REGISTER | LOGIN
Breaking News
News & Analysis

IoT May Need Sub-50-Cent SoCs

Panel previews rough road to 2027 end nodes
10/27/2017 00:01 AM EDT
9 comments
NO RATINGS
Page 1 / 2 Next >
More Related Links
View Comments: Newest First | Oldest First | Threaded View
realjjj
User Rank
CEO
Re: ...
realjjj   10/31/2017 2:43:01 PM
NO RATINGS
Yes harvesting RF energy wouldn't fit every scenario. Wireless connectivity has one weak point, no power delivery and RF power delivery/harvesting would allow for some integration so it seems the obvious choice and an area that should attract investments. Ofc other forms of energy harvesting are more evolved today and it's difficult to predict what solution or mix of solutions we'll see in 2027.

rick merritt
User Rank
Author
Re: Memory Table Update for ReRAM
rick merritt   10/31/2017 1:17:43 PM
NO RATINGS
Thanks for the reality check, Gil. It would be great if Lucien from ARM could jump on and address some of tese questions.

Gil Russell
User Rank
Author
Memory Table Update for ReRAM
Gil Russell   10/31/2017 12:59:57 PM
NO RATINGS
Rick, Thanks for the report. We've come to appreciate the panel's conclusions on Resistive RAM in IoT.

One minor, but important, tick on the memory chart is under the PCM/ReRAM column, row Voltage states 2.5 Volts. I took a quick look at Adesto's Moneta family which lists 1.65 - 2.75 Volts. Core supply is 0.97 - 1.03 Volts. I realize this is from the panel presentation but I think it should represent what is available in today's market.

Gil Russell, Principle Technology Analyst, WebFeet Research, Inc.

rick merritt
User Rank
Author
Re: ...
rick merritt   10/28/2017 1:39:35 AM
NO RATINGS
Thanks for chiming in, Fred. There are a couple questions here for Lucien, too.

fredmartin
User Rank
Rookie
Re: ...
fredmartin   10/27/2017 5:07:19 PM
NO RATINGS
Harvesting of RF energy is certainly one type of harvesting that could be employed.  In my comments on the panel, I was trying to describe a more abstracted approach, where any type of harvested energy could be used.  One of the big challenges with almost any harvester approach is that it works in only a narrow niche of applications and physical settings.  RF harvesting, for example, requires a relatively strong RF signal.  This may be found within a meter or two of a WiFi access point or some other strong source.  More generally, there is just not enough ambient RF energy in most locations to make opportunistic harvesting viable.  I have seen some proposals where a strong RF source is provided intentionally to power multiple nodes in a small area.  This could be interesting for certain apps.

rick merritt
User Rank
Author
Re: ...
rick merritt   10/27/2017 1:57:01 PM
NO RATINGS
Great idea!

realjjj
User Rank
CEO
...
realjjj   10/27/2017 8:15:40 AM
NO RATINGS
Why not try to use RF for power harvesting and/or delivery. If the radio integrates both power and data, there might be substantial savings.

resistion
User Rank
Author
Re: Write speeds in PCRAM/ReRAM
resistion   10/27/2017 7:41:07 AM
NO RATINGS
What would be the read voltage? Would it be so low it's close to noise level?

Shinobee
User Rank
Manager
Write speeds in PCRAM/ReRAM
Shinobee   10/27/2017 12:58:53 AM
NO RATINGS
Interesting article. I am curious about the read speed of ReRAM/PCRAM vs MRAM.  The diagram claims that ReRAM/PCRAM has <1ns read speed while MRAM is <5ns.  Aren't they all resistance sensing devices?  I am not sure why MRAM would seemingly be slower than the other two. At least for PCRAM, it may be worth mentioning the "erase" time as well since that takes significantly more time than to write a bit.

Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed