Analog computing will take over 30 billion devices by 2040. Wtf does that mean? | Hard Reset

Freethink
9 Oct 202311:51

Summary

TLDRこのスクリプトは、最新の技術が実際には非常に古い技術である可能性について述べています。アナログプロセッサは、波形を用いて思考し、コンピューティングの根幹を根本的に見直す可能性を秘めています。アナログプロセッサは、デジタルプロセッサのようにソフトウェアでプログラミングが可能であり、AIや機械学習アルゴリズムを使いながらも、デジタル方式よりもはるかに少ないエネルギーで動作できるとされています。これにより、より効率的で正確な処理が可能になり、デジタルコンピュータとは異なる新しいデバイスを構築できると期待されています。また、アナログプロセッサはデジタルコンピュータを代替するものではなく、より戦略的に使用される可能性があります。この技術は、デジタルとアナログを組み合わせることで、より多くの人々を効果的に監視できるようにするなど、医療や音響デバイスなどの分野で大きな貢献を提供する可能性があります。

Takeaways

  • 🌊 模拟处理器使用波形而非数字信号进行计算,这可能彻底改变我们对计算机的重新思考。
  • 📱 数字计算由软件驱动,而模拟处理允许我们以最原始的形式理解和推断数据。
  • 🔍 模拟处理器可以像数字处理器一样通过软件编程,这是一个重大的进步。
  • 💡 使用模拟处理器的新设备可能仅使用数字设备的千分之一能量,同时保持高效能和准确性。
  • ❓ 提出了是否所有处理都必须是数字的问题,并探讨了模拟处理器如何改变计算机的工作方式和与我们的互动。
  • 📈 模拟处理器的挑战在于输出的一致性,但Aspinity公司通过软件解决了这一问题。
  • 🔌 模拟处理器可以直接处理来自传感器的信号,无需转换为数字信号,这大大提高了效率。
  • 🌟 Aspinity的产品设计师认为,模拟计算的稳定性和可编程性问题的解决将导致计算架构的快速重置。
  • 🔋 目前,我们将大量能量用于将模拟信号转换为数字信号,而模拟处理器可以更节能地完成这一任务。
  • 🚀 模拟技术并不意味着要取代数字计算机,而是可以更策略性地使用它们。
  • 🔉 模拟处理器可以用于始终开启的计算任务,而数字系统则在必要时唤醒,这样可以节省能源。
  • 🎛️ 模拟计算可以实现更长时间的电池寿命,或者使用更小的电池实现相同的功能。
  • 🔕 模拟处理器还可以用于声音检测等应用,通过特定的过滤器和决策树来区分不同类型的声音。
  • ⚙️ 在大规模社会影响方面,模拟处理器可以用于监控管道、太阳能发电厂、汽车等设备。
  • 🧵 模拟计算机可以指挥数字计算机的活动,这意味着可以在关键地方部署更多的感应系统。
  • 🧘 我们可能永远不会有完全模拟的计算机,但模拟和数字的结合将发挥各自的优势,如心脏监测。
  • 🏠 模拟智能可能会彻底改变智能设备理解我们的方式,实现更低能耗的洞察力收集设备部署。

Q & A

  • アナログプロセッサとは何ですか?

    -アナログプロセッサは、デジタルプロセッサとは異なり、波形を扱って処理を行うプロセッサです。これはコンピューティングの考え方を根本から見直す可能性を秘めています。

  • アナログプロセッサがデジタルプロセッサよりも優れているとされる理由は何ですか?

    -アナログプロセッサは、生のデータそのものを理解し、推論し、洞察を得ることができるため、デジタルプロセッサよりも優れています。また、同じようにソフトウェアでプログラムすることができ、エネルギー効率が非常に高いとされています。

  • Aspinityという会社は何を提供していますか?

    -Aspinityは、ソフトウェアでプログラム可能なアナログプロセッサを提供しています。これにより、アナログ信号をデジタル信号に変換する必要がなく、より効率的なコンピューティングが可能になります。

  • アナログプロセッサがデジタルプロセッサを置き換えるのでしょうか?

    -アナログプロセッサはデジタルプロセッサを置き換えるためのものではありません。それよりも、デジタルプロセッサをより戦略的に使用することができるようになるでしょう。

  • アナログプロセッサが有効な応用分野は何ですか?

