Intel Reports Q2 2023 Financial Results with Strong Guidance!

[ta152h]

View attachment 1321

Operating Profit/Loss in 2nd Quarter 2023:
- TSMC operating profit: US$6.58 billion
- Intel: 5th straight operating loss: (US$1.02 billion)
- Samsung chip division: 2nd straight operating loss: (₩4.36 trillion) (US$3.32 billion)

But, another way to look at it is, Intel improved not only in operating profit/loss, but also revenue from Q1 to Q2. TSMC?

I'm up almost 50% on my Intel investment, while TSMC is down, so I'm not complaining. And this is verifiable, since I was saying I was long on Intel months ago, when they were well under $30. Snapshots like this aren't meaningless, but they also don't mean all that much when a company is pivoting, and essentially entering a new, and extremely expensive, market.

Even so, Intel is doing poorly, just not as poorly. I also found it a little ironic that when Intel cut their dividend, they were saying they can't invest all this money and being paying such a high dividend percentage. The only reason that percentage was high was because the stock price collapsed! The payout percentage wasn't high until that happened, so Pat is a bit slick. Ok, a lot slick.

Their revenue is what concerns me the most. Yeah, it's up Q to Q, and higher than expected, and they are guiding up for Q3. But, they're still down from Q2 2022, which was down from Q2 2021, and both by a considerable degree. So, it's better than expected, yeah, and they're out of the bottom, but things are still far from rosy. Even so, the stock is still priced to reflect that.

 

[hist78]

View attachment 1321

Operating Profit/Loss in 2nd Quarter 2023:
- TSMC operating profit: US$6.58 billion
- Intel: 5th straight operating loss: (US$1.02 billion)
- Samsung chip division: 2nd straight operating loss: (₩4.36 trillion) (US$3.32 billion)

Since 2009, Intel net profit margin was between 20% and 30% then dropped to negative in 2023. TSMC net profit margin was running between 30% and 35% but started moving up toward 43% since 2020.


Intel net profit margin:

TSMC net profit margin:

Source: www.macrotrends.net

 

[blueone]

Since 2009, Intel net profit margin was between 20% and 30% then dropped to negative in 2023. TSMC net profit margin was running between 30% and 35% but started moving up toward 43% since 2020.


Intel net profit margin:

View attachment 1324


TSMC net profit margin:

View attachment 1325

Source: www.macrotrends.net

Exactly. One of the reasons I originally bought a block of TSMC stock. Not a very good investment by my standards though. I sold the entire position recently on an up day. ASML too, which was even a worse investment.

 

[ta152h]

Exactly. One of the reasons I originally bought a block of TSMC stock. Not a very good investment by my standards though. I sold the entire position recently on an up day. ASML too, which was even a worse investment.

I'm so tempted to short sell AMD today, before earnings. What do you think?

 

[blueone]

I'm so tempted to short sell AMD today, before earnings. What do you think?

IMO, AMD is waaay over-valued, but that's not really the salient point, is it? And it's up ahead of the earnings announcement. The question is, what will institutional holders think about earnings? It's a tough call. I'm a buy and hold kind of guy, I never short, and I suspect the big holders will like what they see, so it may go up even more. Look at Arista Networks today. Just ridiculous. But AMD is over-valued this morning, no less after a bump. I wouldn't short AMD, but I can see where you might want to.

 

[ta152h]

IMO, AMD is waaay over-valued, but that's not really the salient point, is it? And it's up ahead of the earnings announcement. The question is, what will institutional holders think about earnings? It's a tough call. I'm a buy and hold kind of guy, I never short, and I suspect the big holders will like what they see, so it may go up even more. Look at Arista Networks today. Just ridiculous. But AMD is over-valued this morning, no less after a bump. I wouldn't short AMD, but I can see where you might want to.

I rarely short, but I have shorted AMD twice this year and made money.

I'm thinking with Intel eating into their market share, it might be ugly for AMD. And certainly, with all their discounts on products, one after another, they aren't looking good. But, you're completely right, you look at something and it's horrible, but then the market reacts differently. I think sometimes it's because "analysts" don't want to admit they were wrong.

I'll probably skip it, because I hate shorting, but my mind is really telling me to. But, my heart just doesn't want to. Normally, when I follow my intellect I'm better off than my intuitions, which generally suck.

