Let's Build a Chip - We lay out the costs of building a chip - with spreadsheets!
Research and publish the best content.
Get Started for FREE
Sign up with Facebook Sign up with X
I don't have a Facebook or a X account
Already have an account: Login
WHY IT MATTERS: Digital Transformation
214.7K views |
+0 today
Get weekly or monthly digest of all posts in your inbox: https://fmcs.digital/wim-subscribe
Curated by
Farid Mheir
 Your new post is loading...
Scoop.it!
McDonald’s Drive-Thrus are a baseline for the entire industry and seeing them install LCDs that basically blow up and fail totally is something everyone wants to understand.
Farid Mheir's insight:
WHY IT MATTERS: found this example of displays that fail at McDonalds drive-thrus. It highlights the fact that digital transformation often means deploying new hardware in places where there not in the past. And that creates a huge number of issues in terms of design, maintenance, operations, etc. Very often not a big deal is planned but very much a big deal if not accounted for in the beginning. So think about cooling, electricity availability, repairs, software version updates, security, content upload, etc. 
Scoop.it!
What is wafer-scale integration? A wafer is a circle, today about 12 inches in diameter. A chip is typically a square no more than an inch on a side — much smaller than a wafer. Many copies of the same chip are printed onto the wafer. They are arranged into a grid with spaces called scribe lines between them. The wafer is then cut along those scribe lines into individual chips. The chips are tested, and after any defective ones are discarded, the rest can be packaged and sold. Wafer-scale integration is the idea that you make a single chip out of the whole wafer. You skip the step above concerning cutting the wafer up — with one chip there is nothing to cut.
Farid Mheir's insight:
WHY IT MATTERS: this new technology enables new possibilities for machine learning applications such as autonomous driving. Tesla has recently announced that every car since 2019 contains a special chip that enables autonomous driving. This wafer-scale technology takes the same concept to another level.
Scoop.it!
The WSE (pronounced “wise”) is the largest commercial chip ever manufactured, built to solve the problem of deep learning compute. The WSE is 1.2 trillion transistors, packed onto a single 215mm x 215mm chip with 400,000 AI-optimized cores, connected by a 100Pbit/s interconnect. The cores are fed by 18 GB of super-fast, on-chip memory, with an unprecedented 9 PB/s of memory bandwidth.
Farid Mheir's insight:
WHY IT MATTERS: this article and accompanying PDF presentation dive into the technical details of how it is now possible to put 1 trillion transistor on a chip designed for the sole purpose of machine learning. This is a BIG DEAL. Slides and word document are here:
Scoop.it!
Superlatives abound at Cerebras, the until-today stealthy next-generation silicon chip company looking to make training a deep learning model as quick as buying toothpaste from Amazon. Launching after almost three years of quiet development, Cerebras introduced its new chip today — and it is a doozy.
Farid Mheir's insight:
WHY IT MATTERS: this is a BIG deal in hardware design. Being able to fabricate such a large system enables capabilities to a scale previously unheard of. Read this article for a high level overview, watch the next blog post for an in depth description.
Scoop.it!
Microchips - are indeed can be considered a black box - as long as it's working you normally don't look inside. But what if you want to?
Farid Mheir's insight:
For those of you that wonder what lies underneath those black squares that power your computers and cell phones, this short paper will show you.
WHY THIS IS IMPORTANT Pretty much everything you own these days is powered by a thing like that. You need to know what's inside, if only just to know.
Scoop.it!
Think about this scenario: You see someone at a party you like; his social profile is immediately projected onto your retina—great, a 92% match. By staring at him for two seconds, you trigger a pairing protocol. He knows you want to pair, because you are now glowing slightly red in his retina screen. Then you slide your tongue over your left incisor and press gently. This makes his left incisor tingle slightly. He responds by touching it. The pairing protocol is completed.
Farid Mheir's insight:
A very good review of the human computer interfaces and what the future may hold for us.
WHY THIS IS IMPORTANT We cannot hold a mouse and keyboard in our hands to interact with physical world objects that have been digitized. Think of a warehouse employee that needs to interact with an automated shelves to fetch a specific product. Today we would use buttons on a mobile phone touchscreen. But would it not be simpler if we could simply interact with the shelves itself in ways that our body would recognize, then send appropriate commands to the devices we interact with?
In the article, the far fetched story about the couple in a bar may not be very realistic but it provides some very useful insights about how to leverage our bodies to do things we cannot with our devices today, for privacy reasons for example. Very interesting.
Scoop.it!
Now anyone can make cool hardware and, thanks to Arduino, it is easier than ever to connect your devices to the Internet and take in data from the outside world. The ESLOV IoT Invention Kit is an official Arduino product that adds Internet of Things capabilities to your hardware products. Trying to build a connected fridge to beat Samsung? Go for it with the ESLOV. Want to knock Sony off its perch? Try popping an ESLOV unit onto your Arduino board and take on major manufacturers from your living room.
Farid Mheir's insight:
Describes a new set of hardware components to build IoT solutions.
WHY THIS IS IMPORTANT It is with hardware devices like these that the penetration of digital in our daily lives will grow. In recent years, we've seen Arduino, Raspberry Pi and other hardware toolkits that simplify the prototyping and innovation to move the web, sensors, cameras, and other digital devices that bridge the real world with virtual one.
|
Curated by Farid Mheir
Get every post weekly in your inbox by registering here: http://fmcs.digital/newsletter-signup/
|
WHY IT MATTERS: my master and first jobs where building ASICs in the 1980s-1990s. So articles like these resonate as they show with numbers the very high cost of creating your own silicon. And if I understand it, 2B$ is the cost of silicon *production*, not the cost of building the foundry - which I assume is an order of magnitude or more expensive. But when companies like Apple sell billions of phones or Tesla sells millions of cars, the economics of build-vs-buy can tip in the favor buidling. Beware: the article is back of the envelope / educated guesses. Nevertheless fascinating!