    -アナログプロセッサは、センサー信号の収集や解釈、音響デバイス、心拍数モニタリング、および家での音声コンピューティングなど、エネルギー効率が求められる分野で有効です。

  • アナログプロセッサの導入により、遠隔地でセンサーデータを監視することが現実的になる理由は何ですか?

    -アナログプロセッサは、低エネルギーで動作し、デジタルシステムよりもはるかに少ないエネルギーで現実世界の感覚を監視できるためです。これにより、ディジタルシステムでは扱いが困難な場所での監視が可能になります。

  • アナログプロセッサが持つ課題とは何ですか?

    -アナログプロセッサの課題は、同じ回路でも異なるシリコンで異なる出力を得ることがある「電圧オフセット」です。Aspinityは、ソフトウェアを通じて微調整を行い、同じ出力を保証する手法を開発しました。

  • デジタル信号とアナログ信号の間にはどのような違いがありますか?

    -デジタル信号は0と1の二値表現を使用して情報を処理しますが、アナログ信号は連続的な波形を用いて情報を処理します。アナログ信号は、現実世界の様々な現象を直接的に表現できます。

  • アナログプロセッサがもたらす可能性として、リモコンの電池寿命が延びるというのはどういう意味ですか?

    -アナログプロセッサは常に待機状態で低電力で動作し、必要に応じてデジタルシステムに切り替わるため、リモコンの電池消費を劇的に減らすことができます。これにより、同じ電池で長時間使用できるようになる、または小さな電池で同じ機能を実現できるようになります。

  • アナログプロセッサが持続可能な社会に与える可能性とは何ですか?

    -アナログプロセッサは、エネルギー効率の良いセンサーシステムを導入し、産業機械、太陽光発電所、自動車などの監視と最適化を可能にします。これにより、持続可能な社会の構築に貢献することが期待されます。

  • アナログプロセッサがもたらすコンピューティングアーキテクチャの変化とは何ですか?

    -アナログプロセッサは、より多くの価値をもたらす洞察を得るために、より少ないリソースで行う新しいコンピューティングアーキテクチャを創造します。これにより、センサーデータをより効果的に収集し、解析することが可能になります。

Outlines

00:00

🤖 デジタルからアナログへ、コンピューティングの再定義

デジタルプロセッサは0と1で思考しますが、アナログプロセッサは波形で思考します。アナログプロセッサはコンピューティングを根本から再考できる可能性を秘めています。アナログプロセッサはソフトウェアでプログラミングでき、AIや機械学習アルゴリズムを使いながらも、デジタルプロセッサよりも1/1000のエネルギーで動作することが可能です。これはコンピュータがどのように機能し、私たちとどのように相互作用するかを変える可能性があります。

05:03

🔋 エネルギー効率とアナログコンピューティングの未来

アナログ信号をデジタル信号に変換するために無駄に使われている大量のエネルギーを減らすために、アナログコンピューティングが注目されています。デジタルコンピュータのエネルギー消費は、私たちがエネルギーを生成する速さよりも迅速に増加しています。アナログコンピュータは、センサーからのデータを直接解釈し、無関係なデータをフィルタリングすることで、より効率的にデータを処理することができます。これはリモコンの電池が長持ちするだけでなく、より小さな電池で同じ機能を実現できることを意味しています。

10:05

🌐 アナログとデジタルの融合、新たなコンピューティングアーキテクチャ

アナログコンピュータは、デジタルコンピュータの活動をオーケストレーションし、現実世界の感覚をより少ないエネルギーで行うことができます。アナログコンピュータは、心拍数などのアナログ信号を効率的に監視し、異常を検出するとデジタルコンポーネントに警告することができます。これにより、より多くの人々が効果的に監視され、クラウドや医療プロバイダーに連絡することができます。アナログセンサーの種類が増えることで、より多くの人々が低エネルギーで洞察を集めるデバイスを利用できるようになります。

Mindmap

Keywords

💡アナログプロセッサ

アナログプロセッサとは、デジタルプロセッサとは異なり、波形を扱って処理を行うコンピュータの部品です。ビデオでは、アナログプロセッサがコンピューティングの基礎を根本的に再考できる可能性を秘めていると述べています。これは、原始的なデータの形での洞察を得ることができるため、デジタルコンピューティングに比べてエネルギー効率が高く、処理速度が早くなります。

💡デジタルプロセッサ

デジタルプロセッサは、コンピュータや携帯電話に搭載されており、1と0の二進法で思考します。ビデオでは、デジタルプロセッサがソフトウェアによって駆動されてきたことと、それに比べてアナログプロセッサの利点が強調されています。

