It's a common phenomenon that things like hard drive sizes, RAM sizes
and processor speeds tend to grow at a superlinear rate over time in
home computers. (I won't say "exponential" because someone recently
complained about that, and I don't have hard data to back up that claim.)
Very typically, for example, each new hard drive one buys tends to be
larger than the combined sizes of all hard drives that person has bought
in the past. (I'm of course talking about the average computer user rather
than those who buy hard drives like candy, eg. because they need them for
multimedia projects or something.)

  Monitor screen sizes do not seem to follow this trend, however, and
seem to grow a lot more linearly.

  The very first monitor I owned (which I bought with my own money, and
hence was fully mine) was a 15-inch monitor. The next one was a 17-inch,
and the next one a 19-inch monitor (all of them CRTs). I recently put that
latter one out of its misery and replaced it with a 24-inch LED monitor.
(Although this last one is a 19:6 monitor while all the previous ones were
4:3, this last one is a bit larger even on the vertical dimension than the
old 19-inch one.)

  While I skipped the 21-inch step in between the last two, the growth has
nevertheless been very linear. (In fact, I had the 19-inch monitor for quite
a long time. I did not want to jump to LCDs before they had fully matured.
Finally I decided it was time. While this has a few drawbacks compared to
my good old 19-inch CRT, I'm quite happy with it. It's also great for
watching movies and playing games. The difference in size and aspect ratio
is just amazing.)

  Physical weight has been more varied. The weight clearly increased with
monitor size, but then had a huge drop with the last one. (The 19-inch CRT
is so heavy that you could break your back trying to carry it, while this
24-inch LED monitor is so light that you can *easily* carry it with one
hand.)

-- 
                                                          - Warp

Post a reply to this message

On 03/07/2012 02:36 PM, Warp wrote:
>    It's a common phenomenon that things like hard drive sizes, RAM sizes
> and processor speeds tend to grow at a superlinear rate over time in
> home computers. (I won't say "exponential" because someone recently
> complained about that, and I don't have hard data to back up that claim.)
> Very typically, for example, each new hard drive one buys tends to be
> larger than the combined sizes of all hard drives that person has bought
> in the past. (I'm of course talking about the average computer user rather
> than those who buy hard drives like candy, eg. because they need them for
> multimedia projects or something.)
>
>    Monitor screen sizes do not seem to follow this trend, however, and
> seem to grow a lot more linearly.
>
>    The very first monitor I owned (which I bought with my own money, and
> hence was fully mine) was a 15-inch monitor. The next one was a 17-inch,
> and the next one a 19-inch monitor (all of them CRTs). I recently put that
> latter one out of its misery and replaced it with a 24-inch LED monitor.
> (Although this last one is a 19:6 monitor while all the previous ones were
> 4:3, this last one is a bit larger even on the vertical dimension than the
> old 19-inch one.)
>
>    While I skipped the 21-inch step in between the last two, the growth has
> nevertheless been very linear. (In fact, I had the 19-inch monitor for quite
> a long time. I did not want to jump to LCDs before they had fully matured.
> Finally I decided it was time. While this has a few drawbacks compared to
> my good old 19-inch CRT, I'm quite happy with it. It's also great for
> watching movies and playing games. The difference in size and aspect ratio
> is just amazing.)
>
>    Physical weight has been more varied. The weight clearly increased with
> monitor size, but then had a huge drop with the last one. (The 19-inch CRT
> is so heavy that you could break your back trying to carry it, while this
> 24-inch LED monitor is so light that you can *easily* carry it with one
> hand.)
>

Pretty much the path I took as well ... one big difference! At first I 
thought it was absurd to have two displays, but now I find I couldn't 
even imagine settling for just one display. The extra "real estate" is a 
must for development work.

Post a reply to this message

James Holsenback <nom### [at] nonecom> wrote:
> Pretty much the path I took as well ... one big difference! At first I 
> thought it was absurd to have two displays, but now I find I couldn't 
> even imagine settling for just one display. The extra "real estate" is a 
> must for development work.

  I remember the time when a graphics card with support for two displays
was luxury (and a selling point). Nowadays they probably don't even make
graphics cards without support for (at least) two monitors. (And this has
been so for a pretty long time now.)

  In fact, many modern graphics cards seem to have support for 3 monitors
(typically with 2 DVI ports and one HDMI port).

-- 
                                                          - Warp

Post a reply to this message

On 3/7/2012 11:36 AM, Warp wrote:
>    It's a common phenomenon that things like hard drive sizes, RAM sizes
> and processor speeds tend to grow at a superlinear rate over time in
> home computers. (I won't say "exponential" because someone recently
> complained about that, and I don't have hard data to back up that claim.)

Yay! My complaining finally had an actual effect on things!  Although as 
far as I'm aware you would have been correct in using the term 
exponential here.  If I recall correctly the growth of many computer 
components has been following a roughly exponential growth pattern.  You 
can see some examples here:

http://en.wikipedia.org/wiki/Moore's_law#Other_formulations_and_similar_laws

>
>    Monitor screen sizes do not seem to follow this trend, however, and
> seem to grow a lot more linearly.

