Synopsys Amazon AI Deal
Synopsys Secures $1 Billion Amazon Chip Deal Alongside New AI Agent Launch
Anonymous
Key Points
1
Unlike command-by-command AI assistants, AgentEngineer can receive an engineering goal, break it into tasks, devise a plan and use Synopsys tools to execute the work. Its applications include verification coverage closure, optimizing performance, power and area in physical design, and manufacturing tasks such as mask synthesis.
2
Synopsys says Amazon will be the lead customer for application-optimized silicon IP, reflecting a shift toward IP tailored to particular workloads and system designs. The company says its IP model is evolving to combine licenses with royalties as production volumes grow.
3
The Amazon collaboration builds on more than 15 years of work together and covers Amazon’s Nitro, Graviton and Trainium chip portfolio. The companies also plan to advance multiphysics solutions on Trainium and Graviton and explore custom AI agents for Amazon engineering teams.
4
Amazon said its chip business had reached a $25 billion annual revenue run rate, with triple-digit growth; the figure includes product lines such as Trainium AI accelerators and Graviton processors.
Synopsys signed a multiyear deal worth more than $1 billion with Amazon to power the tech giant's custom chip strategy, marking a major shift in how AI-driven semiconductor design works Yahoo Finance. The agreement expands Amazon's use of Synopsys design software, simulation tools, and specialized chip building blocks—called silicon IP—for Amazon's Nitro, Graviton, and Trainium processor families.
The partnership reflects Synopsys' push into autonomous chip engineering through its new AgentEngineer platform, which uses AI agents to plan and execute complex design tasks without human direction at each step BigGo Finance. Unlike older AI assistants, these agents receive a goal, break it into steps, devise a plan, and run the engineering work themselves across design, testing, and manufacturing.
AI Agents That Work Without Constant Commands
AgentEngineer marks a sharp break from traditional AI design tools. Instead of responding to one command at a time, the new agents receive an engineering goal—like closing verification coverage or optimizing chip power use—and then independently break it into tasks, create a plan, and execute it Yahoo Finance. Early trials with partners like Intel showed 10% to 30% productivity gains in logic design and verification workflows.
The platform tackles verification coverage closure, performance and power optimization, and manufacturing tasks like mask synthesis BigGo Finance. Synopsys is working with more than 50 customers ahead of a planned year-end launch, partnering with companies ranging from Samsung and Nvidia to Fujitsu and MediaTek.
Amazon's $25 Billion Custom Chip Business Powers the Deal
Amazon's chip division has exploded into a major revenue driver. The company said its custom silicon business reached a $25 billion annual revenue run rate with triple-digit growth Streamline Feed. This includes Trainium AI accelerators for training machine learning models and Graviton processors for general computing in data centers.
The Synopsys deal builds on over 15 years of collaboration between the companies Yahoo Finance. Amazon will serve as the lead customer for Synopsys' new application-optimized silicon IP—building blocks tailored to specific workloads rather than generic designs. The two companies also plan to develop custom AI agents specifically for Amazon's engineering teams.
Synopsys Shifts to Royalty-Based Revenue Model
The Amazon deal signals a fundamental change in how Synopsys makes money. Instead of one-time upfront licensing fees, the company is moving toward a model combining licenses with royalties tied to how many chips get produced Yahoo Finance. This mirrors the high-margin recurring revenue model used by chip designer Arm Holdings.
CEO Sassine Ghazi called Amazon the "lead customer for this next phase of growth" as Synopsys positions itself at the center of AI-driven chip creation Yahoo Finance. The company also announced partnerships with OpenAI, signaling its bet that autonomous engineering will reshape how semiconductors get designed as AI demand explodes.
Sentiment
Mostly PositiveNEGATIVENEUTRALPOSITIVE
Publishers
25
Articles
140
Reach
165
A
AnonymousAuthor
