Тёмный
No video :(

Dual Edge TPU: Double Power, Same Slot 

Data Slayer
Подписаться 212 тыс.
Просмотров 32 тыс.
50% 1

For 20% Off Brilliant.org Membership! 👉 brilliant.org/...
Full Tutorial Blog Post 👉 / how-to-setup-the-coral...
Hardware Links
Webcam 👉 amzn.to/49RgZw9
Coral AI Dual Edge TPU 👉 amzn.to/3TwHUaY
Zima Board 👉 amzn.to/3VhFdeK
Dual Edge TPU Adapter Low Profile 👉 www.makerfabs....
Discover the revolutionary Coral AI Dual Edge Accelerator, a game-changer in the realm of artificial intelligence and machine learning. This compact yet powerful device integrates not one, but two Tensor Processing Units (TPUs) into a single PCIe M.2 chip, offering unmatched efficiency and performance for your AI projects. Designed for enthusiasts and professionals alike, the Coral Dual Edge Accelerator shines with its cost-effectiveness and low power consumption, making it an investment that pays for itself over time.
Unlike traditional USB accelerators or solo PCIe solutions, this cutting-edge device doubles your machine learning capabilities without occupying extra space, thanks to its innovative use within the same M2 slot. Its exceptional performance is why Google incorporates it into their enterprise conferencing products, achieving lightning-fast local AI processing speeds. Surprisingly, this $25 Coral TPU can outperform a $2,000 CPU, making it a favorite among the Home Assistant community for powering Frigate's NVR surveillance. Capable of processing tensorflow lite models at an astonishing rate of 800 frames per second, it's an ideal solution for real-time image analysis for security and monitoring purposes, effectively serving as a more versatile and powerful alternative to commercial products like Amazon Ring.
Expanding your surveillance capabilities is straightforward with more TPUs allowing for additional camera connections. However, setting up this system comes with its challenges, particularly in navigating dual bus PCIe requirements and motherboard compatibility for optimal performance. But worry not, as this guide will walk you through the intricacies of making the most out of the Dual Edge TPU, from selecting the right adapters to configuring your motherboard for bifurcation, ensuring you unlock the full potential of this technology.
In this comprehensive walkthrough, you'll learn about the Coral AI Dual Edge Accelerator's setup, from installing it on compatible motherboards to the final steps of configuring your Frigate NVR surveillance system. Discover how to overcome common setup obstacles, choose the right adapters with proven track records, and streamline your installation with handy tips and resources. Whether you're upgrading your home lab or building a state-of-the-art surveillance system, the Coral AI Dual Edge Accelerator offers an unparalleled opportunity to enhance your AI capabilities with ease and efficiency.

Опубликовано:

 

5 сен 2024

Поделиться:

