Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > aus.computers > #55156
| From | "Rod Speed" <rod.speed.aaa@gmail.com> |
|---|---|
| Newsgroups | aus.computers |
| Subject | Re: Device advice... |
| Date | 2016-09-08 11:12 +1000 |
| Message-ID | <e3bs8cFtd7rU1@mid.individual.net> (permalink) |
| References | (13 earlier) <nqje8n$565$1@gioia.aioe.org> <e36394Fibh7U1@mid.individual.net> <nqktqi$rp3$1@gioia.aioe.org> <e36gcuFl6nhU1@mid.individual.net> <nqq70h$10ve$1@gioia.aioe.org> |
"Computer Nerd Kev" <not@telling.you.invalid> wrote in message
news:nqq70h$10ve$1@gioia.aioe.org...
> Rod Speed <rod.speed.aaa@gmail.com> wrote:
>> "Computer Nerd Kev" <not@telling.you.invalid> wrote
>>> Rod Speed <rod.speed.aaa@gmail.com> wrote:
>>>> "Computer Nerd Kev" <not@telling.you.invalid> wrote
>>>>> Rod Speed <rod.speed.aaa@gmail.com> wrote:
>>>>>> "Computer Nerd Kev" <not@telling.you.invalid> wrote
>>>>>>> Rod Speed <rod.speed.aaa@gmail.com> wrote:
>>>>>>>> "Computer Nerd Kev" <not@telling.you.invalid> wrote
>>>>>>>>> Rod Speed <rod.speed.aaa@gmail.com> wrote:
>>>>>>>>>> "Computer Nerd Kev" <not@telling.you.invalid> wrote
>>>>>>>>>>> Rod Speed <rod.speed.aaa@gmail.com> wrote:
>>>>>>>>>>>> news16 <news16@woa.com.au> wrote
>>>>>>>>>>>>> Computer Nerd Kev wrote
>>>>>>>>>>>>
>>>>>>>>>>>>>> Looking at the specs for those things just confuses me.
>>>>>>>>>>>>>> Their limitations are so obviously artificial considering
>>>>>>>>>>>>>> the state of technology. Makes you wonder how much
>>>>>>>>>>>>>> a $1000 laptop really costs to make.
>>>>>>>>>>>>
>>>>>>>>>>>>> The same as the $40 ones.
>>>>>>>>>>>>
>>>>>>>>>>>> Wrong with the cpu used alone.
>>>>>>>>>>>
>>>>>>>>>>> For the manufacturer of the laptop, yes. But for the
>>>>>>>>>>> manufacturer
>>>>>>>>>>> of the CPU, one wonders whether, with the cost savings of mass
>>>>>>>>>>> production, it actually costs them much more to make the die of
>>>>>>>>>>> a high spec. consumer CPU compared to a low spec. one.
>>>>>>>>>>
>>>>>>>>>> Corse it does.
>>>>>>>>>
>>>>>>>>> Oh right. What extra costs are involved in the mass manufacture
>>>>>>>>> of high end consumer CPUs then?
>>>>>>>>
>>>>>>>> Obviously the number of active components in the
>>>>>>>> cpu where the failure of them renders the cpu useless.
>>>>>>
>>>>>>> So what is the difference in failure rate
>>>>>>> between high end and low end CPUs then?
>>>>>>
>>>>>> Massive.
>>>>>
>>>>> How massive?
>>>>
>>>> Massively massive.
>>>
>>> How big is that?
>>
>> Massively massive, as I said.
>
> Is that a quote from someone who's seen the figures?
>
>>>>>>> What's to say that the types of failures they experience
>>>>>>> are mostly constrained to single components rather
>>>>>>> than larger defects in the silicon die?
>>>>>>
>>>>>> Nothing. In fact most of the failures that make a particular cpu
>>>>>> only good for the bin are not the failure of a single component.
>>>>>
>>>>> So the number of active components in the CPU is not directly
>>>>> related to the number of failures during manufacture.
>>>>
>>>> Wrong, of course it is.
>>>
>>> But if most failures affect multiple components, that means that
>>> means that they aren't caused by a risk in every component of it
>>> being defective, adding up as the number of comonents increases,
>>> but a wider issue in the manufacture of the die.
>>
>> It actually means that a defect in the die
>> affects more than just one component.
>>
>>> That issue may not relate directly to the
>>> number of active components in the CPU.