I guess we'll see soon enough. I just have the feeling all the analysts are going to spin it so it appears to be far better than it is.

 

[blueone]

I'll probably skip it, because I hate shorting, but my mind is really telling me to. But, my heart just doesn't want to. Normally, when I follow my intellect I'm better off than my intuitions, which generally suck.

I know the feeling well. When INTC was under $26 for a short time, my mind was telling me B-U-Y!! But I know many of the senior players well, too well perhaps, and I've seen them make too many d-u-m-b decisions. I didn't buy, but obviously I wish I did. A few folks asked me for advice, and I told them I thought INTC was way over-sold. They took my advice and told me when it hit $35 I was a genius. No, I wasn't. Not then, not now.

 

[blueone]

AMD's results announced today don't impress me much, but the stock is up about 6% in after hours trading. I thought this might happen.

 

[Daniel Nenni]

AMD's results announced today don't impress me much, but the stock is up about 6% in after hours trading. I thought this might happen.

Gaming sector revenue was down 4% from last year to $1.6 billion, while data center fell 11% to $1.3 billion. The client segment, which includes personal computing, saw a 54% revenue decline to $998 million while embedded segment revenues rose 16% to $1.5 billion.

Looking into the current financial year, AMD said it sees first quarter revenue in the region of $5.7 billion, plus or minus $300 million, with gross margins of around 51%. Refinitv forecasts had the third quarter revenue forecast at $5.82 billion.

“We delivered strong results in the second quarter as 4th Gen EPYC and Ryzen 7000 processors ramped significantly,” said Su. “Our AI engagements increased by more than seven times in the quarter as multiple customers initiated or expanded programs supporting future deployments of Instinct accelerators at scale."

"We made strong progress meeting key hardware and software milestones to address the growing customer pull for our data center AI solutions and are on-track to launch and ramp production of MI300 accelerators in the fourth quarter,” she added.

 

[ta152h]

AMD's results announced today don't impress me much, but the stock is up about 6% in after hours trading. I thought this might happen.

Yes, the earnings were absolutely dismal, but the stock is up. Horrible earnings. Revenue down a lot from last year, and flat from Q1, while Intel revenue was up from Q1.

But, people spin things so they don't look like they were wrong. And just say AI somewhere, and then you look like you have a great future. What utter nonsense.

 

[Piefkee]

So does Intel and the rest of the market.
In general Intel and AMD having the same market issues, but personally intel has a lot more do to then AMD.
Products:
- DC currently is not competitiv against AMD or ARM
- GPU devision is a mess
- client, competitive against AMD and favour of a lot of capacity
Process:
- No "real" big foundry costumers yet
- a lot of investment needed
- Intel 4nm, 3nm and smaller exist only on powerpoint slides

The big question is can Intel execute on their roadmap to stay competitive on products and process?
Intel needs to worry on two things and AMD only at one thing
Thats the difference...

 

[Piefkee]

Also to add if a foundry costumer is putting eggs in Intels prozess, how can intel ensure that their own products not cannibalize the costumers products if capacity limited? That is big red flag

 

[ta152h]

So does Intel and the rest of the market.
In general Intel and AMD having the same market issues, but personally intel has a lot more do to then AMD.
Products:
- DC currently is not competitiv against AMD or ARM
- GPU devision is a mess
- client, competitive against AMD and favour of a lot of capacity
Process:
- No "real" big foundry costumers yet
- a lot of investment needed
- Intel 4nm, 3nm and smaller exist only on powerpoint slides

The big question is can Intel execute on their roadmap to stay competitive on products and process?
Intel needs to worry on two things and AMD only at one thing
Thats the difference...

So, Intel can make money from customers who don't even buy their CPUs, and that's a bad thing?

Intel has historically had no problems staying competitive on products and processes. They did fall behind on processes, kind of, since their processes are still the best for their client group, and have been, but overall, yes. Their design has not had any problems at all.

It's not one person dividing themselves in two, which is how this is often categorized. It's a company, and companies have multiple people, and those people can focus on their business segment. In fact, one could easily argue that IFS would be in horrible shape if not for the fact it's part of the same company, which prevented it from great harm.

