14 DAY TRIAL // The graphics solution manufacturer Nvidia has been through a rough period lately, with lawsuits filed against it on the earlier reported card failures that affected notebook users. Also, the growing competition with AMD/ATI made Nvidia cut the prices for its products, which resulted in lower quarterly revenue. Even so, the company is still devoted to developing quality products and believes that its graphics boards will continue to be used by many customers, said Jen-Hsun Huang, president and CEO of Nvidia, in an interview with Digitimes.
Jen-Hsun Huang admitted that the faulty chips brought damages to the company and said that the failure causes are under investigation, while reassuring that only a small number of notebook models were affected by the problem. According to him, Nvidia has never tried to minimize its responsibility for the situation and has a conscious attitude towards finding a solution as to also prevent something similar. “Nvidia is a responsible company and we are willing to spend US$200 to solve a problem caused in a GPU worth US$20. Currently, we are negotiating with partners to recall products and hope to give consumers a satisfactory resolution,” he said.
When asked about a possible transition to the 40nm fabrication process, Huang said that the company would adopt it when the “appropriate time” comes. The main focus is on maturity and yield rates, and the transition to the latest process nodes is carefully looked into.
The GPU market is considered by some as being a mature one. As Huang stated, graphics processors will expand into other industries. Nvidia presented at Nvision the possibilities GPUs are able to bring to the industry. He also said that Nvidia partners with software makers like Adobe, Microsoft and Apple to help bring professional image production to the mainstream.
When asked about the integration of the graphics processors into CPUs, Huang said, “I don't really care if they put GPUs into chipsets or CPUs. Nvidia's GPUs still own advantages in design and performance, so even after Intel and AMD launch combined CPU and GPU products, vendors will still choose Nvidia's IGP chipsets and GPUs in order to fulfill market demand. Nvidia will devote our full strength to make sure our GPUs remain in high demand in the market”.
In Huang's opinion, GPUs will remain unchanged, at least in the near future. Video cards will continue to perform graphics calculations and will not compete against CPUs. “However, as to what different technologies can be integrated into future GPUs, that is top secret right now, consumers will find out more when the time comes,” he added.
According to him, IGP chipsets can easily perform tasks like game playing and image decoding, and all three major industry players are able to provide solutions for these tasks. Nvidia has an advantage over competition, Huang says, and that is the CUDA technology, which is able to provide 10%-100% more performance in the area of parallel calculations, while the PhysX technology can make games look more realistic. His belief is that Nvidia is able to deliver technologies unmatched by competition.
"High-end graphics technology is not something you can achieve by just stepping through the door" “I understand that Nvidia doesn't have the advantage of having our own platform, but this also means that the technological capabilities and performance of our GPUs must surpass that of our competitors. This is why we have invested heavily in the development of CUDA and PhysX,” he added. The fact that CPUs will come with integrated GPUs will not affect Nvidia's market share. The most important thing is for Nvidia to have strong graphics technologies.
“As long as our GPUs lead in technology and performance, our competitors' integrated GPU solutions will just be a waste of resources and people will still continue to adopt Nvidia IGP chipsets and GPUs,” Huang said. “I really don't worry about the IGP chipset market, since Intel's own IGP chipsets account for 99% of its platform and we have a less than 1% share of the market. We are performing stronger on the AMD platform side, but since the scale of this market is much smaller than the Intel one, once Intel and AMD start to integrate GPUs into their CPUs there won't be any major affect on our market. What our competitors do is not the problem; the performance we deliver with our GPUs is the key factor.”
As for the growing notebook market, Huang said that Nvidia does not have a strong share at this moment, as it invested mainly in AMD platforms and less in Intel ones. The company hopes to raise its share in the area at over 50% courtesy of continuous improvements in performance. Yet, it is rather skeptical whether the dual-core GPUs will prove to be a better solution than single-cores.
Also, Nvidia does not plan to turn to x86 CPU design as Intel did with Larrabee. “The most important thing to remember is that Intel is the leader in the CPU market and so it's better that Nvidia focuses on what we do best. To cut in to the x86 CPU market would just be a waste of time and resources,” Huang revealed. “However, I would like to comment on Intel's push to using x86 architecture for GPUs. High-end graphics technology is not something you can achieve by just stepping through the door. Intel's hope of changing the whole GPU industry is not going to be something easily achieved.”
Huang said that the graphics chip manufacturer does not want to be involved in other projects than the current ones, at least for the time being: “Nvidia's GPU journey is not yet complete. I believe the GPU market can continue to grow for another 15 years. Currently, I only want to focus on our current product lines”. He revealed that “I expect the mobile device market to have large business potential in the future. Currently, consumers have at least one handset and with Internet applications continuing to expand, multifunction mobile devices which integrate telecommunication capabilities will become standard equipment in the future”.
In Huang's opinion, mobile devices have a great advantage in the fact that “they offer PC functionality and yet are easy to carry”. He said that “the reason Nvidia adopted the ARM architecture [for the Tegra CPU] is that it fulfills the basic requirements of small size and low power. x86 architecture designs require devices to be much larger and heavier. For this reason x86 will never become the mainstream in this market”.
Huang concluded that “I personally like GPUs better [than CPUs], not only because they bring new visual pleasure to PC products, they are also the key to Nvidia's current success. I don't want to see GPUs become a sub-product line. However, this is just my personal opinion, not the official stance of the company. As CEO, of course, I hope mobile device CPU revenues can pick up quickly and reach their maximum potential. All Nvidia's product lines are the result of each team's hard work and I cannot comment on which will be our major product line in the future since both still have potential”.