Naren Vallepalli

What is a System on Chip (SoC) in semiconductors? SoC (System on Chip) is an integrated circuit (IC) that combines all the essential components of a computer or other electronic system into a single chip. It’s designed to be a complete system housed within one semiconductor die, optimized for performance, power efficiency, and compactness. 🔧 Main Components of a SoC (with examples) 1. CPU (Central Processing Unit) • Function: The main processor that performs general-purpose computing tasks. • Example: ARM Cortex-A78 core in Qualcomm Snapdragon chips. 2. GPU (Graphics Processing Unit) • Function: Handles graphics rendering and image processing, crucial for gaming and UI. • Example: Adreno GPU in Snapdragon; Apple uses custom GPUs in their A-series chips. 3. Memory (RAM / Cache) Controller • Function: Manages access to memory for faster data exchange between components. • Example: LPDDR5 controller in Samsung Exynos SoCs. 4. DSP (Digital Signal Processor) • Function: Processes signals such as audio, video, and sensor input efficiently. • Example: Hexagon DSP in Snapdragon processors. 5. ISP (Image Signal Processor) • Function: Optimizes camera data, including autofocus, HDR, and low-light enhancements. • Example: Google Pixel’s Tensor SoC includes a custom ISP. 6. NPU / AI Engine • Function: Accelerates machine learning tasks such as voice recognition and real-time translation. • Example: Apple Neural Engine (ANE), Qualcomm Hexagon AI Engine. 7. Modem (Cellular & Connectivity) • Function: Enables 4G/5G, Wi-Fi, Bluetooth, and other communication protocols. • Example: Snapdragon X70 5G modem integrated in Snapdragon 8 Gen 2. 8. Security Module • Function: Handles encryption, secure boot, biometrics, etc. • Example: Apple’s Secure Enclave or ARM TrustZone. 9. Power Management Unit (PMU) • Function: Manages power distribution to different components for efficiency. • Example: Custom PMUs in wearable SoCs to maximize battery life. 10. IO Interfaces / Peripherals • Function: Includes USB, HDMI, PCIe, SPI, I2C, etc. to connect with external devices. • Example: USB 3.1 controller in AMD Ryzen Embedded SoCs. ✅ Why SoCs are Important • Compactness: One chip = smaller devices • Power Efficiency: Optimized for mobile and embedded systems • Performance: Tight integration = fast communication between components • Cost-Effective: Fewer parts = lower bill of materials For all semiconductor and AI related content, follow TechoVedas If you are looking to invest in semiconductors, and need expert insights- drop us a DM.

471

11 Comments

Very fundamental concept in digital design and timing. 1. Synchronous Digital Design • In a synchronous design, a single global clock drives all sequential elements (flip-flops, registers, etc.). • Data is transferred on clock edges, and all timing is measured relative to this same clock. • Characteristics: • Easier to design and verify (since all timing paths are referenced to the same clock). • Common in CPUs, GPUs, and many SoCs. • The speed is limited by the slowest path (critical path) in the circuit. • Example: A CPU pipeline running on a single 2 GHz clock. ⸻ 2. Source-Synchronous Digital Design • In a source-synchronous system, the transmitter (source) sends both data and a clock (or strobe) to the receiver. • The receiver uses this forwarded clock/strobe to capture incoming data instead of its own global clock. • Why used? • Useful for high-speed interfaces where the global clock skew/uncertainty would be too large (because data would arrive too fast for a shared clock domain). • The clock and data signals experience similar delays across the channel, so timing between them stays valid. • Examples: • DDR SDRAM interfaces → The memory chip sends/receives data with DQS (data strobe). • High-speed parallel buses between chips (like FPGA-to-FPGA). ⸻ 3. Asynchronous Digital Design • In an asynchronous system, there is no global clock. • Instead, circuits use handshaking protocols (request/acknowledge) or self-timed logic to communicate. • Characteristics: • Potentially very power-efficient (no clock tree → less switching power). • Can be faster in theory (not limited by worst-case global clock). • But much harder to design and verify, since timing is based on delays and handshakes rather than a fixed reference. • Examples: • Asynchronous FIFOs between two unrelated clock domains. • Fully asynchronous processors (rare, but exist in research). Visual comparison of the three styles: 1. Synchronous → Data is aligned and sampled with a single global clock. 2. Source-Synchronous → The strobe/clock is forwarded along with data, so the receiver samples relative to it. 3. Asynchronous → No clock at all, data transfer happens using request/acknowledge handshakes.

