From mboxrd@z Thu Jan 1 00:00:00 1970 MIME-Version: 1.0 In-Reply-To: References: <969a33151abb4d97c8e84351602bfc85@gmx.de> <94dacb4b1569c9a8fbd0afce950a4436@ladd.quanstro.net> <24b3015141c19676572513e89c82fdd6@ladd.quanstro.net> Date: Tue, 7 Jun 2011 17:06:11 +0200 Message-ID: From: =?ISO-8859-1?Q?Gabriel_D=EDaz_L=F3pez_de_la_llave?= To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net> Content-Type: multipart/alternative; boundary=20cf3074d4987ed50c04a5208f45 Subject: Re: [9fans] Hey, new to this. Trying to get plan9 to work in a VM. Topicbox-Message-UUID: ecf4d8c0-ead6-11e9-9d60-3106f5b1d025 --20cf3074d4987ed50c04a5208f45 Content-Type: text/plain; charset=ISO-8859-1 Hello AMD and IBM/Sony think a bit different with their Fusion and Cell processors+gpu integrated, no? slds. On Tue, Jun 7, 2011 at 4:38 PM, Josh Marshall < joshua.r.marshall.1991@gmail.com> wrote: > A system running more hardware will, for all practical purposes, use more > energy. What this would do is increase the efficiency of that power use. > Say you're using a single threaded indexing program, and its indexing a very > slow medium. Why use a CPU processor when you can idle them, and idle most > of the other GPU processors and just use the one? This is mainly for max > hardware utilization though. > > In the VERY long run, I'm seeing thing trending towards very distributed > models. As system resources grow, I believe it will become practical to > "network" within a system. This can manifest itself in two ways. First, is > that due to multi-core systems slowly changing to many-core systems, a > networking model is very scalable and with so many things to break, the > fault tolerance will become a must. This could allow then for computer > systems to continue their march towards a more biological like organization, > like a multi-cellular organism. This will likely be abstracted to > programmers and users, but on a hardware level, it allows for variable > redundancy, extreme fault tolerance, internal and external networking > models, and any few components which break will have no or minimal impact on > the stability and usability of the system. This is WAY WAY in the future, > but that's where I imagine it going and this could be a step in that > direction. Was that as coherent as it should be? I'm still playing with > this in the back of my head, so its by no means well planned :P I'd be more > than happy to talk to someone about this, because no one at my university > knows this area--our math and CS/CIS departments are feeble. > > > On Mon, Jun 6, 2011 at 9:19 PM, erik quanstrom wrote: > >> > Well, two reasons come to my mind immediately. First, I'd be cool. >> Second, >> > the wattage you listed is the max wattage, not the idle or light load >> > wattage which would likely be used. Per processing element, GPUs use >> less >> > power, and you get more processing power per watt than a CPU under >> certain >> > loads. >> >> i'd sure like a reference to a case where a system with a gpu draws less >> power than the same system without. it's not like you can turn the cpu >> off. >> >> > This concept could be taken as far as to bring all processing off >> > specialized areas for general purpose use, allowing potentially for an >> > internally distributed system with high regularity, fault tolerance, >> etc. >> > That's on the far end, but not to be totally discounted. >> >> please explain. how is a machine more of any of these things than >> a regular multi-core machine? >> >> - erik >> >> > --20cf3074d4987ed50c04a5208f45 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Hello

AMD and IBM/Sony think a bit different with their = Fusion and Cell processors+gpu integrated, no?

sld= s.

On Tue, Jun 7, 2011 at 4:38 PM, Josh M= arshall <joshua.r.marshall.1991@gmail.com> wrote:
A system running more hardware will, for al= l practical purposes, use more energy.=A0 What this would do is increase th= e efficiency of that power use.=A0 Say you're using a single threaded i= ndexing program, and its indexing a very slow medium.=A0 Why use a CPU proc= essor when you can idle them, and idle most of the other GPU processors and= just use the one?=A0 This is mainly for max hardware utilization though.
In the VERY long run, I'm seeing thing trending towards very distri= buted models.=A0 As system resources grow, I believe it will become practic= al to "network" within a system.=A0 This can manifest itself in t= wo ways.=A0 First, is that due to multi-core systems slowly changing to man= y-core systems, a networking model is very scalable and with so many things= to break, the fault tolerance will become a must.=A0 This could allow then= for computer systems to continue their march towards a more biological lik= e organization, like a multi-cellular organism.=A0 This will likely be abst= racted to programmers and users, but on a hardware level, it allows for var= iable redundancy, extreme fault tolerance, internal and external networking= models, and any few components which break will have no or minimal impact = on the stability and usability of the system.=A0 This is WAY WAY in the fut= ure, but that's where I imagine it going and this could be a step in th= at direction.=A0 Was that as coherent as it should be?=A0 I'm still pla= ying with this in the back of my head, so its by no means well planned :P= =A0 I'd be more than happy to talk to someone about this, because no on= e at my university knows this area--our math and CS/CIS departments are fee= ble.


On Mon, Jun 6, 2011 at 9:19 PM, erik quanstr= om <quanstro@quanstro.net> wrote:
> Well, two reasons come to my mind immediately. =A0First, I'd = be cool. =A0Second,
> the wattage you listed is the max wattage, not the idle or light load<= br> > wattage which would likely be used. =A0Per processing element, GPUs us= e less
> power, and you get more processing power per watt than a CPU under cer= tain
> loads.

i'd sure like a reference to a case where a system with a gpu dra= ws less
power than the same system without. =A0it's not like you can turn the c= pu
off.

> This concept could be taken as far as to bring all processing off
> specialized areas for general purpose use, allowing potentially for an=
> internally distributed system with high regularity, fault tolerance, e= tc.
> That's on the far end, but not to be totally discounted.

please explain. =A0how is a machine more of any of these things than<= br> a regular multi-core machine?

- erik



--20cf3074d4987ed50c04a5208f45--