Skip to main content

How Intel Mismanaged its Mobile Chip Unit

Forbes reports that financial analysts have always assumed that Intel's cell phone related division has been a money pit. Now the company has proved them correct, dumping the unit to Marvell Technology for the fire-sale price of $600 million.

The deal comes after the company invested nine years and likely more than $3 billion into cracking the billion-unit market for its cell phone division. It ultimately accomplished little more than illustrating how difficult it was for the chip giant to stray from its core business.

In 1997, after an ugly lawsuit, Intel bought the chip unit of Digital Equipment for $700 million, which included a small microprocessor chip that fits inside cell phones and other smaller electronics. Two years later, the company paid another $1.6 billion for an Israeli cell phone-chip company called DSP Communications.

But the company never won big chunks of business from the behemoth cell phone makers like Motorola, Nokia or Samsung. Those companies each ship hundreds of millions of units per year, but they have loyal suppliers like Freescale Semiconductor, Texas Instruments and Qualcomm that were not easy to displace.

But Intel continued to pour money into the business. In a Marvell conference call, it was clear how bloated the Intel division had become. It has 1,400 employees but annual sales of only $400 million. While Marvell officials didn't spell out how much money the unit was losing, they did say the deal would depress their earnings for the next year.

Popular posts from this blog

Why AI Budgets Miss a $9.2 Billion Storage Problem

The worldwide external enterprise storage systems market reached a key point in the first quarter of 2026, and the implications for organizations running Applied-AI initiatives are impossible to ignore. IDC's latest market study reveals that systems grew 22.7 percent year-over-year to $9.2 billion, a dramatic acceleration from the 3.9 percent full-year 2025 growth rate. For enterprise leaders who have spent the past two years prioritizing GPU clusters and server infrastructure while treating storage as a secondary consideration, this data should serve as actionable guidance. The storage bottleneck in AI deployment is no longer theoretical, and the market dynamics now unfolding will directly shape the cost, timeline, and architectural viability of AI programs through at least 2027. Storage Systems Market Development The first quarter of 2026 produced several milestones that demand the attention of any organization with active or planned AI infrastructure investments. All Flash Array...