And if you don't think Intel dominating AMD in desktops and laptops has nothing to do with their fabs, I'd disagree. Having their own fabs not only allows them to make higher performing chips, but also to win in price wars because they don't to pay for TSMC's margins.

But, bottom line, Intel revenue up, AMD's not, on a quarter where AMD is historically up, after a horrible Q1. I mean, how could they not show revenue growth from Q1? But, they didn't, they were break even. Guidance was below expectations, Intel's was above. Intel talks about market share gains, AMD about how they are going to get aggressive in AI.

It's very possible for Intel to treat external customers and internal the same, but they have to show they can and will. So, I get your concern there, and it's valid, but I also believe it's something they can address, and better address. Or they shouldn't even try. I think Pat and the rest of them are very smart, and very aware of this concern, and will make sure it gets the proper attention, and is handled appropriately.

 

[delong.height]

- GPU devision is a mess

I don't really think that their GPU is a mess any more. Many software updates have been delivered to date, which provided a huge performance boost and stability since day launched.

- DC currently is not competitiv against AMD or ARM

Wait until Sierra Forrest and Granite Rapids come out early next year. At least what we got is that their Intel 3 has met their projected milestone on time.

No "real" big foundry costumers yet

No external big customers, but plenty of internal demand to get that IDM advantage.

a lot of investment needed

They are spread out across next couple of years, they are building shells, and fill the fab with equipment once there is sufficient demand. Also, there's chip act money from US and EU. Customers also need to purchase for capacity just like how they dealt with TSMC. So I wouldn't worry too much about it.

- Intel 4nm, 3nm and smaller exist only on powerpoint slides

There has been a number of good articles here and on internet which talks about Intel 4, 3, 20A, 18A. Also, you need to omit that 'nm' since no foundry in this industry still uses that term for their leading edge nodes (TSMC N3E, TSMC N7). Follow what the foundry players give. If Intel gave Intel 4, use Intel 4 (no 4nm as that create more confusion among people),

 

[Piefkee]

But, bottom line, Intel revenue up, AMD's not, on a quarter where AMD is historically up, after a horrible Q1. I mean, how could they not show revenue growth from Q1? But, they didn't, they were break even. Guidance was below expectations, Intel's was above. Intel talks about market share gains, AMD about how they are going to get aggressive in AI.

Both beat market expectation. They talk about market share gains but the data indicates otherwise.
Don't get me wrong i cheer for AMD and Intel, but the next year for Intel a make or break year.
Talk is cheap, look back to 10nm

I don't really think that their GPU is a mess any more. Many software updates have been delivered to date, which provided a huge performance boost and stability since day launched.

Wait until Sierra Forrest and Granite Rapids come out early next year. At least what we got is that their Intel 3 has met their projected milestone on time.

No external big customers, but plenty of internal demand to get that IDM advantage.

GPU? Head of devision left, basically all dGPU programm canceled. Competing with NV and AMD with double the amount of Die Area on same perf?
Rialto Bridge canceled, Falcon Shores only as GPU, not as APU(XPU) in 2025

Sierra Forrest in H1/24 as sad in their call, is losing to already launched bergamo parts.

 

[ta152h]

Both beat market expectation. They talk about market share gains but the data indicates otherwise.
Don't get me wrong i cheer for AMD and Intel, but the next year for Intel a make or break year.
Talk is cheap, look back to 10nm


GPU? Head of devision left, basically all dGPU programm canceled. Competing with NV and AMD with double the amount of Die Area on same perf?
Rialto Bridge canceled, Falcon Shores only as GPU, not as APU(XPU) in 2025

Sierra Forrest in H1/24 as sad in their call, is losing to already launched bergamo parts.

Expectations are meaningless when talking about whether a company is doing well or not. It just indicates if they are doing worse than expected by talking heads.

Bottom line, Intel Q2 revenues up sequentially, AMD even, in a quarter they historically are up, from a disastrous Q1.

Data does not indicate otherwise regarding whether Intel gained market share in the client group. AMD didn't even suggest otherwise.

The discrete GPU division is completely operational, and working Battlemage. New drivers are coming out all the time, with performance improvements, on existing models. I'm not sure where you're getting this from, but it contradicts reality in a pretty obvious way.

