SAN JOSE, Calif. -- Nvidia Corp. plans to take a charge from $150-to-$200 million to cover anticipated warranty, repair, return, replacement and other costs arising from a ''weak die/packaging material'' in select devices.
At first, Nvidia--which now faces a huge shortfall in sales for the quarter--pointed the finger at one of its foundry partners: Taiwan Semiconductor Manufacturing Co. Ltd. (TSMC). The silicon foundry giant apparently does some level of the IC-packaging process for Nvidia. The ''packaging (material) was supplied by TSMC,'' according to a spokesman for Nvidia, in an e-mail when first explaining the problem.
Then, in another e-mail, Nvidia did an about-face. ''Bottom line: We take responsibilty for this,'' the Nvidia spokesman said. ''We worked closely with TSMC on packaging and the material.''
The suspect material was used in select graphics processors and MCPs in notebook systems. Certain notebook configurations with the suspect devices ''are failing in the field at higher than normal rates,'' according to graphics-chip giant Nvidia (Santa Clara, Calif.).
In subsequent e-mails, Nvidia somewhat backpeddled and changed its original explaination, which blamed TSMC for the problem. ''With regards to TSMC, we are not 'blaming' TSMC,'' the Nvidia spokesman said in the second e-mail. ''Also, to be clear, the material set was co-qualified between [Nvida] and TSMC.''
"This matter concerns complex issues involving multiple parties and all parties are working together to better understand the situation," according to a TSMC spokesman.
The problem could be traced in any number of places. For foundry services, Nvidia uses a number of vendors: TSMC, UMC, Chartered, SMIC, and AMS. For IC-packaging, Nvidia uses a number of subcontractors, such as ASE, Amkor, JSI, KYEC, Siliconware and STATS ChipPAC, according to Nvidia's annual report. And within those vendors, there are a plethora of materials suppliers.
Nvidia, which did not elaborate, will take the charge in the second quarter for the chip snafu. The company will also seek to access insurance coverage for this matter.
''We didn't recall any chips,'' the Nvidia spokesman said in the first e-mail. ''We've replaced the products. We've changed our packaging and we've developed and distributed a software driver to help avoid the failures.''
To date, abnormal failure rates with systems other than certain notebook systems have not been seen.
"Although the failure appears related to the combination of the interaction between the chip material set and system design, we have a responsibility to our customers and will take our part in resolving this problem,'' said Nvidia President and CEO Jen-Hsun Huang, in a statement.
"This has been a challenging experience for us,'' he added. ''As for the present, we have switched production to a more robust die/package material set and are working proactively with our OEM partners to develop system management software that will provide better thermal management to the GPU."
As a result, second quarter revenue and gross margin are expected to be lower than guidance provided during its first quarter financial conference call held May 8.
Total revenue is now estimated to be from $875 million to $950 million. Analysts expected to see sales of $1.1 billion and a profit of $0.34 a share.
''The estimated decrease in revenue and gross margin is due to several reasons: end-market weakness around the world, the delayed ramp of a next generation MCP, and price adjustments of our GPU products to respond to competitive products,'' according to the company.
In any case, the problems at Nvidia opens up the door for AMD, Intel and others in the graphics chip business.