Nvidia's chief executive officer is emphatic that his company has a strategy for building processors beyond its mainstay graphics chips.
During an interview with CNET, Jen-Hsun Huang addressed an issue with the company's chips and spoke about ongoing Intel litigation.
On Thursday, Nvidia reported a second-quarter net loss of $141 million, or 25 cents per share, worse than the net loss of $105.3 million, or 19 cents a share, a year earlier. The graphics processing unit (GPU) supplier--whose chips are found in PCs from Hewlett-Packard, Dell, Acer, Sony, and Toshiba--cited muted demand for consumer graphics chips and economic weakness in Europe and China, which drove consumers to lower-priced products. Nvidia products typically are targeted at the upper end of the market.
In the earnings announcement, the company addressed a longstanding issue--and ongoing financial burden--centered on a defect in some of its earlier GPUs and chipsets. The problem was first cited by Nvidia in July 2008 when it announced a charge ranging from $150 million to $200 million to cover costs to repair and replace GPUs and chipsets due to "weak die/packaging material" in older laptop products. "Die/packaging" essentially describes the chip. Nvidia also announced additional charges after July 2008.
On Thursday, Nvidia said it recorded an "additional net charge" of $193.9 million related to the same problem. "The charge includes additional remediation costs as well as the estimated costs of a pending settlement of a class action lawsuit related to this matter," the company said in a statement. Combined with the $282 million of net charges announced previously, the total net charge related to the issue comes to $475.9 million, the company said.
I asked Huang Thursday if he thought the problem was now largely behind it. "It's our best estimate," he said. "With much, much more understanding and more data than we had two years ago or a year ago. And we've had the opportunity to work with every single PC (maker) out there. So we think it's near the end."
I also asked Huang about the company's strategy for central processing units, or CPUs, used in smartphones and tablets. Nvidia has been supplying its first-generation Tegra chip to portable music device makers such Microsoft, which used Tegra in the Zune HD. The second-generation Tegra 2 is targeted at smartphones and tablets but has yet to make an appearance in a product from a first-tier device maker. All Tegra chips are based on a design from United Kingdom-based ARM.
"Our CPU strategy is ARM," Huang said, referring to the fact that Nvidia was, unit last year, only a supplier of GPUs. "ARM is the fastest growing processor architecture in the world today. ARM supports (Google's) Android best. And Android is the fastest growing OS in the world today," Huang said.
Huang said that its dual-core Tegra 2 chips currently come in two flavors, the AP20 for smartphones and the T20 for tablets. "And both of them are being designed into products," Huang said.
Huang also addressed the FTC ruling against Intel and the ongoing court case with its Santa Clara, Calif., neighbor. "The FTC case is about Intel's monopoly behavior and conduct. Our (court) dispute with Intel is related to a pair of contracts--a license of (data) bus rights to us and the cross license to them," he said. "Intel's conduct as a company informs some of their behavior in our case. However, our case is not related to the FTC. Our case is specifically related to a contract dispute."
One of the consequences of the legal wrangling with Intel is that Nvidia can no longer make chipsets--silicon that accompanies all Intel processors--for Intel's new Core i series of processors and future generations of processors. This is a nontrivial issue for Nvidia since it had been the sole supplier of chipsets for the previous generation of Apple MacBooks.
If the case is resolved in Nvidia's favor, would Nvidia consider returning to the chipset business? "They (Intel) have disrupted our chipset business. The damage has been done. We've been out of the chipset business for well over a year, so if this got resolved we're not expecting to ramp back up the thousand engineers that we had working on chipsets."
Intel declined to comment.