INTC - Intel

[formthirteen]

It's INTC vs. TSMC, AAPL, AMZN, AMD, and NVDA.

 

I don't see how INTC can win other than through financial engineering. They've already lost the hardware-engineering battle. It's "an x86 massacre".

 

The only "good" news this year is Conan's visit to Intel's headquarters:

 

;D

 

https://jamesallworth.medium.com/intels-disruption-is-now-complete-d4fa771f0f2c

https://www.anandtech.com/show/15578/cloud-clash-amazon-graviton2-arm-against-intel-and-amd/9

https://www.phoronix.com/scan.php?page=article&item=amazon-graviton2-benchmarks&num=12

https://www.pcgamer.com/tsmc-confirms-3nm-tech-for-2022-could-enable-epic-80-billion-transistor-gpus/

https://www.tomshardware.com/features/amd-vs-intel-cpus

[fareastwarriors]

Intel falls on report Microsoft plans to design own chips for PCs and servers

 

https://www.cnbc.com/2020/12/18/intel-falls-on-report-microsoft-will-design-own-chips-for-pcs-servers.html

[MattR]

Intel falls on report Microsoft plans to design own chips for PCs and servers

 

https://www.cnbc.com/2020/12/18/intel-falls-on-report-microsoft-will-design-own-chips-for-pcs-servers.html

 

I wonder how TMSC want to keep up with production. They only have a limited amount of manufacturing and there are supply shortages of the AMD Ryzen already due to Apple paying a lot to get all the 5nm goodness that TSMC can provide.

 

I love building PCs and have to build a lot of PCs at work. Intel CPUs are currently too expensive, too hot and have higher power usage. I will buy it under 35$, but not a cent above.

[Jurgis]

I wonder how TMSC want to keep up with production. They only have a limited amount of manufacturing and there are supply shortages of the AMD Ryzen already due to Apple paying a lot to get all the 5nm goodness that TSMC can provide.

 

That's the problem for most near-term bear theses on Intel: AMD simply does not have production to take Intel's market share. It's possible that things will change in couple years, but adding 5nm capacity on the Intel production size is not gonna happen in months.

 

Disclosure: I have some Intel. Longer term the situation is not great.

 

 

[Spekulatius]

I wonder how TMSC want to keep up with production. They only have a limited amount of manufacturing and there are supply shortages of the AMD Ryzen already due to Apple paying a lot to get all the 5nm goodness that TSMC can provide.

 

That's the problem for most near-term bear theses on Intel: AMD simply does not have production to take Intel's market share. It's possible that things will change in couple years, but adding 5nm capacity on the Intel production size is not gonna happen in months.

 

Disclosure: I have some Intel. Longer term the situation is not great.

 

I sold my INTC stock, swapped some of it into SIMO.

 

Ok, I am exaggerating here, but the fact that every Joe Schmoe  - AMD, NVDIA, Apple, Google, Amazon, Alibaba and now Microsoft can design their own silicon and bring it to market using better manufacturing technology than Intel has is not exactly bullish.

 

The WinTel alliance has been in existence for 40 years roughly and the fact that MSFT starts to design their own silicon Tesla’s me that they are scared to fall behind because Intel isn’t cutting it any more.

 

Intel need to do something really quick, or the go they way of IBM. They need a technical wizard on the helm that takes some risk and shakes up their engineering because at that point, the biggest risk is that they keep doing what they have been doing.

[Peregrine]

I just bought a new PC and despite the rave reviews the Ryzen series has gotten, I found it hard to actually find laptops using the latest Ryzen 7 CPU.

[MattR]

I just bought a new PC and despite the rave reviews the Ryzen series has gotten, I found it hard to actually find laptops using the latest Ryzen 7 CPU.

 

Yes because due to the TSMC supply chain shortage. This actually gives Intel about 1.5-2 years until TSMC has enough capacity to also produce enough chips for enough mainstream Laptops. We saw how great the M1 Apple Silicon works and how great AMD Ryzen is. And that all with TSMC. Sadly there seems to be no sign other than a fancy logo -  so that noone knows the brand anymore and some financial shuffling like buybacks instead of proper development and leadership change that I fear the worst for Intel.  They are not on an IBM level, but they seem to get there.

[beerbaron]

I'd like to argue that Intel will get their 7nm process together, they have been doing wafers for 50 years. It's not like they suddenly forgot. They will regain competitiveness in the x86 because well, it's easy for AMD to be ahead because they have a lead on the fabrication side. I'd even argue that Intel on the design side is probably being pressured to be innovative quite a bit more so than before when they had the manufacturing lead, great thing will be seen in 2 years... it's do or die for the design side.

 

On the  supply constrains from TSMC, what we are seeing shows that the fabrication side of Intel might be worth more than expected. It takes a long time to build a plant and major R&D risks are involved. Not a lot of players can enter.

 

My feeling is that in the lab they had a great process to make 10nm and rolled out the process on the production floor. Something went wrong with the process and they have been trying to fix a a bad solution since then. It seems to me they dropped that idea and went back to less risky improvement for the 7nm. Can't really blame them, it's R&D and you can't know what you can't know. You can always screw up even if you do all the right things with the information's you have.

