I have a significant QTY of Sophon SC1 TPUs in deployment for our research customers.
It’s worth noting that the device presents itself to the OS as a Xilinx memory controller with 16GB DDR4.
Bitmain has detailed FP16/32 INT functions available to the Metal API with a theoretical max throughout of 2TFLOPS.
If anyone is interested in getting schedule SSH access to a host with an SC-1, please feel free to PM me.
Topic: 🔴 SyncAi - distributed Artificial Intelligence System - page 14. (Read 19013 times)
What kind of HW is necessary?
Can you give a little more description how it works?
Can you give a little more description how it works?
Now in the AI industry there are some problems: datasets are difficult to prepare and have nowhere to sell. Algorithms are the same. I want to start with simple things and then develop to the marketplace. The most important thing is to create a system in which you could use someone else's experience: the AI chains
Ps what is it - HW?
Have you heard about Pallium? That's a decentralized platform for AI-developers. They make an ecosystem where developers can sell their models, and another people or companies can buy it. Also there are a "marketplace" for data buyers and sellers. I think it can solve the problems you describe!
I did a little analysis and understood that models and data-sets can be downloaded and purchased on many sites) But no one will give the best options. Do you know why? - "pirates") and illegal copying. In Sinkai both data and models and algorithms are not transmitted to anyone from the author. You can process your data with his program and get the result. and that's all. The client receives the result, the author receives money for the work of the program.
What kind of HW is necessary?
Can you give a little more description how it works?
Can you give a little more description how it works?
Now in the AI industry there are some problems: datasets are difficult to prepare and have nowhere to sell. Algorithms are the same. I want to start with simple things and then develop to the marketplace. The most important thing is to create a system in which you could use someone else's experience: the AI chains
Ps what is it - HW?
Have you heard about Pallium? That's a decentralized platform for AI-developers. They make an ecosystem where developers can sell their models, and another people or companies can buy it. Also there are a "marketplace" for data buyers and sellers. I think it can solve the problems you describe!
Nice! I met Sophia once here in London. Cool AI humanoid 
Looking forward to test it
Looking forward to test it
Haven't much, just some samples to try it out. Setting it up for now
Beta version under dev. SyncAi is a “lite version” of SingularityNet
Haven't much, just some samples to try it out. Setting it up for now
I am developing an AI/Deep Learning machine, if I can be of any help, please let me know
Specs:
Amd 16 cores water cooled
128Gb Ram
2 x 256Gb 2GB/s SSD (one Ubuntu 16.04 LTS, one Windows 10 Enterprise LTSB)
Latest mobo
Nvidia 1080Ti 11Gb
AMD Radeon Pro Duo 32Gb
Is built already, just tuning it up with standards DL software
I also have standard mining rigs (multiple AMD, Nvidia cards) if you need to try a software on different machines.
Specs:
Amd 16 cores water cooled
128Gb Ram
2 x 256Gb 2GB/s SSD (one Ubuntu 16.04 LTS, one Windows 10 Enterprise LTSB)
Latest mobo
Nvidia 1080Ti 11Gb
AMD Radeon Pro Duo 32Gb
Is built already, just tuning it up with standards DL software
I also have standard mining rigs (multiple AMD, Nvidia cards) if you need to try a software on different machines.
very good. Did you teach neural networks?
I am developing an AI/Deep Learning machine, if I can be of any help, please let me know
Specs:
Amd 16 cores water cooled
128Gb Ram
2 x 256Gb 2GB/s SSD (one Ubuntu 16.04 LTS, one Windows 10 Enterprise LTSB)
Latest mobo
Nvidia 1080Ti 11Gb
AMD Radeon Pro Duo 32Gb
Is built already, just tuning it up with standards DL software
I also have standard mining rigs (multiple AMD, Nvidia cards) if you need to try a software on different machines.
Specs:
Amd 16 cores water cooled
128Gb Ram
2 x 256Gb 2GB/s SSD (one Ubuntu 16.04 LTS, one Windows 10 Enterprise LTSB)
Latest mobo
Nvidia 1080Ti 11Gb
AMD Radeon Pro Duo 32Gb
Is built already, just tuning it up with standards DL software
I also have standard mining rigs (multiple AMD, Nvidia cards) if you need to try a software on different machines.
new topic: best AI tools
opennmt net
OpenNMT is an open source (MIT) initiative for neural machine translation and neural sequence modeling.
