Tony Chan Carusone

The soon-to-come Ottawa Semiconductor Fab in Canada, a vast III-V semiconductor tech campus, is set to be a hub for semiconductor manufacturing, research, academia, and related amenities. SILICAN's report stresses the need to enhance domestic chip capacity to safeguard Canada's security and economy amidst geopolitical shifts, urging investments to connect chip producers and end-users, while calling for a robust talent pool in the industry. CMC Microsystems emphasizes securing advanced device supply, protecting IP, and boosting onshore manufacturing through initiatives like FABriC, aiming to create intelligent sensor technology and enhance Canada's competitive advantage and supply chain security, particularly in Ottawa's Compound Semiconductor sandbox. Various partnership opportunities are envisioned to propel compound semiconductor innovation, support academia, develop technology like GaN power systems, foster EV battery advancements, and bolster autonomous vehicle technologies within Ottawa's flourishing semiconductor ecosystem. #Canada #semiconductor #chip #GaN #EV #AI #power #research #innovation #manufacturing #technology #sensor #compoundsemiconductor https://lnkd.in/evsN-6HN

16

A very interesting blog by Giuseppe Piccolboni of Weebit Nano Ltd about the use of our #ReRAM for #neuromorphic applications

36

Check out this highlight of a recent study investigating the local 3D dopant distribution within SiC power semiconductor devices using atom probe tomography!

26

5 Common Misconceptions About Microchip Design 1. Misconception: Microchips are all about size. Reality: It’s not just about making them smaller. Performance, power consumption, and heat management are just as critical as size in modern microchip design. 2. Misconception: Only a handful of companies innovate in microchip design. Reality: Innovation is thriving among smaller firms and startups, with 35% of recent breakthroughs in microchips coming from companies with fewer than 500 employees. 3. Misconception: Microchips are standard across all devices. Reality: A microchip designed for a smartphone will not perform optimally in a wearable device. Custom chip designs tailored for specific use cases are now more important than ever. 4. Misconception: Only the latest generation chips matter. Reality: Legacy chips are still in high demand, especially in industries like automotive, where reliability and long product life cycles are key. 5. Misconception: Microchip development takes years to yield results. Reality: With advanced design tools and cloud-based simulations, development cycles have been shortened by 50%, allowing companies to innovate faster.

7

My article on SuperComputing Architecture Shift in EFY's Aug Magazine. "The silicon chipset, which is procured from Intel as of today, is the most crucial component in the high-performance compute server. Everything involved in making a supercomputer like the motherboard, the rack, cooling systems, and the software stack, depends on the architecture and features of the silicon chipset. " Electronics For You #Aum #Supercomputing #HPC

151

4 Comments

ϕPPL has the same execution model as C. So programs that use only scalar processing written in C should run about the same speed as equivalent programs written in ϕPPL. But ϕPPL has parallel features that C lacks. When the processor has supporting parallel hardware then programs written in ϕPPL can run appreciably faster. For example, we can add two vectors together in ϕPPL this way: A ⇐ B + C; Whereas in C you must write: for (i=0; i< Count;++i) A[i] = B[i] + C[i]; Today I read a post comparing the speed of execution of programs written in Rust as compared to their equivalents written in C. C programs ran about 11× as fast as their Rust equivalent. C++ typically runs about 1/3 slower than its C equivalent. It is partly due to its more complicated frame stack. Python is much slower and typically runs at around 1% the speed of C. #computers #programming #rust #c #c++ #python

2

Actually, RAM=Read write Random Access Memory. It is volatile. ROM=Read only Random Access Memory. It is non-volatile. Clearly explained about the above fullforms.This post answers the question "Why RAM is called as Random access memory? " #memories #embeddedsystems #computers #volatileandnonvolatilememories #processors

