On Jan 10, 4:04 am, William Stein <wst...@gmail.com> wrote:
> On Mon, Jan 9, 2012 at 7:12 PM, Dr. David Kirkby
>
> <david.kir...@onetel.net> wrote:
> > I'm not sure if this idea will go down well, but I'll mute it anyway.
>
> > I know William has switched off machines (t2.math for example), as he does
> > not have the power/cooling for them. That seems a real shame for that
> > machine to sit and collect dust, when it could be used.
>
> I have the power now.   The only machine that is off is t2.math.  No
> other machine is off.  That was a temporary measure in light of a
> *major* electrical issue that arose in the UW math server room, which
> was beyond my control.    If anybody will volunteer to maintain
> t2.math at a software level, I'm happy to turn it on and give that
> person root.

I'll do my best in the limited time I have available if you switch it
back on.

> > So would William consider paying the electric costs for those that make
> > servers available, and are not covered by their institution? Would it be a
> > waste of money, or good use of money?

> How much exactly would the electrical costs be?

I have two decent servers decent (both IBM) but want to move CPU from
the 1U unit to the 2U unit, as the 1U unit has faster processors. But
I've had to order the necessary heatsink and VRM to make the change. I
will have a machine with a pair of quad core 3.16 GHz Xeons (8 cores
in total) and a significant amount of RAM - perhaps as much as 48 GB.

Once I get set up, I'll measure the power consumption. Obviously it's
impossible to give an exact figure for electric costs.

But I was more thinking of the general principle that if you want more
buildbots, then at the end of the day someone has to pay for them.
Electricity cost is a significant cost of the ownership of servers.

-- 
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to 
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org

Reply via email to