Jen-Hsun Huang's goal is finally realized with the release of Fermi this week.
Last week Nvidia held an investor meeting where company CEO Jen-Hsun Huang took over the microphone to speak on several interesting points.
One obvious point was the launch and critical reception of the Fermi-based chips, the GeForce GTX 480 and GTX 470. Even casual onlookers will see that the first wave of GF100 GPUs is in short supply. While the first batch of retail cards will be hitting this week, they're still two weeks behind the media launch from March 26. The Nvidia CEO admitted that his company can't take manufacturing advances for granted.
"From a supply perspective, we wish we had more 40-nm capacity," Huang said, according to Venture Beat. "We are working with TSMC really closely. They are doing a fab job. Yields are improving. Capacities are improving. But we are finding it hard to keep up. Everyone is clamoring to have Fermis out the door. We are working really hard to get Fermis out the door."
Nvidia is now taking additional steps to ensure that such a lag will be avoided in the future by dedicating several engineers to manufacturing issues. The company hopes that this will help it ease through the next manufacturing transition that it expects to occur in 18 months.
Even with the problems at the 40nm process, the Nvidia wants to transition its full product line for the sake of higher margins – a positive aspect that comes with process shrinks. The company has a gross profit margin of around 30 percent for 55nm chips, while the 40nm products will result in a 40 percent profit margin.
Despite Nvidia's technical achievements, the GeForce GTX 480 and GTX 470 cards have been met with some less-than-stellar reviews, at least in light of AMD's success with its current ATI GPU offerings. According to Venture Beat, Huang blames some of the criticism on the fact that Nvidia didn't give reviewers enough time to evaluate the product.
Nevertheless, the CEO feels a lot better now that Fermi is out the door. "What’s different now is Fermi is in production," he said. "Two years ago, I was just fantasizing about it."