[Discuss] Linux has 100% of Market Share.

Marco Milano marco.milano at gmx.com
Mon Jun 18 17:26:42 EDT 2018



On 06/18/2018 04:22 PM, Bill Bogstad wrote:
> On Mon, Jun 11, 2018 at 5:59 AM Marco Milano <marco.milano at gmx.com> wrote:
>>
>>
>>
>> On 06/09/2018 12:11 PM, Bill Bogstad wrote:
>>> On Sat, Jun 9, 2018 at 11:11 AM Bill Ricker <bill.n1vux at gmail.com> wrote:
>>>>
>>>> the Top500 statistics include Historical charts. You can see how recently the last 2 AIX systems were pushed out of  Top500, how "mixed OS" had a brief surge early last decade, etc.
>>>>
>>>> https://www.top500.org/statistics/overtime/
>>>
>>> Thanks for the pointer to those charts.  100% Linux was apparently
>>> reached in last November's list so that is a very recent thing.
>>> The dominance of Linux goes back well over a decade with which I was aware.
>>
>> Most of the computational power is coming from the GPU units.
>> You can probably use almost any OS and still come on top on those
>> benchmarks as long as you have sufficient number of GPU units.
>> I don't think the top500 as relevant as before, it is more like a pissing
>> match between USA and China these days.
> 
> I think it is just as relevant as it ever was.  

I disagree, 20 years ago, that was the only option,
now anybody can "create" a virtual supercomputer in the cloud
as needed, as long as they have the funds for it.
So, less relevant now compared to 20 years ago.
If you look at the specs of this latest one, there is nothing special,
anybody with a fat wallet can create one, and that is exactly what this is.
Pissing match... between USA and China.

Systems on this list
> have always been unusual in some way.  And the uses to which they are
> put are far from typical as well.  Massive simulations of various
> sorts remain very important in science, technology, biology, medicine,
> etc. and even the latest entries on these lists still aren't able to
> do as much as some researchers would like.  Whether research in these
> areas is relevant is undoubtedly a question of values.
> 
> Bill Bogstad
> 



More information about the Discuss mailing list