FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Gentoo > Gentoo User

 
 
LinkBack Thread Tools
 
Old 07-26-2012, 08:39 AM
microcai
 
Default new machine : CPU : 22 nm vs 32 nm

2012/7/26 Philip Webb <purslow@ca.inter.net>:
> I've listed what's available at the local store,
> which I trust to stock reliable items, tho' I wouldn't ask their advice.
>
> All the AMD's are 32 nm , while the Intel recommended by one commenter
> -- Core i5-3570 4-Core Socket LGA1155, 3.4 Ghz, 6MB L3 Cache, 22 nm --
> is 22 nm : it costs CAD 230 & they have 3 in stock,
> which suggests demand, but not the most popular ( 9 in stock).
>
> Isn't 22 nm going to be faster than 32 nm ?
>
> In the same price range, AMD offers Bulldozer X8 FX-8150 (125W)
> 8-Core Socket AM3+, 3.6 GHz, 8Mb Cache, 32 nm ( CAD 220 , 2 in stock).
>
> How do you compare cores vs nm ?
> How far is cache size important ( 6 vs 8 MB )?

cache size is always the most important thing. cache miss is the top
reason your application slows down.

>
> When I built my current machine 2007, the CPU cost CAD 213 ,
> so both look as if they're in the right ballpark.
>
> --
> ========================,,======================== ====================
> SUPPORT ___________//___, Philip Webb
> ELECTRIC /] [] [] [] [] []| Cities Centre, University of Toronto
> TRANSIT `-O----------O---' purslowatchassdotutorontodotca
>
>
 
Old 07-26-2012, 10:34 AM
Volker Armin Hemmann
 
Default new machine : CPU : 22 nm vs 32 nm

Am Mittwoch, 25. Juli 2012, 16:24:42 schrieb Philip Webb:
> 120725 Volker Armin Hemmann wrote:
> > Am Mittwoch, 25. Juli 2012, 16:05:29 schrieb Philip Webb:
> >> I've listed what's available at the local store,
> >> which I trust to stock reliable items, tho' I wouldn't ask their advice.
> >> All the AMD's are 32 nm , while the Intel recommended by one commenter
> >> -- Core i5-3570 4-Core Socket LGA1155, 3.4 Ghz, 6MB L3 Cache, 22 nm --
> >> is 22 nm : it costs CAD 230 & they have 3 in stock,
> >> which suggests demand, but not the most popular ( 9 in stock).
> >> Isn't 22 nm going to be faster than 32 nm ?
> >
> > no
>
> In the absence of further explication, I'm likely to go with 22 nm .

because structure size has no influence on the performance - from a user point
of view.

In theory: smaller structers - less power needed - faster switching - so
higher clocks are possible.,

In practice: smaller structures - more leak current - not as much faster
clocks as hoped.

For a user there is no difference between a 3ghz 32nm or a 3ghz 22nm cpu. The
later one MIGHT use less power. But nothing is guaranteed.

>
> >> In the same price range, AMD offers Bulldozer X8 FX-8150 (125W)
> >>
> >> 8-Core Socket AM3+, 3.6 GHz, 8Mb Cache, 32 nm ( CAD 220 , 2 in
> >> stock).
> >>
> >> How do you compare cores vs nm ?
> >
> > who cares?
>
> These answers are not very helpful : does anyone have anything more so ?

because you don't. cores and nm are in no way related.

>
> >> How far is cache size important ( 6 vs 8 MB )?
> >
> > depends on the architecture.
>
> It occurs to me that a larger cache goes with more cores,
> so the last question is not so important.

no, really, this is the only question that makes sense.

And it depends on the cache structure. A 6mb L3 'victim' cache that only
caches stuff that is not in L2 and L1 might be better than a 8mb L3 cache that
also holds the same stuff as L2 and L1.

