Question Intel Q3: Ouch

[jpiniero]

https://d1io3yog0oux5.cloudfront.net/_c675d259e2193ee11667e182e2fab5ca/intel/news/2020-10-22_Intel_Reports_Third_Quarter_2020_Financial_1424.pdf

Desktop volume down 16%, Notebook up 16% but ASP down 7%.

Q4 forecast is bad too, they are expecting 14% decline in revenue compared to Q4 last year.

 

[Vattila]

Two opposing views on Intel from a stock analyst and an investment manager:

 

[Markfw]

No CC on the second video, so I can't comment. But the first one is highly critical of Intel for the same reasons I am.

 

[Vattila]

No CC on the second video, so I can't comment.

Basically, like many, she trusts Intel to sort out their problems and come back strong in due course, considering their size (income, resources, influence), and hence she thinks Intel is a good long-term investment opportunity based on their current low valuation (remarkably low PE ratio compared their peers).

 

[cytg111]

Two opposing views on Intel from a stock analyst and an investment manager:

Those two videos, facts vs. faith. The best way I can describe it.

 

[Hitman928]

Basically, like many, she trusts Intel to sort out their problems and come back strong in due course, considering their size (income, resources, influence), and hence she thinks Intel is a good long-term investment opportunity based on their current low valuation (remarkably low PE ratio compared their peers).

Their valuation right now is definitely low compared to AMD, Nvidia, and others, but they also are not a growth company to begin with so their valuation will of course be based more upon present PE compared to someone like AMD who is growing rapidly, and there obviously just isn't trust in Intel to right the ship. Hearing an analyst say that she trusts that Intel will fix their problems because they're big and make a lot of money just seems so foolish to me when the leadership has given no clear vision for how they will actually fix their issues. Maybe she's right, but having confidence that they will just because they are making money today seems like someone just shutting their eyes so they don't have to watch the forward cars on the train derail to keep pretending everything is fine.

I mean, they basically blamed their recent bad quarter on COVID and server market saturation. If no one else is reporting the same thing, that just points even more to the leadership lacking any kind of vision of what their problems really are and how they fix them. We'll see what the other players in the space say about the economic conditions they have been facing as more earnings reports come out in the next few weeks.

 

[dmens]

The problem with Intel moving to TSMC, is that TSMC has to agree to take them and they've already stated that they aren't willing to be a crutch for Intel, so it's not as obvious as solution as it may seem, even if Intel can port their designs.

Porting designs to TSMC is not a problem. TSMC process is much easier to design to than Intel process: TSMC actually has to satisfy customers.

But when Intel pays TSMC for fab time, they can kiss their margins goodbye. TSMC is not running a charity.

 

[Thunder 57]

I'm seeing a lot of strawman here. Show me who is saying "Intel is dying" or "Intel has no chance", etc. I don't see anyone saying that.

 

[esquared]

Mark and Likelinus.

Full. Stop. Now.

I am getting tired of having to deal with all these reported posts.
Stop responding to each other.

Put each other on ignore. I don't care. Just stop responding.
Ignore each other. Whatever it takes.



esquared
Anandtech Forum Director

 

[mxnerd]

if the TSMC's new fab produce 20,000 12" (300mm) wafers per month , and each chip is 12mm x12mm, according to

Die-Per-Wafer calculator

Die-Per-Wafer Estimator

www.silicon-edge.co.uk

each wafer can have up to 410 chips. And that will be 8,200,000 chips per month!

 

[videogames101]

TSMC is the last game in town for leading edge digital nodes. If I were the US Government I'd be sending more soldiers to Taiwan. Although I'm not sure anyone in D.C. has even an inkling of how important this shift is... Intel missing the mark on fab tech so badly is becoming a national security issue. While it's good for competition now (AMD, Arm, etc), it's not going to be good for the US in the long run. It must be pretty weird being the CEO of TSMC, which now also means being the head of (possibly) the most important private company on earth.