2

5 Reasons Why Legacy Nodes Remain Crucial in the 1nm Era? Legacy nodes Typically refer to older, well-established fabrication processes, often at 45 nm or 28nm and larger. 🚀 Cost-Effectiveness: Manufacturing chips at advanced nodes like 1nm is expensive due to the complexities involved. Legacy nodes are a better choice for cost-sensitive applications. Imagine building a high-performance gaming PC vs. a basic office computer. The gaming PC (1nm chip) is expensive with all the latest tech. But a basic computer (legacy node) is much cheaper to make, perfect for everyday tasks like browsing the web. 🚀 Power Efficiency: Smaller transistors in advanced nodes enable faster performance but often come at the cost of higher power consumption. Legacy nodes, with their larger transistors, can be more power-efficient, a critical factor for battery-powered devices like smartphones and wearables. Tiny transistors in new chips are like powerful engines in cars - they work great but use a lot of fuel (battery power). Bigger transistors in legacy nodes are like fuel-efficient car engines - they might not be the fastest, but they use less battery, ideal for things like smartphones and smartwatches that need to last all day. 🚀 Maturity and Reliability: Legacy nodes have been around for longer and have undergone extensive refinement, leading to highly optimized and reliable processes. This maturity ensures consistent performance and fewer bugs, crucial for applications demanding stability. Think of a recipe you've made many times. You know exactly how much of each ingredient to use and how long to cook it for. Legacy nodes are like that tried-and-tested recipe - they're reliable and we know how to make them work perfectly for everyday electronics. 🚀 Compatibility: Not all applications require the bleeding-edge performance offered by 1nm chips. Legacy nodes can integrate well with newer technologies, enabling efficient design of complex systems with a mix of performance requirements. Imagine building a house - you might need different materials for the roof, walls, and foundation. Legacy nodes can be like those different materials - they work well with newer technologies to create all sorts of electronic devices, like a phone with a powerful processor (1nm chip) but a low-power Bluetooth chip (legacy node) for connecting to headphones. 🚀 Internet of Things (IoT): The vast and growing world of IoT devices, many of them simple sensors, often doesn't need the processing power of 1nm chips. Legacy nodes are perfectly suited for these low-power, low-cost applications. Tiny sensors in things like fitness trackers or smart thermostats don't need the processing power of a fancy 1nm chip. Legacy nodes are the perfect tool for these smaller jobs, keeping costs low and battery life long. Guess, Which Country is investing huge in legacy Nodes? Tell us in the comments. A detailed post is in comments. For all semiconductor and AI content, follow TechoVedas

77

3 Comments

In #SNUGIsrael's second keynote presentation, Google’s Uri Frank explained why the AI-era silicon challenges are much different than we had in the past. For start, AI models require a x10 speedup in compute power every year, much more than Moore's law.   Therefore, the current design flow that requires an average of 3 years for a new major device won’t cut. By the time these chips are ready, the workloads they were planned for are no longer relevant! Uri said that an #evolution won’t be enough- like the #AI-based small improvements offered by the EDA vendors. We need a #revolution - we must define new ways of designing chips:  “What if designing a custom chip took a few people a few weeks?” he asked. Synopsys Users Group (SNUG)

45

9 Comments

“IEEE802.3dj Work on 200 Gbps per Lane and How Different FEC Options Affect SI,” Cathy Ye Liu, Broadcom. Very interesting paper that I recommend to read.

10

Thermal Characterization for AI chips A common requirement for thermal management design efforts of AI chips is the need for a real-time, hardware-oriented, emulation capability. Attempting to use real application chips is difficult to implement. First, waiting until the chips are available means that time-to-market is greatly elongated. Second, when creating a laboratory setup using application chips(s), the instrumentation for thermal management design and development is not a trivial task, in terms of both time and cost. The problem is further complicated when the “application chip” is a big CPU chip surrounded by smaller chips, as in a chiplet package. Each area of the big main chip and each of the surrounding small chips in an AI device usually operates at different power density levels, and implementing these power density levels may be very difficult. These difficulties can be overcome by using a Thermal Test Vehicle (TTV) built around Thermal Test Chips (TTC). The TTV package, which usually matches the physical size and material composition of the application device it is intended to emulate, can contain one or more TTCs each closely matching the size and power dissipation maps of their real-life counterparts in the application device package. Such TTV's can be available long before the actual application device is built. The TTC is an array of Unit Cells (either 1mm square or 2.54mm square) where each cell contains heating resistor(s) and temperature sensing diode(s) (TSD) which can be individually and directly addressed. Power dissipation mapping is accomplished by electrically inter-connecting cells on the package substrate and bringing out the desired connections, and the location-desired TSD connections can be brought out in a similar way to allow for temperature measurements and real-time mapping. Thus, the power density can be controlled, and the temperature can be monitored in real time, down to the Unit Cell level (1x1mm) across each of the TTCs in the package. As a result, the power mapping of the application device can be accurately emulated using the TTV (coupled with actual thermal management solutions, including thermal interface materials and heat sinks, for example), and the temperature evolution across each chip can be closely mapped, for steady state and transient settings. For further information about how the TTC/TTV can enable the development of AI device thermal management solutions, please visit www.thermengr.net. #SemiconductorPackaging #HeterogeneousIntegration #CHIMES #semiconductormanufacturing #semiconductors #semiconductorindustry #semiconductor #advancedpackaging #thermalmanagement #thermalanalysis #thermalengineering #systemintegration #microelectronics #3d #packaging #computing #Chiplets, #SoC, #SiP #chips #TSV and 3D Packaging Technologies #Automotive Electronics Packaging #Semiconductor #advancedmanufacturing

45

3 Comments

1st #ICCAD2025 keynote by Philip Wong on „Design and Design Automation for a Future Generation of Chips“ just started. Looking forward to great insights. #iccad

15

#Canada is bolstering its #semiconductor industry with a $120 million investment from the federal Strategic Innovation Fund (SIF) to CMC Microsystems. This funding, part of CMC's $223-million Fabric program, will provide affordable tools, training, and prototype production to hardware firms and researchers. The initiative aims to bring new semiconductors and advanced technologies to market, supporting startups with prototypes to attract investors and customers. Projects in key areas like #semiconductors and #photonics will receive up to $10 million each, highlighting Canada's commitment to strengthening its tech capabilities and reshoring supply chains. #SemiconductorStartups #CanadianTech #StartupSupport #Innovation #TechDevelopment #CMCMicrosystems

13

1 Comment

Our latest performance update for Llama70B inference today achieves over 2,100 tokens per second, making it the fastest inference, period. And 16x faster than the fastest GPU solutions today. All made possible by the Cerebras Systems Wafer Scale Engine chip, that's 56x larger than traditional GPU chips. To try it out and learn more how this accelerate your #ml and #ai applications, go to https://lnkd.in/gYx4Y5XT

27

Great article with insights by Bernard Murphy at SemiWiki.com, "Arteris Frames Network-On-Chip Topologies in the Car." "... while there is plenty of complexity in evolving car architectures and the consequent impact on subsystem chip/chiplet designs, connected on-chip through both coherent and non-coherent networks, the intent behind these systems is quite clear. We just need to start thinking about total car system architecture rather than individual chip functions." #Arteris #NetworkOnChip #Automotive https://bit.ly/3UiKqls

25

Tackling Data Retention in SSDs (by Anil Burra from Intelligent Memory) 🤩 "SSDs have conquered the market due to their speed, durability, energy efficiency, and small form factor. However, write endurance and data retention are concerns in #embedded and #industrial applications, especially as the technology moves to ever smaller nodes." (Anil Burra) With the scaling of #NAND #flash #technology to higher densities (e.g., Triple-Level-Cell or Quad Level-Cell) and smaller process nodes, data retention becomes more challenging due to increased cell-to-cell interference and reduced margins for charge storage. Therefore, newer #technologies may have shorter specified data retention periods compared to older, more robust #SLC (Single-Level-Cell) or #MLC (Mul ti-Level-Cell) #NAND products. Two key factors affect data retention, the period over which #flashmemory can store #data without loss or corruption, in #NANDflash cells: program/erase (P/E) cycling and operating temperature. A very big thank you to Anil Burra from Intelligent Memory for this insightful article via Electronic Specifier Ltd with more background and insights via the link below 💡🙏👇 - A very good read to an important TOPIC 🤩 https://lnkd.in/eqdH7-N2 #semiconductorindustry #semiconductormanufacturing #innovation #it #ai #ssd #datastorage #automotiveindustry #embeddedsystems #ic #computer #computing #iot #memory #chip #chips #technology #tech

50

1 Comment