Seems like a lot of people are confused between the concept of edge computing and the concept of conventional personal computers. Conventional PCs are something that you can with have a lot of freedom and flexibility. Essentially you can do whatever you want with your PC - re-program it, install software on it, play with it... However an edge computing device is something that is more likely administered by a central service provider (i.e. Amazon Alexa offered by Amazon) that you cannot easily tamper with, but it does have enough processing power to do many things locally rather than always having to talk to the Cloud. A prime example is a self-driving car. A self-driving Tesla is an epitome an "edge computer" which drives your car without having talking to the Cloud - and quite conceivably so. Moreover, a Tesla self-driving car receives upgrades from Tesla every once in a while - a sign of being administered by Tesla. Last but not least, you can't re-program a Tesla to let it ignore pedestrians in front of itself otherwise it'd become a really cheap assassin... Hope it helps.
Very nice video on wdge computing that gives an crystal clear explanation on the topic without making us confuse between concepts like embedded devices and etc.Thanks for the video to both IBM and Rob High.
my guess is that he writes it normally on the glass, then they flip the video horizontally to make the words readable again you can also see he's wearing his watch on the right wrist
Sounds like "Edge Computing" is just normal computing. Our premature adoption of cloud technology has left us in a situation where we realize what we needed all along is what we had before investing into the cloud. Ironic really.
True and we do see that the cloud is also useful thus more hybrid deployments are occurring. Remember the whole point of edge computing is to bring time sensitive programs and applications closer to the end users/devices to avoid latency to increase productivity and network resilience. So the cloud is great for storage and NOC visibility of a large infrastructure deployments but, not so good for engineering design tools, shop floor control systems, Docker platforms, the list goes on.
For sure. But However I think the big difference is that is not about computers anymore, but things with a computer. I am no expert here, but that is how I see it.
You just hit me with the piece that I was missing. I was sitting here watching video after video trying to figure out what the heck edge computing is, and you answered it. It's just over-thought non-cloud computing, connected to the cloud.
That's what I had sensed when I first read about it. I think the only difference is that what we now see as IOT devices will have more dynamic computing abilities.
Fantastic explanation, great video. I'd love to see more of these. I'd love some clarification on 2 points though: 1- The 'Network Edge' seems to be quite different conceptually from Edge Devices and Servers. I wonder if this is just a nomenclature thing 2- What kinds of workloads are we thinking of when we refer to Edge Servers or the Network Edge?
How does an analyst perform threat hunting and detect attacks on edge devices such as routers, firewalls, F5, switches, etc.? Is the approach to detecting attacks on edge devices specific to the SIEM (Ex: QRadar) and how to query the logs? If so, how do I find keywords to search, strings to search, Event IDs, etc.? Are the keywords, strings, and event IDs vendor specific to search and detect malicious behavior? So far, the only solutions I have found to detect attacks on edge devices are CVEs and Nessus Plugins. Any assistance would be much appreciated.
hi Rob, thanks for the nice video！ a question：could you like kindly explain what is the different between ‘edge computing’ and ‘distributed computing’？ I am really confuse about these two conceptions KR
Very nice video. It perfectly aligns with what I was thinking in terms of Architecture and Applications. Due to the limited range of the 5G (while research is going on), the applications will be internal network using 5G technology. Eventually, with a higher range 5G (like a 4G LTE), of course, it opens up doors for all interesting use cases. Perhaps at that time, it will be called 6G.
It almost seems like computing has come full circle. From on-premise to cloud to essentially on-premise again, except the computing power is now in devices rather than on server racks. When is this going to end, lol? When we get to full 5G (or better) and much better computing power, will we then decide it's better to (re)locate everything back to a central location?
your comment actually helped me understand this video. lol so thanks for that! i wonder why there was a need for edge computing in the first place. could you help again?
We do not, computing on the device itself is much quicker, what happens now is faster and faster computing on the devices which leads to better performance.
Is edge computing a kind of atavism: the previous local computer's ability is very weak, so the computing tasks are uploaded to the cloud computing, but now we want to reduce the cloud computing and return more computer tasks to the local computers.
Hello sir I am bit confused . Where does the edge computing perform i.e in edge device itself or in edge server. If it performs the computing in edge servers then in the case of autonomous vehicles we cant expect continuous network connection because of its mobility. So in this case if the network connection is lost then how can the edge server communicate with that autonomous vehicle.
great video: few questions: Is edge computing simply passing the data processing/computing loads to edge devices/edge servers? Does this mean that Sensitive data won't be transferred to the data centers at all? If it will still be transferred, then how does that fix the vulnerability issue. Thanks so much!
