Break SoC Memory Bottlenecks
Stop sacrificing latency to maximize bandwidth. Avoid choking your layout with thousands of DRAM controller wires. The MemMax DRAM scheduler uses Virtual Channels to maximize throughput while guaranteeing QoS. Fully utilize DRAM parallelism. Insulate your SoC architecture from DRAM choices.
Expose Multi-core Communication and Sharing
No more untraceable races. No more hidden performance surprises. SonicsMT provides performance monitoring and hardware tracing for critical NoC targets. Debug precise sequences. Characterize specific traffic. Trigger and trace into your ARM CoreSight-compliant system.
Zoom Across Clock and Power Boundaries
Stop fighting timing paths at domain crossings. Avoid netlist re-work to match power domain hierarchy. SonicsExpress crosses clock, voltage, and gated power domain boundaries with flexible distance spanning. Partition your NoC where YOU want. Optimize domain crossings to your floorplan.
NoCs Optimize Shared Memory
SoC performance is often dominated by external DRAM access, particularly in digital consumer devices running high quality video and graphics. This white paper covers optimization of the NoC and memory subsystem to achieve the required throughput
Download the Memory Bottleneck White Paper
Subscribe to “The Ice Breaker”
We’ll keep you up to date on product news, technical issues, and industry trends on a regular basis when you subscribe to Sonics’ blog.
Subscribe here
Request Info; Give Feedback
Send us an email briefly describing your request for more information or constructive feedback and we’ll respond promptly.