> Though I'm not certain, it seems reasonable to think that the key
> temperature is the junction temperature, which is much higher than
> the air temperature. With more efficient cooling, that is,
intel/amd spec chips as "Tcase", which makes sense, since they don't
really wanting you removing th
On Feb 28, 2013, at 7:51 AM, Christopher Samuel wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> On 13/02/13 02:02, Dr Stuart Midgley wrote:
>
>> I've started a blog to document the process I'm going through to
>> get our Phi's going.
>
> Has anyone seen a list of applications th
Jörg Saßmannshausen writes:
> Also, they don't cool
> the board to 20 °C but have it more or less at a higher temperature.
Though I'm not certain, it seems reasonable to think that the key
temperature is the junction temperature, which is much higher than
the air temperature. With more efficien
> Actually, since the thermal conductivity of the liquid is so much better
>than air, you probably don't need finned heat sinks. Heat transfer rates is
well, you don't need AS MUCH surface area and/or can tolerate lower flow.
even given better conductivity, and acknowledging greater heat capacity
On Thu, Feb 28, 2013 at 10:12:44AM +, Jörg Saßmannshausen wrote:
> Dear all,
>
> we are currently seriously looking into that (see my previous posting) and if
> I get the go-ahead the system should be running this summer.
>
> We are planning to convert 5 clusters which will bring down the en
On 02/28/2013 05:00 AM, Hearns, John wrote:
I think this has been discussed here before, but it is a pretty
innovative product:
http://www.theregister.co.uk/2013/02/28/wet_servers_cut_cooling_costs_research_leeds_university/
I really should just save my rants about immersive liquid coolin
> immersion cooling sounds appealingly unconventional, but if you think
> about the heatflow, you've still got to move it around. you still
> need a heatsink on the CPU with fins, some way to move the fluid past
> these fins and get them to the secondary heat exchanger. as with
> airflow man
On 02/28/2013 02:58 PM, Mark Hahn wrote:
>> consumption from 41 kW (electricity, probably same amount again for cooling)
> I really doubt it. there is something profoundly wrong if an
>HPC-type datacenter, with completely conventional servers, air/DX cooling
> runs at a PUE of more than about
Hi Mark,
check out the Iceotope website of how they are doing it:
http://www.iceotope.com/
They are only using convection and the mother board is mounted upright as
opposed to be the more convential horizontal mounting. Also, they don't cool
the board to 20 °C but have it more or less at a high
On 02/26/2013 10:19 PM, Mark Hahn wrote:
>> unless I bought a server as well. It turns out that you need a compatible
>> motherboard. Supermicro has a helpful list of there models here:
>>
>> http://www.supermicro.com/products/nfo/Xeon_Phi.cfm
>>
>> Our six month old systems were not compatible.
>
> At least with the Phi, within a single system, its completely
>cache-coherent... all 60 cores (240 hardware threads).
well, hardware-managed, so it's still time-slicing the core's resources.
(one FPU per core, for instance.)
> Am I stupid in thinking that they would also make very interesting w
> consumption from 41 kW (electricity, probably same amount again for cooling)
I really doubt it. there is something profoundly wrong if an
HPC-type datacenter, with completely conventional servers, air/DX cooling
runs at a PUE of more than about 1.3-1.4.
immersion cooling sounds appealingly
> http://www.theregister.co.uk/2013/02/28/wet_servers_cut_cooling_costs_research_leeds_university/
http://www.leeds.ac.uk/news/article/3374/wet_computer_server_could_cut_internet_waste
fairly ho-hum if you've ever seen an old Cray. OTOH I got a bit of
a chuckle from the very end of the video...
At least with the Phi, within a single system, its completely cache-coherent...
all 60 cores (240 hardware threads).
Am I stupid in thinking that they would also make very interesting web
applications servers - get your database on a solid state array
Beside the Phi and crank up those web server
In theory it is possibly, but I suspect in practise it isn't. The phi's are
diskless and with no way to attach the disk, you will need to do it via NFS or
some other network file system (or write your own file system over the pci
bridge).
You might be able to use iscsi… but I'm not sure… laten
Your guess is wrong.
Our experience with the Phi is streaks ahead of our experience with K10 (our
code is single precision). In the space of a few hours, you can have your
codes running on the phi in native mode. It is just a little computer sitting
along. ssh to it and your away. You can g
Dear all,
we are currently seriously looking into that (see my previous posting) and if
I get the go-ahead the system should be running this summer.
We are planning to convert 5 clusters which will bring down the energy
consumption from 41 kW (electricity, probably same amount again for cooling
I think this has been discussed here before, but it is a pretty innovative
product:
http://www.theregister.co.uk/2013/02/28/wet_servers_cut_cooling_costs_research_leeds_university/
Dr John Hearns | CFD Hardware Specialist | McLaren Racing Limited
McLaren Technology Centre, Chertsey Road, Woking,
18 matches
Mail list logo