Please I have a question, I'm having issue on what area to settle into, I have a background on python and the supervised learning in Machine learning but to apply my skills in the automotive industry, can u advise me, on how to tailor my part?
I felt this video was filled with too much tech jargon making it difficult for a newcomer to understand. It seemed like an explanation for people who already understand what edge computing is.
Seems like a lot of people are confused between the concept of edge computing and the concept of conventional personal computers. Conventional PCs are something that you can with have a lot of freedom and flexibility. Essentially you can do whatever you want with your PC - re-program it, install software on it, play with it... However an edge computing device is something that is more likely administered by a central service provider (i.e. Amazon Alexa offered by Amazon) that you cannot easily tamper with, but it does have enough processing power to do many things locally rather than always having to talk to the Cloud. A prime example is a self-driving car. A self-driving Tesla is an epitome an "edge computer" which drives your car without having talking to the Cloud - and quite conceivably so. Moreover, a Tesla self-driving car receives upgrades from Tesla every once in a while - a sign of being administered by Tesla. Last but not least, you can't re-program a Tesla to let it ignore pedestrians in front of itself otherwise it'd become a really cheap assassin... Hope it helps.
you're describing an edge device, not an edge server
You are the only channel that i know off that tells the description in a vary clear and understandable manner. Thanks
Very nice video on wdge computing that gives an crystal clear explanation on the topic without making us confuse between concepts like embedded devices and etc.Thanks for the video to both IBM and Rob High.
Like how you guys are able to mirror write so smoothly.
my guess is that he writes it normally on the glass, then they flip the video horizontally to make the words readable again
you can also see he's wearing his watch on the right wrist
Sounds like "Edge Computing" is just normal computing. Our premature adoption of cloud technology has left us in a situation where we realize what we needed all along is what we had before investing into the cloud. Ironic really.
Agreed, Sounds like client server with data warehouse
True and we do see that the cloud is also useful thus more hybrid deployments are occurring. Remember the whole point of edge computing is to bring time sensitive programs and applications closer to the end users/devices to avoid latency to increase productivity and network resilience. So the cloud is great for storage and NOC visibility of a large infrastructure deployments but, not so good for engineering design tools, shop floor control systems, Docker platforms, the list goes on.
For sure. But However I think the big difference is that is not about computers anymore, but things with a computer. I am no expert here, but that is how I see it.
You just hit me with the piece that I was missing. I was sitting here watching video after video trying to figure out what the heck edge computing is, and you answered it. It's just over-thought non-cloud computing, connected to the cloud.
That's what I had sensed when I first read about it. I think the only difference is that what we now see as IOT devices will have more dynamic computing abilities.
Always a pleasure to learn from Rob High.
this video, and this channel in general, would be better if they included more specific use cases
Fantastic explanation, great video. I'd love to see more of these.
I'd love some clarification on 2 points though:
1- The 'Network Edge' seems to be quite different conceptually from Edge Devices and Servers. I wonder if this is just a nomenclature thing
2- What kinds of workloads are we thinking of when we refer to Edge Servers or the Network Edge?
Okay I just have to say unless some video-magic is occurring, his backwards-writing skills are impeccable
Or maybe they just flipped the video.
That's exactly what we did, Debashish! 👏 More details here 👉 ibm.co/3gDrpDY
I'd also been thinking about how the video was achieved... 🤣
Here is a blog post we wrote that explains how we make our videos! ➞ ibm.co/2LTPMjo
@@IBMTechnology the link provided doesn’t work. Can you please examine and send a link that can be viewed by public.
Really like these education series from IBM, thanks!
Glad you like them!
Excellent overview of Edge Computing!
Thank you, Eric! Glad you enjoyed the video!
Go Edge! Thx for video.
Great quality explanation.
Very good video. Thanks for creating it.
Just commenting so that algo suggests me more of such videos
Ha, ha, great trick! 😉 Feel free to also browse through our channel to see other videos like this, we have quite a lot of them. 💪
excellent video thanks! keep posting videos like this one
There's a project called StrongNode that uses Edge computing. Check their litepaper if you are interested. Really fascinating stuff.
Hi.. How i get that paper
thank you! The info is nicely explained that you don't have to be an expert to understand.
Very Informative Video
I understood it and I’m a beginner in a lot of this stuff!
