Intel beats earning estimate, but down

[ta152h]

You can look at things from different angles and say Intel TAM is higher than TSMC. Well GloFlo TAM is the same as TSMC but there is a big difference in ability to actually seize that.

Personally I think it’s not just a matter of Intel having the wrong products. They have the wrong business model, and having the wrong products is an outcome of having the the wrong business model.

Way too general for me to really directly reply to.

But, I'd argue GF doesn't have the same TAM as TSMC, since they aren't a leading edge fab anymore.

But, because something isn't 100% of the picture doesn't mean it isn't relevant. Intel having more ways to make money is a positive, but it surely doesn't mean they will. My big problem is the valuations. Is TSMC worth 4x Intel? In my book, not even close. TSMC is doing well, but that's already priced into their stock. If they were valued lower, yeah, they'd be more interesting.

I'd argue Intel has a perfect business model, finally. As fabs got more expensive, using them in-house only became less viable, and now they finally are making a real effort. Their prior leadership, particularly Krzanich, was atrocious. He missed every opportunity, and jumped and wasted money on things that weren't.

TSMC is in for a real fight, against a competitor that can do things they can't. Intel has IP that can be used for their fab customers, and their logic is best in class. TSMC doesn't have that. TSMC does what they do very well, currently better than Intel, overall, but that's hardly guaranteed going forward. Imagine being able to go to a fab and pick up blocks of IP and use them to develop your processor. It's going to save a lot of development costs and time. And Intel's processors have the best single-threaded performance, overall, so it's not like they're getting low grade stuff.

Where they run into trouble is, Intel will be competing against its own customers. Will AMD want to use Intel processes? Doubtful. So, there's plusses and minuses, but a company having their own fabs is a huge benefit, just not always.

But, let's see how it plays out. Nothing is certain. AMD's quarter will tell us something, and a year from now, we'll know a lot more, although by no means everything, since their fabs are going to take a long time to really get momentum. I see a really good future for them, but I don't pretend to have a crystal ball.

 

[ta152h]

Gross margin 38%, Jeeeeeezus

A product of high fixed costs . But again, don't view it without the lense of what was expected, and what's happening in the industry in general. It's bad out there.

 

[TICin852]

As you mentioned, @ta152h you have positions in Intel. I'd recommend taking a step back and just analyzing why your view may be overly optimistic and what you could have gotten wrong. The fact is that people tend to justify and defend their past decisions more than they objectively evaluate them.

One thing I will say is this: if everyone expects you to lose 1 billion dollars, and you only lose 500 million, you still lost 500 million. Intel is still losing lots of money, even though they lost less than people expected them to lose.

 

[Daniel Nenni]

Interesting process update:

Relative to five nodes in four years, notably, two out of these five nodes, Intel 7 and Intel 4, are now essentially done. Intel 7 is in high-volume manufacturing, and Meteor Lake on Intel 4's ramp in production wafer starts today for a second-half product launch.

We are quickly mastering EUV technology with Intel 4 as our first EUV node. As we focus on the next three nodes, Intel 3 is on track and we highlighted in our recent DCAI webinar, Sierra Forest will begin shipping in the first half of '24, with Granite Rapids shortly thereafter, both on Intel 3.

We also have significant milestones planned in Q2 for Intel 3, Intel 20A and Intel 18A, and look forward to providing more details as we execute. Overall, we are squarely on track to deliver five nodes in four years. We understand that our foundry ambitions will not be realized overnight, building a vibrant foundry ecosystem will take time. But we also understand our foundry success is vitally important to establishing a geographically diverse and secure supply of semiconductors.

We took a major step forward in building our ecosystem this month when we announced a multi-generation agreement with Arm Holdings. This will enable chip designers to build leading-edge mobile SoC designs on Intel 18A, giving the design community a new foundry alternative for product innovation and fast time-to-market, while also opening up new options and approaches for large ecosystem of Arm customers.

