By Richard McPartland, Technical Marketing Manager, Moortec
As you may have read in our previous blogs, chip designers working on advanced nodes now typically include a fabric of sensors spread across the die to address a number of very specific challenges. In this, the final instalment of a three-part blog series Richard McPartland, Moortec’s Technical Marketing Manager continues to explore some of the key applications and benefits of these types of sensing solutions. In this blog the focus is voltage monitoring and why understanding in-chip supply variation alongside in-die process speed and thermal conditions is essential if you want to maximise performance and power, improve reliability and ultimately reduce costs in your cutting-edge design.
Chip developers continue to migrate to new smaller geometry nodes to take advantage of the considerable benefits associated with higher logic density, faster performance and lower power. These benefits come hand in hand with a raft of new challenges which also need addressing. In the first blog in this series entitled “Key Applications for In Chip Monitoring…Thermal Sensing”, we touched on the “End of Dennard Scaling” as just one of these challenges. (Click HERE to learn more)
Why do we need to monitor voltage Supply?
In-chip voltage monitors provide the means to accurately measure core supply domain voltages on advanced node digital MOS devices and validate the power distribution network which may be deeply embedded in a 2.5D or 3D package. Further, voltage monitors can help understanding of how IR droops vary with software workload and are particularly useful for SoCs implementing voltage scaling schemes, which we touched upon in the last blog “Key Applications for In Chip Monitoring…In-Die Process Detection” (Click HERE to read the blog)
IR Drop Analysis
Embedded voltage monitors are capable of monitoring multiple supply domains and can also be used to generate low voltage alarms as well as providing real-time supply data. Further to the trends of more power per unit area and large chip sizes, comes with each new node a reduction in core VDD voltage and increased interconnect and via resistance. VDD voltages have been steadily dropping with each new advanced node and are now well below 1 volt. With gate delay strongly dependent on VDD, a 100mV IR drop that might have been sustainable on an older node is likely to cause major timing problems on the latest FinFET chips. Many SoCs are now software driven and predicting the worst case software driven change in work load can be very difficult. Often the control software is written by another team perhaps in another company further compounding the difficulty in defining worst case.
To address these challenges, it is strongly advisable to embed voltage sense points especially for all critical circuit blocks. This will give visibility into on chip conditions and help answer a fundamental question: is my power distribution network good enough so that critical circuits are powered correctly and can operate at their intended frequency? In addition voltage monitors can measure the static voltage droop due to changes in workload. These changes can be software driven and difficult to predict in advance. Measuring and checking these voltage drops on silicon is the final proof.
SoC developers are facing a range of new challenges when designing on advance process nodes especially FinFET. These include increased power per unit area on increasingly large die leading to the potential for hotspots/thermal problems and difficult to predict IR drops. Process variation is now significant not just from die to die but even within large die.
To help address these challenges, a common strategy is to embed a fabric of in-chip monitors across the die. These give visibility into on chip conditions which is especially valuable for critical circuit blocks. The latest FinFET SoCs often embed tens of temperature sensors to monitor hotspots and enable thermal management to reduce maximum temperature and temperature gradients across the die.
Multiple Sense Points
Voltage sense points are also very useful to monitor supply voltages at critical circuit blocks and check the power distribution network is well designed and also for difficult to predict voltage droops from software driven workloads. On chip process detectors can provide a quick and independent check of process variation directly at critical circuits and also dramatically speed up testing for speed binning or voltage scaling. In-Chip monitoring is fundamental for all developers who wish to obtain the maximum from their SoC whether it is performance, power consumption or reliability or combination of these.
So, if this blog has got you thinking about the effects of variable voltage supplies in your existing or next design, why not contact Moortec today to find out more.
Look out for Moortec’s next “Talking Sense” blog entitled: “Are you Listening” which will be dropping on the 4th of May. If you have missed any of Moortec’s previous “Talking Sense” blogs, you can catch up HERE