💡ソフトウェアプログラマブル

ソフトウェアプログラマブルなアナログプロセッサとは、ソフトウェアを使ってプログラムできるという意味です。ビデオでは、これは大きな突破であり、アナログプロセッサをデジタルプロセッサと同じ方法でプログラムできるようになることを意味しています。

💡エネルギー効率

エネルギー効率とは、同じ仕事をするのに必要なエネルギーを最小限に抑える能力です。ビデオでは、アナログプロセッサがデジタルプロセッサよりもはるかに少ないエネルギーで動作し、AIや機械学習アルゴリズムを使いながらも、その1/1000のエネルギーで動作できると述べています。

💡センサー

センサーとは、環境からの情報を集め、アナログ信号として伝えるデバイスです。ビデオでは、携帯電話や家、車、職場などにある多くのセンサーがアナログ信号を集め、デジタルコンピュータがそれらを理解するために大量のエネルギーを費やしていると触れています。

💡デジタル信号とアナログ信号

デジタル信号は、0と1の形式で情報を伝える信号であり、一方、アナログ信号は、連続的な波形で情報を伝えるものです。ビデオでは、アナログ信号をデジタル信号に変換するために多くのエネルギーが使われていると指摘し、アナログプロセッサがこの変換を避けることができると述べています。

💡バイナリ

バイナリとは、0と1の2つの値だけを使って情報を表現する方法です。ビデオでは、アナログプロセッサがバイナリとは異なる方法で思考し、波形を扱うことができると説明しています。

💡ハードリセット

ハードリセットとは、コンピュータの内容を初期状態に戻す操作です。ビデオでは、ハードリセットというシリーズが、世界をゼロから再構築するというテーマで、アナログプロセッサがコンピューティングの基礎を再考できる可能性を探求していることを示しています。

💡クラウド

クラウドとは、インターネットを通じて提供されるコンピューティングリソースのことを指します。ビデオでは、アナログプロセッサが異常を検出すると、クラウドや医療プロバイダーと通信するデジタルコンポーネントを起動する可能性があると述べています。

💡機械学習

機械学習とは、コンピュータがデータから学習し、タスクを遂行する能力を獲得するプロセスです。ビデオでは、アナログプロセッサがAIや機械学習アルゴリズムを使いながらも、デジタルプロセッサよりも少ないエネルギーで動作できると述べています。

💡パワーインテリジェント

パワーインテリジェントとは、少ないエネルギーで効率的に情報を処理できることを意味します。ビデオでは、アナログプロセッサがパワーインテリジェントであり、センサーの入力でデータをリアルタイムで判断し、その後の処理にその効率性を活用できると説明しています。

Highlights

The next big technology could be an analog processor, which operates on waves rather than binary digits.

Analog processing can provide insights from data in its rawest form, unlike digital computing which is software-driven.

Analog processors can still be programmed with software, offering a significant advantage.

New devices using AI and machine learning algorithms with analog processors could use 1/1000th of the energy of digital counterparts.

Aspinity has developed a software-programmable analog processor that maintains analog benefits while being easily programmable.

Aspinity's technology solves the 'voltage offset' issue in analog processing, ensuring consistent outputs across different chips.

Analog processors allow signals from sensors to be used directly without conversion to digital, simplifying the process.

The development of analog computing could lead to a 'quick reset' in the architecture of computing.

Digital signal conversion consumes significant energy, and analog computing could reduce this consumption dramatically.

Analog computing could make it feasible to monitor more systems effectively, such as in industrial machinery or vehicles.

Analog processors are not intended to replace digital computers but to be used strategically in combination with them.

Analog computing can enable 'always-on' functionality with minimal energy use, extending battery life in devices.

The technology could allow for more specific and low-energy sensors, enhancing personalized insight-gathering devices.

Analog intelligence could change how intelligent devices in our lives understand and interact with us.

Aspinity's analog processors have the potential to make therapy more affordable and accessible through online services like BetterHelp.

The future of computing may involve an orchestration of analog and digital computers for optimized efficiency and sensing.

Analog computing could lead to the creation of new kinds of sensors and an explosion of innovation in the field.

Heart monitoring is an ideal application for analog computers, potentially allowing for more effective health monitoring with less power.

