Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/semiwatch-apple-iphone-x-impact-on-chip-demand.10113/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021770
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

SemiWatch: Apple Iphone X impact on chip demand

Daniel Nenni

Admin
Staff member
Samsung profit is all semi, But Capex down, second order derivative trading- Rubber ceiling, Samsung had yet another fantastic quarter with semiconductor profits far above and beyond other divisions which paled or dropped by comparison. Samsung also passed Intel to become the number one chip maker, which was already a foregone conclusion. Samsung is also not shy about supporting its golden goose by more than doubling spending last year. So much so that they alone counted for one out of every three dollars spent on capital equipment last year.

While not giving numbers, Samsung did say that capital spending would be down this year. The question on everyone's mind is "how far down?".

It is also a foregone conclusion that Iphone X production is down according to everyone in the supply chain. Again the question is how far down as this will clearly impact not just NAND demand, of which the iPhone X is a NAND pig, but also logic which will drive foundry demand, especially at the leading edge.

Lastly, chip equipment stocks have again bounced off of and retested prior high limits that memory worries had created in the end of last year. This time, however, stocks are bouncing off those same limits after reporting very good quarters. This could be an even more ominous sign as compared to just the "fear factor" of slowing memory. When stocks go down on good news.......god help you if you report just in line or heaven forbid a miss!!!

The triple Threat - Samsung , Apple & Selling on an earnings beat

Samsungs Spending

View attachment 21093
Chart from IC Insights

This chart seems to say it all with a picture. We all know the numbers but a picture speaks volumes. If we were to take away the labels and stopped someone on the street and ask if 2017 was likely a sustainable trend I think I know what their answer would be after they stopped laughing. Yet many industry analysts are talking about capex being up again with Samsung leading the charge.

Even if Samsung is down only 10% or 20% ($2.6B or $5.2B) who is going to make up for that? We don't think it will be Intel. While TSMC recently announced a large new fab, the spending is spread out over many years. The problem is that Samsungs spend is so large, one of every three dollars, that it alone makes up the marginal dollar difference that has driven every chip equipment companies upside.

In a worst case scenario if Samsung went back to its prior trendline of about $13B per year it would mean a 50% drop.

A best case scenario would be a drop of only 10% which would still mean that other companies would have to kick up their spend significantly just to keep total capex flat year over year.

We think Samsung will continue to spend biggly on NAND production, may slow DRAM at some point and will keep up the foundry/EUV spending, just clearly not at last years rate. So while we don't have an answer as to how far down Samsung will be, we do know its down which means total capex growth is going to be difficult to come by.

Apple iPhone X and chip demand

We do know that both Apple and Samsung are the two biggest users of semiconductor devices. Apple seems to be driving the leading edge of Moore's law by demanding ever better transistor performance.

Now that the iPhone X seems to be slowing, what will that mean as to Apple's rate of consumption of chips? When you think about the Iphone X, its higher end model uses 256GB of NAND or roughly the size of an average SSD. When I bought an iPhone X I opted for the 256GB with the thought of "in for a penny, in for a pound", if you are dropping a $1000 on a phone just go all the way....

This also means that if iPhone X production drops, so does a lot of demand for NAND. How much is hard to say, I guess we will find out from Apple.

Second order derivative trading and transparent rubber ceilings

We have long been a proponent of second order derivative stock trading. That is that its doesn't matter whether business is improving or falling but rather what the rate of increase or decrease is doing. Whether our quarter over quarter is increasing or decreasing, even if still positive.

Its clear that the second order derivate is indeed negative and the rate of increase has slowed over last years ludicrous rate. This is also one of the main reason's that stocks have sold off in the face of great quarters and up guidance. Basically the up guidance is not as up as it was previously.
We also find it interesting that the stocks which recovered all they lost in the memory scare only to bounce off the same transparent rubber ceiling stock prices seems to be telling us that investors are resistant to ever increasing values we saw last year.

Its unclear how many times we will retest these limits or if these will prove to be peak values but for now the ceiling is very solid.

Micron and Memory
Even though we are worried about capex spend and overall chip demand we still think there is enough room left and Micron stock seems to still be way cheap compared to its peers. Micron obviously had a positive reaction to Samsung's comments about memory pricing and supply/demand balance.

We think that Micron still has a lot of profit in front of it and remains a good stock bet. We think that Samsungs slowing capex is good for memory capacity as it means they are being much more conservative than the bad old days where memory makers keep spending until they drove off a cliff.

OPEC and OMEC

We tend to view the memory market much like another commodity market such as the petroleum market. After many years of stupid behavior, oil producing countries banded together to limit and control supply of oil in order to control pricing and profitability of the ultimate commodity. Perhaps memory makers should get together and form OMEC- the Organization of Memory Exporting Companies. Samsung could play the role of the 800lb gorilla, Saudi Arabia with the rest of memory makers jousting for their place in the industry. The key here is preventing stupid self harming behavior.

