Intel (INTC) is trying to make a mountain out of a mole hill – trying to place the company's ill-conceived Atom chip in every conceivable market to validate its existence to shareholders. Now according to Bloomberg, Anand Chandrasekher, who lead the chipmaker’s faltering attempt to get into the mobile-phone market, has resigned from the company.
Atom got off to a rocky start with Intel. The company correctly positioned the processor in the nascent netbook market. I noted back in January 8, 2009, in a Seeking Alpha article that Intel lost as much a $1 billion misjudging the success of the netbook and switching production from more lucrative notebook chips like the Penryn to make up for a shortfall in Atom chips.
Much of the ballyhoo in the netbook debate over market growth and Atom versus ARM Holdings (ARMH) processors is history. The netbook market is dead. Out of the ashes came the media tablet. The debate between Atom and ARM has now merely been moved to a different platform.
Atom held a dominate market position for the netbook because it could compute intensive tasks such as MS Office functions using Microsoft Windows software. ARM was delegated primarily to Internet surfing and apps. With the netbook market dead, the problem for Intel is that consumers are not purchasing the media tablet as a lower cost, more portable netbook to tote around. They are buying it for surfing the web and Internet apps.
Now comes a major blow to Intel’s dreams of its Atom – Windows 8 will run on ARM. While Windows CE supported ARM devices, it offered lackluster performance. Microsoft’s (MSFT) Windows 8 move is likely due to the market momentum of Google’s (GOOG) Android. So, for any media tablet owner who does want to do computer intensive tasks, Windows 8 on ARM offers a viable opportunity.
The resignation of Anand Chandrasekher is due to Intel’s belief that it will compete successfully against ARM in the smartphone market. The problem is ARM owns the Mobile Internet Device space. They own 95% of the mobile phone market and 85% of the smartphone market in unit shipments. ARM processors are being manufactured in the best semiconductor facilities. Companies that are currently or formerly ARM licensees include Alcatel (ALU), Atmel (ATML), Broadcom (BRCM), Cirrus Logic (CRUS), Digital Equipment Corporation, Freescale, Intel (through DEC), LG Group, Marvell Technology Group (MRVL), NEC, NVIDIA (NVDA), NXP (previously Philips), Oki, Qualcomm (QCOM), Samsung (SSNLF.PK), Sharp (SHCAY.PK), ST Microelectronics (STM), Symbios Logic, Texas Instruments (TXN), TSMC, VLSI Technology, Yamaha (YAMCY.PK) and ZiiLABS.
Intel continues to pour money into the Atom. Intel announced last month it will spend $5 billion on a new fab in Oregon, to make 14-nanometer chips – smaller chips to compete against ARM chips already smaller and already dominating the same market Intel is going after.
Now Intel wants to stick its Atom in computer servers. ARM rolled out its A15 core in 2010 with enhanced memory support for servers and networking gear. To me it makes sense, because high- performance servers require increasingly power-hungry processors – namely Intel Xeon processors. Even though these operate at 20 watts, the Atom-based server will operate at under 10 watts. That’s great if you are a data center. But what does that do for Intel. An ARM-based server from startup Calxeda is expected to operate at only 5 watts. More importantly, the price tag of an Atom is well below $100 each while the price of its 20 watt e3-1220l server processor is about $300 each.
I said in an October 2, 2009, Seeking Alpha post, “Intel Facing a Formidable Foe in ARM,” that Intel needs to buy ARM Holdings as the only solution to being a dominant player in the mobile market. Intel keeps skirting the issue. It purchased Infineon for $1.4 billion last year because Infineon’s chips are used in smartphones. Even more interesting, Intel’s acquiring Infineon’s business came four years after selling its ARM business to Marvell (MRVL), another misfire.
It’s not clear what the purchase price of ARM would be, but the $5 billion spent by Intel on a new fab to build chips to compete against ARM would probably do the trick. Intel doesn’t have much time. Once Windows 8 comes out, it will put the company further behind.