And with Intel 3, the positive updates that we've given on Granite and Sierra Forest for next year, the volume sampling that I've already referred to gives us a lot of confidence that, that is now coming along very nicely. Both Intel 4 and Intel 3 are EUV nodes. As you say, as we go to 20A and 18A, the two major innovations are the RibbonFET, the gate all around transistor architecture and the backside power.

Given the uniqueness of the backside power, as you indicated, we had an internal node that we didn't expose to products or externally to derisk that node, and that went extremely well. We had very good results from the backside power, the power delivery, the routability improvements that gave. And as I was -- as one proof point of that -- the Arm announcement was one that demonstrated significant benefits of backside power that we were able to do.

So 20A and 18A are the next ones up. 20A will be primarily a client node as we ramp our Arrow Lake products in '24 and '25. 18A will be everything. We will have server products, client products, networking products and many foundry products. We also noted that this was the quarter that we have our first foundry test chips coming out. And so some of the test chips for external customers on 18A are now popping out a fab and being tested by them.

So good affirmation from them, you also mentioned, I think it's actually a very insightful question, Matt, the cost structure. And one of the things that we've put a lot of emphasis on with 18A is getting to structural cost parity with what we believe is the best in the industry at that point. So, we view this as not just getting to power and performance parity, but also area parity and cost structural parity as we get to 18A.

And we believe, as we've benchmarked ourselves against the industry best, we believe we're on track to do that in the 18A timeframe. And that's part of why we talk about in the internal foundry model as being able to start really measuring the P&L right? I'm really viewing it as the industry price for wafers is understood, and we have to benchmark ourselves against that and deliver margin structure at the wafer level that's competitive with that.

 

[blueone]

With chiplet designs, Intel can realistically get the processor design teams to get on a foundry model, and reduce the number of steppings they need. I'm sure that'll be a very significant cost savings for future CPUs. Chiplets should also improve Intel's design schedule integrity. Now if they could only reduce the level of tracking and management layer bureaucracy.

 

[hist78]

Gross margin 38%, Jeeeeeezus

38.4% is Intel Q1 2023 Non-GAAP gross margin. Intel's GAAP gross margin dropped to 34.2% for the same period of time.

The Non-GAAP operating margin is -2.5% and GAAP operating margin dropped to -12.5%.

Net cash flow from operating activities for Q1 2023 is -$1.785 billion vs $5.891 billion Q1 2022. It's very bad.

Additionally, during Q1 2023 Intel got cash injection of $10.968 billion from "Issuance of long-term debt, net of issuance costs". This is a huge amount and is almost as big as the combined revenue of Intel CCG, DCAI, and NEX groups!

Source: https://d1io3yog0oux5.cloudfront.net/_9ffaaa3a9984d36dd2ad28487bcbe79f/intel/db/887/8943/earnings_release/Q1+23_EarningsRelease+(004).pdf

 

[SailorBob]

This write up over at nextplatform was interesting. This bit jumped out at me:

According to Intel’s 10-Q filing with the US Securities and Exchange Commission, and picked up by our good buddy Aaron Rakers over at Wells Fargo, Intel’s Xeon XP server shipments were off 50 percent in the quarter. By comparison, server CPU shipments in all of 2022 were off 16 percent.

The Beginning Of The Bottom For Intel’s Datacenter Business

Intel can talk all it wants about how it beat its own expectations or those of Wall Street, but the fact remains that the first quarter of 2023 was

www.nextplatform.com

 

[Brady]

This write up over at nextplatform was interesting. This bit jumped out at me:

According to Intel’s 10-Q filing with the US Securities and Exchange Commission, and picked up by our good buddy Aaron Rakers over at Wells Fargo, Intel’s Xeon XP server shipments were off 50 percent in the quarter. By comparison, server CPU shipments in all of 2022 were off 16 percent.

The Beginning Of The Bottom For Intel’s Datacenter Business

Intel can talk all it wants about how it beat its own expectations or those of Wall Street, but the fact remains that the first quarter of 2023 was

www.nextplatform.com

It’s very, very ugly out there for DCG… and with Ming Chi Kuo saying that MSFT cut Xeon orders for 2023 by 50-70% it’s going to get worse before it gets better.