Interestingly, the above page lists monitors as also improving 
exponentially as measured in pixels per dollar.  I wouldn't have guessed 
this though, since my intuition on the matter was the same as yours.

Post a reply to this message

Am 07.03.2012 20:57, schrieb James Holsenback:

> Pretty much the path I took as well ... one big difference! At first I
> thought it was absurd to have two displays, but now I find I couldn't
> even imagine settling for just one display. The extra "real estate" is a
> must for development work.

Same here. The only substitute for screen real estate is screen real estate.

With one 24" 16:10 (1920x1200) and one 19" 4:3 (1280x1024) at my 
disposal, I found that trying to work on a single 17" 4:3 display gives 
me claustrophobia these days. No kidding.

Post a reply to this message

Am 07.03.2012 21:15, schrieb Warp:
> James Holsenback<nom### [at] nonecom>  wrote:
>> Pretty much the path I took as well ... one big difference! At first I
>> thought it was absurd to have two displays, but now I find I couldn't
>> even imagine settling for just one display. The extra "real estate" is a
>> must for development work.
>
>    I remember the time when a graphics card with support for two displays
> was luxury (and a selling point). Nowadays they probably don't even make
> graphics cards without support for (at least) two monitors. (And this has
> been so for a pretty long time now.)

Anyone recall the days when a dual-screen workplace meant plugging in a 
monochrome graphics adapter next to your VGA card?

And yes, I did that back then. It made a nice debug output device - 
Turbo Pascal 6.0 even included native support for that.

Post a reply to this message

clipka <ano### [at] anonymousorg> wrote:
> 19" 4:3 (1280x1024)

  With a 19-inch 4:3 monitor one should not be content with anything
less than 1600x1200 (at 85 Hz if it's a CRT).

-- 
                                                          - Warp

Post a reply to this message

On 03/07/2012 05:12 PM, Warp wrote:
> clipka<ano### [at] anonymousorg>  wrote:
>> 19" 4:3 (1280x1024)
>
>    With a 19-inch 4:3 monitor one should not be content with anything
> less than 1600x1200 (at 85 Hz if it's a CRT).
>
LOL ... you're the most consummate dissenter I've EVER met ... you slay 
me dude.

Post a reply to this message

Am 07.03.2012 23:12, schrieb Warp:
> clipka<ano### [at] anonymousorg>  wrote:
>> 19" 4:3 (1280x1024)
>
>    With a 19-inch 4:3 monitor one should not be content with anything
> less than 1600x1200 (at 85 Hz if it's a CRT).

It's okay for a secondary display, given that it's a few years old 
already, and had served me as my primary display before I retired my 17" 
4:3 in favor of the 24" 16:10. (A 1600x1200 would indeed be better 
suited of course, as it would then have the same vertical resolution as 
the primary display, but given that I'm rather short on money these days 
it doesn't warrant buying a new display.)

Post a reply to this message

You also missed the drop (sic!) in resolution lately: It went all the
way to 1920x1200 (limits of DVI-I bandwidth unless it's a dual link:
(Single) WUXGA (1,920 × 1,200) @ 60 Hz).
But it now is reduced to 1920 x 1080, because this is the resolution
needed by HDTV... you would have a very hard time finding a 1200 now.

Also, DVI is now tagged as obsolet and would be replaced by HDMI (with
identification & all the DRM) and DisplayPort, should the industry
succeeds... (yet, there is DP 1.0, 1.1, 1.2 and also DP++ (passive
compatibility with single link DVI), which means the consumers will
still get fooled by naughty sellers: Yes, it's HD-ready... bummer!)

Post a reply to this message

>    Monitor screen sizes do not seem to follow this trend, however, and
 > seem to grow a lot more linearly.

Assuming the supported resolution is changing proportionately, this is 
actually quadratic growth in terms of number of pixels.

This is probably limited by graphics hardware in some way (although I'm 
not sure how fast that grows in "power"). It's also limited by the fact 
that popular OSs such as Windows still use pixels as a unit of 
measurement, so there's limited benefit to higher resolutions as it 
shrinks the size of important objects on screen.

  - Slime

Post a reply to this message

Slime <pov### [at] slimelandcom> wrote:
>  >    Monitor screen sizes do not seem to follow this trend, however, and
>  > seem to grow a lot more linearly.

> Assuming the supported resolution is changing proportionately, this is 
> actually quadratic growth in terms of number of pixels.

  The standard measurement for screen size is the length of the diagonal.

> This is probably limited by graphics hardware in some way (although I'm 
> not sure how fast that grows in "power"). It's also limited by the fact 
> that popular OSs such as Windows still use pixels as a unit of 
> measurement, so there's limited benefit to higher resolutions as it 
> shrinks the size of important objects on screen.

  Resolution has little to do with screen size.

-- 
                                                          - Warp

Post a reply to this message

I don't know about physical size, but there's something I've always 
wondered about:

Let me pick an example at random. When USB flash drives first came out, 
64MB was about the biggest drive you would possibly buy. Today you can 
easily pick up a 4GB drive that costs less than the price of having it 
physically delivered to your house.

