AWS CEO Matt Garman on AI Roadmap, Customer Innovation, and Cloud Expansion
Bloomberg PodcastsMay 30, 202515 min244 views
32 connectionsΒ·40 entities in this videoβAWS CEO's First Year and Key Achievements
- π Matt Garman reflects on his first year as AWS CEO, highlighting customer innovation and adoption of new technologies as the biggest achievement.
- π‘ He notes a significant explosion of AI and Agentic technologies, with more customers migrating entire estates to AWS.
Generative AI Revenue and Impact
- π° AWS has achieved a multi-billion dollar run rate for its AI business, a mix of customers running their own models and using Amazon Bedrock with first-party and third-party models.
- π― Garman emphasizes that AI is in its earliest stages and will fundamentally transform every business, industry, and job.
- π While Amazon uses AI internally for optimization, customer-facing applications, and Alexa, AWS customers are also leveraging AI across various services, from contact centers to custom chip development.
Training vs. Inference Workloads
- π Initially, AI workloads were dominated by training, but inference usage is now exploding and is expected to constitute the vast majority (80-90%) of future usage.
- π§© Inference is described as a core building block for applications, similar to compute or storage, making it difficult to isolate AI-driven revenue.
- β Present-day usage indicates that inference already exceeds training in terms of volume.
Project Rineer and Custom Silicon
- π€ Project Rineer is a collaboration with Anthropic to build the largest compute cluster for training their next-generation Claude models on AWS's custom Tranium 2 accelerators.
- β‘ Tranium 2 servers are landing and in operation, showing impressive performance and cost-effectiveness, crucial for bringing down the high costs of AI.
- π‘ Innovation in silicon, software, and algorithms is key to reducing AI expenses and enabling broader adoption.
Market Position and Customer Choice
- π AWS sees massive opportunity in the AI space, with room for both its Tranium 2 and NVIDIA's technologies, emphasizing customer choice.
- π οΈ AWS is a design partner with NVIDIA, ensuring the latest technology is available, while also promoting its own accelerators like Tranium 2.
- π AWS is committed to being the best place for every workload, including supporting models like Anthropic's Claude on AWS, while acknowledging their availability elsewhere.
Global Cloud Expansion
- π AWS is aggressively expanding data center capacity globally, with significant growth in Latin America (new regions in Mexico and Chile, existing region in Brazil).
- πͺπΊ In Europe, AWS is launching the European sovereign cloud by the end of the year, designed for critical EU-focused sovereign workloads, addressing data sovereignty concerns.
Knowledge graph40 entities Β· 32 connections
How they connect
An interactive map of every person, idea, and reference from this conversation. Hover to trace connections, click to explore.
Hover Β· drag to explore
40 entities
Chapters8 moments
Key Moments
Transcript58 segments
Full Transcript
Topics16 themes
Whatβs Discussed
AWSArtificial IntelligenceGenerative AIAI RoadmapCustomer InnovationCloud MigrationAmazon BedrockAmazon QTraining WorkloadsInference WorkloadsProject RineerAnthropicTranium 2NVIDIAData Center ExpansionEuropean Sovereign Cloud
Smart Objects40 Β· 32 links
CompaniesΒ· 4
ProductsΒ· 13
ConceptsΒ· 18
PeopleΒ· 3
LocationΒ· 1
EventΒ· 1