> The Linux HPC market is on the market demand track from CDC -> Cray ->
>> Vax -> Unix Workstation -> Linux . It has grown modestly, but not
>> enormously. Dual CPUs have a bigger growth market and potentially the
>> Microsoft HPC could help develop 4-way and 8-way processors as standard
>> Technical issues will not maintain the market for cluster computing. Nor
>> will technical best-of-breed win over uninterested markets. The past is
>> full of superior technical solutions losing to market accessible
>> good-enough solutions ( eg betamax vs VHS and token ring vs Ethernet ..
>> and the beat goes on...)
>> To paraphrase, I know (some) Linux, Linux is a "friend" of mine, Windows
>> is "No Linux" - but I doubt I am the market segment that Windows is "good
>> enough" for.
>> So please desist discussing technical superiority. Technical issues are
>> not the key issue in market segment growth of HPCs. So lets see who wants
>> CPU power with windows integration and discover how big that market is.
>> Tom Pierce
>> Bldg 7/ Rm 207D - Spring House, PA
>> 215-641-7475 - Office _______________________________________________
>> Beowulf mailing list, Beowulf at beowulf.org
>> To change your subscription (digest mode or unsubscribe) visit
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics LLC,
email: landman at scalableinformatics.com
web : http://www.scalableinformatics.com
phone: +1 734 786 8423
fax : +1 734 786 8452 or +1 866 888 3112
cell : +1 734 612 4615
More information about the Beowulf