Rumours or truth?
While we all like the idea of hot lovers, GPU hotspots are a killer. Each generation of chip gets more and more complicated, creating unwanted ‘electrical friction’ – with billions of microscopic components interacting in unpredictable ways to generate current loss and heat. However, give an engineer time and he will find ways to distress the GPU design with the expert touch of a Swedish Masseuse. Will nVidia’s GTX485 be able to woo the undecideds? If you have a production line with one person, and that person works a little harder, then you get a little more product at the end of the day. Pack the line with 480 people, and even a relatively small increase in productivity can deliver clearly measurable results. Now spend time moving the production lines around and installing better air conditioning, then you might find that they produce more and in a nicer, cooler environment. KitGuru has heard that nVidia’s been through the first iteration of the GTX480 with a fine tooth comb, to create a GTX485 product that can run at least 20MHz more on the core clock, with improved cooling and performance. It’s also worked through some neat driver improvements that will also help. Nice. How has this improvement been made possible and why has it taken so long? Bill Dally is a smart cookie. One that spent 10 years at MIT creating massive, experimental, parallel processing systems. He’s used to increasing the smoothness of a machine’s operation by reducing the effort needed to synchronize disparate components. Useful, now that he has Fermi to tame.
So what is the GeForce GTX 485? Simply put, it’s what Fermi should have been at launch.