Transcripts

play00:00

- What if the next big technology

play00:02

was actually a really old technology?

play00:06

The processors in your computer or phone are digital.

play00:09

They think in ones and zeros.

play00:12

But this processor thinks in waves.

play00:15

This is an analog processor,

play00:18

and it might allow us to totally rethink computing

play00:21

from the ground up.

play00:22

Oh wow, it can cut through anything.

play00:24

- For the last few decades,

play00:26

digital computing has always been driven by software.

play00:29

And it became very easy to use and very easy

play00:31

to build applications.

play00:33

But all of the information in the world is natively analog.

play00:37

And so what analog processing allows us to do

play00:40

is understand and potentially inference and gather insights

play00:44

from that data in its rawest form.

play00:46

- But just like the digital processors you know,

play00:48

it can still be programmed with software,

play00:50

and that's kind of a big deal.

play00:52

- Imagine being able to build new devices

play00:54

that use AI and machine learning algorithms,

play00:57

but yet, use 1/1000 of the energy that they use

play01:00

from a digital perspective.

play01:01

And so you get all those benefits of analog-

play01:05

the power, the efficiency, the accuracy-

play01:07

but now you can use a software to program it very quickly.

play01:12

- Does all processing need to be digital?

play01:14

How will analog processors change the way our computers work

play01:17

and interact with us?

play01:19

This is "Hard Reset,"

play01:21

a series about rebuilding our world from scratch.

play01:25

Hey, everyone.

play01:26

Real quick, we're trying something new at "Hard Reset."

play01:29

This is a quick message about our sponsor

play01:31

for this episode, BetterHelp.

play01:32

Now, I wanna make sure you know "Hard Reset"

play01:34

will never take sponsorship money from our subjects,

play01:37

like Aspinity, which is the company we're profiling

play01:39

in this episode.

play01:40

But sponsorships from other companies, like BetterHelp,

play01:43

will make it easier for us to tell more stories

play01:45

we know you will love.

play01:46

And BetterHelp is a company we're happy to talk about.

play01:49

Frankly, at "Hard Reset," one of the main things

play01:51

we're always asking ourselves as we develop stories is:

play01:54

What are the assumptions about the way the world works

play01:57

that we should rethink?

play01:58

And personally, I think we need

play01:59

to rethink our cultural taboos around mental health.

play02:03

Your mental health is super important-

play02:05

and it's something worth taking care of.

play02:08

I'm someone who has benefited from good therapy,

play02:10

and many of my dearest and closest friends have as well.

play02:13

I think we can all agree that the last few years

play02:15

have been pretty wild.

play02:17

And one of the things I learned through therapy

play02:19

is that it's totally normal for all of that to get to you

play02:21

from time to time.

play02:23

And just knowing that

play02:24

has helped make things more manageable for me.

play02:26

BetterHelp's mission is to make therapy more affordable

play02:28

and more accessible.

play02:30

Basically, it's an online service

play02:31

that matches you up with therapists based

play02:33

on your preferences and their availability.

play02:36

BetterHelp makes it easy for you to take that first step

play02:39

towards improving your mental health.

play02:41

It's easy to do, and it's worth doing.

play02:44

Taking care of your mental health isn't just nice to do.

play02:47

It's vital to making sure you're living a life

play02:50

that is fulfilling and satisfying for you.

play02:52

If you do decide to try out BetterHelp,

play02:55

you can get 10% off your first month

play02:57

by going to betterhelp.com/freethink.

play03:01

Okay, back to nerding out about analog computers.

play03:04

The first computers ever made were analog.

play03:07

Think Ancient Greece. Later on, we had slide rules,

play03:11

and in the 19th century, we developed analog computers

play03:14

that calculated how things like the tides would change

play03:16

over time or solve differential equations.

play03:19

We met the folks from Aspinity to learn

play03:21

about analog processors.

play03:23

- Analog has not gone away.

play03:25

I mean, every iPhone uses analog to collect data

play03:28

and analog to transmit data in the form of RF.

play03:31

So analog has always been there,

play03:32

but it's bene relegated to those primary capabilities.

play03:37

- Analog processors were surpassed

play03:39

by digital processors,

play03:41

mainly because digital could be manufactured more easily

play03:44

and they could be programmed with software.

play03:46

This made them far more flexible and adaptable.

play03:51

But what about a software-programmable analog processor?

play03:54

Well, that's what this is.

play03:57

- The challenge with analog has always been

play04:00

when you put a certain input in,

play04:02

for the same circuit in a different silicon,

play04:06

you might get a slightly different output.

play04:08

And we call that 'voltage offset.'

