Hardware Development Trends

China just bent the rules of electronics — literally. Facinating? Chinese and global researchers are advancing Metal-Polymer Conductors (MPCs) — circuits made from liquid metals like gallium–indium embedded in elastic polymers — that defy traditional rigid wiring by remaining conductive even when stretched up to 500% or more. Why this is a big deal: 🔹 High Stretchability: Certain liquid-metal conductors maintain electrical conductivity even when stretched 5× their original length. 🔹 Durability: Printable metal-polymer conductors can withstand over 10,000 cycles of stretching with minimal resistance change (<3%). 🔹 Conductivity: Hybrid conductors based on indium alloys can achieve extremely high conductivity (~2.98 × 10⁶ S/m) with minimal resistance change under extreme strain. 🔹 Fine Feature Sizes: Advanced techniques can pattern circuits as small as 5 micrometers, rivaling conventional PCBs. Market Insight: The global market for wearable and flexible devices is expected to surge into the hundreds of billions of dollars, with advanced stretchable materials at the core of the next wave of innovation. (Wearable tech projected >US$150B by 2026 in soft electronics growth — wearable industry data) Where AI Fits In: AI is not just hype — it’s accelerating how we design and discover materials like MPCs. AI/ML models help predict material properties — like conductivity and mechanical resilience — before physical prototypes are made. Computational simulations can evaluate thousands of polymer + metal combinations far faster than physical testing alone. AI-assisted optimization reduces lab iterations, cutting time and cost in early-stage development. In other words: AI + materials science = faster discovery of smarter, stretchable electronics. Potential Applications: Soft robotics that mimic human motion Wearables that feel like fabric Artificial skin with embedded sensing Health monitoring devices that conform to the body On-skin motion recognition and bioelectronics. The era of electronics you can twist, stretch, and wear is here — and AI is helping make it a reality. #FlexibleElectronics #MaterialsScience #AIinInnovation #SoftRobotics #WearableTech #DeepTech #FutureOfElectronics #Innovation

Don't reduce the carbon footprint of your products without understanding all the possible trade-offs. You could end up increasing your environmental impact instead. Here are 3 things to consider when designing sustainable sound experiences: ⚠ Lowest footprint ≠ Winning concept Successful circular products don’t have the lowest environmental burden by default. Modularity is considered a circular design practice, but it also contributes to increased carbon footprint and depletion of materials (mostly gold, beryllium, and neodymium). A modular product containing electronics has roughly 10% higher impact for both GWP and ADP. 🛠 UX plays a core role as much as CMF and ID Functionalities and usability have their footprint: removing a battery from earpods charging case and using the smartphone battery instead decrease hardware volume and materials footprint (-25%) . The same works for magnets: fashionable to have an earpod snapping to the charging case, until you realize that 1/3 of the overall material impact is due to neodymium. 🔄 Trade-offs are inevitable It is better to design for one core circular principle than having a concept that mediocrely covers all of them. A concept can successfully be repairable and fit a circular ecosystem, but it will hardly be repairable, modular, recyclable, refurbishable, low-carbon, low-resource, long-lasting, energy-efficient, biodegradable, compostable and fit a circular ecosystem. Sustainable design isn’t about ticking every box. It’s about making informed choices that truly minimize impact. ➡What’s your take? Which design principle would you prioritize for a truly circular product? Drop your thoughts below and let’s discuss! #sustainabledesign

Operating our data centers more sustainably means being thoughtful about every step - including the materials we use for things like circuit boards and hardware devices. Copper is one of those essential materials, and now there’s a way to source it that supports our goal of The Climate Pledge. Amazon Web Services (AWS) is the first buyer of copper produced from Rio Tinto's innovative Nuton technology. It's a breakthrough process that uses microorganisms - or "bioleaching" - to extract copper from sulfide ores (which are traditionally hard to process and often become waste). Why does that make a difference?   It removes the need for traditional concentrators, smelters, and refineries. The process uses up to 80% less water usage than traditional mining methods. It also has a carbon footprint well below the global average. It significantly shortens the mine-to-market supply chain.   This innovation is another example that solutions exist, and forward momentum continues. Amazon is working across our entire value chain - from steel and concrete to copper - to source materials differently, and I'm thrilled to see AWS leading the industry in the right direction! Learn more about our work on copper in this The Wall Street Journal article by Ryan Dezember: https://lnkd.in/g9AgshDn  

What happens to Google’s hardware when its 'first life' in the data center is over? ⚙️ A decade ago, we began imagining a system that allows our decommissioned servers to get a second life. Today, that vision is a global reality: In 2024 alone, we successfully recovered 8.8 million components from our data centers, including over 3 million hard drives. Through reusing, repairing, or recycling hardware, we can reduce material costs and associated carbon emissions for data centers. We've learned a lot along the way, and we're proud to share our insights in a new report. Check out our "Bridging the Gap" analysis and share it with colleagues who are working to advance operational circularity: goo.gle/3O8dlIG