 

[scannall]

I really don't know enough about their current leadership to know if they have the right people in place. That said, yes, they have significant overhead compared to AMD. That said, they also still have 10X profit and 80 market share. The whole "Intel is on their deathbed" is way silly. They have to dig themselves out of the hole but it just amazes me that people forget history.

We have AMD in our house hold and was actually playing with them this weekend L4D and CSGO, old school!, and it was fun. But I'm not such a fan that I can't realize the benefits of both.

It's funny when you literally show someone that the developers and users are having issues with AMD CPUs on certain software, but those people recommend AMD without any research or even recognizing the issues from the developers mouth!!! Not even asking a single question about usage in some threads! It's just AMD is best, buy it!

Whatever man, it is what it is. Intel isn't going anywhere soon.

The thing many are missing is momentum. Does that make Intel's products better? No. Are Intel's current products better? No. And it doesn't matter. For now. When a salesperson makes a call, it isn't an all bizz thing. Salesmen if they don't suck make relationships.

 

[zinfamous]

This is what Intel gets for making me choose between 6 slower cores for $700 and 4 faster ones for $500. They can suck it.

you won't buy AMD, though.

You still owe us an eaten box, lest you forget (I never forget)

 

[zinfamous]

Two opposing views on Intel from a stock analyst and an investment manager:

Intel pays a 2,7% dividend right now, I think?

That's a bit of a trap that I'd hate to be in on their side to decide between reducing or ending that...or buying back more stock. It's desperation for the next couple of months (years, really), because their brains are stuck on floating "perceived stock value," which is fundamentally stupid, all while they are paying out cash that they don't have to keep lazy fun-holders happy...none of them really knowing, probably, that they hold Intel dividends.

Intel needs to bring in a soulless engineer-person with actual brains to cut the chaff, eliminate the dividend, sell off the fab, and make them a contemporary circuit design company. That won't happen in the next 5 or 7 years, though, because they are very old and very stupid people, wearing their stupid blue ties, live in grey parking lot cubicles, and are primarily run by "industry" tech insiders and CEOs of airline and service industry companies.

Good luck, Intel.

 

[gorobei]

Dr Cutress's breakdown of the report. data center is telling, more than a few contradictions on the nand sale and aurora pointed out.

 

[Spartak]

I Have to say I'm surprised seeing AMD still having only 20% of the desktop market and even more surprised mobile where the advantage is much less pronounced has the same marketshare. Since there is no DIY mobile market AMD's OEM market share for desktops must be even lower than mobile. Can't wrap my head around that.

if we assume the DIY market is indeed about 80% AMD and 20% Intel (as MindFactory and Amazon figures suggest) this means, if we assume OEM vs DIY volume is about 90% vs 10%, 87% of all OEM desktop systems is still Intel. if the DIY fraction is larger Intels OEM desktop market share goes very quickly towards 100%.

Since a large proportion of OEM desktops are gaming PCs I find this very hard to believe...something is off in that desktop market share chart.

 

[jpiniero]

Since a large proportion of OEM desktops are gaming PCs

It's not. OEM gaming PCs is probally growing in share but it's still mostly boring corporate desktops. Intel's actually fortunate the volume decline is only 18% if you think about it (and no ASP decline), considering the virus.

 

[DrMrLordX]

meh. they have cartoons on cpu boxes now. It'll be fine

AVENGERS DISSEMBLE!

Wait, that's not how it's supposed to go . . .

How is owning more of your company a bad thing?

If you buy shares at $X to prop up value while your stock is in decline, you wind up losing money. If you buy shares at $Y after hitting a bottom, you stand to gain money if/when your shares go back up in value. Typically, stock buybacks take place before the tumble, rather than after. It's basically free money for whoever decides to sell their shares after the buyback (taking advantage of an artifically-inflated stock price).

It's cliché time. Betamax, DEC Alpha, or BeOS were advanced tech, but didn't break into the market.

