Samsung Poised to Become World's Largest Semi Supplier in 2Q17

[Daniel Nenni]

Intel could yield the #1 position it has held since 1993.

After nearly a quarter of a century, the semiconductor industry could see a new #1 supplier in 2Q17. If memory market prices continue to hold or increase through 2Q17 and the balance of this year, Samsung could charge into the top spot and displace Intel, which has held the #1 ranking since 1993. Using the mid range sales guidance set by Intel for 2Q17, and a modest, yet typical, 2Q sales increase of 7.5% for Samsung, the South Korean supplier would unseat Intel as the world’s leading semiconductor supplier in 2Q17 (Figure 1). If achieved, this would mark a milestone achievement not only for Samsung, specifically, but for all other competing semiconductor producers who have tried for years to supplant Intel as the world’s largest supplier. In 1Q16, Intel’s sales were 40% greater than Samsung’s, but in just over a year’s time, that lead may be erased and Intel may find itself trailing in quarterly sales.

View attachment 19737

Samsung’s big increase in sales has been driven by an amazing rise in DRAM and NAND flash average selling prices (Figure 2). IC Insights expects that the tremendous gains in DRAM and NAND flash pricing experienced through 2016 and into the first quarter of 2017 will begin to cool in the second half of the year, but there remains solid upside potential to IC Insights’ current forecast of 39% growth for the 2017 DRAM market and 25% growth in the NAND flash market.

View attachment 19738

As shown in Figure 3, Intel has been locked in as the world’s top semiconductor manufacturer since 1993 when it introduced its x486 processor and soon thereafter, its revolutionary Pentium processor, which sent sales of personal computers soaring to new heights.

View attachment 19734

Over the past 24 years, some companies have narrowed the sales gap between themselves and Intel, but never have they surpassed the MPU giant. If memory prices don’t tank in the second half of this year, it’s quite possible that Samsung could displace Intel in full-year semiconductor sales results as well. Presently, both companies are headed for about $60.0 billion in 2017 semiconductor sales.

Report Details: The 20th Anniversary 2017 McClean Report

Further details and rankings of top semiconductor suppliers as well as overall market trends within the IC industry are provided in the 2017 edition of The McClean Report—A Complete Analysis and Forecast of the Integrated Circuit Industry, which is IC Insights’ flagship report covering the IC market. A subscription to The McClean Report includes free monthly updates from March through November (including a 250+ page Mid-Year Update), and free access to subscriber-only webinars throughout the year. An individual-user license to the 2017 edition of The McClean Report is priced at $4,090 and includes an Internet access password. A multi-user worldwide corporate license is available for $7,090.

To review additional information about IC Insights’ new and existing market research reports and services please visit our website: IC Insights | Semiconductor Market Research.

PDF Version of This Bulletin
A PDF version of this Research Bulletin can be downloaded from our website at http://www.icinsights.com/news/bulletins/

 

[Daniel Payne]

It's with some irony that Samsung could pass Intel on the basis of the DRAM segment, because Intel was founded on the DRAM product initially but then exited that business because it couldn't maintain a #1 or #2 position in it. I will never forget working at Intel on DRAM design at a time when the Japanese competitors had just surpassed all technical specs of Intel, because our culture was filled with denial and claims of unfair trade practices when in fact the Japanese had applied the principles taught by W. Edwards Deming on continuous quality improvements.

 

[Arthur Hanson]

Daniel N. How large do you feel the market will be for 3dXpoint memory? Could this move the needle for Intel and Micron? Is there any competition to this technology on the horizon?

 

[Daniel Nenni]

Daniel N. How large do you feel the market will be for 3dXpoint memory? Could this move the needle for Intel and Micron? Is there any competition to this technology on the horizon?

I have no idea, memory is not my thing. It will be epic though if Intel loses both the largest semiconductor company award and the smallest process award in the same year. Times are certainly changing!

 

[Daniel Payne]

Intel lost a lot of credibility when they pre-announced 3Dxpoint memory technology, because what they actually delivered was off the mark in several metrics by 10X.

 

[benb]

I'm skeptical that cyclical changes in the DRAM and NAND market will permanently change the overall semiconductor market positions of Intel and Samsung, because DRAM and NAND prices basically fall 20% a year, on average. Yearly shrinks are required just to stay even.

Here's a chart of a Crucial 4GB DDR4 DRAM module. The price fell in 2015, bottomed out in late 2016, rose from August 2016 through February 2017, and then basically flattened out since then.

Here's a chart of a high-volume Samsung 250GB SSD drive. The price shows the same pattern; falling in 2015 and much of 2016, then bouncing up.

So Samsung has unusual tailwinds providing a boost at the moment.

Intel may be experiencing the opposite. Intel Core i5-6500 prices were stable, slowly declining, with a bit of bigger dip recently. Microprocessor market dynamics are different than commodity DRAM or NAND. It is basically a stable pricing market with improvements in performance or core counts driving this.

Samsung of course has the foundry and LSI business but it's tiny compared to NAND and DRAM. As it grows, it builds a case for replacing Intel as the leader. Because the leader has for decades had the largest exposure to the performance driven, stable-pricing market, the high-performance microprocessor market.

 

[count]

3DXpoint is basically DOA. It doesn't deliver on the promised performance improvements, and its flawed with durability issues. I think it will be regulated to niche markets and eventually fade into obscurity. Reminds me of the RAMBUS debacle.

 

[rhueber]

Xpoint has a few interesting features, the ability to randomly write without having to do a block erase means you can sustain lots of small random writes much better than a NAND-based device. That said, if the endurance is barely better than NAND, it's not clear how useful that is.

 

[benb]

DRAM scaling is nearly finished--two nodes left (1y and 1z). 1z may end up being 12-14nm. It's nail-biting time in DRAM land.

With Xpoint there is this sense of really, this is the future? It has flaws and doesn't really fit as DRAM replacement or NAND replacement. But if it at least can scale, vertically I guess, then it at least has a future.