Overview
The TCI LPDDR PHY is a high-performance, scalable system using a radically new architecture that continuously and automatically adjusts each pin individually, correcting skew within byte lanes. This state-of-the-art tuning acts independently on each pin, data phase and chip select value. Read gate and data eye timing are also continuously adjusted. Automatic training is included for multi-cycle write leveling and read gate timing, read/write data eye timing, and PHY Vref and DRAM Vref settings. Remarkable physical flexibility allows the PHY to adapt to each customer’s die floorplan and package constraints, yet is delivered and verified as a single unit for easy timing closure with no assembly required. The PHY is DFI 5.1 compliant, and when combined with an appropriate LPDDR memory controller, a complete and fully-automatic LPDDR system is realized.
Learn more about LPDDR IP core
JEDEC has just released the LPDDR6 specification that is expected to take the LPDDR DRAM market to new heights with data transfer speeds that can reach up to 14.4Gbps, which is a 50% improvement over LPDDR5X speeds. LPDDR6 device density can range from 4 GB to 64 GB.
This solution brief highlights how Samsung’s mainstream 8nm process node and LPDDR5X (LP5X) memory together deliver an optimized balance of performance, bandwidth, power efficiency, and cost. The combination enables scalable AI inference across diverse computing environments, from edge devices and AI appliances to emerging inference-focused AI infrastructure, making advanced AI deployment more accessible for startups, fabless innovators, and OEMs alike.
Next-generation automotive systems are advancing beyond the limits of currently available technologies. The addition of advanced driver assistance systems (ADAS) and other advanced features requires greater processing power and increased connectivity throughout the vehicle.
Understand memory-aware ASIC design for AI workloads, from memory hierarchy and dataflow to NoC, DMA, and hardware-software co-design.
As AI workloads continue to diversify, the systems that support them are evolving just as quickly. AI is no longer confined to the hyperscale data center. It is moving to the factory floor, into vehicles, and increasingly to the edge, where power, cost, and form factor constraints can matter just as much as raw performance.