The future of semiconductor chips is shifting from simply making transistors smaller to reducing the distance data must travel inside a computing system. Modern AI chips depend not only on fast calculations but also on high speed data movement between memory, logic, processors, and input output systems. As artificial intelligence workloads grow larger, chip design is becoming less about one giant processor and more about chiplets, advanced packaging, 3D integration, memory bandwidth, power efficiency, heat control, and faster communication between specialized components. Odisha’s advanced glass semiconductor packaging unit represents a major step in India’s semiconductor ecosystem, with glass substrate based packaging and heterogeneous integration helping multiple chiplets work together like one coordinated system. This approach is important for AI computing, data centers, 5G, defense electronics, high performance computing, and next generation consumer technology, where speed, energy efficiency, cooling cost, and system design matter as much as raw processor power. #technology #tech #india #semiconductor
ADVERTISEMENT
This is a second channel. You will get different type of topics covered here.
To be honest I personally attended the seminar of 3d glasses chip in bhubaneswar. It was wonderful 😅❤
I hardly ever think those hungry tech giants would ever lower the prices as time passes.
Jay Jagannath ⭕❗⭕
As an ece student it felt fascinating to me 😅
The probleuis there is threshold to reducing size of semiconductors at certain limit we can not make more smaller semiconductors. So thet is the reason we are optimising it in another way...
Jai maa kali 😊
SOC(System on a chip) ❌ COC (Computer on a chip) ✅ ✋🏻😂🤚🏻
MADE in INDIA ❣️🇮🇳☠️🔥
Love from odisha ❤
Jay Jagannath From Odisha ❤🙏
I only eat lays chips😎
U da real AI
Informative video 📹 ❤
Bro is now ai arc
Respect for government ❤ 📈
India day by day ☠️
Maintenance cost reduced by 0.5 rupee😂
That's why arm with Apple's unified memory design is the future in my opinion Arm for high efficiency and less heating & unified memory for no copy r/w between cpu and gpu since the physical ram itself acts as vram for the gpu. Apple's mlx already provides APIs for utilising this design architecture
That's why macbook have unified memory 😊