--
#163933
 
Old 07-26-2012, 01:50 PM
Michael Mol
 
Default new machine : CPU : 22 nm vs 32 nm

On Thu, Jul 26, 2012 at 9:43 AM, Евгений Пермяков <permeakra@gmail.com> wrote:
> On 07/26/2012 12:05 AM, Philip Webb wrote:
>>
>> I've listed what's available at the local store,
>> which I trust to stock reliable items, tho' I wouldn't ask their advice.
>>
>> All the AMD's are 32 nm , while the Intel recommended by one commenter
>> -- Core i5-3570 4-Core Socket LGA1155, 3.4 Ghz, 6MB L3 Cache, 22 nm --
>> is 22 nm : it costs CAD 230 & they have 3 in stock,
>> which suggests demand, but not the most popular ( 9 in stock).
>>
>> Isn't 22 nm going to be faster than 32 nm ?
>>
>> In the same price range, AMD offers Bulldozer X8 FX-8150 (125W)
>> 8-Core Socket AM3+, 3.6 GHz, 8Mb Cache, 32 nm ( CAD 220 , 2 in
>> stock).
>>
>> How do you compare cores vs nm ?
>> How far is cache size important ( 6 vs 8 MB )?
>>
>> When I built my current machine 2007, the CPU cost CAD 213 ,
>> so both look as if they're in the right ballpark.
>>
> If you're building new, performance-oriented box, you should take latest
> intel with AVX because of AVX. As I recall, recent gcc has support for avx,
> so some performance gain may be achieved.
> If you want home box, you may be interested in AMD A8 and similar chips, as
> they are reasonably fast and very chip

AMD parts have had AVX since the Bulldozer core release in Q3 2011.

>
> In any case, I'd put most of my money in 2-4 big 3Tb HDD's for media and 8+
> Gb fast memory, as modern browsers eat memory like crazies and CPU is
> usually fast enough. Decoding HDTV mkv's should occur on gpu block in any
> case, so general performance for most uses is irrelevant, as it was fast
> enough four yesrs earlier. Simply check, that you can offload HDTV decoding
> to GPU in your config.

Here, you're talking about either VDPAU or VAAAPI support. VDPAU is
only offered by nVidia cards, and even then you need to run the
proprietary driver. VAAPI is supported by Intel graphics and ATI's
proprietary driver. There's talk about using VDPAU as a backend to
VAAPI, but everything I read on the subect says things like
'potentially' and 'could be'.

--
:wq
 
Old 07-26-2012, 01:55 PM
Michael Mol
 
Default new machine : CPU : 22 nm vs 32 nm

On Thu, Jul 26, 2012 at 4:39 AM, microcai <microcai@fedoraproject.org> wrote:
> 2012/7/26 Philip Webb <purslow@ca.inter.net>:
>> I've listed what's available at the local store,
>> which I trust to stock reliable items, tho' I wouldn't ask their advice.
>>
>> All the AMD's are 32 nm , while the Intel recommended by one commenter
>> -- Core i5-3570 4-Core Socket LGA1155, 3.4 Ghz, 6MB L3 Cache, 22 nm --
>> is 22 nm : it costs CAD 230 & they have 3 in stock,
>> which suggests demand, but not the most popular ( 9 in stock).
>>
>> Isn't 22 nm going to be faster than 32 nm ?
>>
>> In the same price range, AMD offers Bulldozer X8 FX-8150 (125W)
>> 8-Core Socket AM3+, 3.6 GHz, 8Mb Cache, 32 nm ( CAD 220 , 2 in stock).
>>
>> How do you compare cores vs nm ?
>> How far is cache size important ( 6 vs 8 MB )?
>
> cache size is always the most important thing. cache miss is the top
> reason your application slows down.

Generally speaking, but it does depend on your workload; if you're
processing and referencing the same piece of memory over and over,
cache shines. If you're streaming through a lot of data...not so much.

Certainly, though, the former behavior is far more common.

