Author's Latest Posts


AI Data Centers And Auto Industry Converge On Same Issues


Key Takeaways:   AI data centers need power from a range of sources, including batteries, to safeguard against blackouts, transient voltage spikes, and grid demand spikes.   As with regenerative braking and bidirectional charging in electric vehicles, data centers could feed power or heat back into the grid for public use, but the immediate goal is to disrupt the grid as little ... » read more

Fine-Tuning Humanoid Vision And Movement


Key Takeaways: Humanoids and autonomous vehicles share a zonal architecture, a need for FuSa, and a reliance on radar technology to see around corners and through objects. Multiple cameras with different fidelities can better replicate a human’s field of view; for example, some can provide wide fields of view at low grade, low fidelity, while others strive for infinite fidelity in a... » read more

I/O Design Challenges Grow In AI Data Centers And HPC Clusters


Key Takeaways: A designer’s choice of I/O connectors and interconnect protocols can be the difference between a massively profitable AI chip and a flop. I/O tradeoffs impact airflow, cooling, rack design, power coming into the rack, and other critical aspects of HPC chip design. Reliability is paramount, so standards must be followed, and I/Os need redundant pins. Other innovations... » read more

Orbital Data Centers Are Souped-Up Satellites – For Now


Key Takeaways: Today’s orbital data centers are better described as compute centers in space, as they resemble satellite constellations more than terrestrial data centers. The most common power solution is sun-synchronous solar at the poles, but this requires a multi-hop data relay from power-hungry compute to standard satellite constellations in the low-orbit mesh, then to Earth. ... » read more

AI-Defined Vehicles Increase Pressure On Auto Ethernet Reliability


Key Takeaways: For AI-defined vehicles and onboard agentic AI, Automotive Ethernet provides high bandwidth for sensor data fusion, TSN ensures low latency and synchronization for real-time decisions, and MACsec secures the data link. Time-sensitive networking (TSN) is an essential protocol for ensuring 10BASE-T1S delivers data to where it needs to go on time. Still, it becomes less esse... » read more

Swapping Out Chiplets: I/Os Vs. Compute


Key Takeaways: Companies can save time and money by swapping out a compute, memory, or I/O chiplet to gain technology improvements, while keeping the other dies stable. Chip architects may choose to keep their I/Os stable and swap out compute to move from a 5nm process node to 3nm to achieve performance and power improvements, or swap out memory from LPDDR5X to LPDDR6. Swapping out... » read more

Confusion Grows With More Interconnect Options And Tradeoffs


Key Takeaways: Designers are frequently evaluating 5 or more different interconnects in a single system, each with a distinct purpose. While chip-to-chip (PCIe) and die-to-die (UCIe, BoW) technologies seem to be solving a similar problem, in practice they bring different challenges. PCIe, CXL, NVLink, and UALink are all active in the hyperscaler space, but Ethernet-based technologies... » read more

Humanoid Touch And Voice Are Improving Rapidly


Key Takeaways Humanoid robots are rapidly expanding beyond factories and logistics toward broader, general-purpose roles (including in-home assistance), driven by advances in AI and sensing. Compared with vision and language, touch (haptics) and hearing/voice in real environments remain the hardest — and most commercially important — sensing challenges, requiring fast sensor fusio... » read more

Next-Gen Batteries Require Impedance Data And Active Balancing


Key Takeaways Electric vehicles and energy storage systems using LFP batteries require more sophisticated diagnostic methods because they exhibit very flat cell voltages across various charge levels, making it harder to estimate usable energy. Battery management systems are leveraging new components, AI/ML, digital twins, and other techniques to get more accurate, real-time data, includ... » read more

NoC Coherency Challenges Balloon With AI SoCs And Chiplets


Key Takeaways Data movement, congestion, and energy efficiency are key determiners of whether compute is usable. Different processors bring various coherency challenges. For example, a cache-coherent NoC for CPUs is expensive and harder to verify than an I/O-coherent NoC for an accelerator. Designers need to balance top-down performance with bottom-up physical engineering to effect... » read more

← Older posts