GDDR6 Beyond Graphics : Memory for AI,VR, and Autonomous Driving
Modern computer applications rely heavily on graphics processing and rendering which involve a lot of simultaneous mathematical calculations. A typical CPU is not suitable for jobs that require simultaneous processing, which is why the concept of a dedicated Graphics Processing Unit (GPU) was introduced. The GPU has found its scope not only in graphics processing but also several emerging applications like AI, machine learning, VR, autonomous driving, and network routing.
GPU’s require memory which can offer much higher throughput than conventional memories like DDR, since it processes massive chunks of data all at once. The memory must also be capable of providing minimal latency, along with the possibility of simultaneous write/read. As a result, Graphics Double Data Rate (GDDR) memory, a dedicated type of SGRAM for the GPU, came into the picture.
To read the full article, click here
Related Semiconductor IP
- NPU IP Core for Mobile
- NPU IP Core for Edge
- Specialized Video Processing NPU IP
- HYPERBUS™ Memory Controller
- AV1 Video Encoder IP
Related Blogs
- Autonomous Vehicles: Memory Requirements & Deep Neural Net Limitations
- SSD Interfaces and Performance Effects
- What’s on the Horizon for NAND and DRAM?
- DDR3/DDR2 price crossover reached
Latest Blogs
- Securing The Road Ahead: MACsec Compliant For Automotive Use
- Beyond design automation: How we manage processor IP variants with Codasip Studio
- Cadence Extends Support for Automotive Solutions on Arm Zena Compute Subsystems
- The Role of GPU in AI: Tech Impact & Imagination Technologies
- Time-of-Flight Decoding with Tensilica Vision DSPs - AI's Role in ToF Decoding