Seeking Alpha
Long/short equity, research analyst, tech, alternative energy
Profile| Send Message| ()  

Intel (INTC), which is the world's biggest semiconductor company, is trying to enter the mobile devices market in a big way. The company has made serious investments in R&D and equipments to become a large supplier of chips to the massively growing tablet and smartphone market. Intel has an 85% plus share in the PC and server processor market due to its strong semiconductor manufacturing and technological expertise. However, the PC market is expected to decline by more than 10% this year, making it imperative for Intel to look at other adjacent markets. Intel is already producing mobile processors using its 5-year-old Atom architecture on some of its older process nodes (32 nm). These mobile chips have managed some design wins with Asian companies such as Lenovo, Samsung (SSNLF.PK) and Asustek. However, these chips are not as good as competitor chips made by companies such as Mediatek and Qualcomm (QCOM). Intel also lacks the communication expertise of its rivals in the 4G LTE space which is a handicap in selling these processors. Intel will unleash a new version of its Atom architecture in late 2013. Silvermont will enable Intel to produce mobile processors at the world's leading edge 22 nm process node. This will allow Intel to pack transistors more densely into existing silicon wafers which will improve performance and reduce costs. These chips will have a big performance and cost advantage over rivals. A recent Antutu benchmark test gives the first suggestion that Intel will start dominating the tablet in the next year. I remain positive on Intel's stock given its low valuation and launch of products for the mobile market in the next one year.

Intel already leads others in Manufacturing cost and Chip performance

Intel is the only company to make chips on the 22 nm node while others are still producing chips on the 28 nm node. Intel has already started working on the 14 nm node and has a well-defined roadmap for making chips on the 14 nm fab next year. The company is also accelerating the pace of technology development by investing heavily in Dutch equipment maker ASML (ASML) for further node shrinks. While the world's biggest foundry TSMC (TSM) has announced plans to produce chips on the 20 nm node, there are doubts whether it will be able to master the FinFET transistor technology. If TSMC and other foundries fail to reach the 20-22 nm nodes on time, then Intel will have a huge advantage going into 2014.

Where Intel stands now in the Mobile Processor race

I have been highly positive about Intel's prospects in the mobile chip market and think there is a possibility that Intel could gain a monopoly marketshare in this market just like the PC and server market. Intel has got a big manufacturing advantage over mobile processor makers because it has the best factories for making chips in the world. Intel makes silicon using the 22 nm 3D FinFET process, while its rivals are still stuck making chips using the 28nm planar process. Intel has not been able to use its manufacturing advantage in the mobile processor node as it was stuck with a 5-year-old Atom architecture. But that will change when mobile processors using Intel's new architecture start being produced at 22 nm later this year. Intel's current "Clovertrail+" mobile chips are being made at the 32 nm node and they are not doing too badly against rival Nvidia (NVDA) and QCOM chips. Competitors will bear the full brunt of Intel's process manufacturing lead, when the 22nm "Baytrail" and "Merrifield" chips hit commercial production over the next year.

Recent Benchmarking test is terrible news for Intel rivals

Extremetech recently reported results from an Antutu benchmark test that shows a "Baytrail" 1.1 GHz SoC outperform the best 2.1 GHz Qualcomm chip by more than 40%. The results are alarming for its rivals, given that Baytrail was running at half the clock speed of the QCOM chip. What this means is that an equivalent Baytrail chip could possibly outperform the best ARM designed QCOM chip by 3:1 later this year. The results are even more dramatic when compared to the test results of Samsung and Nvidia processors. This is not a surprise given that Qualcomm makes the best mobile SoCs. This is a huge lead and could see Intel capture a huge share of the table SoC market. I am sure that mobile SoC makers using ARM (ARMH) designs will also advance their technology and processing in the next 6 months. But overcoming a 200% lead seems improbable in my view. There is little reason to doubt this report and it seems logical in the light of the earlier Intel claims about Silvermont and Intel's process advantage.

The first benchmarks of Intel's upcoming Bay Trail SoC have appeared online - and it's good news for x86 fans, but terrible news for ARM: Bay Trail-T, clocked at just 1.1GHz, is around 30% faster than Qualcomm's Snapdragon 800 clocked at 2.3GHz, the fastest ARM chip on the market. This data comes from an Antutu benchmark, which shows a device with the code name byt_t_ffrd10 -Bay Trail-T Form Factor Reference Device - racking up the magnificent score of 43,416. This is apparently at 1.1GHz, while running Android 4.2.2 Jelly Bean (it is probably the Android reference tablet that Intel showed off at Computex last month). By comparison, the latest Galaxy S4 with LTE-Advanced support, with the Snapdragon 800 SoC, scores just shy of 30,000. To round out the comparison, last year's Exynos 4 (in the Galaxy S3) scores around 20,000, and Tegra 3 scores around 15,000. In short, the Bay Trail-T SoC is an absolute monster that's 30% faster than the top competitor and three times the speed of Tegra 3.

Source - Extremetech

Where Intel could slip

These tests do not mean that Intel has already won the tablet market, as it will still not execute on its roadmap and win customers. Apple (AAPL), which is the world's biggest tablet maker with a 40% plus global marketshare, uses modified ARM design chips and gets it manufactured through foundries. The company recently signed a 3 year deal with the Taiwanese foundry TSMC to produce its next generation of "A" series processors from 2014 onwards. Intel is trying to get this deal as it would give it a huge boost in mobile chip revenues. However, if Intel can deliver on the performance/watt with Baytrail, then Apple may not have an option but to turn to Intel for its tablet SoCs.

Intel's Merrifield chips, which are targeted towards smartphones, have not been seen in benchmarks yet. Communication capabilities are also more important for smartphones than tablets where QCOM has an edge. The smartphone market is larger than the tablet market, with an estimated 900 million smartphones shipments compared to the 150-200 million tablet shipments in 2013. To become a serious mobile device player, Intel will need to substantially beat the competition through "Merrifield" as well.

Intel's Stock Performance and Valuation

Intel stock is consolidating around the $24 level, after breaking the $20-22 range in which it was stuck for a long time. Intel's valuation remains low with a forward P/E of just ~12x and a dividend yield of 3.75%. The company's stock remains almost ~20% below the average valuation of the broader market. The reason for the low valuation is the decline of the PC market growth. However, as Intel shows traction in the mobile market, the valuation should increase.

Summary

Intel has managed to remain the No.1 semiconductor company for a long time by making smart strategic choices. It has crushed competitors such as Advanced Micro Devices (AMD) in the past through its sheer size and spending power. The company has now brought its technological and manufacturing expertise to focus on the mobile market. The company has spent a tremendous amount of money in building the world's best fabs and used its $10 billion R&D spend to change its product focus towards producing power efficient chips for the mobile market. The results will be apparent over the next year and investors should position themselves to take advantage.

Source: Intel's New Tablet Processor Beats The Best ARM Chip By A Huge Margin