 

I personally do not own shares but I've been eye balling INTC for a while and there is certainly lots of ways to unlock value. Here is what I'd see:

1- Split the fab and design side.

2- On the design side - put x86 in one division and ARM division into another

3- Increase spending on the GPU / NN side to become competitive in that space

 

BeerBaron

[perulv]

I bought INTC back in august at ~$48, and I must say I am becoming more and more skeptical about owning this. But then again, if I sell now, am I selling at a time when all the bad stuff and then some is baked into the price? After all, I knew much of the same in august. I don`t know if this falls into the "cheap for a reason" category or not, and I am equally afraid of selling based on gut feeling and bad news, and not selling based on some endowment-effect. Perhaps selling some, or half, is the best choice, at least for my peace of mind? :-\

[debtvulture]

Then maybe sell a call a little OTM on half of your position?

[beerbaron]

I'm trying to assess the value of INTC production capacity. They have 6 wafer production site but I have a hard time finding how many 12 inch wafer equivalent those represent. TSMC for example says they have a capacity  of 10.1M/Y... it would be interesting to know INTC number. Anybody has something along those lines?

 

BeerBaron

[scorpioncapital]

"cheap for a reason" category or not"

I think you  need some perspective here. You bought 4 months ago and the stock is 4% below your purchase price appx. It pays a 3% dividend. This is peanuts. Look at Buffett , he held IBM for over 5 years before dumping it in favour of a better horse - Apple. If IBM had doubled, I'm sure he would have thought twice about dumping it, unless Apple was also dirt cheap. But it didn't play out that way...Time, research, and contemplation will tell you the answer. As another reference point there are money losing bio-tech/pharma stocks losing tens and hundres of millions a year trading at multi-billion dollar valuations. In comparison , even a declining Intel seems positively juicy.

As for cheap for a reason, yes I have to agree with this because the market is usually quite efficient and prices in all known information, it smells competitive issues restraining growth and lower ROIC. But whether the market is wrong in the longer term as new information becomes available or if inflation picks up is something that will take a while to play out. In a world of less than free money, alot of malinvestment, or excess investment occurs.

 

[Spekulatius]

I'm trying to assess the value of INTC production capacity. They have 6 wafer production site but I have a hard time finding how many 12 inch wafer equivalent those represent. TSMC for example says they have a capacity  of 10.1M/Y... it would be interesting to know INTC number. Anybody has something along those lines?

 

BeerBaron

 

This report gives you some idea:

https://www.asminternational.org/web/fort-wayne-chapter/asm-industry-news/-/journal_content/56/10180/39738678/NEWS

 

INTC is at 817k/200mm wafers month. That would be 0.817M*(8/12)^2*12=4.36M 12” wafer equivalent/year.

 

I don’t think that just the wafer throughout is the right matrix without taking into account the process used. S9me fabs are old and almost worthless. There are still fabs that use 50 or 65 nm processes which are useless for leading edge CPU’s, but can be used for automobile electronics etc. Analog chip plants sometimes use decades old equipment because smaller feature size isn’t necessarily better (if you need certain current outputs etc).

 

Just eyeballing I don’t think that INTC manufacturing capacity is worth close to INTC EV, if that’s your ultimate thesis.

[beerbaron]

I'm trying to assess the value of INTC production capacity. They have 6 wafer production site but I have a hard time finding how many 12 inch wafer equivalent those represent. TSMC for example says they have a capacity  of 10.1M/Y... it would be interesting to know INTC number. Anybody has something along those lines?

 

BeerBaron

 

This report gives you some idea:

https://www.asminternational.org/web/fort-wayne-chapter/asm-industry-news/-/journal_content/56/10180/39738678/NEWS

 

INTC is at 817k/200mm wafers month. That would be 0.817M*(8/12)^2*12=4.36M 12” wafer equivalent/year.

 

I don’t think that just the wafer throughout is the right matrix without taking into account the process used. S9me fabs are old and almost worthless. There are still fabs that use 50 or 65 nm processes which are useless for leading edge CPU’s, but can be used for automobile electronics etc. Analog chip plants sometimes use decades old equipment because smaller feature size isn’t necessarily better (if you need certain current outputs etc).

 

Just eyeballing I don’t think that INTC manufacturing capacity is worth close to INTC EV, if that’s your ultimate thesis.

 

Thanks for the input. What I'm trying to do is to assess the break up value of Intel. Because, if there is a huge value gap... it will happen. If not it will not.

 

For example TSMC has let's say 10M capacity and is worth 470 Billions. INTC manufacturing with 4.3M should be worth 95G. Haircut it by 30% for temporary technological lag and call it 65G. If I look at Wikipedia most of their fabs are 14nm or better which does not show a portfolio that is that old.

 

Now, we would be left with the question of, is INTC design and IP worth 189G-65G=124G$? This one is harder to evaluate but can be somewhat approximated. We could say their X86 IP is worth as much or more than AMD x86 and use that as a a base. Then, bundle everything else into others. If others is free we have a pretty good deal.

