The Alphawave Semi AI Platform

Pathway 1:
64G UCle, 224G SerDes and 800G/1.6T UALink and UEC Controllers

Scaling connectivity efficiently and effectively to support the growth of AI clusters requires high-speed, low-latency interconnects to handle the massive data throughout required by AI applications.

Alphawave Semi has customized silicon solutions optimized for diverse workloads across 64G UCle, 224G SerDes, 800G/1.6T UALink and 800G/1.6T UEC controllers.

Pathway 2:
1.6T and 3.2T I/O Multi-Standard SerDes Chiplets for UALink and UEC Controllers

Reducing time-to-market with readily available 1.6T and 3.2T I/O Chiplets in addition to customizable silicon solutions for compute-to-compute, memory, or I/O.

These include die-to-die (UCle) and off-chip-I/O multi-standard SerDes for pairing with connectivity subsystems to solve scale up (UALink) and scale out (UEC).

Server Room with Rows of Servers and Blue Lights - Realistic Photo

Scale-Up and Scale-Out Networks

Create hyper-optimized scale-up networks that link over 1000 GPUs acting as one. Connect and orchestrate these over dedicated scale-out network channel for low latency, low power, and low collisions

64G UCle

For die-to-die connectivity - including custom HBM

224G SerDes

With the industry's highest shoreline density and power efficiency

800G/1.6T UALink

For ultra-efficient, -reliable, -efficient connections

800G/1.6T UEC

For ultra-fast server-to-server connections

"Alphawave Semi's foundational IP for AI platforms is helping advance the high-speed connectivity essential for next-generation AI infrastructure."

Kurtis Bowman - UALink

"The AI platform is a strong example of how advanced process technology and packaging can come together to enable the next wave of AI and data center innovation."

Lipen Yuan - TSMC