Anthropic has disclosed that it has committed to spend close to $900 billion on AI infrastructure in the coming years, and that its spending on computing power is growing faster than its revenue.
Anthropic · AI infrastructure
Anthropic commits close to $900B to AI infrastructure as compute costs outpace revenue
The AI developer has disclosed a very large capital commitment to chips, data centers and computing capacity. Its compute spending is growing faster than its revenue, raising the question of how the gap will be closed.
~$900B
Committed to AI infrastructure in the coming years
Faster
Compute spending is growing faster than revenue
The scale of the commitment
Each block represents $100B, so nine blocks equal the roughly $900B figure. The blocks show scale only, not how the money is split.
$100B
$100B
$100B
$100B
$100B
$100B
$100B
$100B
$100B
~$900B
Total committed over the coming years, shown as nine $100B blocks. Timing and allocation across chips, data centers and cloud capacity were not specified.
How the commitment depends on future revenue
Step 1
Compute commitments
Multiyear spending on hardware and capacity
→
Step 2
Revenue projections
Income must grow to support the spending, and may take years
→
Step 3
Margins
Whether usage turns into margins that keep pace with outlays
The open questions
The disclosure does not say how the gap between spending and revenue will be closed.
Timeline: the exact schedule of the commitments is not specified. Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article...