Block’s Dessa explores conversation without starting from a text model
Speech-only training and simultaneous listening and speaking are the research claims. Selected demos, laptop inference and a 72-GPU training rack need different interpretations.
Topic archive
On-device AI
Follow8 stories
NVIDIA’s Nemotron Bolt puts short answers and tool calls on a tight memory budget
The local model targets interactive assistants. Its 4GB positioning does not describe the memory needs or privacy of an entire voice system.
LattePanda Sigma 2 makes local AI a question of memory, ports and integration
The compact board’s 180 TOPS headline needs software context. Its expansion links are the more practical starting point.
Ubuntu 26.10 arrives with local dictation and a nine-month support window
Myna begins as an optional preview; the release also changes foundational software and documents a limitation for encrypted-system upgrades.
RTX Spark reaches its scheduled launch day with software at the center of the pitch
Nvidia’s October 16 laptop launch puts local agents alongside games and creative apps, but some advertised software improvements are still forthcoming.
Nvidia’s new Nemotron models split the agent race between servers and local PCs
Super VL and Bolt target different deployment sizes, making memory requirements and task reliability central to the comparison.
gFiber’s Wi-Fi 8 gateway has three promises. Only one is a speed number
The planned gateway combines broadband, wireless reliability and local AI controls. Certification and territory still matter.
M5 Pro and M5 Max put local AI in the professional laptop pitch
Apple's March 2026 announcement linked new silicon and faster storage to on-device workloads, with vendor claims still requiring workload-specific testing.