Using evolutionary programming with OpenEvolve (my open-source implementation of DeepMind's AlphaEvolve), I successfully optimized Metal kernels for transformer attention on Apple Silicon, achieving 12.5% average performance improvements with 106% peak speedup on specific workloads. What makes this particularly exciting: 🔬 No human expert provided GPU programming knowledge - the system autonomously discovered hardware-specific optimizations including perfect SIMD vectorization for Apple Silicon and novel algorithmic improvements like two-pass online softmax 📊 Comprehensive evaluation across 20 diverse inference scenarios showed workload-dependent performance with significant gains on dialogue tasks (+46.6%) and extreme-length generation (+73.9%), though some regressions on code generation (-16.5%) ⚡ The system discovered genuinely novel optimizations: 8-element vector operations that perfectly match Apple Silicon's capabilities, memory access patterns optimized for Qwen3's 40:8 grouped query attention structure, and algorithmic innovations that reduce memory bandwidth requirements 🎯 This demonstrates that evolutionary code optimization can compete with expert human engineering, automatically discovering hardware-specific optimizations that would require deep expertise in GPU architecture, Metal programming, and attention algorithms The broader implications are significant. As hardware architectures evolve rapidly (new GPU designs, specialized AI chips), automated optimization becomes invaluable for discovering optimizations that would be extremely difficult to find manually. This work establishes evolutionary programming as a viable approach for automated GPU kernel discovery with potential applications across performance-critical computational domains. All code, benchmarks, and evolved kernels are open source and available for the community to build upon. The technical write-up with complete methodology and results is published on Hugging Face. The intersection of evolutionary algorithms and systems optimization is just getting started. Links in first comment 👇 #AI #MachineLearning #GPUOptimization #PerformanceEngineering #OpenSource #EvolutionaryAlgorithms #AppleSilicon #TransformerOptimization #AutomatedProgramming

AI performance is won between the specs. A GPU may advertise enormous compute, but real-world AI speed depends on how effectively the entire system moves data, uses memory, serves tokens, and distributes work across hardware. The hardware layer begins with CUDA cores for parallel computation, Tensor Cores for matrix operations, and HBM for feeding data to the GPU. In many inference workloads, memory bandwidth—not peak compute—becomes the actual constraint. As systems scale, communication matters just as much. ↳ NVLink connects GPUs directly ↳ InfiniBand links GPU servers into high-speed clusters ↳ Data movement determines whether expensive compute stays productive Measurement reveals what specification sheets cannot. FLOPS describe theoretical performance. MFU shows how much of that compute a training run actually uses. Compute-bound versus memory-bound analysis identifies whether the workload is waiting for calculations or data. For inference, users experience two metrics immediately: ↳ Time to First Token ↳ Tokens per Second Precision techniques such as mixed precision, quantization, FP8, and FP4 help models use less memory and achieve higher throughput. Then come the techniques that make inference more efficient: ↳ KV Cache avoids recomputing previous tokens ↳ Continuous batching keeps GPUs productive ↳ Flash Attention reduces slow memory operations ↳ PagedAttention improves KV-cache allocation ↳ Speculative decoding accelerates generation ↳ Disaggregated serving scales prefill and decode separately At larger scale, data, tensor, pipeline, and expert parallelism distribute models and workloads across multiple GPUs. The real lesson is simple: AI compute is not one GPU, one benchmark, or one performance number. It is a complete system of compute, memory, networking, precision, inference, and parallelism. Understanding these 25 terms helps teams make smarter decisions about speed, scalability, infrastructure, and cost.

Quick sustainability win of the week: Start tracking peripheral purchases. You’d be amazed how few organisations do this! We've just wrapped up a review across five large orgs (each with 25,000+ employees). Every single one had the same approach with new starters: onboarding kits were given by default, including a keyboard, power blocks, mouse, headset, docking station, cables, bag, plus sometimes even phone cases. And in every case, 50 to 70% of that kit went unused. Straight into drawers, or binned after a year and straight to landfill. Often because the gear was cheap or the user already had better. There was nearly always also a constant churn of replacement accessories being ordered via internal "shops" with very little oversight. New chargers, random adapters, yet another headset. One organisation was spending over $5 million a year on peripherals alone. That’s $5 million in Scope 3 emissions and plastic waste that is totally invisible, unmanaged, and unnoticed. This isn't procurements fault, they are only following a plan, it’s actually more of a cultural and process issue. TBH, if we’re actually serious about doing something positive with sustainability, this kind of waste has to go. I'd personally recommend a simple approach like: 1) Ditch the onboarding kits, just ask what people actually need. 2) Track peripherals separately from core assets. 3) Introduce a reuse-before-rebuy policy (refurb stuff is awesome). 4) Audit what’s in stock before raising a new PO. Small fix. Big impact. Less plastic, less carbon, less water usage, more $$$$ saved. 😃