>>
>> Of course it does when the highest performance
>> cpus have a lot more active components on them
>> than the lowest performance cpus.
>
> But if the defects aren't related to individual components,
> what is the actual cause of failure?
A defect in the die.
> How does it increase as complexity increases,
Obviously a particular defect in the die will affect more
components when the components are denser on the die.
> and to what extent?
>>>>>>>> And clearly it costs a lot more to make a cpu that produces
>>>>>>>> massive amounts of heat when running flat out, to get rid
>>>>>>>> of that heat compared with a much lower performance cpu
>>>>>>>> that doesn?t need anything like the same need to get rid of
>>>>>>>> that heat.
>>>>>>
>>>>>>> That's mainly a problem with heatsinking and fan cooling outside
>>>>>>> the CPU than something to increase the cost of the CPU itself.
>>>>>>
>>>>>> Wrong, as always. With low performance cpus that need no external
>>>>>> cooling at all, nothing adds to the cost of the cpu itself.
>>>>>>
>>>>>> With high performance cpus that will die instantly without
>>>>>> very effective external cooling, much of the cost of the cpu
>>>>>> is providing a mechanism for all that heat to be removed
>>>>>> for the cpu
>>>>
>>>>> That just sounds like a challenge in CPU design, how does
>>>>> it increase the manufacturing cost?
>>>>
>>>> Because none of that is there on the cpus that don?t need cooling.
>>>
>>> What is there to add to the manufacturing cost of CPUs that do need
>>> cooling?
>>
>> Like I said, getting the heat from where it is generated
>> to where it can be got away from where its generated.
>
> What method used to achieve that task results in the increased
> manufacturing cost?
Whatever method is used to get the heat away from where it is generated.
>> And there is an obvious cost in generating less
>> heat for the same performance in the first place.
>
> In design.
In manufacturing.
>>>>>> and with the automatic shut down that all high
>>>>>> performance cpus have that stops them bursting into flames
>>>>>> if the fan stops or the heatsink comes off etc etc etc.
>>>>
>>>>> A built-in temperature sensor costs amost nothing.
>>>>
>>>> It isnt the temperature sensor that's the cost.
>>>
>>> The other circuitry for automatic over-temperature shutdown
>>> wouldn't cost any more than the temperature sensor.
>>
>> Wrong, as always. Even just what is required to allow the cpu
>> to be turned off costs a lot more than a temperature sensing diode.
>
> Why is that?
Because that is the way it works.
>>>>>>> Otherwise the average cost of PC CPUs would have gone up over time
>>>>>>
>>>>>> They have.
>>>>>
>>>>> Unfortunately I can't find any accurate tables of historical
>>>>> CPU price data to disprove you with.
>>>>
>>>> There is no such thing.
>>>
>>> Actually I found some nice data sets which may be collated into
>>> historical price data, but would require a script to be written
>>> to process them. I might have a go at it one day, but not
>>> just for this argument.
>>
>> Your problem.
>>
>>>>> However the cost of computer equipment as a
>>>>> whole has been on a steep decline since the 90s:
>>>>> http://www.bls.gov/opub/ted/2015/long-term-price-trends-for-computers-tvs-and-related-items.htm
>>>>
>>>> Irrelevant to the cost of the cpu.
>>>
>>> Not entirely.
>>
>> Entirely.
>>
>>> It suggests a trend in that area of technology,
>>
>> But says nothing about what the price of the
>> cpu is doing with the highest performance cpus.
>
> It describes a trend which you say CPUs contradict
I didn’t say it as absolutely as that.
JUST with the higher performance CPUs.
Same with the higher performance video cards too, even
someone as stupid as you should have noticed that the
highest performance gaming video cards cost MUCH
more than they did when the first PCS showed up,
even than the highest performance early video cards
like the Hercules etc.
> (and not just high performance, as CPUs have increased in heat output,
Like hell they have with the lowest heat output cpus. Those
do in fact produce much less heat than the early PC cpus.
> you say that they should all have become more expensive).
Wrong, as always.
> I have data for the trend,
Like hell you do.
> do you have data for the contradiction?
Yep.
>> Even a terminal fuckwit such as yourself should be
>> able grasp that cpus with multiple cores have to
>> cost a lot more than cpus with single cores.
> Yep, but as technology has progressed over the years the
> cost of making CPUs with multiple cores is now lower than
> it was to produce single core CPUs back when they were
> most common.