Putting on my speculation hat, I’d say it’s two things: SPR being 2 years late and mostly uncompetitive with Genoa in $/core or $/compute, and Sierra Forest being pulled in and having a bit of an Osbourne Effect for cloud providers. One of those things is known and very bad, one is not really known and could have an upside to it.

Either way, at this rate DCG earnings ain’t recovering until 2H24… do they have the money to survive DCG bleeding out while they attempt to ramp a new business in the most capital intensive industry on earth? Will CCG keep them afloat? I believe in Intel’s roadmap but yikes this is really driving on the cliff’s edge here.

 

[ta152h]

It’s very, very ugly out there for DCG… and with Ming Chi Kuo saying that MSFT cut Xeon orders for 2023 by 50-70% it’s going to get worse before it gets better.

Putting on my speculation hat, I’d say it’s two things: SPR being 2 years late and mostly uncompetitive with Genoa in $/core or $/compute, and Sierra Forest being pulled in and having a bit of an Osbourne Effect for cloud providers. One of those things is known and very bad, one is not really known and could have an upside to it.

Either way, at this rate DCG earnings ain’t recovering until 2H24… do they have the money to survive DCG bleeding out while they attempt to ramp a new business in the most capital intensive industry on earth? Will CCG keep them afloat? I believe in Intel’s roadmap but yikes this is really driving on the cliff’s edge here.

As you mentioned, @ta152h you have positions in Intel. I'd recommend taking a step back and just analyzing why your view may be overly optimistic and what you could have gotten wrong. The fact is that people tend to justify and defend their past decisions more than they objectively evaluate them.

One thing I will say is this: if everyone expects you to lose 1 billion dollars, and you only lose 500 million, you still lost 500 million. Intel is still losing lots of money, even though they lost less than people expected them to lose.

Oh no, I don't think that's it. But, I do appreciate your concern.

Let's look at it a different way. How could I possibly be so sanguine on Intel, without investing a lot in it? I think you have the tail wagging the dog, but, who knows, maybe I'm too close it. I do tend to be an optimist by nature, and I do believe in people and their ability to get things done when they put their mind to it. We all view life through a different lense, but being afraid of admitting mistakes is not a flaw I have. Being overly optimistic and believing people can do a lot when they are focused could very well be one.

But, nothing in their earnings concerns me, because none of it relates to long term problems. Wait until AMD does a release, and we can talk more about it. We'll have better clarity. And we'll revisit it as time goes on.

Oh, and for the record, I'm already up on my Intel investment (but barely). So, it would be pretty easy to walk away from it, if I really didn't believe in them. I do, I think this is a very unusual situation with enormous potential rewards. I am looking to buy more, mainly by selling puts (I've to 20 contracts out now, and will be trying to sell more if the stock dips).

Either way, I do appreciate your concern. I hope it's misplaced, but you may be right.

 

[ta152h]

It’s very, very ugly out there for DCG… and with Ming Chi Kuo saying that MSFT cut Xeon orders for 2023 by 50-70% it’s going to get worse before it gets better.

Putting on my speculation hat, I’d say it’s two things: SPR being 2 years late and mostly uncompetitive with Genoa in $/core or $/compute, and Sierra Forest being pulled in and having a bit of an Osbourne Effect for cloud providers. One of those things is known and very bad, one is not really known and could have an upside to it.

Either way, at this rate DCG earnings ain’t recovering until 2H24… do they have the money to survive DCG bleeding out while they attempt to ramp a new business in the most capital intensive industry on earth? Will CCG keep them afloat? I believe in Intel’s roadmap but yikes this is really driving on the cliff’s edge here.

Intel has tons of money, and has access to tons of loans if they want them. I wouldn't worry.

