Graphcore vs NVIDIA
Side-by-side comparison of AI visibility scores, market position, and capabilities
Graphcore makes Intelligence Processing Units (IPUs) — AI chips purpose-built for graph-based ML workloads — acquired by SoftBank in 2024 for ~$500M and reinvesting $450M+ toward AGI-era silicon.
About
Graphcore is a British semiconductor company that designs Intelligence Processing Units (IPUs), a class of AI accelerator chips architected specifically for the graph-based computation patterns underlying machine learning workloads. Founded in 2016 in Bristol, UK, by Nigel Toon and Simon Knowles, Graphcore raised hundreds of millions of dollars from investors including Sequoia, Microsoft, and BMW before being acquired by SoftBank Group in July 2024 for approximately $500 million.
H200/GB200/Blackwell GPU family powering 90%+ of AI training workloads; $130B+ quarterly revenue run-rate; $3T+ market cap; 85% of revenue from AI compute. Every major AI company — OpenAI, Anthropic, Google, Meta, xAI — runs on NVIDIA hardware.
About
NVIDIA Corporation is a Santa Clara, California-based semiconductor and AI computing company — publicly traded on the NASDAQ (NASDAQ: NVDA) as an S&P 500 Information Technology component and member of the Dow Jones Industrial Average — designing and supplying graphics processing units (GPUs), AI accelerators, networking infrastructure, and computing platforms for data center AI training and inference, gaming, professional visualization, and automotive applications through approximately 36,000 employees worldwide. In fiscal year 2025 (ending January 2025), NVIDIA reported revenues of $130.5 billion (+114% year-over-year) — driven by unprecedented demand for H100 and H200 AI GPU clusters from hyperscale cloud providers (Microsoft Azure, Amazon Web Services, Google Cloud), AI-native companies (OpenAI, Anthropic, xAI, Cohere), and enterprise AI deployments — making NVIDIA the fastest-growing large-cap company in recorded history and the third-most-valuable company globally (market capitalization exceeding $3 trillion in 2024-2025). CEO Jensen Huang has led NVIDIA's transformation from a gaming GPU company into the foundational infrastructure provider for the artificial intelligence economy: NVIDIA's CUDA (Compute Unified Device Architecture) software platform — developed since 2006 — has accumulated 4+ million developers, 4,000+ GPU-accelerated applications, and a decade of AI research papers, libraries, and frameworks (PyTorch, TensorFlow, cuDNN) optimized for NVIDIA hardware, creating the most powerful software moat in technology. The Blackwell GPU architecture (B100, B200, GB200 — launched 2024, ramping production in 2025) delivers 5x training performance improvement over the H100, sustaining NVIDIA's generational performance advantage that justifies continued AI capital expenditure at $300-500 billion annual industry pace.
Key Details
Capabilities & Ecosystem
Capabilities
Integrations
Also Compare
Track AI Visibility in Real Time
Monitor how your brand performs across ChatGPT, Gemini, Perplexity, Claude, and Grok daily.