 

BeerBaron

 

 

 

 

BeerBaron

[Value_Added]

I’m curious if anyone has looked into Intel’s new Optane Persistent Memory option as a potential unidentified catalyst for future growth?  I was reading their 10k’s and upon researching some of their market offerings, I found this to be very interesting to say the least.  I am into technology and thoroughly understand it on the consumer level, but for those who understand the commercial technology with the cloud and servers, what are your thoughts on this?  From my understandings from my research into Micron, they are quickly reaching the theoretical limit to efficiently scale memory. 

 

Optane Persistent memory seems as though it could serve as a much needed bridge for the gap that is currently in the market.  However, it doesn’t t seem as though intel is really touting it and it seems like they are still performing real world testing which makes me wonder if they’re waiting on definitive proof of concept.  In short, it can act as DRAM or SSD depending on the configuration used.  It is mostly used in the DRAM configuration and is installed in traditional memory ports and recognized by the server as actual memory.  While it’s not true DRAM, it bridges the gap in a very cost effective manner.

 

From the research I’ve seen, under certain configurations, this could be a game changer and serve as a very effective catalyst for their future growth.  They recently announced sale of their NAND to Sk Hynix but excluded the Optane line from the sale.  This is what initially peaked my interest and if it can be effective in real world implementation as the research shows, i feel as though it makes the share price even more attractive.  The question is, will they be able to market it and sell it as a memory solution?

[Liberty]

https://www.cnbc.com/2021/01/13/intel-ceo-bob-swan-to-step-down-in-february.html

 

"Intel CEO Bob Swan to step down"

 

[kab60]

Intel up 10 pct., AMD down 3 pct. - nice farewell party, Bob

[mwtorock]

Well, they are getting a product guy in there. May be worth looking again...

[Jurgis]

My INTC  :D

My VMW  :'(

 

Or maybe it's the other way round.

[Liberty]

https://stratechery.com/2021/intel-problems/

[Candyman1]

Wow, just listened to the INTC call. The expectation is that by the time INTC is onto 7 nm, TSMC will be onto 3 nm. That assumes that INTC is able to stick with the current time line. Nothing guarantees this. 

 

Yes, PC business had increased demand due to Covid. How long does that last with vaccine on the way? Also will we sit on excess number of second hand PCs in a year?

 

Shortages for Ryzen will abate in 2021.

 

Risk runs at less than 50% of power consumption of X86 with same capacity in working laptop with no need for a fan, probably somewhat better. MSFT, Adobe, etc ... are all pushing for Risk. Imagine what Apple will be able to do with the next evolution of the M1 dye if it is willing to cool it for desktops? How long before Dell announces it will support Risk in the laptop environment?

 

Given the M1 dye performance from Apple my guess is that Intel's Apple business will be dead way faster than many believe. Shortages of M1 dyes is their only savior now. Not much to count on.

[MattR]

Wow, just listened to the INTC call. The expectation is that by the time INTC is onto 7 nm, TSMC will be onto 3 nm. That assumes that INTC is able to stick with the current time line. Nothing guarantees this. 

 

Yes, PC business had increased demand due to Covid. How long does that last with vaccine on the way? Also will we sit on excess number of second hand PCs in a year?

 

Shortages for Ryzen will abate in 2021.

 

Risk runs at less than 50% of power consumption of X86 with same capacity in working laptop with no need for a fan, probably somewhat better. MSFT, Adobe, etc ... are all pushing for Risk. Imagine what Apple will be able to do with the next evolution of the M1 dye if it is willing to cool it for desktops? How long before Dell announces it will support Risk in the laptop environment?

 

Given the M1 dye performance from Apple my guess is that Intel's Apple business will be dead way faster than many believe. Shortages of M1 dyes is their only savior now. Not much to count on.

 

Yes it is looking pretty bleak -  yet while Intel Stock is down, is isn't nearly as much down as it should be. The new CEO is paying engineers to come back out of retirement and hopefully change the culture. Friend of mine got 3 times the money working as a software developer for a bank, than what he got from Intel.  However I doubt it will be a quick turnaround. Probably will take around 5 years or more -  and in that time I think the price will go down even more.

[formthirteen]

New Intel CEO Making Waves: Rehiring Retired CPU Architects:

https://www.anandtech.com/show/16438/new-intel-ceo-making-waves-rehiring-retired-cpu-architects

 

The new CEO has started to turn around the ship which still is on the path of financially-engineered obsolescence.

[Spekulatius]

I don’t know how I would feel about rehiring retired design engineers. The rehiring of the Intel veteran from VMW goes along the same lines.

It is almost like the line in the Blues Brothers Movie “ We are bringing the old band together:

 

I guess the junior design engineers at Intel now are all getting the message and their resumes ready.

[Castanza]

I agree with the above. Hard to believe the culture is that bad there that they need to bring people out of retirement. The New CEO was a good move but I was hoping for someone a bit more exciting. If you can't attract young talent then you'll have issues for quite some time. I'll be selling out today and taking what little profit I can get. Simply put there plenty of better values out there.