IoT: little devices in huge numbers = huge data. Many challenges (security, security, & security are top 3!). Era of instrumenting the world is just beginning. Collect data at vast scale to answer truth # 1.
-
Show this thread
-
Tech leaders are proclaiming disruption, revolution. But it is this AND a natural evolution. AI interest exploding because it is better than any previous methods at insight/foresight. Quality of data vastly important as well.
1 reply 0 retweets 1 likeShow this thread -
Growth of deep learning due to 3 factors: 1. explosive growth in data, of all types 2. tremendous growth in computing power 3. algorithmic advances in artificial neural networks
1 reply 0 retweets 1 likeShow this thread -
ML process: training phase, use (inference) phase. Training very expensive, inference fast. Story: engineer wants to prevent cats from peeing on his yard by creating AI that recognized cats & would automatically turn on sprinklers to chase them away.
1 reply 0 retweets 1 likeShow this thread -
Trained algorithm on pics of cats from the internet, but it worked only 65% of the time. Turns out he was training algorithm on cute cats. With further non-cute cat input increased success to 90%.
1 reply 0 retweets 1 likeShow this thread -
Innovating requires ever more capability. Modeling/simulation enable design according to natural laws, virtual engineering, etc. ML & DL enable discovery & innovation where laws of nature don't apply or are non-linear/difficult to simulate accurately. May be quicker/cheaper.
1 reply 0 retweets 1 likeShow this thread -
AI can't replace all simulation, but it can augment, guide, reduce cost.
1 reply 0 retweets 1 likeShow this thread -
HPC = high performance, not mainstream. All about performance. Important in AI because of massive computations necessary. Lots of linear algebra.
1 reply 0 retweets 1 likeShow this thread -
Data center/cloud aren't going anywhere. Entering an era with billions of autonomous devices on the edge that can asynchronously connect to the data center/cloud.
1 reply 0 retweets 1 likeShow this thread -
These include: machines predicting their future failure & calling for maintenance; detecting intruder in your house; turning on/off lights, etc.
1 reply 0 retweets 1 likeShow this thread
To solve incompatibility between IoT systems, Linux Foundation's EdgeX Foundry Project. Dell is supporting this project.
-
-
Smart Cities: 1 example of bringing it all together. Predictive analytics & predictive simulations to optimize traffic, safety, energy/water utilization, & more.
1 reply 0 retweets 1 likeShow this thread -
AI is not a trend, it is here to stay & will evolve rapidly. HPC is fundamental in this.
0 replies 0 retweets 1 likeShow this thread
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.