Perhaps in a case of Darwinism we have already culled out the stupid memory makers and are left with a gene pool of smarter memory makers. Spending less money on capex is a sign of intelligence and planning to keep supply from getting out of hand and causing harm as it has in the past. While we not say that every memory maker has suddenly gotten religion, we do think on average we are way better off and way more sustainable, thus would not be surprised to see modulation of spend to keep things in balance.

In the long run, more modulated memory spend is also good for equipment makers as it lowers the volatility of cycles.

About Semiconductor Advisors LLC
Semiconductor Advisors is an RIA (a Registered Investment Advisor), specializing in technology companies with particular emphasis on semiconductor and semiconductor equipment companies. We have been covering the space longer and been involved with more transactions than any other financial professional in the space. We provide research, consulting and advisory services on strategic and financial matters to both industry participants as well as investors. We offer expert, intelligent, balanced research and advice. Our opinions are very direct and honest and offer an unbiased view as compared to other sources.
 
Last edited:
Dan, what's your opinion on the shift away from disk drives driving demand for memory and how it relates to the memory market? I see the disk drive disappearing and being replaced by SSDs for power reasons among others. I also feel the build out of data centers and communications infrastructure will continue to be major drivers, especially with 5G coming on. Also the markets for robotics and automation are expanding at a fast clip. The fact that all companies and processes are becoming tech companies should continue to be a major driver. We will have pauses along the way for adjustment, but this is a long term trend and picking up speed. The companies that don't make this transition get left behind and this trend continues. Most of the world is not yet highly automated and they have no choice, but to become tech companies in one fashion or the other. It even surprised me that GM is bringing out an autonomous car with no steering wheel or pedals in 2019 and they usually keep to their time lines. Just a few thoughts and any comments or inputs appreciated.
 
Dan, what's your opinion on the shift away from disk drives driving demand for memory and how it relates to the memory market? I see the disk drive disappearing and being replaced by SSDs for power reasons among others. I also feel the build out of data centers and communications infrastructure will continue to be major drivers, especially with 5G coming on. Also the markets for robotics and automation are expanding at a fast clip. The fact that all companies and processes are becoming tech companies should continue to be a major driver. We will have pauses along the way for adjustment, but this is a long term trend and picking up speed. The companies that don't make this transition get left behind and this trend continues. Most of the world is not yet highly automated and they have no choice, but to become tech companies in one fashion or the other. It even surprised me that GM is bringing out an autonomous car with no steering wheel or pedals in 2019 and they usually keep to their timelines. Just a few thoughts and any comments or inputs appreciated.

I have had an SSD for many years and cannot believe regular disks still exist. It is SO much faster and more reliable.

In my opinion the cloud will continue double digit growth for many years to come. My wife, like so many others, is a digital hoarder. She will go from 16GM storage on her phone to 256GB with the iPhone X and her iCloud storage will increase accordingly. And our devices will continue to generate hoards of data that will be mined for profit by the likes of Google and Amazon making personal security impossible.

I'm impressed with autonomous car advancements and feel it is for the greater good fo the semiconductor industry, absolutely. Making more reliable semiconductors is critical to reverse the trend of disposable mobile phone chips. Unfortunately automotive customers are slow to adopt. My wife and I keep our cars for an average of 10 years so we are very slow adopters. The national average is closer to 6 years. And I do not see fully autonomous cars working with the majority of cars being manually operated. If you think online bullying is bad, wait until you see manual drivers bullying the autonomous ones.
 
What do you feel medical will do to semi demand, especially with the Bezos/Buffet/Dimon team using technology to drive down insane, irrational medical costs?
 
What do you feel medical will do to semi demand, especially with the Bezos/Buffet/Dimon team using technology to drive down insane, irrational medical costs?

I'm a big Jeff Bezos fan for his disruptive business models. Medical is definitely next! I am so sick of (no pun intended) technology lagging doctors and hospitals. Google, Apple, Amazon, Tesla, all for the greater good, absolutely! And semiconductors are a key component of course (no pun intended).

Do you have money in the SOX?
 
HDD may well disappear from laptops/desktops, but replacing these with (invariably smaller) SSDs means people still need somewhere to put their big files (collections of photos and videos), and these will then move to the cloud -- which just then means that the data centres need more HDDs because cost per bit is more important than the highest possible speed, and most of this consumer data is cold.

So as far as the consumer is concerned HDD might be dead, but not as far as the industry is concerned -- they might be slow compared to SSD but they are still a lot cheaper per bit, and there's no sign of the costs converging any time soon.

Of course SSDs will also proliferate in data centres for warm/hot online storage where IOPS matter, but not for semi-archival storage which is where the big volume customer-created data is.
 
HDD may well disappear from laptops/desktops, but replacing these with (invariably smaller) SSDs means people still need somewhere to put their big files (collections of photos and videos), and these will then move to the cloud -- which just then means that the data centres need more HDDs because cost per bit is more important than the highest possible speed, and most of this consumer data is cold.

So as far as the consumer is concerned HDD might be dead, but not as far as the industry is concerned -- they might be slow compared to SSD but they are still a lot cheaper per bit, and there's no sign of the costs converging any time soon.

Of course SSDs will also proliferate in data centres for warm/hot online storage where IOPS matter, but not for semi-archival storage which is where the big volume customer-created data is.