So... why didn't they just make the 4GB drives to start with? Why did 
they have to start by making 64MB drives, and then starting to make 
128MB drives, and then moving on to 256MB drives, and so forth? Why 
couldn't they just go directly to 4GB? What enables them to make those 
today but prevented them from making them back then?

Why do you have to design, test, manufacture and sell a 64MB drive 
before you can attempt to make a 128MB one? How does the former help you 
do the latter? Why can't you just jump straight to 4GB? (Or perhaps even 
more than that?)

I can't think of any /technical/ reason. (Besides "that's how it's 
done".) The only rational reason I can think of is that if you keep 
putting out slightly better devices year after year, people are going to 
keep upgrading their stuff, and that gives you income. If you just went 
straight out and sold the best possible device, then once everyone has 
got one, you'd have nothing new to sell to them, and you'd have no money.

Then again, cars don't improve in performance at all. Today's cars have 
performance within a few percent of cars made 40 years ago. And yet, 
people still buy cars.

I suppose a car is different to a piece of technology. Cars wear out. 
Cars are status symbols. Cars are fashion accessories. Some of that 
applies to a very limited extent to computers or phones, but it really 
doesn't apply to something like a USB flash drive.

On the other hand, pens and pencils don't improve in performance either, 
and they still sell plenty of those. Flash drives could be considered as 
being like pens. (Hell, they even /call them/ "pen drives" sometimes...)

Post a reply to this message

Invisible <voi### [at] devnull> wrote:
> I don't know about physical size, but there's something I've always
> wondered about:
>
> Let me pick an example at random. When USB flash drives first came out,
> 64MB was about the biggest drive you would possibly buy. Today you can
> easily pick up a 4GB drive that costs less than the price of having it
> physically delivered to your house.
>
> So... why didn't they just make the 4GB drives to start with? Why did
> they have to start by making 64MB drives, and then starting to make
> 128MB drives, and then moving on to 256MB drives, and so forth? Why
> couldn't they just go directly to 4GB? What enables them to make those
> today but prevented them from making them back then?
>

When they started making flash memory chips, the manufacturing technology was
not as advanced as it is now.
Structure sizes were much bigger then, meaning for each mm² of chip area you'd
have much less memory cells.
Also the manufacturing yield was probably much less due to contamination in the
manufacturing process.
What that means is that the bigger the area of a single chip, the higher the
probability of contamination and the fault of the chip. This has a parabolic
relation. This meant that chip sizes were kept small to increase yield.
Nowadays processes have much better yield and much smaller structure sizes which
means you can have a higher storage density.

Invisible <voi### [at] devnull> wrote:
> Why do you have to design, test, manufacture and sell a 64MB drive
> before you can attempt to make a 128MB one? How does the former help you
> do the latter? Why can't you just jump straight to 4GB? (Or perhaps even
> more than that?)

If I'm correct you do work in a research facility. So even you're not a
researcher you should know how research works.
That's exactly what happens with chip technology.
You start researching, get a workable product with low performance. To finace
further research, you start selling it, knowing that you can do better.
Performance keeps improving, prices drop and so on.

Just compare the chip size of an old simm memory card to a DDR3 memory module.
The chip case will be about the same, but the contained memory is magnitudes
lager.

Post a reply to this message

>> So... why didn't they just make the 4GB drives to start with?
>>
>
> When they started making flash memory chips, the manufacturing technology was
> not as advanced as it is now.

So the question "why wasn't hardware more advanced back then?" has the 
answer "manufacturing technology was not as advanced back then". Which 
immediately provokes the question "why wasn't the manufacturing 
technology more advanced back then?"

> Nowadays processes have much better yield and much smaller structure sizes which
> means you can have a higher storage density.

I'm fairly sure that back when they were selling 64MB flash drives, it 
was already common for a PC to have multiple GB of RAM. So clearly the 
feature sizes already existed.

> If I'm correct you do work in a research facility. So even you're not a
> researcher you should know how research works.
> You start researching, get a workable product with low performance. To finance
> further research, you start selling it, knowing that you can do better.
> Performance keeps improving, prices drop and so on.

I can understand how something /complicated/ - like, say, a processor - 
would take time. Inventing a new feature takes a lot of R&D. But just 
making the transistors smaller? I don't really see why they have to 
shrink in tiny little baby steps. Why not just go the whole hog straight 
away?

Part of the answer is manufacturing process, which just takes us back to 
where we started - now we're arguing about a different bit of hardware. ;-)

Post a reply to this message

Invisible <voi### [at] devnull> wrote:
> So... why didn't they just make the 4GB drives to start with?

  You could just as well ask why don't they make 64TB drives right now.

-- 
                                                          - Warp

Post a reply to this message

On 08/03/2012 11:59 AM, Warp wrote:
> Invisible<voi### [at] devnull>  wrote:
>> So... why didn't they just make the 4GB drives to start with?
>
>    You could just as well ask why don't they make 64TB drives right now.

Yep, I'd be interested to hear the answer to that one too.

And why they aren't making petabit network cards yet. And why nobody 
sells gigabit USB connectors. And why there's no 64-core CPUs yet. [That 
last one at least has a plausible explanation: There's no software yet.]

Basically, why do all technological devices have to improve in small 
increments? Why can't they take big strides?