178

1 Comment

Top 10 Semiconductor Technology Trends Shaping 2026 & Beyond 🚀  Semiconductors are no longer just components—they’re the infrastructure of AI, EVs, and the digital economy.   Here are 10 key trends that will define the next chapter of the industry:  → AI Compute & Custom Silicon   Enterprise spend keeps tilting toward accelerators and in-house chips. NVIDIA’s data-center revenue alone hit $39.1B in Q1’26, +73% YoY.  → Chiplets, 3D IC & Advanced Packaging   CoWoS and SoIC capacity is expanding fast—expected to hit ~75k wafers/month by 2025, with even more tied to Blackwell-class GPUs.  → High Bandwidth Memory (HBM)   HBM is the scarce currency of AI. It already represents 20% of DRAM revenue and will exceed 10% of bits shipped in 2025.  → Automotive Chips   By mid-decade, each vehicle will carry $2000+ worth of silicon. Regulations like EU’s GSR II and US AEB mandates are accelerating adoption.  → The 2nm Race & Angstrom-Class Roadmaps   TSMC’s N2 enters high-volume in 2025; Intel’s 18A promises +25% performance or -36% power. Backside power delivery becomes reality.  → Supply Chain Geopolitics   The US has awarded $30.6B across 19 firms—TSMC, Intel, and Samsung each securing >$6B. Semiconductor capex announcements now lead FDI.  → Photonic & Quantum Integration   NVIDIA’s Spectrum-X Photonics targets 100–400 Tb/s switch throughput, with Co-Packaged Optics (CPO) expected to hit volume by 2027.  → Edge AI & Domain-Specific Processors   AI-capable PCs will be ~57% of shipments by 2026. Counterpoint forecasts >400M GenAI smartphones by 2025.  → Wide-Bandgap Power (SiC & GaN)   From EVs to renewables, SiC and GaN have gone mainstream. Navitas alone reported $450M in GaN design wins and 50% YoY growth in 2024.  → Sustainability   One fab can use up to 5M gallons/day of ultrapure water. TSMC consumed 101M m³ in 2023. Water, waste, and energy efficiency are becoming competitive differentiators.  Key takeaway:   By 2026, the semiconductor industry won’t just be about Moore’s Law—it will be about materials, architectures, geopolitics, and sustainability all colliding at once.  P.S. If you want more deep-dive explainers on semiconductors, check out my blog *The Semiconductor World* — link in comments.  #Semiconductors #AI #Chiplets #Packaging #2nm #HBM #SiC #GaN #Sustainability #TechTrends #TestFlow

136

6 Comments

PwC predicts that the semiconductor industry will enter its largest expansion in history. $1.5 trillion in fab investments will be made by 2030. What’s driving it? 1. AI workloads - AI accelerators are projected to account for 50% of data centre semiconductor revenue by 2030 2. Power semiconductors will make up more than half of EV chip costs as sales approach 50% of global car sales 3. AR/VR, robotics, and smart devices adding new layers of demand 4. Government subsidies are fueling the boom By the end of the decade, servers & networks will surpass all other sectors to become the largest semiconductor market. That shift means not only new capacity but also new risks, from supply chain dependencies to massive energy demands. https://lnkd.in/etgDfFEH

186

10 Comments

Ever wonder what solar panels could do if they had brains? That’s basically what HTEC and Infineon Technologies pulled off. Their AI-powered system, running on Infineon’s PSOC™ Edge platform, uses real-time voltage, current, and temperature data to tweak power output on the fly—no fancy irradiance sensors needed. ⚡ 23× faster response 🔋 42× less energy used 🔧 Built-in predictive maintenance Smarter energy. Less waste. More uptime. This is the kind of tech that moves us forward. Curious how it works? 🔗 Check out the whitepaper: https://lnkd.in/gKqhM_F4