A lot went bad for Intel. Sapphire Rapids didn't help, but it's a pretty competitive chip now, and Emerald Rapids should be too. But, I really agree with you, it's going to be 2024 before we see any hope of market share gains. And it's not even certain then. But, their competitive stance will be much better. I think Sierra Forest is the sleeper really. Granite Rapids will be needed as well, but I think Sierra Forest allows them to play in scenarios they weren't in before.

These concerns are why the stock is dirt cheap. I mean, when a stock is super cheap, there has to be something that is bothering people. It's nice to see these forums so I can see how the other side views things, and I don't mean that sarcastically. If someone says something that really disturbs me, I will probably reevaluate my position and views, but so far, nothing does. But, that doesn't mean I'm right either. Time will tell.

 

[hist78]

Intel has tons of money, and has access to tons of loans if they want them. I wouldn't worry.

A lot went bad for Intel. Sapphire Rapids didn't help, but it's a pretty competitive chip now, and Emerald Rapids should be too. But, I really agree with you, it's going to be 2024 before we see any hope of market share gains. And it's not even certain then. But, their competitive stance will be much better. I think Sierra Forest is the sleeper really. Granite Rapids will be needed as well, but I think Sierra Forest allows them to play in scenarios they weren't in before.

These concerns are why the stock is dirt cheap. I mean, when a stock is super cheap, there has to be something that is bothering people. It's nice to see these forums so I can see how the other side views things, and I don't mean that sarcastically. If someone says something that really disturbs me, I will probably reevaluate my position and views, but so far, nothing does. But, that doesn't mean I'm right either. Time will tell.

”Intel has tons of money, and has access to tons of loans if they want them. I wouldn't worry."

Net loss, negative cash flow from operations, shrinking revenue, shrinking gross margin, and one of the lowest "current ratio" among major Intel competitors are all bad signs for Intel's financial standing.

Considering additional factors such as the high capital intensity, paying large amount of dividends, the so called Semiconductor Co-Investment Program (SCIP), and the heavy investment needed for IFS and IDM 2.0, Intel doesn't have a lot money floating around like it used to have. Intel's ability to borrow cheap money is also dwindling.

 

[ta152h]

”Intel has tons of money, and has access to tons of loans if they want them. I wouldn't worry."

Net loss, negative cash flow from operations, shrinking revenue, shrinking gross margin, and one of the lowest "current ratio" among major Intel competitors are all bad signs for Intel's financial standing.

Considering additional factors such as the high capital intensity, paying large amount of dividends, the so called Semiconductor Co-Investment Program (SCIP), and the heavy investment needed for IFS and IDM 2.0, Intel doesn't have a lot money floating around like it used to have. Intel's ability to borrow cheap money is also dwindling.

Except it's transient. That's the part that makes it unimportant. That's how you make money, let the people panic, and make money off of them while they do.

Revenue is projected to go up this quarter, and up each succeeding this year. Gross margin is related to that, since they have high fixed costs. And the high capital intensity is already part of their earnings, it's not like it's going to get worse, and their costs are frontloaded.

No one has cheap money like it used to, I mean, the interest rates are way up. But, Intel has a lot of cash on hand, and is still the highest rated credit tier, so yeah, they can get what they need. Also, the dividend was slashed pretty aggressively, so it is much less of a concern.

One point I'm not sure you're factoring in is, even if PC sales don't improve, Intel will generate a lot more revenue. Intel only sold around 80% of the processors that were consumed, and that inventory correction can't run indefinitely. They are saying it should be complete in Q2. And with Q2 guided upward, the problems aren't that severe. Q3, with the inventory correction over with, I have to imagine things will be significantly better. And Q4 should be their strongest quarter, as it typically is.

So, nothing there bothers me. It's great it bothers other people though, this way I can keep getting the stock cheap. But, I look at market share gains in client, maintaining share in servers, and all their nodes on schedule or ahead as very positive. Transient losses while they pivot to a new model and burn inventory don't scare me at all. I'm far more positive after the earnings than I was before, and I was pretty positive even then. Everything is going as planned, or ahead, outside of the macro-environment, which will find its footing eventually. But, can they continue to execute? I think so. I guess you don't. We'll know in time.