play04:10

So 2.1 and 1.9 aren't exactly the same.

play04:15

You want it to be two.

play04:16

And Aspinity has solved a major hurdle,

play04:20

and through our software,

play04:20

we're able to make fine-tune adjustments

play04:23

so that that chip gets a 2 at the output,

play04:26

this chip gets a 2 at the output.

play04:29

But we also solved a key problem with regards to use model

play04:32

and how easy it is to use.

play04:33

And so we're now able to program an analog processor

play04:37

from Aspinity in the same way that we do digital.

play04:40

- That means the signals from all the sensors

play04:42

that these chips are connected to don't need

play04:44

to be converted into ones and zeros-

play04:47

they can just go straight in.

play04:48

And you can write the software that directs the chip

play04:51

about how to interpret those signals.

play04:53

That's sorta huge.

play04:55

But don't take my word for it.

play04:57

Jared here has a little more insight

play04:59

as a product designer at argodesign here in Austin, Texas.

play05:03

- Aspinity has slayed a dragon here,

play05:06

and I've been going along blissfully thinking

play05:08

like the history of computing is set and settled.

play05:11

We will do digital.

play05:12

We will go binary.

play05:14

Turns out that all along,

play05:16

analog has been slowly creeping along.

play05:20

Now that we've solved the stability

play05:22

and programmability issues of analog computing,

play05:26

we're gonna see a quick reset

play05:29

to the entire architecture of computing.

play05:33

- Right now, we spend tons of energy

play05:35

breaking analog signals into digital signals,

play05:37

just so that our computers can understand them.

play05:40

And by tons of energy, we're not being hyperbolic.

play05:43

Think about all the sensors in your phone,

play05:45

your house, your car, at work.

play05:48

Most of them collect analog signals

play05:50

that have to be converted into digital signals

play05:53

and usually back into analog signals

play05:55

so that we can hear them or see them.

play05:57

- We expect 30 billion devices like this by the end

play06:00

of the decade.

play06:01

Every one of these will be computing on data,

play06:04

and it takes tremendous amounts of energy

play06:07

to have these always on computing.

play06:10

- All that energy

play06:11

we spend converting signals adds up.

play06:14

In fact, the energy consumption of digital computers

play06:17

is rising faster than the rate at which we create energy-

play06:20

that's a problem.

play06:21

- It allows these products to be more power-intelligent.

play06:25

So we can really understand right at the sensor input

play06:29

whether the data is relevant or not.

play06:32

And then that drives everything after that

play06:34

to take advantage of that efficiency.

play06:36

- This technology isn't meant

play06:37

to replace digital computers,

play06:39

but it could mean we use them much more strategically.

play06:42

- We can still have a combination of analog

play06:45

and digital locally,

play06:47

but we wanna be able to use that analog

play06:48

as your always-on computing and then wake up the digital

play06:52

as necessary or wake up and send data

play06:54

to the cloud as necessary.

play06:56

- Think of it this way:

play06:57

How often do you change the batteries

play06:58

in your smart TV's remote control?

play07:00

Well, even those batteries are slowly being nibbled away at

play07:03

by the digital processors that sit and listen

play07:06

for your voice to command them.

play07:07

- And all this is just so you can say, you know,

play07:09

"Hey TV, I wanna watch a rom-com with Matthew McConaughey.

play07:13

- You really can't make it five minutes

play07:14

into a conversation in Austin

play07:16

without someone bringing up Matthew McConaughey.

play07:18

- 'All right, all right, all right.

play07:20

- Oh Christ.

play07:23

- How you doin'?'

play07:24

- Anyway-

play07:25

- With analog computing, it can be listening continuously

play07:28

for that wake word that's using almost no electricity,

play07:31

never doing analog-digital conversion,

play07:33

and then it can wake up a more expensive system

play07:35

to do the work of the request.

play07:37

- That means that same remote control

play07:39

could last 10 or even 100 times longer on the same batteries

play07:43

or you could just use much smaller batteries

play07:45

and get the same functionality.

play07:49

- And you can virtually do any audio, right?

play07:51

This is a glass break detector.

play07:53

This also is for automotive glass break as well.

play07:55

- Right.

play07:56

So you can potentially build the kind of filters

play07:58

and decision tree that can say,

play07:59

"That was glass breaking.

play08:01

That's a window.

play08:02

This is a car window, tempered glass breaking."

