NVIDIA: The Impact of Process Technology on Kepler’s Efficiency

So, what makes NVIDIA’s current midrange card, the GK104-based GTX 680 way more powerful than their previous top dog GTX 580 and even AMD’s HD 7970? A close, three year collaboration with foundry partner TSMC developing the 28nm technology and the GPU architecture together, as explained by Joe Greco, Kepler lead engineer in his NVIDIA blog. In fact, he likens it to developing a new jet engine using exotic materials which are still in development. Doesn’t sound easy.

Since power use is the limiting factor of today’s processors, his team concentrated on wringing out as much efficiency from the manufacturing process and those tiny transistors as possible, “Today, the primary constraint on processor performance is the power consumption budget. So our goal is always to develop solutions that deliver the highest performance within a fixed power budget. Having a more efficient process enabled us to add more processing cores, thus increasing performance. Put simply, greater efficiency equals greater performance and optimal performance per watt.”