eReaders, CPU die-shrinks and battery life.

[ottdmk]

Well, I believe Hrafn's point was that 28nm chips have been around for at least three years. It would be interesting to discover how much price difference there is between a three-year-old 28nm cpu and the even older 40nm variety currently in wide use in e-readers.

[rkomar]

Perhaps it's just too painful to re-tune the custom linux systems for new components and drivers. They are quite aggressive in putting stuff to sleep to save on battery power. New hardware + new drivers may mean going back to square one in power management. I don't know, I'm just guessing why they seem to be so conservative with the hardware (I'm really thinking of PocketBook, here).

[Hrafn]

Originally Posted by j.p.s

The 9.7" eReaders also display 2.6x more characters per page, so the 6" eReaders need 2.6x more page turns per hour, so the power per hour for page turns is the same.

Except that the exemplars I used (last generation pre-Android Onyx readers) had their battery lives listed in thousands of page turns, not time.

[AnemicOak]

Originally Posted by Hrafn

Are 28nm eReaders already out there (just not with Onyx-Boox)?

I don't think so, not from what I can tell at least. Kindle Voyage and PW2 use the Freescale i.MX6, the Kobo Aura H20 uses a Freescale i.MX507 and their brand new Glo HD uses an i.MX6, and the current Nook uses the same TI OMAP 3621 the previous one did AFAIK.

The i.MX processors are very popular for readers since they have the eInk controller on-board (so does the OMAP 3621).

Originally Posted by ottdmk

Well, I believe Hrafn's point was that 28nm chips have been around for at least three years. It would be interesting to discover how much price difference there is between a three-year-old 28nm cpu and the even older 40nm variety currently in wide use in e-readers.

But is there a 28nm chip that includes the epd controller?

[Teknikal]

Personally I'd like to see an eink reader with a modern SOC and backed up by about 2 of ram.

There were plenty of occasions on my T1 that the device would crash just trying to look through my library I actively had to avoid books with too big a filesize because I knew it would have issues. It really felt to me like the hardware wasn't up to the purpose I've never had any issues like that on even a modest spec phone/tablet.

[Hrafn]

Originally Posted by AnemicOak

The i.MX processors are very popular for readers since they have the eInk controller on-board (so does the OMAP 3621).

But is there a 28nm chip that includes the epd controller?

The 28nm Freescale i.MX 7 includes an epd controller (Freescale doc, page 5).

[On closer examination, this chip appears to have been cancelled -- the 2011 i.MX 6 remains their most recent chip.]

The Rockchip RK3026 used in most 6" Onyx Booxes (the larger screens use the i.MX 6) does appear to have a 28nm replacement in the RK3188/RK3188T. (I don't know what OB does for epd control for these readers.)

[HarryT]

Originally Posted by Hrafn

The 28nm Freescale i.MX 7 includes an epd controller (Freescale doc, page 5).

[On closer examination, this chip appears to have been cancelled -- the 2011 i.MX 6 remains their most recent chip.]

The Rockchip RK3026 used in most 6" Onyx Booxes (the larger screens use the i.MX 6) does appear to have a 28nm replacement in the RK3188/RK3188T. (I don't know what OB does for epd control for these readers.)

Is the EPD controller integrated with the CPU? I wasn't aware of that.

[Jason90]

The i.mx7/8 hasn't been cancelled as was thought previously and is slated to come out later this year,however freescale seems to be one of the most secretive of companies,whether or not we will actually see ereaders with said cpu i'm not sure,but seeing as almost all ereader's use freescale cpu's it seems likely they will transition to the new ones at some point.

The i.mx7 uses 2 cortex A7's running at 1.5 ghz and they are targeted at the ereader market (or were according to their 2013 presentation) we currently have android based ereaders running dual core i.mx 6 cpu's which are cortex A9 at 1/1.2 ghz,the A7 is actually slower than the A9 but makes up for it in the higher clock speed.

So i would of thought at 28nm plus the transition to a more energy efficient architecture would result in some healthy battery gains and perhaps a slight performance boost.

I would also expect a new eink display sometime next year as eink carta is getting pretty old,they must be working on a next gen display tech...

[tnforpaul45]

I use a lot of PDFs on my eReading devices. I am certain that chews up a lot of CPU cycle time and battery as I page turn those. Sometimes, ePub and .mobi just does not do it for me.

