AI News · Hot news ·
Amazon raises EC2 Capacity Blocks for ML prices by roughly 15%
Amazon announced a roughly 15% price increase for its EC2 Capacity Blocks for ML service, according to MarketWatch reporting cited by Quartz. The service lets customers reserve Nvidia chip capacity, covering hardware from the A100 to the B300. Updated rates are scheduled to take effect the following week.
Key points
- Amazon announced a roughly 15% price increase for EC2 Capacity Blocks for ML.
- The service lets customers reserve Nvidia chip capacity, ranging from A100 to B300 hardware.
- Updated rates are scheduled to take effect the following week.
- Exact chip-specific rates and treatment of existing reservations were not reported.
What happened: Amazon has announced a roughly 15% price increase for EC2 Capacity Blocks for ML, according to MarketWatch reporting cited by Quartz. The service lets customers reserve access to Nvidia chips for machine learning workloads. Updated rates are scheduled to take effect the following week, making the change relevant to teams preparing to reserve capacity or approve spending for new workloads.
The details: The service covers Nvidia hardware ranging from the older A100 to the newer B300. The reported increase concerns EC2 Capacity Blocks for ML, rather than a stated change across all Amazon computing services. Exact updated rates for individual chip types were not reported, nor was a chip-by-chip breakdown of the roughly 15% increase. That leaves customers without enough detail in the report to calculate a precise revised budget for a particular reservation.
Who it affects: The announcement is most directly relevant to businesses using this service to secure Nvidia chip capacity. Teams planning reservations should refresh their cost estimates and compare alternatives before committing to new workloads. The report did not say how the change would apply to reservations already made, so existing customers should not assume that their current commitments will be repriced.
What to watch: The immediate questions are the updated rate for the hardware a team needs and which reservations will be subject to it. Although the change was described as taking effect the following week, a calendar date was not reported. Amazon’s reason for raising prices was also not reported, leaving the cause of the increase unclear.
Our take
Teams reserving GPU capacity should refresh cost estimates and compare alternatives before committing to new workloads.