the research consultancy for energy technologies

Search results for: “semiconductor semiconductors”

  • Groq: AI inference breakthrough?

    Groq: AI inference breakthrough?

    Groq has developed LPUs for AI inference, which are up to 10x faster and 80-90% more energy efficient than todayโ€™s GPUs. This 8-page Groq technology review assesses its patent moat, LPU costs, implications for our AI energy models, and whether Groq could ever dethrone NVIDIAโ€™s GPUs?

    Read more

  • Data-centers: the economics?

    Data-centers: the economics?

    The capex costs of data-centers are typically $10M/MW, with opex costs dominated by maintenance (c40%), electricity (c15-25%), labor, water, G&A and other. A 30MW data-center must generate $100M of revenues for a 10% IRR, while an AI data-center in 2024 may need to charge $5/EFLOP of compute.

    Read more

  • Computer memory: aide memoire?

    Computer memory: aide memoire?

    Three types of computer memory dominate modern information processing: Flash, DRAM and SRAM. This 5-page note simply covers each one, how it works, what it costs, advantages, disadvantages, market sizes and leading companies. AI likely boosts all three, but can more SRAM unlock big efficiency gains?

    Read more

  • Energy and AI: the power and the glory? ย 

    Energy and AI: the power and the glory? ย 

    The power demands of AI will contribute to the largest growth of new generation capacity in history. This 18-page note evaluates the power implications of AI data-centers. Reliability is crucial. Gas demand grows. Annual sales of CCGTs and back-up gensets in the US both rise by 2.5x? This is our most detailed AI report to…

    Read more

  • Bill of materials: electronic devices and data-centers?

    Bill of materials: electronic devices and data-centers?

    Electronic devices are changing the world, from portable electronics to AI data centers. Hence what materials are used in electronic devices, as percentage of mass, and in kg/kW terms? This data-file tabualates the bill of materials, for different devices, across different studies.

    Read more

  • Energy intensity of AI: chomping at the bit?

    Energy intensity of AI: chomping at the bit?

    Rising energy demands of AI are now the biggest uncertainty in all of global energy. To understand why, this 17-page note is an overview of AI computing from first principles, across transistors, DRAM, GPUs and deep learning. GPU efficiency will inevitably increase, but compute increases faster. AI most likely uses 300-2,500 TWH in 2030, with…

    Read more

  • Internet energy consumption: data, models, forecasts?

    Internet energy consumption: data, models, forecasts?

    This data-file forecasts the energy consumption of the internet, rising from 800 TWH in 2022 to 2,000 TWH in 2030 and 3,750 TWH by 2050. The main driver is the energy consumption of AI, plus blockchains, rising traffic, and offset by rising efficiency. Input assumptions to the model can be flexed. Underlying data are from…

    Read more

  • Energy transition market sizing: hydrocarbon, new energies, capital goods and materials?

    Energy transition market sizing: hydrocarbon, new energies, capital goods and materials?

    This data-file contains energy transition market sizing analysis, for hydrocarbons, new energies, capital goods and materials in $bn pa, integrating over 1,000 items of energy transition research and our latest roadmap to net zero. In aggregate, energy, materials and transition-related markets double from $25 trn pa to $50 trn pa.

    Read more

  • Wind and solar capacity additions?

    Wind and solar capacity additions?

    Global wind and solar capacity additions reached 600GW pa (AC-basis) in 2024, which is 2x higher than in 2021 and 10x higher than in 2011. The pace of gross wind and solar capacity additions can rise by a further 5x by 2050, bringing wind and solar to 60% of a greatly expanded global power grid…

    Read more

  • Photovoltaic silicon: the economics?

    Photovoltaic silicon: the economics?

    This model breaks down the costs of photovoltaic silicon, which explains $0.1/W of a $0.3/W solar panel. There is no way silicon producers are making economic returns below $12.5/kg mono-crystalline polysilicon prices. The average kg of PV silicon in a solar panel is also most likely associated with 140kg of direct CO2.

    Read more

Content by Category