Nvidia Corp.’s Tegra family of system-on-chip (SoC) devices gas not been exactly successful: shipments of actual devices were postponed and the number of design wins is hardly impressive. Nevertheless, Nvidia pins a lot of hopes on the second-generation Tegra SoC that is due to be announced early in January. According to the chip designer, Tegra 2-based devices will emerge already in 2010.

“At CES we are going to make a major announcement about Tegra family. It is highly possible that we will see some very interesting form-factors coming out at the same time. [There will be products] shown by our partners using the next-generation Tegra device. You are going to see roll-outs and deployments of tablet PCs, smartbooks, netbooks, MIDs throughout the first half [of the year]; and then you will see major roll-outs of smartphones in the second half,” said Michael Hara, senior vice president of investor relations and communications at Nvidia, at Barclays Technology Conference last week.

Few details are known about Tegra 2 at the moment, but at various events Nvidia’s high-ranking executives said that the next-gen Tegra will be at least two times more powerful than the predecessor. In particular, the SoC will feature a dual-core ARM general-purpose processor as well as more powerful graphics technology inside. Perhaps, Tegra 2 will also use faster memory and will have other improvements. Michael Hara said that Nvidia Tegra 2 will be made using 40nm low-power process technology.

Nvidia claims that with the next-generation Tegra its SoC sales will become considerably larger and once smartphones powered by Tegra are released, this will have even more significant impact in the second half of 2010. Nvidia hopes that by the end of its fiscal 2011, Tegra revenues per quarter will hit $100 million.

Although Nvidia pins a lot of hopes onto its Tegra business, it should be noted that the company has virtually no experience in non-branded businesses. Nobody cares what is inside a new smartphone and cell phone manufacturers do not advertise components they use, but concentrate on feature-set. Without Nvidia Tegra logotype on the box of Microsoft Zune HD hardly a lot of consumers will ever learn what chip powers their personal digital media player. Nvidia hopes that its approach of crating compute-intensive chip for handhelds will enable experience that will differentiate appropriate devices for Nvidia.

“We want to deliver a Web computing experience that is better than on your PC. Today, if you take your PC and go to the Internet, you want to see high-definition in everything. You want to have fast response times and switching between your windows, you want to see high-definition videos, you want to see high-definition images, so, your experience is about HD Internet. Our objective with Tegra is to deliver the same experience to your handheld devices,” explained the senior vice president of investor relations and communications at Nvidia.

At present companies like Texas Instruments, Qualcomm and others tend to integrate all the functions of a mobile device – including computing and communication functions – into the same piece of silicon in order to make them slimmer and more affordable. However, Nvidia believes in “compute intensive” Tegra SoCs.

“The main difference between what we are developing within the Tegra architecture compared to our competition is that we are building a computer on a chip. The baseband guys take a different approach. They talk about integration. It is ironic is that they talk about their advantage is the same exact argument we have in the PC space regards to Intel. The question what you have to ask yourself is based on what users want, what users demand, ‘is it a time to integrate technology or is it a time to innovate and make things better?’ Our premise is […] that this is a wrong time to be integrated and the right time to be compute-intensive,” said Mr. Hara.

Tags: Nvidia, Tegra, 40nm