NVIDIA Launches SUSE AI Factory; Cerebras and Google Chip Competition Heats Up
NVIDIA partnered with SUSE to launch SUSE AI Factory, a unified AI stack offering pre-validated blueprints, zero-trust security and simplified workflows, with general availability slated for late 2026. Rivals including Cerebras (2025 revenue $510M, 76% growth) and Google’s upcoming inference chips are intensifying chip market competition.
1. SUSE AI Factory Partnership
SUSE announced SUSE AI Factory with NVIDIA, a unified software stack designed to help enterprises build, deploy and scale AI applications securely across data centers and edge environments. The solution integrates SUSE’s open-source infrastructure with NVIDIA’s AI technology, offering pre-validated blueprints, zero-trust security and simplified deployment workflows; a preview at SUSECON 2026 precedes general availability later in 2026.
2. Cerebras IPO Intensifies Competition
Cerebras Systems filed for a Nasdaq IPO under ticker CBRS after generating $510 million in 2025 revenue, a 76% year-over-year gain, and securing major customers including OpenAI, Amazon and Meta. With a recent $20 billion investment commitment from OpenAI, Cerebras is positioning its wafer-scale AI computing chips as a direct challenge to NVIDIA.
3. Google Inference Chips Challenge
Google is preparing to introduce new AI inference chips dedicated to running trained models, aiming to expand on its December release of Trainium3 and further challenge NVIDIA’s dominance in the fast-growing AI semiconductor category. Leading AI developers are already adopting Google’s chips, signaling heightened competition in inference workloads.
4. Anthropic’s AWS Trainium Agreement
Amazon invested an additional $5 billion in Anthropic, raising its total to $13 billion, while Anthropic committed to spend over $100 billion on AWS cloud services through 2036. The deal secures up to 5 GW of computing capacity for Trainium2 through Trainium4 chips, positioning Amazon’s custom accelerators against NVIDIA GPUs in AI training and inference.