Imagination Technologies created new E-Series GPU chips that run pictures and AI tasks with less power. These chips handle from 2 to 200 TOPS of math work using small numbers. With the same chip design, they support phone language helpers, factory cameras, and self-driving cars. The system uses Neural Cores for AI speed and Burst Processors that save 35 percent more battery life.
The chips keep all graphics skills from older models but add faster AI math inside every core. This makes them work four times better than past D-Series parts at AI jobs. They use local memory first, which saves trips to outside memory that waste power. Developers can use regular tools like OpenCL or oneAPI to write programs for these chips without learning new skills.
The GPU works with sixteen different tasks at once through virtual machines that never slow each other down. Computer makers can run many graphics screens or AI programs without extra helper chips. The first E-Series parts ship in fall 2025, with car, home, desk, and phone versions coming later.
The chips keep all graphics skills from older models but add faster AI math inside every core. This makes them work four times better than past D-Series parts at AI jobs. They use local memory first, which saves trips to outside memory that waste power. Developers can use regular tools like OpenCL or oneAPI to write programs for these chips without learning new skills.
The GPU works with sixteen different tasks at once through virtual machines that never slow each other down. Computer makers can run many graphics screens or AI programs without extra helper chips. The first E-Series parts ship in fall 2025, with car, home, desk, and phone versions coming later.
