FACT ALERT 🚨 : In modern agentic coding, 42% of the time is spent on CPU doing tool use such as editing files, running Bash scripts, running lints, etc. The economy of traditional cloud computing charges at $ per cpu core. In the economy of agents, the business model is $ per token thus to increase token revenue, you need to increase the amount of CPUs power u have so that you can generate your tokens.
Goldman Sachs on AI power demands:
Outlook
"We expect US data center power demand to increase from 31 GW in 2025 to 41/66 GW in 2026/27 assuming a capacity utilization rate of 70%. By 2027, we expect data centers to account for 8.5% of total US power demand."
Their Trade
"We continue to recommend hedging against both upside power price risks from surging data center power demand and downside power price risks where regional supply growth may temporarily outpace demand growth."
TokenSpeed is a brand new inference engine purpose built for speed-of-light agentic workloads.
Read their blog to learn more about its advanced KV cache management, safe and efficient scheduler, and pluggable layered kernel system designed for multi-silicon support. Plus, it also has the fastest MLA attention kernel on NVIDIA Blackwell.
Congrats to @lightseekorg on the launch!
In the early stages, ODM server assembly mainly focused on manufacturing. ODM produced standardized racks, motherboards, and server systems on a large scale. Their primary advantages were cost efficiency, capacity, and yield.
In the AI era, IT racks have become much more complex. GPU/ ASIC, high-power systems, liquid cooling, high-speed connections, and rack management all need to work together within the rack. To simplify cabling and maintenance, cableless designs may also become more common.
As a result, ODM are no longer just manufacturers. They are evolving into partners in design, integration, and mass production. Moving forward, they will support various GPU / ASIC platforms and data center designs, and help vendors build the broader AI infrastructure ecosystem.
Crypto card spending is the next big trend:
Crypto card spending volume has surged +500% since September 2024, now running at $600 million per month.
As a result, stablecoin-linked payment cards are now one of the fastest growing businesses on the blockchain, with 90% of transactions captured by Visa, $V.
Visa's strategy has centered around partnering with emerging infrastructure providers which reduces reliance on traditional sponsor banks.
The growth comes amid the launch of Jupiter Global which returns 4-10% cash back to crypto cards and has seen +660% MoM volume growth in April.
Crypto card adoption is growing.
The Internal Pricing mechanism will be updated to use an EMA time constant of 30 minutes. This change will go into effect today at 4PM ET and will apply to all XYZ markets.
One can expect the oracle to be more responsive, and for the funding to normalize more quickly during internal pricing sessions.
To learn more about this change and how the internal pricing mechanism supports 24/7 price discovery, refer to our documentation here: https://t.co/LDQBbH9vIf