 

[HYB]

Except it's transient. That's the part that makes it unimportant. That's how you make money, let the people panic, and make money off of them while they do.

Revenue is projected to go up this quarter, and up each succeeding this year. Gross margin is related to that, since they have high fixed costs. And the high capital intensity is already part of their earnings, it's not like it's going to get worse, and their costs are frontloaded.

No one has cheap money like it used to, I mean, the interest rates are way up. But, Intel has a lot of cash on hand, and is still the highest rated credit tier, so yeah, they can get what they need. Also, the dividend was slashed pretty aggressively, so it is much less of a concern.

One point I'm not sure you're factoring in is, even if PC sales don't improve, Intel will generate a lot more revenue. Intel only sold around 80% of the processors that were consumed, and that inventory correction can't run indefinitely. They are saying it should be complete in Q2. And with Q2 guided upward, the problems aren't that severe. Q3, with the inventory correction over with, I have to imagine things will be significantly better. And Q4 should be their strongest quarter, as it typically is.

So, nothing there bothers me. It's great it bothers other people though, this way I can keep getting the stock cheap. But, I look at market share gains in client, maintaining share in servers, and all their nodes on schedule or ahead as very positive. Transient losses while they pivot to a new model and burn inventory don't scare me at all. I'm far more positive after the earnings than I was before, and I was pretty positive even then. Everything is going as planned, or ahead, outside of the macro-environment, which will find its footing eventually. But, can they continue to execute? I think so. I guess you don't. We'll know in time.

I think in the short to medium term (5-7 years out) its transient as semiconductors, especially memory, tend to bullwhip and have these large upturns and downturns.

As a secular trend, or 10-20 years out, PC TAM seems to be stagnating to the tune of very low single digit growth. Data Center and HPC growth is strong (10-15%) and could possibly even be accelerating depending on AI and other compute. The problem is that data center share seems to be getting spread more and more as competition from everywhere comes in due to it being a very lucrative market. Ampere, AWS Graviton are examples of ARM server chips that have been successful so far with many more to come from Google, Microsoft, Meta, Alibaba, etc. Obviously AMD and NVIDIA are also big players. Its almost the opposite of the memory market which is not as lucrative becoming an oligopoly with Samsung, Micron, and Hynix. Whereas the data center market is lucrative to the point where it invites a lot of players, spreading the market share more thin.

Overall I am very bullish on Intel because of foundry and believe that foundry revenues will become the main driver in the long term. I think thats how Intel will win big in the long term, as they can capture a share of every chip no matter which companies become dominant. I think Intel share price is not really even being driven by their core business, but rather on news of node and IFS progress.

 

[SailorBob]

It’s very, very ugly out there for DCG… and with Ming Chi Kuo saying that MSFT cut Xeon orders for 2023 by 50-70% it’s going to get worse before it gets better.

Putting on my speculation hat, I’d say it’s two things: SPR being 2 years late and mostly uncompetitive with Genoa in $/core or $/compute, and Sierra Forest being pulled in and having a bit of an Osbourne Effect for cloud providers. One of those things is known and very bad, one is not really known and could have an upside to it.

Either way, at this rate DCG earnings ain’t recovering until 2H24… do they have the money to survive DCG bleeding out while they attempt to ramp a new business in the most capital intensive industry on earth? Will CCG keep them afloat? I believe in Intel’s roadmap but yikes this is really driving on the cliff’s edge here.

We'll know allot more when AMD reports tomorrow. In their Q4 CC they had said they expect 1Q23 DC revs to be down double digits. If it's low double digits then Intel's problem is their products, if it's high double digits like Intel, then macro is the primary problem I think...

 

[SailorBob]

Intel has tons of money, and has access to tons of loans if they want them. I wouldn't worry.