Not with the highest performance cpus we are discussing.
Same with the highest performance video cards.
>>> so if you claim something breaks that trend
>>> you should supply data to support your claim.
>>
>> Even a terminal fuckwit such as yourself should be
>> able grasp that cpus with multiple cores have to
>> cost a lot more than cpus with single cores.
>>
>>>>> So I think it's unlikely that CPUs prices happen to have been
>>>>> going up while the cost of everything else has been going down.
>>>>
>>>> They have anyway with the highest performance desktop cpus.
>>>>
>>>>> Time for a citation from you.
>>>>
>>>> Nope.
>>>
>>> When is it time for a citation from you?
>>
>> Never. You are free to find whatever data you like
>> that shows what you claim. That?s how it works.
> I never claimed anything.
Everyone can see for themselves that you did and are still doing that.
> I suggested a possible situation ("makes you wonder").
You went on to make claims we can still see in the quoting.
>>>>>>> because old CPUs like the 386 produced far less heat than
>>>>>>> modern ones. In fact they often didn't even require heatsinking.
>>>>>>
>>>>>> And they are MUCH cheaper because of that.
>>>>>
>>>>> They're not made at all now.
>>>>
>>>> Even sillier and more pig ignorant than you usually manage.
>>>>
>>>> What the fuck do you 'think' is used in the lowest
>>>> performance tablets ? Intel i7 cpus that haven't done
>>>> very well in the quality checks ? Even a terminal fuckwit
>>>> such as yourself should be able to open one and
>>>> see that the cpu used is nothing even remotely
>>>> like an i7.
>>>
>>> Read what you responded to.
>>
>> Go and fuck yourself, again.
>
> It's that attitude that leads to these poor arguments.
The only poor arguments are yours.
>>> You said that old CPUs like the 386 "are MUCH cheaper".
>>> Naturally you can't buy a new 386, 486 or 268 these days.
>>
>> Irrelevant whether they are new or not.
>
> Oh yes, your confusion is always irrelevent.
You never could bullshit your way out of a wet paper bag.
>>>>> But they _weren't_ MUCH cheaper
>>>>> compared to today's CPUs either.
>>>>
>>>> Wrong, as always.
>>>
>>> Unfortunately the CPU price data I found only goes back to 2000,
>>
>> So is completely fucking useless.
> Nope,
Yep.
> it covers major PC CPU models as their heat output increased.
But doesn’t cover the highest performance cpus
being discussed.
>>> but I know they weren't MUCH cheaper,
>>> especially with inflation accounted for,
>>
>> Wrong, as always.
>>
>>> so unless you have some data to back
>>> you up this isn't going to get anywhere.
>>
>> You never ever get anywhere.
>
> Where have you got then?
Pissing on you from a great height, as always.
>>>>>>>> If they all cost the same to make, even someone as stupid
>>>>>>>> as you should be able to grasp that all laptops and tablets
>>>>>>>> and phones would all have the highest performance cpus
>>>>>>>> because there would be no point in the separate manufacturing
>>>>>>>> lines that are used to produce the much cheaper cpus.
>>>>>>
>>>>>>> Yes there would. So that consumers who can afford it are
>>>>>>> enticed to buy more expensive models with their fancier CPUs.
>>>>>>
>>>>>> Even sillier than you usually manage.
>>>>
>>>>> For someone so abusive, it's amazing how you
>>>>> appear to make a very honest businessman.
>>>>
>>>> You never could bullshit your way out of a wet paper bag.
>>>
>>> I see you're just as modest here as in our other little thread.
>>
>> You never could bullshit your way out of a wet paper bag.
>
> Yay! I can push that button here as well!
You never could bullshit your way out of a wet paper bag.
>>>>>>> Much of the extra price they pay being profit
>>>>>>
>>>>>> Even more pig ignorant than you usually manage.
>>>>>>
>>>>>>> because the actual manufacturing cost of the
>>>>>>> components may not be much higher.
>>>>>>
>>>>>> It is in fact MUCH higher.
>>>>>
>>>>> Why?
>>>>
>>>> Because they are MUCH harder to make
>>>
>>> Why?
>>
>> Because it is much more difficult to make
>> them so small so they will fit and still work
>> and much harder to get all that heat out too.
> To what extent does that affect manufacturing cost?
Its obviously much harder to make them and have them still work.
> Is it to a similar proportion to the increased
> cost of the CPU to consumers?