Also, don't forget the Cellular connections in Kindles too. I would bet the average user does not turn that off when not in use (if they have it, not all do).

I would love to know (but never will) the efficiency of the code in the Kindle and Kobo OS'es. That's another big CPU hog.

All in all, its a multi-faceted issue.

[HarryT]

Originally Posted by tnforpaul45

Also, don't forget the Cellular connections in Kindles too. I would bet the average user does not turn that off when not in use (if they have it, not all do).

I suspect that only a tiny proportion of people buy the 3G Kindle; the vast majority are WiFi only.

[fjtorres]

Originally Posted by Jason90

So i would of thought at 28nm plus the transition to a more energy efficient architecture would result in some healthy battery gains and perhaps a slight performance boost.

I would also expect a new eink display sometime next year as eink carta is getting pretty old,they must be working on a next gen display tech...

Any gains from more efficient electronics might be traded off with smaller, lighter battery packs. We might not see any battery life gains at the product level.

As for a newer eink generation, Carta is only now starting to filter down to the third tier vendors and with declining sales volumes there is no telling how much R&D eInk might be willing to devote to ereaders instead of alternate markets. I wouldn't hold my breath for a newer screen tech from eink this year. I think the 300 dpi Carta is tops for this year.

[Jason90]

Originally Posted by fjtorres

Any gains from more efficient electronics might be traded off with smaller, lighter battery packs. We might not see any battery life gains at the product level.

As for a newer eink generation, Carta is only now starting to filter down to the third tier vendors and with declining sales volumes there is no telling how much R&D eInk might be willing to devote to ereaders instead of alternate markets. I wouldn't hold my breath for a newer screen tech from eink this year. I think the 300 dpi Carta is tops for this year.

It's hard to see how they can improve the screen beyond what is used in the kindle voyage,It's pretty much perfect and reads like a printed book...even better i might add,the next step needed are for larger eink devices in the 10" range,a 300ppi 9.7" eink mobius display based on carta would probably make for my ideal ereader,i say mobius because the glass based current 9.7" ereader are far too heavy,the switch to mobius would reduce the device weight by about half judging by how much the DPT-S1 weighs and it's size. Maybe we could expect such a device at the end of this year or next year...but i suspect unless amazon bring back the DX line and introduce a successor i don't see eink developing a specialized high PPI large screen for anyone else.Amazon seem to be pushing the technology along with everyone else getting it at a later date.

[Hrafn]

Originally Posted by Jason90

It's hard to see how they can improve the screen beyond what is used in the kindle voyage,It's pretty much perfect and reads like a printed book...even better i might add,the next step needed are for larger eink devices in the 10" range,a 300ppi 9.7" eink mobius display based on carta would probably make for my ideal ereader,i say mobius because the glass based current 9.7" ereader are far too heavy,the switch to mobius would reduce the device weight by about half judging by how much the DPT-S1 weighs and it's size. Maybe we could expect such a device at the end of this year or next year...but i suspect unless amazon bring back the DX line and introduce a successor i don't see eink developing a specialized high PPI large screen for anyone else.Amazon seem to be pushing the technology along with everyone else getting it at a later date.

Although I think the Kindle Voyage screen is excellent, I think it can still be substantially improved by moving to a lighter, more durable substrate. This would both yield direct benefits, and allow a thinner and lighter chassis because it no longer needs so much rigidity to protect it. Although Wexler FlexONE was perhaps a flawed design, it offers some perspective to how light an eReader could be (110g).

On larger eReaders, I would only expect slow progress, as (i) they are a niche submarket even within the niche market of eReaders, and (ii) this is a niche where the greater versatility of LCD/OLED tablets gives them the greatest competitive advantage (disincentivising eReader manufacturers from spending too much money on this subniche).

The period of rapid expansion of eReader ownership is over, and I think with it the period of large-scale investment in eReader technologies. I suspect any further investment will be small, carefully calculated, and more likely to be targeted at incremental improvements rather than revolutionary ones.

[Jason90]

Think you may be right,i don't think amazon could justify the next voyage costing the same as the current one with new competition out there (the kobo glo hd) so they will need something to differentiate it if they wan't to keep the $200 price tag,A substantially lighter more compact design would be a pretty good reason.

We'll find out in a few months

[Lemurion]

I think the real driver on die-shrinks is cost. It's not going to become worth it until the savings from more chips per wafer outweigh the cost of retooling to a new process.