--
:wq
 
Old 07-26-2012, 02:48 PM
Michael Mol
 
Default new machine : CPU : 22 nm vs 32 nm

On Thu, Jul 26, 2012 at 10:18 AM, Евгений Пермяков <permeakra@gmail.com> wrote:
> On 07/26/2012 05:50 PM, Michael Mol wrote:
>>
>> On Thu, Jul 26, 2012 at 9:43 AM, Евгений Пермяков <permeakra@gmail.com>
>> wrote:
>>>
>>> On 07/26/2012 12:05 AM, Philip Webb wrote:
>>>>
>>>> I've listed what's available at the local store,
>>>> which I trust to stock reliable items, tho' I wouldn't ask their advice.
>>>>
>>>> All the AMD's are 32 nm , while the Intel recommended by one commenter
>>>> -- Core i5-3570 4-Core Socket LGA1155, 3.4 Ghz, 6MB L3 Cache, 22 nm --
>>>> is 22 nm : it costs CAD 230 & they have 3 in stock,
>>>> which suggests demand, but not the most popular ( 9 in stock).
>>>>
>>>> Isn't 22 nm going to be faster than 32 nm ?
>>>>
>>>> In the same price range, AMD offers Bulldozer X8 FX-8150 (125W)
>>>> 8-Core Socket AM3+, 3.6 GHz, 8Mb Cache, 32 nm ( CAD 220 , 2 in
>>>> stock).
>>>>
>>>> How do you compare cores vs nm ?
>>>> How far is cache size important ( 6 vs 8 MB )?
>>>>
>>>> When I built my current machine 2007, the CPU cost CAD 213 ,
>>>> so both look as if they're in the right ballpark.
>>>>
>>> If you're building new, performance-oriented box, you should take latest
>>> intel with AVX because of AVX. As I recall, recent gcc has support for
>>> avx,
>>> so some performance gain may be achieved.
>>> If you want home box, you may be interested in AMD A8 and similar chips,
>>> as
>>> they are reasonably fast and very chip
>>
>> AMD parts have had AVX since the Bulldozer core release in Q3 2011.
>
> Are they already available in reasonable numbers on market?

http://www.cpubenchmark.net/cpu.php?cpu=AMD+FX-8120+Eight-Core

At $150, fitting into existing Socket AM3+ boards, that looks like the
best part for my money right now.

>>> In any case, I'd put most of my money in 2-4 big 3Tb HDD's for media and
>>> 8+
>>> Gb fast memory, as modern browsers eat memory like crazies and CPU is
>>> usually fast enough. Decoding HDTV mkv's should occur on gpu block in any
>>> case, so general performance for most uses is irrelevant, as it was fast
>>> enough four yesrs earlier. Simply check, that you can offload HDTV
>>> decoding
>>> to GPU in your config.
>>
>> Here, you're talking about either VDPAU or VAAAPI support. VDPAU is
>> only offered by nVidia cards, and even then you need to run the
>> proprietary driver. VAAPI is supported by Intel graphics and ATI's
>> proprietary driver.
>
> I do not see any problems with this. A blob in system is not best practice,
> of course, but it does not need any configuration and is not a performance
> bottle-neck, so there is no reason to care.

I only bring it up because some people do care. I'm running fglrx at
home right now. When I run nVdia, I run the nVidia drivers. In part
because I like accelerated video decoding (which a Geforce 210 does
wonderfully), in part because the nv, nouveau and radeon drivers
historically worked very poorly for me in 2D performance when faced
with multiple 1080p displays. They're always getting better, of
course.

>
> I personally would prefer AMD A8 if I can offload decoding to GPU unit there
> (not sure if I can, so won't change my box till next summer), but discrete
> video card will not be the most costly part in good non-gaming box, hard
> drives will, so again, what the matter?

Computer usage breaks down into more than gaming and non-gaming. My
"non-gaming" boxes at home tend to have their CPU, RAM or NICs as
their most expensive components, because that's where I need them to
perform better.


--
:wq
 
Old 07-28-2012, 12:30 AM
microcai
 
Default new machine : CPU : 22 nm vs 32 nm

CPU speed does not matter. what matters most is the I/O speed.