AIRA: SCALING HEAT PUMPS RESPONSIBLY - A Call for Sustainable Practices The news surrounding Aira, the Swedish company aiming to become a leading heat pump installer, has sparked considerable discussion. While their ambition to rapidly scale heat pump adoption across Europe is commendable a questions arise about the sustainability of their chosen installation methods. It's no secret that manufacturing processes carry an environmental footprint. In the plumbing and heating industry, this is particularly evident in the production of pipe fittings, often forged in energy-intensive furnaces across Europe. Here in the UK, our industry has a long-standing tradition of working with materials efficiently, a practice deeply ingrained in the training of our skilled plumbers. For decades, UK apprentices have been taught the art of bending copper pipe, specifically the commonly used R250 (Table X). This allows for pipework configurations such as 90-degree bends, offsets, and passovers. Achieved using hand-held benders. Alongside this practical skill, environmental awareness and material conservation are core tenets of their training. This brings me to a critical point: why is AIRA seemingly bypassing this established and sustainable practice by exclusively relying on fittings? Reports suggest their training focuses on rapid installation, potentially at the expense of teaching pipe bending skills. While speed is undoubtedly a factor in scaling, the long-term environmental implications of this approach cannot be ignored. As AIRA aims for widespread adoption across Europe, the sheer volume of fittings they will require is staggering. This translates to a significant and potentially unnecessary carbon footprint from the increased activity in those very forges we mentioned. The plumbing and heating industry has a proud history of self-regulation and a willingness to call out unsustainable practices, a tradition dating back to the medieval Guilds. It's in this spirit that I urge AIRA to reconsider their approach and embrace the established, environmentally conscious methods prevalent in markets like the UK. Our European counterparts' ambition to scale heat pump installations is laudable. However, true sustainability lies not just in the end product, but also in the processes used to achieve it. By integrating pipe bending into their training and practices, AIRA can not only reduce their environmental impact but also cultivate a workforce equipped with valuable, time-honored skills. Let's work together to forge a future where the growth of green technologies is underpinned by genuinely sustainable practices. AIRA has the potential to be a true leader in this transition; listening and adapting to established best practices will be key to realising that potential responsibly. Michael Costain Guy Newey Dr Matthew Aylott Madeleine Gabriel Joe Dart #sustainability #heatpumps #plumbing #environmentalawareness #greenenergy #UK #Europe #skills

Sustainability isn’t a coat of paint. It’s part of the blueprint. In digital health transformation, “green” has moved from a nice-to-have to a core part of responsible change. And lately, it’s a recurring topic in many meeting rooms. Ignoring sustainability in transformation isn’t just bad for the planet, it exposes organizations to rising energy costs, regulatory penalties, and reputational risk. Every transformation decision, from strategy to procurement, deployment to retirement, carries an environmental footprint. Treating sustainability as an afterthought leads to waste: 🔸 Systems overbuilt for prestige rather than need 🔸 Infrastructure running far below capacity 🔸 Devices replaced on schedule, not condition I’ve seen entire racks of perfectly good hardware decommissioned, not because they failed, but because refresh cycles didn’t account for reuse or repurposing. It’s a reminder that sustainability isn’t always obvious at first glance. In one study comparing two T-shirts: 🔹 The one labelled as “sustainably produced” wore out quickly, requiring multiple replacements. 🔹 The other, not marketed as green, lasted far longer, and over its full lifecycle, had a smaller environmental footprint. Digital transformation works the same way. True sustainability comes from durability, efficiency, and total lifecycle impact, not just how “green” it looks at launch. Embedding sustainability means building it into every phase of transformation: 1️⃣ Strategy & design Set sustainability goals alongside clinical and operational goals.  Select cloud providers with renewable energy commitments. 2️⃣ Build & deploy Use modular architectures to extend system life.  Prioritize energy-efficient code, devices, and configurations. 3️⃣ Operate & maintain Monitor resource usage, consolidate storage, and optimize workloads for off-peak energy demand. 4️⃣ Retire & replace Plan for secure decommissioning, refurbishment, and recycling from the outset. Before approving your next transformation initiative, run it through the "Green Lens": ✅ Can we meet the need with fewer resources? ✅ Can this run on renewable-powered infrastructure? ✅ Can we extend the life of what we already have? If the answer is “no” across the board, you don’t have a sustainable transformation plan. If you’re leading digital transformation today, are you building it for the next launch… or the next generation? 💡This post is part of 'Rethinking Digital Health Innovation' (RDHI), empowering professionals to transform digital health beyond IT and AI myths. 💡The ongoing series and additional resources are available at www•enabler•xyz 💡Repost if this message resonates with you!