Bad analogy. AMD has more market exposure than those, especially BeOS.

SPR should finally replace those 14nm abominations

That's gonna be a real problem for Intel if Sapphire Rapids isn't ready until 2022.

However, is there really no way they will fix that in late 2021?

We don't know. Intel still won't dish on 7nm. TSMC won't save their core business unless they do something radical, like cough up fabs and start fully integrating TSMC processes into their own (which is something they should have at least considered when they canned Murthy).

But it sure helps out the VP's and c-suite dwellers.

Exactly. After all, whom do you think makes such decisions anyway?

Well, you could have bought AMD. Nobody put a gun to your head and forced you to buy intel.

ShintaiDK is pretty sneaky. You don't know where he was or what he might have been doing with firearms.

As far as Intel dying, that overstates the case. They are losing market share, profit, and product respect. That does not mean they are doing to die, just go downhill for a while. And AMD is taking a bigger piece of the pie.

I don't know why everybody thinks that if Intel is doing bad right now, they are finished. They are in a slump, they will recover, but the question is how much do they loose before they come back ?

Of course Intel isn't going to disappear, but the Intel we know today probably won't be the Intel of tomorrow.

I want to address these three ideas.

Intel will not die, but they may be finished as a producer of cutting-edge x86 CPUs. Right now they have two or three products with maybe a strong future - Altera FPGA, Mobileye self-driving car tech, and Xe. Those are products they can move to other fabs if necessary (see below). Intel has maybe one more chance to retake momentum in the DCG product category (read: Xeon) with Sapphire Rapids. If that shows up too late or not at all, you can stick a fork in em, they're done. If I'm Bob Swan or (perish the thought) Raja Koduri, I'm having a meeting with my fab team managers and CPU design team managers, asking what it is that they're doing right now that's actually going to make Intel any money? Nothing Intel has done since Cascade Lake-SP hit the market has amounted to anything in the server room. Intel was actually improving its revenue through Q2 2020 on a design from two years ago which itself was a respin of 2017's Skylake-SP. You can thank their sales team for that. All of that. Cascade Lake-AP? Dead. Cooper Lake? Niche, mostly dead (props for bfloat16 though). Ice Lake-SP? Delayed until 2021, which is when Sapphire Rapids was supposed to launch. The revenue streams have started to tilt downward, indicating that their brilliant-if-coercive sales department can no longer badger people in continuing to buy Cascade Lake-SP ad infinitum.

When Tiger Lake-U hit, Intel demonstrated that they had worked out some of the kinks in their 10nm process by fielding CPUs that can hit clockspeeds closing in on 5 GHz. Great! But they're only 4c chips, with 8c chips delayed until . . . 2021. Indicating that 10SF and 10SFE may have done little-to-nothing to alleviate the yield issues that plagued Ice Lake-U and that have (presumably) kept Ice Lake-SP off the market. Sapphire Rapids allegedly sidesteps this problem using small chiplets and advanced interconnects. Interconnects which are largely untested on the open market outside of Kaby Lake-G and Lakefield. Does anyone honestly have much confidence in Intel's advanced interconnect technology right now? Before we get a chance to see Alder Lake-S in action? Thought finished below.

The problem with Intel moving to TSMC, is that TSMC has to agree to take them and they've already stated that they aren't willing to be a crutch for Intel, so it's not as obvious as solution as it may seem, even if Intel can port their designs.

TSMC is the last game in town for leading edge digital nodes. If I were the US Government I'd be sending more soldiers to Taiwan. Although I'm not sure anyone in D.C. has even an inkling of how important this shift is... Intel missing the mark on fab tech so badly is becoming a national security issue. While it's good for competition now (AMD, Arm, etc), it's not going to be good for the US in the long run. It must be pretty weird being the CEO of TSMC, which now also means being the head of (possibly) the most important private company on earth.