As far as I can tell, AMD chip suffered with a lot of I/O. Their
Hyper-transport seems not competitive with Intel's ring bus


2012/7/26 Michael Mol <mikemol@gmail.com>:
> On Thu, Jul 26, 2012 at 10:18 AM, Евгений Пермяков <permeakra@gmail.com> wrote:
>> On 07/26/2012 05:50 PM, Michael Mol wrote:
>>>
>>> On Thu, Jul 26, 2012 at 9:43 AM, Евгений Пермяков <permeakra@gmail.com>
>>> wrote:
>>>>
>>>> On 07/26/2012 12:05 AM, Philip Webb wrote:
>>>>>
>>>>> I've listed what's available at the local store,
>>>>> which I trust to stock reliable items, tho' I wouldn't ask their advice.
>>>>>
>>>>> All the AMD's are 32 nm , while the Intel recommended by one commenter
>>>>> -- Core i5-3570 4-Core Socket LGA1155, 3.4 Ghz, 6MB L3 Cache, 22 nm --
>>>>> is 22 nm : it costs CAD 230 & they have 3 in stock,
>>>>> which suggests demand, but not the most popular ( 9 in stock).
>>>>>
>>>>> Isn't 22 nm going to be faster than 32 nm ?
>>>>>
>>>>> In the same price range, AMD offers Bulldozer X8 FX-8150 (125W)
>>>>> 8-Core Socket AM3+, 3.6 GHz, 8Mb Cache, 32 nm ( CAD 220 , 2 in
>>>>> stock).
>>>>>
>>>>> How do you compare cores vs nm ?
>>>>> How far is cache size important ( 6 vs 8 MB )?
>>>>>
>>>>> When I built my current machine 2007, the CPU cost CAD 213 ,
>>>>> so both look as if they're in the right ballpark.
>>>>>
>>>> If you're building new, performance-oriented box, you should take latest
>>>> intel with AVX because of AVX. As I recall, recent gcc has support for
>>>> avx,
>>>> so some performance gain may be achieved.
>>>> If you want home box, you may be interested in AMD A8 and similar chips,
>>>> as
>>>> they are reasonably fast and very chip
>>>
>>> AMD parts have had AVX since the Bulldozer core release in Q3 2011.
>>
>> Are they already available in reasonable numbers on market?
>
> http://www.cpubenchmark.net/cpu.php?cpu=AMD+FX-8120+Eight-Core
>
> At $150, fitting into existing Socket AM3+ boards, that looks like the
> best part for my money right now.
>
>>>> In any case, I'd put most of my money in 2-4 big 3Tb HDD's for media and
>>>> 8+
>>>> Gb fast memory, as modern browsers eat memory like crazies and CPU is
>>>> usually fast enough. Decoding HDTV mkv's should occur on gpu block in any
>>>> case, so general performance for most uses is irrelevant, as it was fast
>>>> enough four yesrs earlier. Simply check, that you can offload HDTV
>>>> decoding
>>>> to GPU in your config.
>>>
>>> Here, you're talking about either VDPAU or VAAAPI support. VDPAU is
>>> only offered by nVidia cards, and even then you need to run the
>>> proprietary driver. VAAPI is supported by Intel graphics and ATI's
>>> proprietary driver.
>>
>> I do not see any problems with this. A blob in system is not best practice,
>> of course, but it does not need any configuration and is not a performance
>> bottle-neck, so there is no reason to care.
>
> I only bring it up because some people do care. I'm running fglrx at
> home right now. When I run nVdia, I run the nVidia drivers. In part
> because I like accelerated video decoding (which a Geforce 210 does
> wonderfully), in part because the nv, nouveau and radeon drivers
> historically worked very poorly for me in 2D performance when faced
> with multiple 1080p displays. They're always getting better, of
> course.
>
>>
>> I personally would prefer AMD A8 if I can offload decoding to GPU unit there
>> (not sure if I can, so won't change my box till next summer), but discrete
>> video card will not be the most costly part in good non-gaming box, hard
>> drives will, so again, what the matter?
>
> Computer usage breaks down into more than gaming and non-gaming. My
> "non-gaming" boxes at home tend to have their CPU, RAM or NICs as
> their most expensive components, because that's where I need them to
> perform better.
>
>
> --
> :wq
>
 
