The performance gains from Nvidia's Hopper to Blackwell GPUs come from increased size and power, not efficiency. This signals a potential scaling limit, creating an opportunity for radically new hardware primitives and neural network architectures beyond today's matrix-multiplication-centric models.

The progress in deep learning, from AlexNet's GPU leap to today's massive models, is best understood as a history of scaling compute. This scaling, resulting in a million-fold increase in power, enabled the transition from text to more data-intensive modalities like vision and spatial intelligence.

The progression from early neural networks to today's massive models is fundamentally driven by the exponential increase in available computational power, from the initial move to GPUs to today's million-fold increases in training capacity on a single model.

The plateauing performance-per-watt of GPUs suggests that simply scaling current matrix multiplication-heavy architectures is unsustainable. This hardware limitation may necessitate research into new computational primitives and neural network designs built for large-scale distributed systems, not single devices.

AI labs like Anthropic find that mid-tier models can be trained with reinforcement learning to outperform their largest, most expensive models in just a few months, accelerating the pace of capability improvements.

Top-tier kernels like FlashAttention are co-designed with specific hardware (e.g., H100). This tight coupling makes waiting for future GPUs an impractical strategy. The competitive edge comes from maximizing the performance of available hardware now, even if it means rewriting kernels for each new generation.

Model architecture decisions directly impact inference performance. AI company Zyphra pre-selects target hardware and then chooses model parameters—such as a hidden dimension with many powers of two—to align with how GPUs split up workloads, maximizing efficiency from day one.

AI progress was expected to stall in 2024-2025 due to hardware limitations on pre-training scaling laws. However, breakthroughs in post-training techniques like reasoning and test-time compute provided a new vector for improvement, bridging the gap until next-generation chips like NVIDIA's Blackwell arrived.

Instead of using high-level compilers like Triton, elite programmers design algorithms based on specific hardware properties (e.g., AMD's MI300X). This bottom-up approach ensures the code fully exploits the hardware's strengths, a level of control often lost through abstractions like Triton.