A lot went bad for Intel. Sapphire Rapids didn't help, but it's a pretty competitive chip now, and Emerald Rapids should be too. But, I really agree with you, it's going to be 2024 before we see any hope of market share gains. And it's not even certain then. But, their competitive stance will be much better. I think Sierra Forest is the sleeper really. Granite Rapids will be needed as well, but I think Sierra Forest allows them to play in scenarios they weren't in before.

These concerns are why the stock is dirt cheap. I mean, when a stock is super cheap, there has to be something that is bothering people. It's nice to see these forums so I can see how the other side views things, and I don't mean that sarcastically. If someone says something that really disturbs me, I will probably reevaluate my position and views, but so far, nothing does. But, that doesn't mean I'm right either. Time will tell.

IIRC, Sierra Forest is their dense chip, and is made up of their single threaded small cores, Gracemont? I don't see how that competes with Bergamo or Turin dense, since those are full Zen4 or Zen5 cores which just have less cache.

 

[ta152h]

IIRC, Sierra Forest is their dense chip, and is made up of their single threaded small cores, Gracemont? I don't see how that competes with Bergamo or Turin dense, since those are full Zen4 or Zen5 cores which just have less cache.

That's kind of the point. They are really going to give them new scenarios, where as Granite Rapids will competing where their conventional server products go. So, it's not that I disagree with you, Granite Rapids is a very necessary product, but Sierra Forest will just be better at certain workloads than anything AMD has, or Intel has with their performance parts. In my mind, it's really more targeted against ARM, which plays well in scenarios where x86 simply has no entries.

AMD can use their big cores and dress them in drag, but they still won't be ideal in that role. Removing cache will make them cheaper to make, but the performance per watt metric matters a whole lot in servers. AMD processor benefit a lot from TSMC's less power hungry nodes, so far, so although they suffer lower clock speeds, their performance per watt is often higher than Intel's.

Sierra Forest is simply a much more efficient architecture, Zen 4 is still Zen 4, and wasn't made with size/power use as the leading goals. It has to perform as well as Intel p-Cores (it doesn't), and be as efficient as Intel e-Cores (it isn't). But, it performs better than e-Cores, and is generally more efficient than p-Cores, so it's got an important spot. Intel moving to I3 for Sierra Forest is the wild card. Will it be as power efficient as the node AMD is using? Not sure. History suggests it won't be, but history is pretty much starting over for Intel with their new goals. So, who knows? But, with a design significantly less focused on power efficiency, if the nodes are even close, Sierra Forest will have a lot of usage scenarios where it's significantly better. And probably even more where it isn't. But, that's the point, you have two products that are designed from the beginning to be focused on what they are best at. AMD has one, and decided to put a dress on it, and say it identifies as a low power CPU. It's not convincing, and it won't give them access to "real" low power scenarios, but it probably will cover a segment of the market. But, not the segment Sierra Forest is going after, which is pretty much new ground for x86. Whether AMD can be successful by pairing down their processors and doing the best they can, that's anyone's guess. I think they will do OK though, it's a good architecture, and they probably can increase the range a bit. But, not like a "from the ground" processor designed for it.

Also keep in mind, 144 cores is just Intel being conservative, and apparently a demand from a customer that this part be ready sooner rather than later. The cores are tiny, and about 1/4 the size of the P-Cores. So, even double that on i3 wouldn't be a particularly huge CPU. But, it's new, and they are taking a conservative/iterative approach, which works better for them. We all know how 10 nm worked out, when they got too ambitious. And we see how well their new roadmap is working out with their nodes being much more iterative.

ARM is lurking. x86 needs an answer to it. That answer isn't a diminished big core, in my opinion. It's a core designed from the beginning to compete in the same workloads.

Of course, I could be wrong. I THINK this is going to fill a nice gap, but I'm far from sure, and there are questions of whether this is a big market, and whether Intel can compete effectively against ARM even if it is. Both have to be something of a yes, or it's probably not going to be profitable. Although, the slot and platform are shared with Granite Rapids, and it does give Intel an opportunity to lead with a lower volume product on that platform, and I3. So, I think it's a good idea, for a few reasons, but only time will tell how well it is adopted.