Sometimes there's a plausible answer. Nobody needs it. There's no 
software for it. The laws of physics make it difficult to cross this 
particular boundary. But a lot of the time, it just seems arbitrary...

Post a reply to this message

Invisible <voi### [at] devnull> wrote:
> On 08/03/2012 11:59 AM, Warp wrote:
> > Invisible<voi### [at] devnull>  wrote:
> >> So... why didn't they just make the 4GB drives to start with?
> >
> >    You could just as well ask why don't they make 64TB drives right now.

> Yep, I'd be interested to hear the answer to that one too.

  Did you know that technology has to first be developed before it can
be used?

-- 
                                                          - Warp

Post a reply to this message

>>>     You could just as well ask why don't they make 64TB drives right now.
>
>> Yep, I'd be interested to hear the answer to that one too.
>
>    Did you know that technology has to first be developed before it can
> be used?

OK. So why spend time and money developing a 32TB drive when you could 
just go develop a 64TB drive instead?

Post a reply to this message

Invisible <voi### [at] devnull> wrote:
> >>>     You could just as well ask why don't they make 64TB drives right now.
> >
> >> Yep, I'd be interested to hear the answer to that one too.
> >
> >    Did you know that technology has to first be developed before it can
> > be used?
>
> OK. So why spend time and money developing a 32TB drive when you could
> just go develop a 64TB drive instead?

So why didn't Mr Benz build a 500HP S-Class with surround audion, ABS and 20
Airbags in 1886?
Just in case you're wondering:
http://en.wikipedia.org/wiki/Benz_Patent-Motorwagen
Because he didn't know how to.

Why aren't you able to play beethoven's 7th when you were 5 years old?

That's basically the same question and the same answer: You didn't know how to.
The engineers didn't know how to either.
They had to spend decades to perfect the processes.

Post a reply to this message

Aydan <hes### [at] hendrik-sachsenet> wrote:
> The engineers didn't know how to either.
> They had to spend decades to perfect the processes.

  To be fair, they could build a 64TB USB drive. It's just that nobody
would buy it because it would probably cost more than your year's salary.
Hence it makes no sense to make one until the technology becomes
significantly cheaper.

-- 
                                                          - Warp

Post a reply to this message

On 08/03/2012 01:08 PM, Warp wrote:

>    To be fair, they could build a 64TB USB drive. It's just that nobody
> would buy it because it would probably cost more than your year's salary.
> Hence it makes no sense to make one until the technology becomes
> significantly cheaper.

Right. /That/ at least makes some sense. Nobody would buy such a 
product, so there's no point trying to make it.

If there was a 64TB HD, I'm fairly certain that somebody would want to 
buy it. The only reason it wouldn't sell is if it was too expensive. 
Obviously everything starts off expensive - there are design costs to 
recoup. But would it be prohibitively expensive?

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
> Aydan <hes### [at] hendrik-sachsenet> wrote:
> > The engineers didn't know how to either.
> > They had to spend decades to perfect the processes.
>
>   To be fair, they could build a 64TB USB drive. It's just that nobody
> would buy it because it would probably cost more than your year's salary.
> Hence it makes no sense to make one until the technology becomes
> significantly cheaper.
>
> --
>                                                           - Warp

A 64GigaByte (Note: You wrote TeraByte) flash drive 12 years ago (wikipedia
claims the first ones were sold in 2000) would not only have been extremely
expensive but extremely bulky as well. I'd say it would have been as big as a
whole computer case. As I said, storage density was low in those times. And I'd
suspect they wouldn't have had a controller capable of adressing that much
memory either.

And terabytes have only become possible since the advent of 64bit addressing I
think.

Post a reply to this message

Aydan <hes### [at] hendrik-sachsenet> wrote:
> And terabytes have only become possible since the advent of 64bit addressing I
> think.

  64-bit CPUs were first developed in the early 90's.

-- 
                                                          - Warp

Post a reply to this message

On 03/08/2012 08:13 AM, Invisible wrote:
> On 08/03/2012 01:08 PM, Warp wrote:
>
>> To be fair, they could build a 64TB USB drive. It's just that nobody
>> would buy it because it would probably cost more than your year's salary.
>> Hence it makes no sense to make one until the technology becomes
>> significantly cheaper.
>
> Right. /That/ at least makes some sense. Nobody would buy such a
> product, so there's no point trying to make it.
>
> If there was a 64TB HD, I'm fairly certain that somebody would want to
> buy it. The only reason it wouldn't sell is if it was too expensive.
> Obviously everything starts off expensive - there are design costs to
> recoup. But would it be prohibitively expensive?

