Smarter cities and spaces
Real time understanding makes cities safer and more resilient
Подробнее
Huge potential
AI is limited
Innovation is hampered by Today’s digital AI processors are extremely expensive to develop and based on traditional computer architectures, which limits innovation to only the largest technology companies.
End-device interface is a challenge End-device interfaces require low latency, low power and must be economical and compact. Digital processors simply cannot meet these complex needs of peripheral AI.
Analog computing is ideal for peripheral AI Frontier AI devices based on analog computing meet all stringent requirements. Legacy digital architectures simply can’t take it – the future is analog.
10x
Lower Power vs GPU
< 0.5W
TDP Power Draw
80M
On-Chip Weights
0
External DRAM
Extended AI output for peripheral devices
A leap forward in performance and power Volga analog matrix processors (Volga AMP™) offer tremendous advantages in power, performance and cost. They lower the barriers to innovation, taking powerful AI-based solutions to the edge.
Unique analog computing architecture Volga’s unique analog computing architecture performs the trillions of operations required for AI logic output inside dense arrays of flash memory. This innovative approach represents a significant advantage over typical digital architectures.
Designed for the most demanding applications With Volga’s integrated development environment, AI developers can quickly deploy even the most complex deep neural networks, confident that they will perform effectively in their product.
Markets
Intelligent machines and robots
Plants and farms maximize resource efficiency in response to the changing environment
Подробнее
Smart Home
Smarter home with AI powered appliances
Подробнее
AR/VR and Metaverse
Track your body's movement! Process interactive actions as fast as anyone can move.
Подробнее
Drones and the aerospace industry
Satisfy stringent bandwidth requirements with high data throughput
Подробнее
Media Center
View all news
