DESIGN
EW: OPEN SOURCE
different approach with its open source LLaMA models. The recent turmoil in the AI industry only underscores this tension. NVIDIA’s stock took a hit after DeepSeek-V2 demonstrated that powerful AI models could be developed with significantly less processing power, disrupting assumptions about who holds the edge in AI. This case proves that open source initiatives have the potential to challenge even the biggest giants, though who will ultimately lead the AI revolution remains to be seen. The question now is whether AI should follow the path of open source software and hardware, enabling broader innovation and scrutiny, or whether closed models will dominate. In a world of rapid technological change, businesses need hardware that can evolve alongside new demands. The traditional model of fixed-function electronics, locked into rigid supply chains, is no longer viable. Open-source hardware offers a solution – allowing companies to build adaptable, upgradeable products that remain relevant rather than becoming obsolete. By removing dependence on proprietary ecosystems, businesses can innovate faster and respond more effectively to market shifts. The electronics industry is reaching an inflection point. Proprietary systems dominate today, and they offer advantages. But the trade-offs are becoming harder to ignore. Companies that double down on closed ecosystems will continue to face higher costs, supply chain risks, and increasing pressure from consumers and regulators. Meanwhile, those that embrace open source design, repairability, and modularity are creating a different kind of value – one that prioritises longevity, adaptability, and innovation at scale. The question is not whether open source hardware will become dominant overnight, but rather how much of the industry will be forced to adopt its principles to stay competitive.
The semiconductor industry is also undergoing a shift toward more open, flexible, and modular hardware. Projected to reach $68.44 billion in revenue in Europe by 2025 (Statista), the industry’s growth is being fuelled by the increasing demand for customisable components that can adapt to evolving technological needs. For years, developers were locked into proprietary CPU architectures like Arm, limiting flexibility and creating dependency on expensive licensing models. But RISC-V, an open source instruction set architecture, is changing that. By eliminating vendor lock-in and allowing companies to design custom chips without licensing fees, RISC-V is driving a wave of innovation across industries – from AI accelerators to embedded computing. With companies like Qualcomm, Western Digital, or Alibaba Cloud integrating RISC-V into their products, open hardware is becoming a core part of the semiconductor industry's evolution. AI is facing the same open vs. closed debate. While OpenAI’s ChatGPT remains proprietary, Meta has taken a
Visit Red Pitaya at embedded world: 4-242
38 ELECTRONICSPECIFIER.COM
Powered by FlippingBook