Cheaper tokens excite me less when the work stays stuck. NVIDIA announced Rubin in January 2026, promising substantially lower inference costs than Blackwell, with partner products expected in the second half of the year. Good. I want that improvement to reach a task I can actually finish.
The frustration I know best is closer to home. I have had agents running on my own projects in Claude Code stop mid-task because the usage limit ran out, with Cursor and Codex available too. The choice was to wait for the reset or move the task to another tool.
That episode tells us nothing about Rubin's performance, but it explains why I distinguish purchased capacity from capacity I can use. A hardware announcement describes part of the bill; the service determines how that capacity reaches users. My own organization still comes after that.
The first thing I want when an agent stops is the reason. Did its allowance run out? Is accessible information missing? Does another task need to finish? Each answer changes the next step. Calling everything slowness makes complaining easy and leaves the queue untouched.
Several tools give me options, but moving work requires context. A task tied to a particular integration may take more effort to move than to resume later. I want to examine the prepared task, work out what it needs, and then choose where to execute it.
I also need to watch the urge to fill every spare bit of capacity. Requesting more investigations is easy; reading the results still takes time. If a price reduction produces only a larger pile of answers, I have created another queue myself. Then I have to ask how much of that material helps with a decision that was pending.
Better hardware deserves attention, including the conditions behind NVIDIA's comparison. To judge its effect on my work, the closer measure is how long a prepared task sits idle and how much effort restarting it takes.