Intel has the fab capacity that TSMC needs to a). diversify geopolitically outside of Taiwan and b). feed Intel. The problem for Intel is that TSMC will kill off Intel's foundry business by essentially taking it over. Intel has to make a choice: die in pursuit of foundry resurgence or survive as a design house with a not-insignificant stake in what could eventually become TSMC North America. As it stands, Intel is their foundry business. Xeon is the lynchpin of how they turn their advanced foundry processes into money. Everything flows from that. Intel is at significant risk of losing both within a year or two. That doesn't mean the company goes away, but it does mean Intel as an IDM goes away, one way or another. It will take a foundry-based miracle to get Sapphire Rapids launched on-time and at an acceptable profit margin, followed by a 7nm process which Intel has repeatedly signaled is badly in peril of never seeing the light of day. Otherwise we wouldn't be having these conversations about Intel trying to take 5nm wafers from TSMC.

TSMC will never be able or willing to provide Intel with enough 5nm wafers to satisfy their current Xeon production requirements. And if they did, Intel would be stuck in the onerous position of paying off TSMC's long-tail capex requirements AND their own capex on their own fabs which they would not be using to sell their darling Xeons. That is a recipe for disaster.

Intel's other option is to let Xeon and their foundry business die, and carry on with lucrative sales of FPGAs and automotive-computing products. And maybe dGPUs/AI accelerators. They could dump a lot of capex, shed a lot of engineers, and be a leaner and more-focused company. The idea of Intel leaving their traditional markets may seem absurd, but when the only functioning and profitable design to come out since really 2017 is Tiger Lake-U, you have to think that Intel's future is not in their x86 products.

 

[moinmoin]

Dr Cutress's breakdown of the report. data center is telling, more than a few contradictions on the nand sale and aurora pointed out.

It has to have been on purpose that Ian wore a Commodore shirt in that video.

 

[Doug S]

if the TSMC's new fab produce 20,000 12" (300mm) wafers per month , and each chip is 12mm x12mm, according to

Die-Per-Wafer calculator

Die-Per-Wafer Estimator

www.silicon-edge.co.uk

each wafer can have up to 410 chips. And that will be 8,200,000 chips per month!

So what's your point? Like I said, that is a pretty small fab. Apple sells twice that many iPhones per month when averaged over a year, and the PC market sells between 2-3x that many laptops/desktops per month.

 

[fusionTi]

Basically, like many, she trusts Intel to sort out their problems and come back strong in due course, considering their size (income, resources, influence), and hence she thinks Intel is a good long-term investment opportunity based on their current low valuation (remarkably low PE ratio compared their peers).

PE ratio isn't a good metric to use here, it is backwards looking and Intel's earnings are shrinking and margins are dropping. Their forward PE will be lower than their PE.

 

[dmens]

TSMC won't save their core business unless they do something radical, like cough up fabs and start fully integrating TSMC processes into their own (which is something they should have at least considered when they canned Murthy).

LOL. If Intel designed first to TSMC then tried to port back over to Intel in-house process, it is game over. Those fabs would be useless.

 

[NTMBK]

What freaking CPUs are the 'Leadership products' in 21/22?
I've never even heard of a 'Leadership' product before.

They're products that "ship for revenue", solely so that Intel leadership can get their bonuses...

 

[positivedoppler]

Anyone remember the troll ShintaiDK and a few others? That was only three years ago.

And I agree with everyone who understands the military importance of Taiwan.

 

[positivedoppler]

Future battlefield will be drones and AI, we cannot allow Taiwan to fall to China

 

[Ajay]

Anyone remember the troll ShintaiDK and a few others? That was only three years ago.

And I agree with everyone who understands the military importance of Taiwan.

Yeah, ShintaiDK was a mix of good info and bad (and belligerent, hence BANNED).
The military importance of Taiwan stem from our economic dependence on the country.
Impossible to defend against a Chinese invasion though, which would be stupid of China.