I wouldn't be able to cite and articles on the subject, but it seems to 
me that the mechanical HD is headed the way for the dodo bird

Post a reply to this message

On 03/08/2012 08:35 AM, James Holsenback wrote:
> On 03/08/2012 08:13 AM, Invisible wrote:
>> On 08/03/2012 01:08 PM, Warp wrote:
>>
>>> To be fair, they could build a 64TB USB drive. It's just that nobody
>>> would buy it because it would probably cost more than your year's
>>> salary.
>>> Hence it makes no sense to make one until the technology becomes
>>> significantly cheaper.
>>
>> Right. /That/ at least makes some sense. Nobody would buy such a
>> product, so there's no point trying to make it.
>>
>> If there was a 64TB HD, I'm fairly certain that somebody would want to
>> buy it. The only reason it wouldn't sell is if it was too expensive.
>> Obviously everything starts off expensive - there are design costs to
>> recoup. But would it be prohibitively expensive?
>
> I wouldn't be able to cite and articles on the subject, but it seems to
> me that the mechanical HD is headed the way for the dodo bird

LOL ... finger don't fail me now ... of the dodo bird

Post a reply to this message

Am 08.03.2012 10:31, schrieb Invisible:

> So... why didn't they just make the 4GB drives to start with? Why did
> they have to start by making 64MB drives, and then starting to make
> 128MB drives, and then moving on to 256MB drives, and so forth? Why
> couldn't they just go directly to 4GB? What enables them to make those
> today but prevented them from making them back then?
>
> Why do you have to design, test, manufacture and sell a 64MB drive
> before you can attempt to make a 128MB one? How does the former help you
> do the latter? Why can't you just jump straight to 4GB? (Or perhaps even
> more than that?)
>
> I can't think of any /technical/ reason. (Besides "that's how it's
> done".) The only rational reason I can think of is that if you keep
> putting out slightly better devices year after year, people are going to
> keep upgrading their stuff, and that gives you income. If you just went
> straight out and sold the best possible device, then once everyone has
> got one, you'd have nothing new to sell to them, and you'd have no money.

Oh, Andy... *Sigh*

How about... sheer device size? You know, last time I heard, you need 
transistors for flash memory cells, and you need silicon real estate for 
transistors. I guess you /did/ hear about Moore's Law some time?

Post a reply to this message

On 08/03/2012 01:35 PM, James Holsenback wrote:

> I wouldn't be able to cite and articles on the subject, but it seems to
> me that the mechanical HD is headed the way for the dodo bird

What, forced to extinction by loss of habitat and alien predator 
species? ;-)

Post a reply to this message

On 03/08/2012 08:56 AM, Invisible wrote:
> On 08/03/2012 01:35 PM, James Holsenback wrote:
>
>> I wouldn't be able to cite and articles on the subject, but it seems to
>> me that the mechanical HD is headed the way for the dodo bird
>
> What, forced to extinction by loss of habitat and alien predator
> species? ;-)
>

It was just a metaphor ... I supposed the alien predator would be the 
solid state drive, and now that I think of it the loss of habitat would 
be the new ultra-books ... oh and the MacBook Air

Post a reply to this message

>>> I wouldn't be able to cite and articles on the subject, but it seems to
>>> me that the mechanical HD is headed the way for the dodo bird
>>
>> What, forced to extinction by loss of habitat and alien predator
>> species? ;-)
>>
>
> It was just a metaphor ... I supposed the alien predator would be the
> solid state drive, and now that I think of it the loss of habitat would
> be the new ultra-books ... oh and the MacBook Air

So... oddly apt then? ;-)

I don't know, man. SSD is still too expensive for anyone except 
performance freaks. More bafflingly, even though the performance of SSD 
should be orders of magnitude superior to HD, apparently some of the 
best HDs can actually surpass SSD. That shouldn't be possible, but 
somehow it is. I find that utterly bizarre, but those are apparently the 
numbers.

Whether SSD will /become/ cheap enough remains to be seen. It's now 
available in capacities large enough to actually be useful, so now we 
just need the prices to come down far enough.

Post a reply to this message

Invisible <voi### [at] devnull> wrote:
> I don't know, man. SSD is still too expensive for anyone except
> performance freaks. More bafflingly, even though the performance of SSD
> should be orders of magnitude superior to HD, apparently some of the
> best HDs can actually surpass SSD. That shouldn't be possible, but
> somehow it is. I find that utterly bizarre, but those are apparently the
> numbers.

Well I have two SSDs and I'm not a performace freak.

> More bafflingly, even though the performance of SSD
> should be orders of magnitude superior to HD, apparently some of the
> best HDs can actually surpass SSD. That shouldn't be possible, but
> somehow it is. I find that utterly bizarre, but those are apparently the
> numbers.

Depends on what key parameters youre looking at. In sheer bulk transfer rate and
especially write speed, it's possible for a mechanical drive to outperform an
SSD.
But for random access you'll be hard pressed to find a mechanical drive
outperforming an SSD.
And that's the main reason I have my SSDs. For the random access speed.

My SSDs have a random access speed in the low megabyte range, but thats still 10
times faster than my mechanical ones (late model 1 & 2 TB Western Digital SATA2
drives).

Post a reply to this message

On 08/03/2012 12:42 PM, Invisible wrote:
>>>> You could just as well ask why don't they make 64TB drives right now.
>>
>>> Yep, I'd be interested to hear the answer to that one too.
>>
>> Did you know that technology has to first be developed before it can
>> be used?
>
> OK. So why spend time and money developing a 32TB drive when you could
> just go develop a 64TB drive instead?