178

4 Comments

🔍 The Formal Verification Cold Start Problem You’ve read the spec. You’ve discussed with the designer. You think you understand the design… But now you’re staring at a blank SVA file wondering: “Where do I even begin?” If that feels familiar, welcome to the cold start problem of formal verification—and yes, we’ve all been there. 💡 The Fix: Start with Cover Properties Forget trying to prove everything from day one. Your goal is to observe behavior before you start proving constraints. 🎯 Real Example: FIFO Design When I was verifying a simple synchronous FIFO, I started like this: cover property (@(posedge clk) wr_en && !full); ✅ Covers a write happening when FIFO isn’t full cover property (@(posedge clk) rd_en && !empty); ✅ Covers a read when FIFO isn’t empty cover property (@(posedge clk) wr_en ##[1:5] rd_en); ✅ Covers a write followed by a read within 5 cycles That’s it. From there, I built ordering checks, flag assertions, and finally full correctness proofs. The key was just starting. 🧪 Design Exercise Plan (for You or Your Team) Try this simple cold-start exercise on a FIFO design: 🔹 Step 1 – Cover Activity Cover a write when not full Cover a read when not empty Cover write followed by read 🔹 Step 2 – Observe State Cover full and empty being reached Cover FIFO staying non-empty for N cycles Cover pointer wraparound 🔹 Step 3 – Simple Assertions assert property: write only when not full assert property: read only when not empty assert: empty and full never both high 📌 Just like in software: don’t over-engineer your first line of code. In formal, don’t overthink your first property. Momentum matters more than perfection. Start anywhere. Learn as you go. #FormalVerification #VLSI #FIFO #SVA #SystemVerilog #DesignVerification #Semiconductors #HardwareDesign #VerificationEngineer #FormalMethods #ColdStartProblem #TechTips

107

6 Comments

This week, Qualcomm reported our Q3FY25 earnings report.   We reported another quarter of strong results, with Non-GAAP revenues up 10% YoY to $10.4B and Non-GAAP EPS up 19% YoY to $2.77. QCT revenues are up 11% YoY to $9.0B and EBT up 22% to $2.7B, driven by strength across Handsets, IoT, and Automotive. Combined QCT Automotive and IoT revenues are up 23% YoY to $2.7B Automotive: We continue to deliver strong results for Qualcomm. Quarterly automotive revenues are up 21% YoY to $984M, with strong momentum on all fronts - #SnapdragonRide ADAS/AD Platform with 20+ global OEM programs —most launching in the next 18 months.   Also very excited as we get ready for BMW Group upcoming Neue Klasse launch, featuring our jointly developed global ADAS stack on our Snapdragon Ride platforms. More to come at IAA MOBILITY in September   Industrial and Embedded IoT:  IoT revenues are up 24% YoY to $1.681B. We continue to build on our broad technology and product roadmap enabling vertical blueprints for industries including Retail, Home and Life, Mining, Oil and Gas, Energy and Utilities, Public Safety, Robotics and Drones.

393

7 Comments

The Fab Whisperer: The Most Misleading Metric in Fabs Every fab runs on metrics: OEE, cycle time, line yield, on-time delivery, cost per wafer/layer. But there’s one metric that, more often than not, misleads fabs into thinking they’re winning when in reality they’re falling behind: “Utilization”. Why it matters? On the surface, utilization feels like the right goal: keep tools busy, maximize output. But fabs are not simple factories — they’re complex, interconnected systems with bottlenecks and long cycle times. When fabs chase utilization too hard, it usually leads to: WIP bloats — tools run wafers just to look “busy,” flooding the downstream line with lots. Cycle time explosion — bottlenecks starve while non-bottlenecks overproduce. False sense of efficiency — high utilization ≠ high throughput. Technician overload — more setups, more firefighting. The paradox is that the harder a fab pushes utilization, the slower and more expensive it often gets. From the field: I once worked with a fab proudly reporting 95% utilization on bottleneck tools. Management celebrated. But wafer cycle times were climbing, and customers were missing deliveries. Tracing WIP revealed that the implant area was flooding the downstream Litho bottlenecks. Lots piled up in front of scanners, miss-managing critical recipes' priorities and inflating cycle time by weeks. Litho Utilization looked world-class. Actual fab delivery performance was collapsing. When we shifted the focus to throughput at the bottleneck and WIP turns at non-bottlenecks, cycle time improved 18% in just three months, and on-time delivery jumped from 70% to 93%. The fix: Stop chasing utilization as a headline fab metric. Focus on throughput at the bottleneck — the true limiter of output. Use metrics for bottleneck tools such as % Idle with WIP, throughput/tool/day or per shift. Focus on Dynamic Cycle Time (DCT) and WIP turns on non-bottlenecks — they reveal the hidden cost of overproduction. Align fab scheduling with pull principles wherever possible. Focus on FLOW. Teach fab leaders the difference between “busy” and “productive.” The most misleading metric isn’t the one you ignore. It’s the one you celebrate — while it quietly erodes flow. So… drop a comment if in your fab the focus on keeping tools busy trumps delivering wafers on time. What's your most misleading metric? #TheFabWhisperer #Semiconductor #FabOperations #Metrics #Utilization #CycleTime #ManufacturingExcellence

