NVIDIA deployed core ERP, then took aim at its extended supply chain
Staff -- Manufacturing Business Technology, 8/1/2004 12:00:00 AM
Joe Sura knows about the value of ERP software, and the challenges of implementing it quickly. Back in 2000, Sura—VP of Information Technology with high-tech manufacturer NVIDIA—helped deploy a Tier 1 ERP system in just over five months.
But Sura also knows that deployment was only the first phase of achieving what the Santa Clara, Calif.-based company wanted from its enterprise software. NVIDIA supplies graphics and communications microprocessors, and relies on partners to manufacture its products, which end up in PCs, military navigation systems, and video game consoles.
NVIDIA's main objective is to use ERP and extended enterprise applications to adapt quickly to changing supply and demand information. The journey began in 1999, when the company replaced an outdated, Tier 3 enterprise system. "We were on a high-growth trajectory, and we had clearly outgrown our information systems," says Sura. "We needed to get onto a highly scalable, Tier 1 ERP system as quickly as possible".
NVIDIA evaluated leading Tier 1 systems, and chose SAP's suite. The initial five-and-half-month project implemented 13 "core" modules in areas including accounting, sales and distribution, order management, inventory control, and human resources.
The rapid implementation brought immediate improvement to transaction processing and enabled NVIDIA to close its books faster and more reliably, says Sura. After digesting the initial phase, NVIDIA launched into a larger project it called "SAP II." The objective, says Sura, was to move the company from a "basic state" of SAP use to an advanced state in one fell swoop.
Phase II added actual costing, foreign trade compliance, better reporting, advanced planning software, and other improvements. "We were trying to tie together sales, finance, and operations so that all groups were part of the same processes," says Sura. "We wanted to abandon all the Excel spreadsheets that were previously used."
A big component of SAP II was the deployment of SAP's Advanced Planner and Optimizer (APO) planning engine, and modules handling production planning and forecasting. NVIDIA went live with APO for its board business several months ago, and is planning to go live for its chip business in early summer.
Something the company continually works on is near-total visibility of work-in-process (WIP) and inventory in its outsourced supply chain. Says Sura, "We need to be able to quickly react to changes in our business, and to do that, we need to know where all our material is, day by day, shift by shift."
Achieving this visibility is no small task, says Sura, because of the number of partners NVIDIA has—including foundries that make chips, assembly and final test facilities, and several board manufacturing partners and inventory hubs. "We need total visibility into the movement of our material in our virtual supply chain," says Sura. "It's all about velocity, volume, and our ability to react to rapidly changing supply-and-demand situations. We track lots through the supply chain as if they were under our own roof."
While Sura says SAP's software handles many planning and forecasting functions well, NVIDIA has made some customizations to SAP's software to make it better suited to managing outsourced manufacturing. NVIDIA has gotten SAP to commit to adding this functionality as part of the standard software.
Ultimately, says Sura, NVIDIA would like to have at least 98-percent raw data accuracy when it comes to data from suppliers and partners. "We are at 97-percent-plus raw data accuracy today, and getting the last percent or two is very, very difficult," he says.
With increased visibility, and an integrated planning and transaction system, Sura says NVIDIA is close to its ultimate goals.
"We've successfully expanded the footprint of our systems to track where materials are in our supply chain, helping us make more conscious, intelligent decisions," says Sura. "It's all about very quick reaction to market conditions and to potential margin falloffs. We want to make as many high-margin products as early as possible, and if we can react quickly enough, we can change production over to achieve that."