How does an analyst perform threat hunting and detect attacks on edge devices such as routers, firewalls, F5, switches, etc.? Is the approach to detecting attacks on edge devices specific to the SIEM (Ex: QRadar) and how to query the logs? If so, how do I find keywords to search, strings to search, Event IDs, etc.? Are the keywords, strings, and event IDs vendor specific to search and detect malicious behavior? So far, the only solutions I have found to detect attacks on edge devices are CVEs and Nessus Plugins. Any assistance would be much appreciated.
Wow thanks I understood all of it!
hi Rob, thanks for the nice video！
a question：could you like kindly explain what is the different between ‘edge computing’ and ‘distributed computing’？
I am really confuse about these two conceptions
KR
loved this. subscribed!
Did he wrote all those in reverse facing the camera while recording the video the whole time??? And his handwriting is awesome!
I was also wondering about that, throughout this Vid.
Or maybe just write normally and then flip the video
Very good video, thanks!
Very nice video. It perfectly aligns with what I was thinking in terms of Architecture and Applications. Due to the limited range of the 5G (while research is going on), the applications will be internal network using 5G technology. Eventually, with a higher range 5G (like a 4G LTE), of course, it opens up doors for all interesting use cases. Perhaps at that time, it will be called 6G.
Excellent ,thanks very much.
Very well explained. Thank you.
Nicely done. Would love to see you do another as a follow up including IBM + IBM Ecosystem and a client environment do it.
It almost seems like computing has come full circle. From on-premise to cloud to essentially on-premise again, except the computing power is now in devices rather than on server racks. When is this going to end, lol? When we get to full 5G (or better) and much better computing power, will we then decide it's better to (re)locate everything back to a central location?
your comment actually helped me understand this video. lol so thanks for that! i wonder why there was a need for edge computing in the first place. could you help again?
Following
We do not, computing on the device itself is much quicker, what happens now is faster and faster computing on the devices which leads to better performance.
I am amazed by how he type those words..
Is edge computing a kind of atavism: the previous local computer's ability is very weak, so the computing tasks are uploaded to the cloud computing, but now we want to reduce the cloud computing and return more computer tasks to the local computers.
Hello sir
I am bit confused . Where does the edge computing perform i.e in edge device itself or in edge server. If it performs the computing in edge servers then in the case of autonomous vehicles we cant expect continuous network connection because of its mobility. So in this case if the network connection is lost then how can the edge server communicate with that autonomous vehicle.
it's in the device
what about caching?
and security?
Great video thank you
So, edge computers are not meant to be used by individual users but by big corporations or industries?
4DS ReRam might help when it comes to the huge amount of memory out there on the edge?
WOW! Does this sir wrote things inversely?
Is there a need for edge computing when quantum computing comes out.
great video:
few questions:
Is edge computing simply passing the data processing/computing loads to edge devices/edge servers?
Does this mean that Sensitive data won't be transferred to the data centers at all? If it will still be transferred, then how does that fix the vulnerability issue.
Thanks so much!
Which software was used to create this video?
Please I have a question, I'm having issue on what area to settle into, I have a background on python and the supervised learning in Machine learning but to apply my skills in the automotive industry, can u advise me, on how to tailor my part?
Ok. Thank you very much, but What is actually an Edge Computing?
what is the difference between edge and fog ?
Fog computing is a edge computing concept but purely on Cisco platforms.
How to work on edge cloud computing simulator?
Cool, but after learning what edge computing does, I still don't know what edge computing is.
Hey Awesome Explanation. I'm thinking of using Edge for my use case.
@ibm Can we talk about this ?
Nice concepts but the presentation lacks examples.
wish the audio was louder hehe
Hi thanks for the great video !!
Please suggest me a problem statement for edge wanna implement in my research .
Thanks
Sana: Can you give us more information on what you are trying to do? What sort of research is it? Thank you.
what about coronavirus? Any way that technology can helps people against this disease?
At 4:15 : Those Deep Eyes with White Box Leds...... Ha ha... Amazing video... Now i know
What kind of white board are you using?
Hi thanks for your question. This post explains everything. ibm.co/2SA1vGd Thank you for watching the video.
It could also be Power of IBM Z.
Very engaging concept but this sounds to me like the beginning of data decentralization once again and the slow death of huge data centers!
Am I the only one baffled by his ability of mirror writing?
waaaaaaaaaaaaaaaaaaaaaaaaay better than GS
I felt this video was filled with too much tech jargon making it difficult for a newcomer to understand. It seemed like an explanation for people who already understand what edge computing is.
I'm looking to learn edge computing but I don't like programming. Can someone please tell me does edge computing include programming?
a car has 50 cpus?
Theta 😌
Average car has 50 cpu's, what.
Sounds like a precursor to pied piper's decentralized internet
Too low sound