Old 07-28-2012, 12:56 AM
Dale
 
Default new machine : CPU : 22 nm vs 32 nm

Alecks Gates wrote:
> On Wed, Jul 25, 2012 at 9:32 PM, Dale <rdalek1967@gmail.com> wrote:
>> Michael Mol wrote:
>>> On Wed, Jul 25, 2012 at 8:21 PM, Dale <rdalek1967@gmail.com> wrote:
>>>> The point made about producing less heat with the smaller nm sounds
>>>> reasonable tho.
>>> Less heat with the smaller nm, but only if all other things remain equal!
>>>
>>> In reality, manufacturers use additional margin within their TDP to
>>> improve the product otherwise. Perhaps they increase the clock speed
>>> somewhat. Perhaps they increase the amount of on-die cache. Perhaps
>>> they reduce the instruction pipeline.
>>>
>>> AMD, for example, has tended to maintain keep something in the market
>>> for a 125W, 95W and 65W TDPs for several years. Each year, the
>>> functionality that used to be in a 125W TDP processor shows up in a
>>> 95W TDP processor, and the latest 125W TDP processor beats the pants
>>> off of last years'.
>>>
>>
>> I found this to be plain weird when I built my new rig. My old rig was
>> a AMD 2500+ single core system with 2Gbs of ram. It pulled about 400
>> watts or so for normal desktop use. A little more when compiling and
>> such. My new rig, AMD Phenom II 955 with four cores and 16Gbs of ram.
>> Heck, just a single core is much faster than my old rig. Thing is, the
>> new rig pulls less than half of what the old one pulls, WHILE
>> COMPILING. I can't recall the nm part but I think the CPU I got for my
>> old rig was supposed to be for laptop use.
>>
>> AMD sure is getting more efficient as you point out. I still wonder
>> where we will be in 10 years. Just how fast can they make them?
>>
>> Dale
>>
>> :-) :-)
>>
>> --
>> I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
>>
>>
> Definitely OT but that's surely not because of the CPU, or at least
> not only the CPU. Many people highly underestimate the value of a
> good and efficient power supply, which can make a huge difference.
> This is one of those things that companies such as Dell like to cut
> costs on because the average user neither sees the PSU specifications
> nor knows enough to ask about it. Of course, efficiency within the
> entire computer helps, but a bad power supply can really hurt your
> electric bill.
>
> On topic, AMD is definitely getting more efficient but mostly because
> that's where the technology is headed in general -- Intel seems to do
> a better job at efficiency per core but they also use hyper threading,
> whereas AMD is putting their bets into more physical cores. Yes, I'm
> going to say it again, but AMD is what you want for multitasking.
> They are switching their goals from high-performance cores to
> highly-concurrent CPUs, GPUs, and APUs.
>
> Concurrency is the future, it's just hard for a lot of people to think
> in such a way (and our technology doesn't leverage it to its full
> capacity). Just look at the human brain: "a maximum of 1,000 nerve
> impulses per second is possible. However, firing rates of 1 per second
> to 300-400 per second are more typical."[1] Basically the average
> neuron seems to be about only 300Hz, but there are trillions upon
> trillions of synapses within the brain. I don't know about you, but I
> am, allegedly, a fully-functioning, self-aware, intelligent being.
>
> [1] http://www.noteaccess.com/APPROACHES/ArtEd/ChildDev/1cNeurons.htm
>
>


It may not be JUST the CPU but the CPU is a big part of it. I might
add, I moved one hard drive from the old system to the new one. The
ones in my new rig that were new are about the same power wise, same
brand too. I actually have the same number of drives in my new rig as
was in my old rig. So that balances out. I might also add I have 16Gbs
of ram in my new rig but only 2Gbs of ram in the old one so that doesn't
fit either. As to the power supply, I build my own rig and I always
pick a good power supply that is efficient. The power supply is larger
in my new rig. I was thinking that the new rig would pull a bit more
power so I actually got a power supply that is really a little bit to
big. If anything, that would be a negative on my new rig not a
positive. The mobo is the only thing different other than the CPU
itself. Oh, let's not forget that my new case has those large 230mm
fans. Three of them to be exact. I wouldn't be surprised if they pull
about the same power tho. The CPU fan is larger on my new CPU tho. It
may pull a small amount more but not enough to really worry about much.
My video card is faster in the new rig too.

So, all in all, one would expect the new rig to pull more power not
less. It is a more powerful machine compared to my old rig. I did some
math, my new rig is overall 7 times faster than my old rig. I plan to
upgrade to a newer, faster CPU with more cores when prices come down a
bit more.

