Internal Documents Detail Amazon’s Titus Data Center Modernization Program
Business Insider reports on internal AWS planning documents detailing an Amazon initiative codenamed "Titus" — a multi-year data center modernization program covering power architecture, liquid cooling, server layouts, and deployment timelines, intended to support the next generation of AI workloads and broader customer demand, according to the documents reviewed by the outlet. As clarified by AWS Global Communications, the broader initiative was publicly announced in December 2024, and Titus is a much broader effort than GPU-centric framing alone suggests. The publicly disclosed components of the program include the In-Row Heat Exchanger (IRHX), AWS's in-house direct-to-chip liquid cooling system unveiled in July 2025 and co-engineered with NVIDIA. Business Insider also reports Amazon is planning a record $200 billion in capital expenditures this year, and that AWS continues to promote its in-house AI accelerators alongside third-party GPUs.
UPDATE — May 18, 2026
An AWS spokesperson contacted Let's Data Science with a clarification on this article. According to AWS, the underlying initiative reported here was first announced publicly in December 2024, and "Titus" — the internal codename surfaced through Business Insider's reporting on internal documents — is a much broader initiative than the GPU-cooling framing alone. The headline, summary, and body have been updated to reflect that the program is an established, multi-year data center modernization effort (not a new announcement) and that its scope extends well beyond next-generation GPU systems. AWS provided the following statement on the record:
*"We're always innovating to provide customers the fastest, most resilient, most secure, and most sustainable infrastructure, and our scale lets us do that while keeping costs down. The initiative reported here, which was announced in December 2024, reflects how we're designing our data centers to support the next wave of AI workloads and the growing demands our customers are bringing to us."* — AWS spokesperson
Source: emailed correspondence from Patrick Neighorn, AWS Global Communications, May 18, 2026.
What happened
Business Insider reports on internal AWS planning documents detailing an Amazon initiative codenamed "Titus" — a multi-year data center modernization program covering power architecture, advanced liquid cooling, server and rack layouts, and accelerated deployment timelines, designed to support the next generation of AI workloads and broader customer demand. As clarified by an AWS spokesperson, the broader initiative was publicly announced in December 2024, with Business Insider's reporting drawing on internal documents that describe the codename and detailed implementation.
Scope of the program
Per Business Insider's reporting on the internal documents, Titus reworks generational data center design to accommodate larger GPU racks, denser power and cooling requirements, and faster build cycles. An internal slide quoted by Business Insider states, *"The objective of the Titus portfolio of program variants is to deliver the next AWS generational Data Center design."* The publicly disclosed components of the broader program include the In-Row Heat Exchanger (IRHX), AWS's in-house direct-to-chip rack-level liquid cooling system unveiled in July 2025, which AWS co-engineered with NVIDIA to support its highest-density AI training and inference instances built around NVIDIA Blackwell-class GPUs. AWS has described the broader effort as part of its longer-running data center innovation roadmap covering speed, resilience, security, and sustainability rather than a single product or AI-only initiative.
Capital context
Business Insider also reports Amazon is planning a record $200 billion in capital expenditures this year, alongside continued promotion of AWS's in-house AI accelerators as alternatives to third-party GPUs.
AWS response
An AWS spokesperson provided the following on-the-record statement: *"We're always innovating to provide customers the fastest, most resilient, most secure, and most sustainable infrastructure, and our scale lets us do that while keeping costs down. The initiative reported here, which was announced in December 2024, reflects how we're designing our data centers to support the next wave of AI workloads and the growing demands our customers are bringing to us."*
Editorial analysis — technical context
For practitioners, multi-year data center modernization programs at this scale matter because they signal where the broader infrastructure stack is heading. Companies scaling for modern AI accelerators typically focus on three engineering fronts: increasing facility power delivery and distribution, integrating higher-capacity cooling (often direct-to-chip liquid), and changing rack-level electrical and mechanical designs to handle denser compute units. These patterns align with the changes Business Insider describes in the Titus documents, and with the publicly disclosed IRHX component. The shift toward proprietary, hyperscaler-built cooling infrastructure also has implications for cooling-vendor positioning and for the pace at which density-driven AI workloads can be deployed at scale.
Industry context
Observed patterns in similar transitions: hyperscalers that redesign facilities for high-density GPUs face higher upfront capital and operational complexity, including stricter electrical engineering, thermal-validation testing, and adherence to emissions and noise constraints. Reporting that AWS is both investing heavily in capex and documenting an internal program to adapt facilities fits an industry-wide trend where cloud providers must reconcile hardware vendor roadmaps with in-house silicon efforts.
What to watch
For practitioners and infrastructure planners, relevant indicators include published AWS instance specifications that require liquid cooling or higher power-per-rack envelopes, changes to availability-zone upgrade timelines, and third-party equipment vendor announcements about compatible rack and cooling solutions. Observers should also track official AWS technical guidance and product SKU changes that reflect the new physical constraints described in the Titus documents.
Further reading from AWS
Scoring Rationale
This is a notable infrastructure story: it signals large-scale physical upgrades by a major cloud provider to support advanced GPUs, which matters to practitioners planning deployments, procurement, and system design. The timeliness and Amazon's sizable capex raise the story's importance.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