Based on the Samsung CAPEX and China memory efforts I think we will have a memory glut sometime soon so hopefully we can see the more efficient SSD in the cloud. But I agree, traditional disk drives are here to stay.
 
Dan, don't you think with dropping prices the penetration of memory/SSDs will increase dramatically especially with the speed, power savings and lifespan/latency? Also I feel we will go to PAAS(products as a service) for with the rate of progress many things will become uneconomic/outdated if their utilization isn't increased with new ownership models. Older technology may become illegal before it wears out if not fully utilized. Any comments or thoughts on this appreciated. I have no money in the SOX
 
Dan, don't you think with dropping prices the penetration of memory/SSDs will increase dramatically especially with the speed, power savings and lifespan/latency? Also I feel we will go to PAAS(products as a service) for with the rate of progress many things will become uneconomic/outdated if their utilization isn't increased with new ownership models. Older technology may become illegal before it wears out if not fully utilized. Any comments or thoughts on this appreciated. I have no money in the SOX

Yes, absolutely. Traditional hard drives are barbaric in comparison to SSD. My wife and I upgraded our phone memory from 16GB to 256GB in the latest iteration thanks to increased NAND density and lower pricing. With smartphones as a driver this technology/pricing curve will continue, absolutely.
 
If you think online bullying is bad, wait until you see manual drivers bullying the autonomous ones.

I get the point, and I've no doubt there will be a few cases of that UNTIL the bad manual drivers learn that, oh dear, EVERY damn thing they did while driving badly (from their speed to their tactics) is recorded by smart cars, along with pesky details like license plates...

We've already had a few cases of road rage prosecuted based on dash cam or passenger phone footage, and prosecuting manual drivers who go beyond what is legal will soon become routinized, without even a requirement for intervention by human legal professionals. As for not quite illegal, but definitely being a dick, well, all we need is a similarly automatic system that uploads the tell-tale footage to insurance companies.

In the past we were supposed to live as though god were always watching. In the future god WILL always be watching, on the road as elsewhere. Unclear what the consequences (ALL things considered) will be...

As for iPhone X and NAND demand, I think you're missing just how flexible NAND is as a material. Even assuming lower iPhone X demand for the iPhone X 2018 model (something I consider an extremely dubious claim) there are multiple places where Apple could increase their use of NAND and has not yet done so primarily based on cost or technology, not on desire. For example ultimately (maybe this year, maybe in 2020) I expect to see "significant" amounts of flash in AirPods (say at least the 4GB or so of an early iPod nano), allowing for direct "on-Airpod" playback of material, reducing battery consumption. Likewise there's scope for more than the 8GB of flash on an aWatch once Apple allows for more flexible audio content storage on the device, and so a better untethered audio experience if, for whatever reason, you choose to leave your phone at home (this seems to be big with runners and other athletes).

Likewise there are companies working hard (eg Peter Hoddie's new venture, Moddable (@moddabletech) | Twitter ) to add appropriate displays to low-cost IoT type items that currently suck because of their baroque UI's [think eg dash cams, or smart scales]. Once again, lots of scope for small (but many) uses of flash there...
 
Very true, memory is not my market. I can only speak to my personal usage and my wife and I are buying new Dell laptops this weekend (Presidents Day Sale) with 1TB SSDs...


I get the point, and I've no doubt there will be a few cases of that UNTIL the bad manual drivers learn that, oh dear, EVERY damn thing they did while driving badly (from their speed to their tactics) is recorded by smart cars, along with pesky details like license plates...

We've already had a few cases of road rage prosecuted based on dash cam or passenger phone footage, and prosecuting manual drivers who go beyond what is legal will soon become routinized, without even a requirement for intervention by human legal professionals. As for not quite illegal, but definitely being a dick, well, all we need is a similarly automatic system that uploads the tell-tale footage to insurance companies.

In the past we were supposed to live as though god were always watching. In the future god WILL always be watching, on the road as elsewhere. Unclear what the consequences (ALL things considered) will be...

As for iPhone X and NAND demand, I think you're missing just how flexible NAND is as a material. Even assuming lower iPhone X demand for the iPhone X 2018 model (something I consider an extremely dubious claim) there are multiple places where Apple could increase their use of NAND and has not yet done so primarily based on cost or technology, not on desire. For example ultimately (maybe this year, maybe in 2020) I expect to see "significant" amounts of flash in AirPods (say at least the 4GB or so of an early iPod nano), allowing for direct "on-Airpod" playback of material, reducing battery consumption. Likewise there's scope for more than the 8GB of flash on an aWatch once Apple allows for more flexible audio content storage on the device, and so a better untethered audio experience if, for whatever reason, you choose to leave your phone at home (this seems to be big with runners and other athletes).

Likewise there are companies working hard (eg Peter Hoddie's new venture, Moddable (@moddabletech) | Twitter ) to add appropriate displays to low-cost IoT type items that currently suck because of their baroque UI's [think eg dash cams, or smart scales]. Once again, lots of scope for small (but many) uses of flash there...
 
Back
Top