This is INCREDIBLE work
Cant wait to checkout the RTL and ISA
Introducing Frostbite: a new class of CPU built around interaction nets.
Given only interaction net theory, our agents co-designed the ISA and chip architecture from scratch, running large swarms to design, verify, and optimize the CPU down to the final GDS.
On our first set of target workloads, Frostbite is 4.6x more performant per watt than modern traditional CPUs at similar area.
10
410
Shashvat retweeted
Drones are hard to build and control with their hardware debugging weight constraints and shipment delays. But once you cross that the rewards are so very sweet.
Here's a blog about my journey with drones from being confused to being confused with obsession.
Cerebras is really pushing the boundaries of what's possible with a single silicon wafer.
Can't wait to see what they do next.
Fearless Engineering: The CS-4 is our first engineering documentary.
Through lab footage and conversations with the engineers behind each breakthrough, the film follows the creation of the CS-4, our latest wafer-scale AI system.
0:00 introducing the engineering team
1:09 Wafer-scale history
2:28 Powering the wafer
4:00 The cooling problem
4:18 Designing the cold plate
5:11 Holes in the wafer
6:33 the alignment problem
7:47 The CS-4 PRD
8:35 Designing for deployability and serviceability
11:01 Removing the motherboard
11:25 Removing internal cooling loops
11:41 inventing the power rack
12:35 A one-person lab deployment
14:03 Backpack: 50% fewer components
14:55 Connecting the compute backpack to cooling manifold
17:58 shifting Cerebras’s focus from training to inference
19:09 Building the Cerebras Cloud
20:01 CS-3 vs. CS-4 performance
20:49 3D Stacked DRAM
8
578
Huawei's tau scaling is picking up speed
Yesss @austinsemis it's tau scaling!
> Traditional EDA uses pseudo 3D engines which break down a 3D design into multiple 2D tiers (partitioning)
>After partitioning - each 2D design is forced to be optimised separately
> This leaves a lot of additional optimisation in terms of time and power on the table
> This limitation arises from decades of specialisation in 2D algorithms that operate in only (X ,Y) ordinates
>@archgen_ captures all this additional optimisation through 3D native place and route in (X, Y and Z) which can 10x the benefits from tau scaling
>However 3D designs can also get insanely hot as heat gets trapped between tiers
> Our foundational AI model for thermal simulations makes sure the place and route are thermal aware.
> This gets rid of thermal hotspots that can cause chips to fail in the real world
> Our foundational models are physics based and not transformers
1
6
338
I never knew Morris Chang was such a gangster
Computer Chips Are the Most Magnificent Things Humanity Has Ever Made
newyorker.com/magazine/2026/…
1
3
202
Shashvat retweeted
recently ive been looking back at my past couple of years or so and thought ill write a blog summarizing what ive done so far and what i want to do
5iri.me/blog/building-comput…
simulators become so interesting as complexity keeps on growing
Replying to @archgen_
@archgen_ is building skyscrapers for semiconductor chips
Shrinking transistors is getting harder. Stacking them makes heat harder to manage. We’re building the tools to design chips in three dimensions.
We stack transistors vertically, shorten connections, and optimize timing and heat together.
Our foundational model for thermal simulation, predicts temperature roughly 5,000× faster than conventional EDA tools.
Our placement engine ranked #1 in HRT’s macro-placement challenge
We’re working with some of the largest fabs in the world to tape out our first chip in April 2027
archgenlabs.com
10
284

















