tag: Axiomise

An AI Model Fit For Purpose
By Brian Bailey - 09 Jul, 2026 - Comments: 0

Key takeaways A model can only be used for its intended purpose, in a defined context, without taking unknown risks.  Models must be created using a well-defined process and verified in a way that provides a level of independence.  Deployment requires trust and a way to track the properties of the model. A model captures some kind of behavior exhibited in the real world, b... » read more

Observability Is A Missing Layer In AI-Era Chiplet Design
By Ann Mutschler - 01 Jul, 2026 - Comments: 0

Key Takeaways: In chiplet-based architectures, observability must be designed as a fabric-aligned, cross-die telemetry plane so architects can correlate traffic, latency, congestion, and fault behavior across package boundaries without losing system context. AI can extract value from high-volume silicon telemetry only when the architecture provides consistent instrumentation, near-senso... » read more

Rethinking Chip Verification
By Ann Mutschler - 29 Jun, 2026 - Comments: 0

Key Takeaways: AI and modern tools are easing traditional verification pain, but they're not addressing the underlying bottleneck in complex designs. Work is underway to create a golden, unambiguous spec above RTL, tracing requirements from spec to implementation to verification and checking for gaps, conflicts, and inconsistencies across levels and blocks, often with AI help. Tool c... » read more

I/O Design Challenges Grow In AI Data Centers And HPC Clusters
By Liz Allan - 25 Jun, 2026 - Comments: 0

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

Designing Chips That Can Explain Themselves
By Ann Mutschler - 17 Jun, 2026 - Comments: 1

Key Takeaways: On-die telemetry gives architects a path to replace worst-case design margin with measured silicon behavior, improving PPA without compromising resilience. As monitor density and control-loop speed increase, observability must be architected hierarchically across local hardware response, on-die processing, and fleet-level learning. The real payoff is architectural: str... » read more

Swapping Out Chiplets: I/Os Vs. Compute
By Liz Allan - 28 May, 2026 - Comments: 0

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

Why Your NoC Verification Strategy Must Consider Using Formal
By Ashish Darbari - 28 May, 2026 - Comments: 0

By Ashish Darbari and Bing Xue It’d be inconceivable these days to design a modern high-performance SoC without a network-on-chip (NoC) fabric. AI hyperscalers are inherently multi-threaded and rely on using hundreds of processing elements (PEs). Crossbar-based fabric would just not scale. What also changes with the adoption of the NoC is how to handle coherency between PEs. ACE is no long... » read more

Confusion Grows With More Interconnect Options And Tradeoffs
By Liz Allan - 18 May, 2026 - Comments: 0

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

Chip Industry Week in Review
By The SE Staff - 01 May, 2026 - Comments: 0

Advanced nodes and capacity The US Commerce Dept. told IC equipment makers to stop shipments to Hua Hong Group, China's No. 2 chipmaker, in order to protect America's lead, according to Reuters. Global AI competition is causing wafer and packaging shortages, but capacity increases are expected to come online later this year and in 2027 to ease the crunch, according to TrendForce. Leadi... » read more

Why Proof Convergence Matters
By Ed Sperling - 20 Apr, 2026 - Comments: 0

Achieving a deterministic "yes or no" answer in semiconductor verification is becoming more challenging as chip complexity increases. There are more cores, more potential interactions, and more reliance on AI to build AI chips. Ashish Darbari, CEO of Axiomise, talks about the impact of functional interactions involving safety and security, and where to look for common patterns to prevent bugs f... » read more

← Older posts

  • Sponsors

About

Navigation

Connect With Us

Copyright ©2013-2026 SMG   |  Terms of Service  |  Privacy Policy