Several years ago I used to work for Motorola in one of their 
semiconductor factories. I seem to remember that there was covered 
switch that when thrown would change the factory output from DIL CMOS 
chips to microprocessors without having to go through the process of 
designing the new chips or retooling the factory. It also connected the 
new suppliers. </leg_pull>

-- 
Regards
     Stephen

Post a reply to this message

Am 08.03.2012 12:44, schrieb Invisible:

> I can understand how something /complicated/ - like, say, a processor -
> would take time. Inventing a new feature takes a lot of R&D. But just
> making the transistors smaller? I don't really see why they have to
> shrink in tiny little baby steps. Why not just go the whole hog straight
> away?

Because maybe - just maybe - they have to /find out/ how to make that 
next "tiny little baby step"? After all, they're creating structure 
sizes nowadays where naive classic photolitpgraphy with ~400 nm UV light 
would fail utterly. Remember that old rule that you can't etch 
structures smaller than the wavelength of the light you're using?

Some of the "baby steps" involved:

- Replacing the ~400 nm Hg vapor lamps with ~250 nm KrF excimer lasers. 
Obviously you need to figure out how to build an excimer laser for this 
feat. You may also need to develop new optical materials (in case the 
old ones absorb the new wavelength) and/or new photoresists (in case the 
old ones are insensitive to the new wavelength).

- Replacing the ~250 nm KrF excimer lasers with ~200 nm ArF excimer 
lasers. Obviously you need to extrapolate your excimer laser design to 
use ArF, and again you may need new optical materials and/or photoresists.

- Replacing the liquid etching agent with plasma, to reduce undercutting 
of the photoresist. Obviously you need to figure out that plasma etching 
will have these benefits in the first place, and then find ways to 
handle the plasma in the fab.

- Exposing the photoresist in liquid rather than in air or vacuum, to 
reduce wavelength accordingly. Obviously you need to find a suitable 
liquid with high purity, and figure out how to handle it in the fab.

- Designing masks in such a way that they no longer directly correspond 
to the desired structures, but rather cause interference patterns of the 
desired shape. Obviously you need to develop suitable algorithms to 
generate the masks from the desired structures, and you probably need 
quite some computing power to pull this off.

Other steps may involve improvements of photoresists and application 
techniques (you want a thin yet uniform coating, which adheres well to 
the wafer) plus suitable developer, etching and resist stripping agents; 
improvements in wafer cleaning processes; reduction of mechanical 
tolerances; and other some such. Much of this goes into overcoming 
problems that hadn't been anticipated before.


Last not least, there's also the question of cost: Using cutting-edge 
technology for memory ICs would result in prohibitively high per-bit 
costs; such production lines are therefore almost exclusively used for 
high-margin products such as high-performance CPUs or GPUs, until the 
next generation of cutting-edge fab technology becomes available. The 
older production lines might then be used for another generation to 
manufacture low-cost CPUs and GPUs, and finally for stuff where 
per-transistor costs are more important than die size, such as memory ICs.

Post a reply to this message

On 08/03/2012 02:50 PM, Aydan wrote:
> Invisible<voi### [at] devnull>  wrote:
>> I don't know, man. SSD is still too expensive for anyone except
>> performance freaks. More bafflingly, even though the performance of SSD
>> should be orders of magnitude superior to HD, apparently some of the
>> best HDs can actually surpass SSD. That shouldn't be possible, but
>> somehow it is. I find that utterly bizarre, but those are apparently the
>> numbers.
>
> Well I have two SSDs and I'm not a performace freak.

You clearly have a hell of a lot more money than I'll ever have.

Or rather, you /had/ a lot of money... ;-)

>> More bafflingly, even though the performance of SSD
>> should be orders of magnitude superior to HD, apparently some of the
>> best HDs can actually surpass SSD. That shouldn't be possible, but
>> somehow it is. I find that utterly bizarre, but those are apparently the
>> numbers.
>
> Depends on what key parameters youre looking at. In sheer bulk transfer rate and
> especially write speed, it's possible for a mechanical drive to outperform an
> SSD.
> But for random access you'll be hard pressed to find a mechanical drive
> outperforming an SSD.
> And that's the main reason I have my SSDs. For the random access speed.

You would expect an SSD to blitz a HD for random access. But even just 
for read access, you would have expected a purely electronic device to 
be many billion times faster than a mechanical HD. And they aren't, for 
some reason...

Write performance is another matter. Apparently you have to bulk-erase 
cells before you can write them again. So I can understand that being 
slower. Even so, you'd think some kind of parallel processing 
arrangement could hide some of that...

Post a reply to this message

On 08/03/2012 03:23 PM, clipka wrote:

> Last not least, there's also the question of cost: Using cutting-edge
> technology for memory ICs would result in prohibitively high per-bit
> costs; such production lines are therefore almost exclusively used for
> high-margin products such as high-performance CPUs or GPUs, until the
> next generation of cutting-edge fab technology becomes available. The
> older production lines might then be used for another generation to
> manufacture low-cost CPUs and GPUs, and finally for stuff where
> per-transistor costs are more important than die size, such as memory ICs.

Right. So that's why, even though 40nm technology already exists, they 
don't use it just to build (say) USB flash drives. (?)