Also, Intel guided up for Q3, AMD down. Intel was up after earnings, AMD was down. Of course, the latter is far from definitive, but apparently most people agree that AMD had another hideous quarter.

Servers though, you certainly have some merit in what you are saying. I don't think Sierra Forest is going to be competitive, it's more like a proof of concept. The next iteration looks very, very competitive and competent. But, yeah, 144 small cores isn't really going to take the world by storm in 2024. Even so, it's a stopgap, and it looks like a prelude to very, very good small core chips. But, we'll have to see.

 

[hist78]

Pat Gelsinger's comments regarding lacking activities/public attention on Intel's AI products is interesting.

Intel CEO says 'the worst is behind us'

After posting two straight quarters of losses, Intel (INTC) returned to profitability in the second quarter. In an interview with Yahoo Finance Live, Intel CEO Pat Gelsinger says he thinks "the worst is behind us at this point and our raise to the guide for Q3 really portends to a more...

finance.yahoo.com

 

[mdsomuncu]

Can I ask a question? I'm only a college student so I don't know much but why did network revenue decline that much? From what I know it was the fastest growing sector for intel with double digit growth rate. And now it is the one that declined the most. Is this mostly cyclical or are they losing market share in that sector too?

 

[ta152h]

Can I ask a question? I'm only a college student so I don't know much but why did network revenue decline that much? From what I know it was the fastest growing sector for intel with double digit growth rate. And now it is the one that declined the most. Is this mostly cyclical or are they losing market share in that sector too?

When you say "too", where else are they losing market share? They're steady in servers, finally, and they gained market share in client computing and now have significantly more than 80%, as they had around 83% - 85% before gaining share.

Intel said on the earnings call there was still some inventory that had to be caught up with. Also, earlier this year they exited 5G and LTE, and also exited their network switching business. So, it's unclear if they lost share, but I don't think whatever happened is cyclical in the sense it's expected seasonality. But, I really don't know for sure.

 

[blueone]

Can I ask a question? I'm only a college student so I don't know much but why did network revenue decline that much? From what I know it was the fastest growing sector for intel with double digit growth rate. And now it is the one that declined the most. Is this mostly cyclical or are they losing market share in that sector too?

There are several factors underlying Intel's problems in networking. Like ta152h said, Intel did shutdown the Barefoot Networks switch chip business, though I'm not sure how much of a revenue factor that was. I would guess Barefoot generated very little revenue in its short tenure in Intel, though the Torfino 2 chip struck me as a nicely innovative chiplet-based design fab'd on TSMC 7nm process for the logic chiplets, and I would guess another TSMC process (16nm?) for SERDES and other I/O. Torfino chips are programmable designs based on the P4 network programming language. P4 is also the basis for the AMD/Pensando NICs (excuse me, "DPUs", I hate that silly nomenclature) AMD acquired recently.

Intel's real networking revenue problems look to be in their server adapter business, where their 800-series Ethernet adapters are behind the competition, which is mostly Nvidia.

Intel co-designed a product called an "IPU" NIC with Google, which interestingly has 16 Arm Neoverse cores, and competes with the Nvidia Bluefield 2 DPU, but Intel hasn't said much about IPU revenue. Bluefield 2 has only 8 Arm cores (both products are 2x100GbE or 1x200GbE), but Nvidia has announced and sampled Bluefield 3, which has 16 Arm cores and 2x200GbE or 1x400GbE connectivity, while Intel hasn't announced an IPU follow-on. (I'm not sure if Bluefield 3 is generally available yet.) These IPU/DPU products represent the higher unit volume part of the cloud networking market (compared to switch ASICs), so falling behind here makes a bigger hit to revenue than Barefoot does. Intel definitely looks like it's falling further behind Nvidia. It's difficult to say where AMD is competitively. Microsoft acquired Fungible a while back (another DPU start-up) for a silly price, but I would guess that was more for the chip design team than the DPU product they had.

Given all of this, my outlook for Intel's networking group is dim, unless they have some leadership products in the works they haven't announced yet. (I doubt it.) Client 1GbE and 10GbE are either integrated (1GbE) or cheap (10GbE), so I doubt they generate much in the way of revenue.