92

10 Comments

If you’ve been waiting for the right moment to begin your journey in the semiconductor industry, this is the perfect time. Together, we can shape a strong and rewarding future in VLSI. 🌍 India’s Journey to Become a Global Chip Powerhouse 🌍 The semiconductor industry is at the heart of every technological revolution - from powering AI and 5G to enabling EVs and smart devices. And now, India is stepping onto the global stage with a bold vision. 🚀 Made-in-India Chips by 2025 By December 2025, India will launch its first fully packaged semiconductor chip - a landmark achievement in our pursuit of self-reliance in technology. This milestone will: ✅ Reduce dependency on costly imports ✅ Create thousands of high-tech jobs ✅ Boost India’s standing in the $600B global semiconductor market Recent Developments Driving the Revolution: 🏭 Gujarat Chip Plant: A ₹7,600 crore OSAT facility in Sanand is set to boost assembly and testing by 2026. 🌐 Rare Earth Edge: With China easing export rules, India is ready to strengthen its rare earth material supply. ⚡ New Projects: The government has approved four major fab projects, unlocking new opportunities for local chip manufacturing. Opportunities for Freshers in VLSI & Semiconductors This revolution is not just for industry veterans — it’s a golden opportunity for young engineers and fresh graduates. With the demand for skilled talent in: 🔹 RTL Design & Verification 🔹 Physical Design 🔹 ASIC & FPGA Development 🔹 EDA Tools & Automation 🔹 Testing, Packaging & Manufacturing Freshers can now step into a high-growth industry that promises: ✅ Global career opportunities ✅ Exposure to cutting-edge technologies (AI, IoT, 5G, EVs) ✅ A chance to contribute directly to India’s self-reliant tech ecosystem With the right training, projects, and mentorship, freshers can secure high-paying roles in semiconductor companies and become part of India’s journey to becoming a Global Chip Powerhouse. The dream of a true “Made in India” semiconductor revolution is on the horizon. And together, we can make India a global hub for advanced electronics. 👉 What are your thoughts on India’s upcoming role in the global semiconductor ecosystem? Do you think fresh graduates are ready to take on this challenge? Demo sessions - https://lnkd.in/g_jj43_V Our website - https://provlogic.com/ Join our WhatsApp channel for daily VLSI Updates - https://lnkd.in/gxBHqS7z Follow us on Instagram for Daily VLSI series - https://lnkd.in/gfziv4dg 👉👉 Please do follow ProV Logic & Prasanthi Chanda for daily Interview series #Semiconductors #ChipDesign #MadeInIndia #VLSI #Technology #Innovation #IndiaChips2025 #FreshersJobs

34

India’s Semiconductor Push: Are We Chasing a Mirage or Leading a Revolution? India’s semiconductor ambitions are undeniably bold, but are they grounded in executional reality? Our ₹76,000 crore incentive scheme (~$10 billion) sounds massive… until you compare it to $52B in the U.S. or up to $65B in China and Europe. Are we truly positioned to compete or already outmatched? Major projects have stumbled: • Adani Group’s $10B fab with Tower? Quietly shelved. • Zoho’s ₹700M withdrawal? Another exit from the race. How many more will follow? India’s projected talent shortfall is concerning: 250,000+ professionals by 2027. Without scale-up, will we continue contributing to foreign fabs rather than building our own? Supply chains are equally fragile: China controls 80% of rare-earth minerals, while India holds under 1%. How do we ensure chip sovereignty without material autonomy? Yes, India can lead in chip design, ATMP, and system-level innovation. But large-scale, high-end fabrication demands deep capital, raw materials, and geopolitical steadiness—challenges we can’t ignore.   #Semiconductors #MakeInIndia #ChipDesign #TechPolicy #IndiaTech #FablessDesign #VLSI #ESDM #ElectronicsManufacturing #SupplyChain #Geopolitics #TechSovereignty #EngineeringIndia #AtmanirbharBharat #ChipWar #SemiconductorStrategy #TalentGap #TechLeadership #DigitalIndia

29

3 Comments