On Wed, Nov 5, 2008 at 5:53 PM, ron minnich <rminn...@gmail.com> wrote: > Just booted Plan 9 on a 1024+16 node BG/P this week. .
> All credit to jmk, ericvh, and charles for this fantastic test run and > the existence of this new kernel.
> Plan is to double it just a few times until we hit 65536 or so. Then > the fun begins: turning on all cores, so we get to > 262144 cpus.
> Boots kinda fast, too. It's peppy. No current plans to run acme, > though we have run rio on a bg/l cpu node in 2007, as ericvh related > on his blog.
> Just booted Plan 9 on a 1024+16 node BG/P this week. .
> All credit to jmk, ericvh, and charles for this fantastic test run and > the existence of this new kernel.
> Plan is to double it just a few times until we hit 65536 or so. Then > the fun begins: turning on all cores, so we get to > 262144 cpus.
> Boots kinda fast, too. It's peppy. No current plans to run acme, > though we have run rio on a bg/l cpu node in 2007, as ericvh related > on his blog.
> Just booted Plan 9 on a 1024+16 node BG/P this week. .
I'll echo the congratulations.
> Plan is to double it just a few times until we hit 65536 or so. Then > the fun begins: turning on all cores, so we get to > 262144 cpus.
So how many cores is that for each member of the Plan 9 community? :-) On a slightly more serious note, would it be terribly naive to guess that there would be more cores running Plan 9 in that machine than there are throughout the rest of the world?
Now we just need to run Inferno so it will cover the range from the GameBoy DS all the way up to the BG/P. :-) There would be something perverse but fun for the DS to import resources from the BG/P or vice versa for that matter.
>> Just booted Plan 9 on a 1024+16 node BG/P this week. .
> I'll echo the congratulations.
>> Plan is to double it just a few times until we hit 65536 or so. Then >> the fun begins: turning on all cores, so we get to >> 262144 cpus.
> So how many cores is that for each member of the > Plan 9 community? :-) On a slightly more serious > note, would it be terribly naive to guess that > there would be more cores running Plan 9 in that > machine than there are throughout the rest of the > world?
> Now we just need to run Inferno so it will cover > the range from the GameBoy DS all the way up to > the BG/P. :-) There would be something perverse > but fun for the DS to import resources from the > BG/P or vice versa for that matter.
On Thu, Nov 6, 2008 at 3:53 AM, ron minnich <rminn...@gmail.com> wrote: > Just booted Plan 9 on a 1024+16 node BG/P this week. .
> All credit to jmk, ericvh, and charles for this fantastic test run and > the existence of this new kernel.
> Plan is to double it just a few times until we hit 65536 or so. Then > the fun begins: turning on all cores, so we get to > 262144 cpus.
> Boots kinda fast, too. It's peppy. No current plans to run acme, > though we have run rio on a bg/l cpu node in 2007, as ericvh related > on his blog.
On Thu, Nov 6, 2008 at 9:07 AM, Bruce Ellis <bruce.el...@gmail.com> wrote: > That's nearly as many cores as I have gates! Very good indeed - just > don't say "Hooter's Girls" to Mr Van Hensbergen.
> brucee
> On Thu, Nov 6, 2008 at 3:53 AM, ron minnich <rminn...@gmail.com> wrote: >> Just booted Plan 9 on a 1024+16 node BG/P this week. .
>> All credit to jmk, ericvh, and charles for this fantastic test run and >> the existence of this new kernel.
>> Plan is to double it just a few times until we hit 65536 or so. Then >> the fun begins: turning on all cores, so we get to >> 262144 cpus.
>> Boots kinda fast, too. It's peppy. No current plans to run acme, >> though we have run rio on a bg/l cpu node in 2007, as ericvh related >> on his blog.
"Sole purpose of visit" as someone supposedly quipped in answer to the question on the US entry form: "Is it your intention to overthrow the government of the United States?"
(that response is variously attributed, although both Evelyn Waugh and Gilbert Harding seem plausible, given their temperaments, but it's hard to imagine it successfully breaking the ice at US Immigration. it is sometimes given as "Not sole purpose of visit" which is also effective.)
> So how many cores is that for each member of the > Plan 9 community? :-) On a slightly more serious > note, would it be terribly naive to guess that > there would be more cores running Plan 9 in that > machine than there are throughout the rest of the > world?
not yet. coraid is responsible for ~4k cores running plan 9 right now. of course the maximum concentration in one place is jus a few hundred.
On Thu, 2008-11-06 at 06:54 -0500, erik quanstrom wrote: > > So how many cores is that for each member of the > > Plan 9 community? :-) On a slightly more serious > > note, would it be terribly naive to guess that > > there would be more cores running Plan 9 in that > > machine than there are throughout the rest of the > > world?
> not yet. coraid is responsible for ~4k cores > running plan 9 right now. of course the maximum > concentration in one place is jus a few hundred.
> Just booted Plan 9 on a 1024+16 node BG/P this week. .
> All credit to jmk, ericvh, and charles for this fantastic test run and > the existence of this new kernel.
> Plan is to double it just a few times until we hit 65536 or so. Then > the fun begins: turning on all cores, so we get to > 262144 cpus.
> Boots kinda fast, too. It's peppy. No current plans to run acme, > though we have run rio on a bg/l cpu node in 2007, as ericvh related > on his blog.
On 11/6/08, ron minnich <rminn...@gmail.com> wrote:
> Just booted Plan 9 on a 1024+16 node BG/P this week. .
> All credit to jmk, ericvh, and charles for this fantastic test run and > the existence of this new kernel.
> Plan is to double it just a few times until we hit 65536 or so. Then > the fun begins: turning on all cores, so we get to > 262144 cpus.
> Boots kinda fast, too. It's peppy. No current plans to run acme, > though we have run rio on a bg/l cpu node in 2007, as ericvh related > on his blog.