play08:04

And then based on whether or not it hits one

play08:06

of those priority glass breaks,

play08:08

it can send that signal to the digital chip,

play08:10

which communicates via either Wi-Fi or cell signal,

play08:13

I'm guessing? - Exactly.

play08:14

Once we make the decision, we go ahead and send the trigger-

play08:17

- Right. - Out to whomever,

play08:19

a call center in some cases, directly to the authorities.

play08:22

- Hey, Siri,

play08:25

the glass is broken.

play08:27

You can't fool it.

play08:28

- You wanna hold it? - Yeah.

play08:30

I'll see what happens.

play08:32

All right.

play08:35

What if I'm talking over it while you do it?

play08:37

Now does it filter out?

play08:37

So here I'm gonna talk.

play08:38

I'm gonna say,

play08:40

this is "Hard Reset,"

play08:41

a series about, wow, that really.

play08:43

- Yeah. - I'm kinda surprised

play08:44

that it could differentiate between the two,

play08:46

'cause a lotta people say my voice sounds

play08:47

like breaking glass.

play08:49

- It may not be a major item that gets solved,

play08:52

but all these little things

play08:54

where you're using energy inefficiently add up.

play08:57

- Where this really starts to impact society at scale

play09:00

isn't your remote control.

play09:02

It's your pipes, your solar power plants, your cars.

play09:05

- On a piece of equipment like a stacker,

play09:08

which is one of those conveyor belts that runs dirt up

play09:10

and makes piles and there's something on the order

play09:13

of a half a million of 'em in operation every day,

play09:15

and just, if you have a chip set that can run for years

play09:19

on battery power with enough machine-learning intelligence

play09:23

to listen to a vibration and determine

play09:26

that it's not just a vibration,

play09:28

but the type that indicates a bearing is going out,

play09:30

and then you could use that information

play09:34

to find out what it would cost to repair a bearing.

play09:37

When you think about the whole system,

play09:41

it's not just that the chip set uses less power,

play09:43

it's that because it uses less power,

play09:45

it creates a new computing architecture

play09:47

where we can get more value, the insights,

play09:50

with less resources.

play09:53

- Monitoring all these things with digital systems

play09:55

would be impractical.

play09:56

But with analog systems, it starts to seem achievable.

play10:02

So, picture a scenario where analog computers orchestrate

play10:05

the activity of digital computers.

play10:07

This would mean better sensing systems and more of them

play10:10

in places where it can make a difference.

play10:12

And of course, all the sensing of the real world

play10:14

can be done with far less energy than the digital systems

play10:17

we currently use.

play10:18

You probably won't ever have an all analog computer.

play10:22

Some things, like audio devices, might become mostly analog.

play10:26

That would probably be music to the ears

play10:28

of your vinyl-loving friend.

play10:30

But most implementations would be about blending

play10:33

the best strengths of analog and digital.

play10:36

For example, heart monitoring is perfect

play10:38

for analog computers.

play10:39

Our heartbeat is an analog signal.

play10:42

And with power-efficient, tiny sensors,

play10:44

more people could be monitored effectively.

play10:47

If the analog portion ever detects an anomaly,

play10:49

it can wake up the digital component

play10:51

that communicates with the cloud

play10:53

or contacts a health care provider.

play10:55

- Once you have these kinda chip sets,

play10:57

you're gonna see an explosion of people looking

play11:00

to create new kinds of analog sensors.

play11:03

This would allow me to create an entire architecture

play11:05

around voice computing in the home,

play11:07

because I could look at the faucet and say, "Turn on,"

play11:09

and the stove wouldn't turn on

play11:11

because I wasn't looking at it.

play11:13

- It's this idea of, it'll be a very low-energy deployment

play11:18

of all of these insight-gathering devices

play11:21

that are very specific to you.

play11:23

And there's no middleman going to the cloud necessarily

play11:26

or no middleman that's monitoring it for you-

play11:29

it can come to you.

play11:31

- We're constantly relying on more

play11:33

and more intelligent devices in our lives.

play11:35

But analog intelligence might totally change the way

play11:37

those devices understand us.

Rate This

5.0 / 5 (0 votes)

Related Tags
アナログプロセッサデジタル対比エネルギー効率ソフトウェアプログラミングAIアルゴリズムデジタルシグナル変換センサーデータパワーインテリジェンスデジタルコンピュータコンピューティングアーキテクチャ心電図監視音響デバイス低エネルギーデプロイ個々のニーズ
Do you need a summary in English?