Since its launch in December 2016, OpenNMT has become a collection of implementations targeting both academia and industry. The systems are designed to be simple to use and easy to extend, while maintaining efficiency and state-of-the-art accuracy.
Where i can find an information about the pool?
You can ask me (im author) or wait for site+wp
Where i can find an information about the pool?
Hi,
Before , I must config a computer with ubuntu
On my FB , posted 3 photos ( I hope that it's authorized )
" https://www.facebook.com/photo.php?fbid=1691110300968841&set=pcb.1691111817635356&type=3 "
" https://simpleintelligence.com/home " and " https://vimeo.com/247744147 "
Regards
Before , I must config a computer with ubuntu
On my FB , posted 3 photos ( I hope that it's authorized )
" https://www.facebook.com/photo.php?fbid=1691110300968841&set=pcb.1691111817635356&type=3 "
" https://simpleintelligence.com/home " and " https://vimeo.com/247744147 "
Regards
Only Ubunty?
Hi,
Before , I must config a computer with ubuntu
On my FB , posted 3 photos ( I hope that it's authorized )
" https://www.facebook.com/photo.php?fbid=1691110300968841&set=pcb.1691111817635356&type=3 "
" https://simpleintelligence.com/home " and " https://vimeo.com/247744147 "
Regards
Before , I must config a computer with ubuntu
On my FB , posted 3 photos ( I hope that it's authorized )
" https://www.facebook.com/photo.php?fbid=1691110300968841&set=pcb.1691111817635356&type=3 "
" https://simpleintelligence.com/home " and " https://vimeo.com/247744147 "
Regards
Hi,
You talk about an AI mining project.
"....Not necessary. But if you can connect this system via API to SyncAi, then, perhaps, it will be more profitable than mining on the video card...."
The last week , I receided SC1 card from bitmain (sophon)
Runs under ubuntu 16.04
Sophon team says not for minning
But do you think that a third partie ( api) , would be developped?
Best regards
You talk about an AI mining project.
"....Not necessary. But if you can connect this system via API to SyncAi, then, perhaps, it will be more profitable than mining on the video card...."
The last week , I receided SC1 card from bitmain (sophon)
Runs under ubuntu 16.04
Sophon team says not for minning
But do you think that a third partie ( api) , would be developped?
Best regards
https://sophon.ai/product/sc1.html
Good choice
use case: how to use that with Syncai. make a program or tensorflow module and connect to Syncai ./syncai youprogram options.yml
Sorry im not spec in neuro networks)
Hi,
You talk about an AI mining project.
"....Not necessary. But if you can connect this system via API to SyncAi, then, perhaps, it will be more profitable than mining on the video card...."
The last week , I receided SC1 card from bitmain (sophon)
Runs under ubuntu 16.04
Sophon team says not for minning
But do you think that a third partie ( api) , would be developped?
Best regards
You talk about an AI mining project.
"....Not necessary. But if you can connect this system via API to SyncAi, then, perhaps, it will be more profitable than mining on the video card...."
The last week , I receided SC1 card from bitmain (sophon)
Runs under ubuntu 16.04
Sophon team says not for minning
But do you think that a third partie ( api) , would be developped?
Best regards
Site: syncai.online
Under construction
Will be on next week
+ wiki
Under construction
Will be on next week
+ wiki
the format of the commands will be like that of Docker
after viewing SN codes on githab, I think that the SN will be difficult to configure and manage and connect
SyncAi is a light version of SN)))
You say that SN is difficult to configure, manage and connect and SyncAI is a light version of SN.
There is already a well known company who do searches based on images and multiple companies who do searches based on snippets of music. They are backed by huge companies.
What do you have that gives you an edge over these companies?
How will you compete with those well funded companies and make a dent on the market?
Singularity and Goertzel havent any super algoritms. Only Sophia that can recognize speech. Siri/Cortana/Alexa is more smarter
Syncai was invented independently of Singularity. I did not know about them. And this idea took me a few months and not years) So I'm smarter than them)))
after viewing SN codes on githab, I think that the SN will be difficult to configure and manage and connect
SyncAi is a light version of SN)))
You say that SN is difficult to configure, manage and connect and SyncAI is a light version of SN.
There is already a well known company who do searches based on images and multiple companies who do searches based on snippets of music. They are backed by huge companies.
What do you have that gives you an edge over these companies?
How will you compete with those well funded companies and make a dent on the market?
Up
Jump to: