Seventy-five years is a long time. Fireworks often light up the sky to celebrate such significant milestones. How fitting it was, then, for the 75th Anniversary Celebration of the IEEE Electronics Component Technology Conference (ECTC 2025) to replace those fireworks with a drone show. After all, without the people in this organization, the technology innovations powering these aerial robots wouldn’t exist!
Like many of the conferences I’ve attended in the past year, the focus of ECTC 2025 was finding sustainable solutions to address power-hungry AI applications. AMD’s Sam Naffzigger talked about it in the keynote, remarking that in 2020 AI was “barely here” and now five years later, it’s driving an “insatiable demand for compute.” He said it’s the fastest growth he’s seen of any technology in the 37 years he’s been in the industry.
Naffzigger said some of the key challenges providing opportunities for the advanced packaging community include increasing compute power, improving power delivery and thermal management, and solving the 3D HBM challenges to bring the logic closer to the memory stacks.
Many of the ensuing plenary sessions focused on these challenges – including hybrid bonding, glass core substrates, materials for co-packaged optics, and a user perspective of chiplet architectures. These were so popular, attendees spilled into the hallways.
My aha moments from attending these sessions can be summed up as follows:
- AI is moving faster than expected
- Energy needs can’t keep up
- It’s critical to develop co-packaged optics
- Not everyone think chiplets are the best thing ever – chip designers are among them
- Optical interconnects are the future
- There are challenges to all these technology developments that need solving
- Power density and heat extraction are ongoing issues that need to be solved and hybrid bonding may be the key
ECTC Past, Present and Future
I haven’t been around nearly long enough to witness 75 ECTCs. Lucky for me and other attendees of ECTC 2025, there are a few who have come close. Rao Tummala says he’s attended every single ECTC. I’m assuming that’s since he joined the industry, because if you do the math, he was only 8 years old in 1950. Bill Chen said he missed one due to surgery, and Patrick Thompson has attended each one since he joined the industry except for when his daughter was born during the same week as ECTC. (The audacity!)
Tummala, Chen and Thompson were joined by other past-presidents of the IEEE Electronic Packaging Society (EPS) – John Lau and Kitty Pearsall. Together, they recapped the history of ECTC. In my opinion, this session was the highlight of ECTC 2025. Lucky for you that if you didn’t stick around until Friday, I did.
From Patrick Thompson, the current EPS president, we learned that the first rendition of ECTC was inspired by a need for standardized solutions for electronic components. Materials and technologies being used today did not yet exist, and the outcomes were not consistent or stable. For two days, pioneers in electronics gathered in Washington D.C. to share their ideas and developed a framework for collaboration. It was called the Symposium on Improved Quality Electronic Components and comprised 5 tracks. Over the years it evolved as the Electronic Component Conference (ECC) and then in 2000, it became ECTC as we know it.
75 years after that inaugural event, ECTC 2025 featured 500 papers in 50 sessions, for four days and drew a record 2500 attendees.
Thompson listed by decade, the major packaging innovations of the last 75 years, resulting from papers presented at ECTC:
From Tummala, we learned about his early vision for integrated system package – something he called system-on-package rather than system-on-chip. Tummala is a true visionary who could see back then what it took the rest of the industry years to understand – that packaging isn’t just about protecting the chip. Interconnecting, powering, and cooling chips and systems is the job of advanced packaging. Liquid cooling, stacked image sensors, and glass substrates are some of the leading breakthroughs his team at Georgia Tech worked on that got us where we are today. From here, he predicts solutions to get us to the amount of computing power needed will be photonics-based. But in the end, the ultimate system, which comprises 90B interconnections and needs only 20 watts of power, will always be the human brain, he said.
In determining the technologies he would highlight in his talk, John Lau said there were so many to consider, he set up criteria – it had to have made it to high volume manufacturing. He selected four papers that were key innovations:
What excites Lau right now? Cu-Cu hybrid bonding continues to be the hottest technology, with the most papers presented at ECTC 2022, 23 and 24. 3.5D IC integration, which will create smaller packages with higher electrical performance. He sees great promise for embedded bridges for chiplets using chip-on-wafer-on-substrate (CoWoS) and predicts that TSMC’s CoWoS-L will be in production before the end of the year. Glass packaging is also on his highlight list, including glass core substrates, through glass vias, and 3D embedded chip in glass to seamlessly integrate optical interconnects.
What will they be talking about at ECTC 2050, according to Lau?
- Co-packaged optics in HPC driven by AI and communications.
- 3D IC replacing TSV interposers with glass core substrates
- Cu-cu hybrid bonding and optical interconnects for chip-to-chip communications.
- How advanced packaging will survive in the chilly quantum computing environment
Bill Chen’s milestones took a slightly different path. He talked about how the EPS became the sole sponsor of ECTC in 2010, ending years of dual sponsorship. The deal, engineered by Chen and Phil Garrou, was signed by then ECTC General Chair, Jean Trewhella, and then EPS president, Rolf Aschenbrenner. This milestone “established a deep and enduring partnership between EPS and ECTC, traversing 15 years of collaborative and innovative growth towards the outstanding ECTC 2025 we have today.”
As a result of this collaboration, Chen highlighted two significant achievements: The development of fan-out technology, and the world’s first 2.5D product with four high bandwidth memory stacks. This joint collaboration of AMD and ASE was released in June, 2015. We at 3D inCites remember it well.
Kitty Pearsall’s trip down memory lane also took a different path than Chen, Tummala and Lau. Leading with her materials expertise, she talked about how ECTC helped the industry navigate environmental legislation such as the EU Montreal Protocol, Waste from Electrical and Electronic Equipment (WEEE) and Restriction of Hazardous Substances (RoHS)
As a sign of the times, Pearsall reminisces about how ECTC has evolved its knowledge dissemination over the years from a thick proceedings manual, to memory sticks, and now online content.
As we look to the future, Pearsall said that environmental sustainability, the electric vehicle infrastructure, smart manufacturing, digital twins, and the importance of securing our supply chain will inform our future decisions.
While many conversations I had with our members focused on what’s driving today’s innovations, the ones I enjoyed most were the reflections on highlights throughout the years, and what applications were driving the advancements.
In upcoming episodes of the 3D InCites podcast, we talk about what sessions topics on certain years were so hot, there was standing room only? What exciting technologies made it to mainstream? Which ones never happened? We shared memories of our very first ECTC, and how they’ve evolved. And for the newcomers experiencing their very first ECTC, we learned what brought them there.
At first, I couldn’t even recall my first ECTC, but after scouring the 3D InCites archives, I discovered it aligned with the year we launched, in June 2009. It took place in San Diego. My Eureka moment that year was realizing that of all the technologies and developments presented, only the ones who can be realized in HVM will make it. That year, the excitement was around integrated materials enabling TSVs, and a novel method of merely lining a TSV with copper, versus filling it. You can find that blog post here. It’s fun to see what we were talking about back then. As for what will happen in the next 25 years? Only time will tell. ~ F.v.T.