Irrelevant.
>>>> and require
>>>> MUCH more expensive equipment to make them.
>>>
>>> What equipment?
>>
>> The equipment used to make them, fuckwit.
>
> What equipment is that?
The equipment used to make them, fuckwit.
> How does the cost of it compare to the cost
> of the equipment used to make low-end CPUs?
Much more expensive.
Back to aus.computers | Previous | Next — Previous in thread | Next in thread | Find similar
Device advice... Jeßus <j@j.invalid> - 2016-08-31 15:13 +1000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-08-31 15:51 +1000
Re: Device advice... Jeßus <j@j.invalid> - 2016-09-01 05:29 +1000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-01 09:10 +1000
Re: Device advice... Frank Slootweg <this@ddress.is.invalid> - 2016-09-01 15:38 +0000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-02 05:28 +1000
Re: Device advice... Frank Slootweg <this@ddress.is.invalid> - 2016-09-01 19:57 +0000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-02 10:33 +1000
Re: Device advice... Frank Slootweg <this@ddress.is.invalid> - 2016-09-02 18:50 +0000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-03 11:05 +1000
Re: Device advice... Pelican <water-birds@sea.somewhere.org.ir> - 2016-08-31 16:07 +1000
Re: Device advice... SolomonW <SolomonW@citi.com> - 2016-08-31 19:41 +1000
Re: Device advice... Jeßus <j@j.invalid> - 2016-09-01 05:34 +1000
Re: Device advice... not@telling.you.invalid (Computer Nerd Kev) - 2016-08-31 23:30 +0000
Re: Device advice... Pelican <water-birds@sea.somewhere.org.ir> - 2016-09-01 09:49 +1000
Re: Device advice... Petzl <petzlx@gmail.com> - 2016-09-01 13:05 +1000
Re: Device advice... Jeßus <j@j.invalid> - 2016-09-07 08:53 +1000
Re: Device advice... news16 <news16@woa.com.au> - 2016-09-03 03:06 +0000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-03 15:29 +1000
Re: Device advice... not@telling.you.invalid (Computer Nerd Kev) - 2016-09-04 01:23 +0000
Re: Device advice... news16 <news16@woa.com.au> - 2016-09-04 03:05 +0000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-04 20:13 +1000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-04 20:06 +1000
Re: Device advice... not@telling.you.invalid (Computer Nerd Kev) - 2016-09-04 22:55 +0000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-05 12:17 +1000
Re: Device advice... not@telling.you.invalid (Computer Nerd Kev) - 2016-09-05 02:57 +0000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-05 15:26 +1000
Re: Device advice... Computer Nerd Kev <not@telling.you.invalid> - 2016-09-05 09:35 +0000
Re: Device advice... "SG1" <lost@the.races.com> - 2016-09-05 21:19 +1000
Re: Device advice... Frank Slootweg <this@ddress.is.invalid> - 2016-09-05 19:09 +0000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-06 06:35 +1000
Re: Device advice... not@telling.you.invalid (Computer Nerd Kev) - 2016-09-05 23:07 +0000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-06 10:19 +1000
Re: Device advice... not@telling.you.invalid (Computer Nerd Kev) - 2016-09-07 23:14 +0000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-08 11:12 +1000
Re: Device advice... F Murtz <haggisz@hotmail.com> - 2016-09-06 12:56 +1000
Re: Device advice... not@telling.you.invalid (Computer Nerd Kev) - 2016-09-06 23:09 +0000
Re: Device advice... Jeßus <j@j.invalid> - 2016-09-07 09:11 +1000
Re: Device advice... "Rod Speed" <rod.speed.aaa@gmail.com> - 2016-09-07 10:50 +1000
Re: Device advice... Jeßus <j@j.invalid> - 2016-09-07 08:52 +1000
Re: Device advice... SolomonW <SolomonW@citi.com> - 2016-09-01 22:38 +1000
Re: Device advice... Jeßus <j@j.invalid> - 2016-09-07 08:55 +1000
Re: Device advice... Jeßus <j@j.invalid> - 2016-09-01 05:33 +1000
Re: Device advice... Frank Slootweg <this@ddress.is.invalid> - 2016-09-01 15:20 +0000
Re: Device advice... Jeßus <j@j.invalid> - 2016-09-07 08:57 +1000
Re: Device advice... Jeßus <j@j.invalid> - 2016-09-08 16:10 +1000
csiph-web