Dale

:-) :-)

P.S. I don't have a store bought system. I build mine from scratch.
While I would recommend Dell to someone who can't build their own, I
wouldn't buy one myself.

--
I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
 
Old 07-28-2012, 12:59 AM
Michael Mol
 
Default new machine : CPU : 22 nm vs 32 nm

On Fri, Jul 27, 2012 at 8:30 PM, microcai <microcai@fedoraproject.org> wrote:
> 2012/7/26 Michael Mol <mikemol@gmail.com>:
>> On Thu, Jul 26, 2012 at 10:18 AM, Евгений Пермяков <permeakra@gmail.com> wrote:
>>> On 07/26/2012 05:50 PM, Michael Mol wrote:
>>>>
>>>> On Thu, Jul 26, 2012 at 9:43 AM, Евгений Пермяков <permeakra@gmail.com>
>>>> wrote:
>>>>>
>>>>> On 07/26/2012 12:05 AM, Philip Webb wrote:
>>>>>>
>>>>>> I've listed what's available at the local store,
>>>>>> which I trust to stock reliable items, tho' I wouldn't ask their advice.
>>>>>>
>>>>>> All the AMD's are 32 nm , while the Intel recommended by one commenter
>>>>>> -- Core i5-3570 4-Core Socket LGA1155, 3.4 Ghz, 6MB L3 Cache, 22 nm --
>>>>>> is 22 nm : it costs CAD 230 & they have 3 in stock,
>>>>>> which suggests demand, but not the most popular ( 9 in stock).
>>>>>>
>>>>>> Isn't 22 nm going to be faster than 32 nm ?
>>>>>>
>>>>>> In the same price range, AMD offers Bulldozer X8 FX-8150 (125W)
>>>>>> 8-Core Socket AM3+, 3.6 GHz, 8Mb Cache, 32 nm ( CAD 220 , 2 in
>>>>>> stock).
>>>>>>
>>>>>> How do you compare cores vs nm ?
>>>>>> How far is cache size important ( 6 vs 8 MB )?
>>>>>>
>>>>>> When I built my current machine 2007, the CPU cost CAD 213 ,
>>>>>> so both look as if they're in the right ballpark.
>>>>>>
>>>>> If you're building new, performance-oriented box, you should take latest
>>>>> intel with AVX because of AVX. As I recall, recent gcc has support for
>>>>> avx,
>>>>> so some performance gain may be achieved.
>>>>> If you want home box, you may be interested in AMD A8 and similar chips,
>>>>> as
>>>>> they are reasonably fast and very chip
>>>>
>>>> AMD parts have had AVX since the Bulldozer core release in Q3 2011.
>>>
>>> Are they already available in reasonable numbers on market?
>>
>> http://www.cpubenchmark.net/cpu.php?cpu=AMD+FX-8120+Eight-Core
>>
>> At $150, fitting into existing Socket AM3+ boards, that looks like the
>> best part for my money right now.
>>
>>>>> In any case, I'd put most of my money in 2-4 big 3Tb HDD's for media and
>>>>> 8+
>>>>> Gb fast memory, as modern browsers eat memory like crazies and CPU is
>>>>> usually fast enough. Decoding HDTV mkv's should occur on gpu block in any
>>>>> case, so general performance for most uses is irrelevant, as it was fast
>>>>> enough four yesrs earlier. Simply check, that you can offload HDTV
>>>>> decoding
>>>>> to GPU in your config.
>>>>
>>>> Here, you're talking about either VDPAU or VAAAPI support. VDPAU is
>>>> only offered by nVidia cards, and even then you need to run the
>>>> proprietary driver. VAAPI is supported by Intel graphics and ATI's
>>>> proprietary driver.
>>>
>>> I do not see any problems with this. A blob in system is not best practice,
>>> of course, but it does not need any configuration and is not a performance
>>> bottle-neck, so there is no reason to care.
>>
>> I only bring it up because some people do care. I'm running fglrx at
>> home right now. When I run nVdia, I run the nVidia drivers. In part
>> because I like accelerated video decoding (which a Geforce 210 does
>> wonderfully), in part because the nv, nouveau and radeon drivers
>> historically worked very poorly for me in 2D performance when faced
>> with multiple 1080p displays. They're always getting better, of
>> course.
>>
>>>
>>> I personally would prefer AMD A8 if I can offload decoding to GPU unit there
>>> (not sure if I can, so won't change my box till next summer), but discrete
>>> video card will not be the most costly part in good non-gaming box, hard
>>> drives will, so again, what the matter?
>>
>> Computer usage breaks down into more than gaming and non-gaming. My
>> "non-gaming" boxes at home tend to have their CPU, RAM or NICs as
>> their most expensive components, because that's where I need them to
>> perform better.
>>
>
> CPU speed does not matter. what matters most is the I/O speed.
>
> As far as I can tell, AMD chip suffered with a lot of I/O. Their
> Hyper-transport seems not competitive with Intel's ring bus