Post a reply to this message

On 08/03/2012 3:53 PM, Invisible wrote:
> to be many billion times faster

Do you really mean "billion"? :-P

-- 
Regards
     Stephen

Post a reply to this message

Invisible <voi### [at] devnull> wrote:
> On 08/03/2012 02:50 PM, Aydan wrote:
> > Well I have two SSDs and I'm not a performace freak.
>
> You clearly have a hell of a lot more money than I'll ever have.
>
> Or rather, you /had/ a lot of money... ;-)
>

I spent all of 200€ for both. Not a lot of money for me. I mean I bought am
Gaming rig for about 1000€ a few years back.


> You would expect an SSD to blitz a HD for random access. But even just
> for read access, you would have expected a purely electronic device to
> be many billion times faster than a mechanical HD. And they aren't, for
> some reason...
>
> Write performance is another matter. Apparently you have to bulk-erase
> cells before you can write them again. So I can understand that being
> slower. Even so, you'd think some kind of parallel processing
> arrangement could hide some of that...

TRIM is the keyword. Pages are erased during idle time of the disk.
The transfer speeds come from the Addressing delays and acrual write and read
timing constraints. Yes, silicon memory does have a finite bandwidth.

Post a reply to this message

Invisible <voi### [at] devnull> wrote:
> On 08/03/2012 03:23 PM, clipka wrote:
>
> > Last not least, there's also the question of cost: Using cutting-edge
> > technology for memory ICs would result in prohibitively high per-bit
> > costs; such production lines are therefore almost exclusively used for
> > high-margin products such as high-performance CPUs or GPUs, until the
> > next generation of cutting-edge fab technology becomes available. The
> > older production lines might then be used for another generation to
> > manufacture low-cost CPUs and GPUs, and finally for stuff where
> > per-transistor costs are more important than die size, such as memory ICs.
>
> Right. So that's why, even though 40nm technology already exists, they
> don't use it just to build (say) USB flash drives. (?)

They use 25nm chips for SSDs nowadays.

Post a reply to this message

"Aydan" <hes### [at] hendrik-sachsenet> wrote:
> Invisible <voi### [at] devnull> wrote:
> > On 08/03/2012 02:50 PM, Aydan wrote:
> > > Well I have two SSDs and I'm not a performace freak.
> >
> > You clearly have a hell of a lot more money than I'll ever have.
> >
> > Or rather, you /had/ a lot of money... ;-)
> >
>
> I spent all of 200€ for both.
Just to clarify: Both together cost 200€

Post a reply to this message

On 3/8/2012 1:31, Invisible wrote:
> Today's cars have
> performance within a few percent of cars made 40 years ago.

Uh, what?  40 years ago, 12MPG was a decent mileage, and you changed the oil 
every 3000 miles or so, and added oil in between. 40 years ago, cars didn't 
even have six digits on the odometer, because almost none of them lasted 
100,000 miles.  Electric windows and stereo radios were luxuries.

-- 
Darren New, San Diego CA, USA (PST)
   People tell me I am the counter-example.

Post a reply to this message

On 3/8/2012 7:23, clipka wrote:
> "tiny little baby step"? After all, they're creating structure sizes
> nowadays where naive classic photolitpgraphy with ~400 nm UV light would
> fail utterly.

Somewhat relevant: http://htwins.net/scale2/

-- 
Darren New, San Diego CA, USA (PST)
   People tell me I am the counter-example.

Post a reply to this message

On 3/8/2012 7:56, Invisible wrote:
> Right. So that's why, even though 40nm technology already exists, they don't
> use it just to build (say) USB flash drives. (?)

Because the factories are very expensive to build (like, there's at most a 
handful in the world), and they're busy building other things that give 
bigger profits.

-- 
Darren New, San Diego CA, USA (PST)
   People tell me I am the counter-example.

Post a reply to this message

On 3/8/2012 5:13, Invisible wrote:
> would it be prohibitively expensive?

Would it be more expensive than 4 16TB drives? Most likely.

-- 
Darren New, San Diego CA, USA (PST)
   People tell me I am the counter-example.

Post a reply to this message

> Somewhat relevant: http://htwins.net/scale2/

An alpha helix is of comparable size to a carbon atom?

Really?

Even though an alpha helix is composed of dozens of amino acids which 
are built out of dozens of carbon atoms?

Post a reply to this message

Invisible <voi### [at] devnull> wrote:
> An alpha helix is of comparable size to a carbon atom?

> Really?

> Even though an alpha helix is composed of dozens of amino acids which 
> are built out of dozens of carbon atoms?

  You are being confused by the logarithmic nature of the scale. The
alpha helix and the carbon atom are not at the same zoom level, even
though they are close.

-- 
                                                          - Warp

Post a reply to this message

>> Assuming the supported resolution is changing proportionately, this 
is actually quadratic growth in terms of number of pixels.
 >
 >    The standard measurement for screen size is the length of the 
diagonal.

I think my statement is still true.