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист
Посмотреть позже
Комментарии : 33   
@marioalejandrox
@marioalejandrox 5 месяцев назад
Is there an adapter for the Coral Dual TPU to run on the Pi 5? Thanks for your videos, they are great!
@granatapfel6661
@granatapfel6661 5 месяцев назад
Yes it's possible search for it in the web, I guess it was a polish manufacturer, which offers different HAT's
@DataSlayerMedia
@DataSlayerMedia 3 месяца назад
Yes! Pine Boards has a new Dual TPU Hat 👉pineboards.io/products/hat-ai-dual-for-raspberry-pi-5 Tutorial 👉 ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-QmUJOFHr-No.html
@_____alyptic
@_____alyptic 4 месяца назад
Wait, did you show both TPU's working over a x1 link!? I'm hoping i can have it in a USB4 adapter/case 🤞🏽
@mihirmakwana9326
@mihirmakwana9326 5 месяцев назад
Love you brother. My GCP load balancer is working using terraform. Your gpc load balancer video was amazing. Thanks a lot
@andre-le-bone-aparte
@andre-le-bone-aparte 7 дней назад
Just found your channel - Excellent Content. Another sub for you sir!
@DataSlayerMedia
@DataSlayerMedia 5 месяцев назад
For 20% Off Brilliant.org Membership! 👉 brilliant.org/DataSlayer/
@ssstars2.0
@ssstars2.0 4 месяца назад
really want to use this concep with a 360 camera to stream car cruises and races!! if i can add the Lora v32, i could also track/communicate the cars during cruises or keep track of my own cars
@amdintelxsniperx
@amdintelxsniperx 5 месяцев назад
and now with chromes ai search engine soon being able to make use of the client side ai accelerators instead of just cloud these will get even better
@Base612
@Base612 5 месяцев назад
Does this only work with tensorflow?
@billlodhia5640
@billlodhia5640 5 месяцев назад
Coral only works with TFLite
@spencerking8373
@spencerking8373 28 дней назад
If you have multiple pcie slots and hats can you run models using 4 (or more) tpus? (With multiple coral dual tpu boards)
@Johan-rm6ec
@Johan-rm6ec 19 дней назад
Running Stable diffusion is a better showcase.
@Dvalin21
@Dvalin21 5 месяцев назад
Hey is the higher level of FPS for the cameras a new feature? Before anytime I would go higher than 10, it would break!
@rezcan
@rezcan 3 месяца назад
Appreciate the detailed video and blog post but I am having difficulty installing the TPU drivers (ls /dev/apex_* finds nothing) on CasaOS v0.4.9 with kernel version 5.10.0-10-amd64. Have you been able to set this up on this version? lspci -nn | grep 089a shows the correct TPUs, only the driver doesnt get installed
@galdakaMusic
@galdakaMusic 5 месяцев назад
How to plug in Rpi?
@granatapfel6661
@granatapfel6661 5 месяцев назад
look at the comment above.
@GROW_YT_VIEWS_137
@GROW_YT_VIEWS_137 5 месяцев назад
I love the creativity in your videos. Always entertaining!
@iblackfeathers
@iblackfeathers 5 месяцев назад
what cpu do you have on the zimablade in all your tests? j3455? they switched the cpu to n3450. do all these projects still work out of the box with the slower cpu?
@AndyTanguay
@AndyTanguay 5 месяцев назад
Hey, can I ask you what screen app utility you’re using? Super smooth.
@NunoRibeiro1975
@NunoRibeiro1975 5 месяцев назад
Screen studio
@giosterr44
@giosterr44 4 месяца назад
Thanks!🔥🔥🔥
@ahmetttt10
@ahmetttt10 5 месяцев назад
can u test advanced local LLM models ?
@HamguyBacon
@HamguyBacon 5 месяцев назад
it uses tensorflow lite...
@DonMacKenzie1
@DonMacKenzie1 3 месяца назад
Getting error that kernel 6.8.0 is not supported. Driver is not being installed. Is there a workaround for this error?
@Sydney268
@Sydney268 Месяц назад
Yeah, compile it from source - issue has been fixed on github
@unkillablespirits8582
@unkillablespirits8582 5 месяцев назад
link for the power monitor thing or should i run an Ai on a still frame of the vid to find out lol ?
@DataSlayerMedia
@DataSlayerMedia 5 месяцев назад
amzn.to/3PDmkzf
@ronaldwhite8227
@ronaldwhite8227 4 месяца назад
Promo sm ✔️
@zxcaaq
@zxcaaq 3 месяца назад
docker is horrible
@AdrianoKrauthein-yu5gx
@AdrianoKrauthein-yu5gx 4 месяца назад
U joking when you measure the power consumption from the power supply?! Sure I don't believe you did that.
@sIXXIsDesigns
@sIXXIsDesigns 4 месяца назад
seen enough... you're pushing insane claims on everything and you're not even being realistic about the claims.... complete and utter BS.
@PablloArruda
@PablloArruda 2 месяца назад
not working, sucks tutorial
Далее
I Built the AI Security Camera I Always Wanted
16:22
Просмотров 17 тыс.
Google TPU & other in-house AI Chips
12:06
Просмотров 31 тыс.
when you have plan B 😂
00:11
Просмотров 3,8 млн
ПРОСТИ МЕНЯ, АСХАБ ТАМАЕВ
32:44
Просмотров 1,2 млн
It’s Been a Good Run, Phone Providers.
26:31
Просмотров 4,5 млн
What else is an M.2 WiFi slot good for?
10:24
Просмотров 1,2 млн
Run your own AI (but private)
22:13
Просмотров 1,4 млн
The Truth About SIM Card Cloning
13:04
Просмотров 1,1 млн
This is NOT a Graphics Card - ASUS AI Accelerator
15:24
Don't Overlook This Slot
15:54
Просмотров 575 тыс.
host ALL your AI locally
24:20
Просмотров 1 млн
when you have plan B 😂
00:11
Просмотров 3,8 млн