(please don't top-post, especially if the thread's already been
primarily organized as bottom-post)

I hadn't read that, but remember that HyperTransport is intended for a
mesh architecture. In single-CPU systems, you'll only have one HT
link, the link between your CPU and your north bridge. In multi-CPU
systems, you'll have additional links between the CPUs. In systems
with many CPUs, you may even have a fully-connected mesh.

The I/O characteristics will greatly depend on the topology of your network.

That said, HyperTransport may just be getting old; when it came out,
it (and AMD's crossbar switch for memory management) beat the pants
off of Intel's SMP solution. Intel's solution ran at lower and lower
clock rates the more CPUs you added, and their first pass at multicore
gave each core its own port onto the memory bus, with predictably poor
results. Intel's had plenty of time to catch up, but with their
price-per-part, it's taken me a long time to pay much attention.

(It also doesn't help that Jon "Hannibal" Stokes stopped writing
detailed technical articles for Ars Technica; I sincerely miss him and
the precision and clarity of his writing on such arcane subjects.)

--
:wq
 
Old 07-28-2012, 12:08 PM
Pandu Poluan
 
Default new machine : CPU : 22 nm vs 32 nm

On Jul 28, 2012 8:03 AM, "Michael Mol" <mikemol@gmail.com> wrote:

>

> On Fri, Jul 27, 2012 at 8:30 PM, microcai <microcai@fedoraproject.org> wrote:


--- >8 Major Snippage


> > As far as I can tell, AMD chip suffered with a lot of I/O. Their

> > Hyper-transport seems not competitive with Intel's ring bus

>


Wasn't Intel's answer to HyperTransport is the QuickPath bus? IIRC, the ring bus is internal to a processor. (I could be wrong, though).


> (please don't top-post, especially if the thread's already been

> primarily organized as bottom-post)

>

> I hadn't read that, but remember that HyperTransport is intended for a

> mesh architecture. In single-CPU systems, you'll only have one HT

> link, the link between your CPU and your north bridge. In multi-CPU

> systems, you'll have additional links between the CPUs. In systems

> with many CPUs, you may even have a fully-connected mesh.

>

> The I/O characteristics will greatly depend on the topology of your network.

>

> That said, HyperTransport may just be getting old; when it came out,

> it (and AMD's crossbar switch for memory management) beat the pants

> off of Intel's SMP solution. Intel's solution ran at lower and lower

> clock rates the more CPUs you added, and their first pass at multicore

> gave each core its own port onto the memory bus, with predictably poor

> results. Intel's had plenty of time to catch up, but with their

> price-per-part, it's taken me a long time to pay much attention.

>


Again, I might be mistaken, but IIRC HyperTransport's throughput depends on how many channels are provided, so there's no theoretical limitation to its throughput, just practical considerations. (E.g., tracing issues).



> (It also doesn't help that Jon "Hannibal" Stokes stopped writing

> detailed technical articles for Ars Technica; I sincerely miss him and

> the precision and clarity of his writing on such arcane subjects.)

>


That makes the two of us bro... BTW, my handle there's "pepoluan", just in case you see it in the forums.


Rgds,
 

Thread Tools




All times are GMT. The time now is 03:27 PM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright 2007 - 2008, www.linux-archive.org