On 3/8/2012 1:10 AM, Warp wrote:
> Slime<pov### [at] slimelandcom>  wrote:
>>   >     Monitor screen sizes do not seem to follow this trend, however, and
>>   >  seem to grow a lot more linearly.
>
>> Assuming the supported resolution is changing proportionately, this is
>> actually quadratic growth in terms of number of pixels.
>
>    The standard measurement for screen size is the length of the diagonal.
>
>> This is probably limited by graphics hardware in some way (although I'm
>> not sure how fast that grows in "power"). It's also limited by the fact
>> that popular OSs such as Windows still use pixels as a unit of
>> measurement, so there's limited benefit to higher resolutions as it
>> shrinks the size of important objects on screen.
>
>    Resolution has little to do with screen size.
>

Post a reply to this message

On 07/03/2012 19:36, Warp wrote:
>    It's a common phenomenon that things like hard drive sizes, RAM sizes
> and processor speeds tend to grow at a superlinear rate over time in
> home computers. (I won't say "exponential" because someone recently
> complained about that, and I don't have hard data to back up that claim.)
> Very typically, for example, each new hard drive one buys tends to be
> larger than the combined sizes of all hard drives that person has bought
> in the past. (I'm of course talking about the average computer user rather
> than those who buy hard drives like candy, eg. because they need them for
> multimedia projects or something.)
>
>    Monitor screen sizes do not seem to follow this trend, however, and
> seem to grow a lot more linearly.

I wouldn't have expected the physical diagonal size to increase 
exponentially for obvious reasons, but you might have expected 
resolution to do so (if you assume "pixel rate" or something could 
double every 18 months, then the "dpi" should double every 3 years).

The main problem is that Windows doesn't work nicely with higher 
resolutions.  I had about 140ppi on my old laptop and even then the 
standard windows GUI was getting too small to read.  I experimented with 
the higher dpi settings but lots of software didn't play nicely (missing 
icons etc.) so in the end I just turned the screen resolution down.

Once a new version of Windows (and 3rd party software) is available with 
proper support for higher resolutions, then I believe you will see the 
drive for higher resolutions start for windows machines (not just 
monitors, but graphics cards, display interfaces etc).  Then a 22" 
300ppi (5760x3240) display will be quite standard, and children will 
laugh when told about the days when you could actually see the 
individual pixels on a screen and the text was all a bit blocky or 
blurry depending what setting you used.

Post a reply to this message

> I wouldn't have expected the physical diagonal size to increase
> exponentially for obvious reasons, but you might have expected
> resolution to do so (if you assume "pixel rate" or something could
> double every 18 months, then the "dpi" should double every 3 years).
>
> The main problem is that Windows doesn't work nicely with higher
> resolutions.

/Windows/ doesn't support it? Or just that every 3rd-party crock of junk 
ported from Windows 3.1 or only tested on one PC assumes that everybody 
has exactly the same DPI?

Post a reply to this message

> Why do you have to design, test, manufacture and sell a 64MB drive
> before you can attempt to make a 128MB one? How does the former help you
> do the latter? Why can't you just jump straight to 4GB? (Or perhaps even
> more than that?)

You need to learn how to make 4GB ones, or more correctly, how to build 
a factory to mass produce 4GB ones.  And you're not going to do that 
overnight in the lab if you are currently only making 64MB ones.  It 
simply won't happen, you'll go bankrupt (because your competitors 
figured out how to make 256MB ones and you didn't) before you figure it out.

The only way to do it is to stretch your existing process to make 128MB 
or 256MB ones, trying to fix the odd few problems that will no doubt 
come up.  That is called research and development.

Once you have a stable process for 256MB ones, you can again try and 
push to make 512MB or 1GB ones, you will no doubt come up against 
different problems than you had last time, you might need some fairly 
big steps in technology development, but it's possible.

> Then again, cars don't improve in performance at all. Today's cars have
> performance within a few percent of cars made 40 years ago. And yet,
> people still buy cars.

Compare pretty much any criteria and cars have vastly improved.  MPG, 
power, safety rating, braking distance, amount of rust, noise level, 
pollution level, ride quality.  It is quite an achievement given that 
many of those factors are trade-offs against each other (eg better 
safety means higher weight which means lower MPG, yet MPG has increased 
significantly).

> On the other hand, pens and pencils don't improve in performance either,
> and they still sell plenty of those.

That's just because you don't work with designing pens :-)  I'm pretty 
sure they are continually using the latest materials and inks available, 
you just don't notice that your pen doesn't snap as easily or 
clog-up/stop working as often as it used to.

Post a reply to this message

On 10/04/2012 16:01, Invisible wrote:
>> I wouldn't have expected the physical diagonal size to increase
>> exponentially for obvious reasons, but you might have expected
>> resolution to do so (if you assume "pixel rate" or something could
>> double every 18 months, then the "dpi" should double every 3 years).
>>
>> The main problem is that Windows doesn't work nicely with higher
>> resolutions.
>
> /Windows/ doesn't support it? Or just that every 3rd-party crock of junk
> ported from Windows 3.1 or only tested on one PC assumes that everybody
> has exactly the same DPI?

Maybe it's just so clunky for devlopers to support it they don't bother? 
  I certainly have no idea how any of my programs would look.  MS need 
to make it easier/mandatory in the APIs and development tools to cater 
for higher DPI environments (maybe they did that already for Windows 8, 
haven't checked the latest versions).

Android is an example of a better way to do it (but probably not the 
best way).

Post a reply to this message