News Intel 3Q23 Earnings

[Hitman928]

Will update once Intel posts the results.

@randomhero Saw your post in the other thread. I will respond here just by saying that Intel already pre-announced (a few weeks ago) that they expect to come in at the upper range of their Q3 guidance from last quarter. I'm hoping that beat their estimate entirely though as the guidance (on a historical basis) still wasn't that great.

Results dropped, looks very good compared to recent results, still down historically. Went above guidance from Q2 as I had hoped. Gross margins still not great though. Will dig in later when I have more time.

 

[Hitman928]

Giving Q4 guidance for $14.6B to $15.6B in revenue, small uptick in GM.

 

[H433x0n]

Some big statements on this earnings call.

• ARL 20A running on Windows
• 18A 0.9 PDK release for availability for external customers
• 18A Panther Lake Client & Clearwater Forest test chips in fabs in Q1 2024
• Major customer for 18A & Intel 3 w/ prepayment - said something about big die designs (hint?).
• Proclaimed there are 2 additional 18A customers - so there is a total of 3 customers for 18A.
• MTL initial shipments began in Q3
• EMR hit PRQ
• EUV HVM in both Oregon & Ireland
• GNR production stepping has taped-in and final ES stepping is in the fab now
• Synopsis partnership to make IP available for Intel 3 & 18A.
• They also signed 2 customers for packaging with 6 more in talks. Theres also an additional 18A customer that’s in talks with the deal expected to close by EOY.
• Pat responded to TSMC’s comments about N3P vs 18A and said they’ve come to a different conclusion but tried to be diplomatic.

Will update throughout the call.

 

[Hitman928]

Link to earnings slide show.

https://static.seekingalpha.com/uploads/sa_presentations/920/96920/original.pdf

Consumer market drove both earnings and operating margin. Data center still struggling. Foundry services is up in revenue but not in profits. Foundry revenue from tool sales and packaging work, seems no significant foundry customers (who are actually running product) yet. Intel seems to really be targeting 18a for meaningful foundry customers.

 

[Markfw]

Link to earnings slide show.

https://static.seekingalpha.com/uploads/sa_presentations/920/96920/original.pdf

Consumer market drove both earnings and operating margin. Data center still struggling. Foundry services is up in revenue but not in profits. Foundry revenue from tool sales and packaging work, seems no significant foundry customers (who are actually running product) yet. Intel seems to really be targeting 18a for meaningful foundry customers.

I read that whole thing. I don't see where they have broken the numbers down by department and Y/Y Exactly how is data center doing ? and customer division ?

 

[jpiniero]

Intel actually made more in Notebook compared to last year. Still nowhere near 2019 though.

 

[Hitman928]

I read that whole thing. I don't see where they have broken the numbers down by department and Y/Y Exactly how is data center doing ? and customer division ?

Slide 6 is consumer, slide 7 is data center.

 

[H433x0n]

Intel actually made more in Notebook compared to last year. Still nowhere near 2019 though.

Wonder how much of that is MTL.

 

[jpiniero]

Wonder how much of that is MTL.

Not much, I'm guessing.

 

[Markfw]

Slide 6 is consumer, slide 7 is data center.

Thanks, don't know how I missed that. So desktop and server are both down Y/Y.

They still need to make a decent server chip. I hope that improvements in process tech will enable them to make a decent desktop chip (in efficiency)

 

[Hitman928]

Further segment breakdown:

 

[Geddagod]

ARL 20A running on Windows

MTL took 1.5 years after booting up windows to launch

18A Panther Lake Client & Clearwater Forest test chips in fabs in Q1 2024

Seems like both chips will launch EOY 2025. Appears to be ~1.5 year cadence for server, DMR then appears like it would launch ~1/2 year after Zen 6.

Proclaimed there are 2 additional 18A customers - so there is a total of 3 customers for 18A.

18A looks legit tbh

GNR production stepping has taped-in and final ES stepping is in the fab now

Very much on track, unless there are some very last minute validation screw ups, I think it's looking really good. Appears to be ~8 months from final ES to Product Launch.

 

[H433x0n]

18A looks legit tbh

A prepay for 18A and a total of 3 (potentially 4 customers) by EOY 2023 certainly seems like a good sign. Pat Gelsinger mentioned on the call that they estimate they have the best packaging tech in the world was a bold statement. He was similarly bold on the N3P vs 18A comments as well.

Very much on track, unless there are some very last minute validation screw ups, I think it's looking really good. Appears to be ~8 months from final ES to Product Launch.

I don’t recall him mentioning SRF though so there’s that.

 

[Henry swagger]

18A is looking good wonder which big customers are ? 🤔 ..

 

[randomhero]

Thank you @Hitman928!

I didn't know about preanno

Will update once Intel posts the results.

@randomhero Saw your post in the other thread. I will respond here just by saying that Intel already pre-announced (a few weeks ago) that they expect to come in at the upper range of their Q3 guidance from last quarter. I'm hoping that beat their estimate entirely though as the guidance (on a historical basis) still wasn't that great.

Results dropped, looks very good compared to recent results, still down historically. Went above guidance from Q2 as I had hoped. Gross margins still not great though. Will dig in later when I have more time.

View attachment 87853

Thanks you!

I didn't know about preannouncement. My guess was right. Client segment was down, with caveat that they had better operating income.They managed to clear stock of old products.
Enterprise is abysmal, half billion down in revenue.

Only good news is that their 18A node is in good shape, apparently. I will reserve judgment on that one when we see products on shelves from foundry customers. I wish IFS good luck.

 

[clemsyn]

When will they reveal their foundry customers?

 

[randomhero]

When will they reveal their foundry customers?

That is billion dollar question, literally😉

 

[moinmoin]

Pat Gelsinger mentioned on the call that they estimate they have the best packaging tech in the world was a bold statement.

That's a statement Intel has been making for years even before Pat, think Lakefield etc. I think the issue Intel is facing is that it doesn't need the best packaging tech, but useful economically viable packaging tech, and there Intel itself failed up to now. Will be very interesting to see whether foundry customers can make better use of it.

 

[gdansk]

Only lost a few million and the market is quite happy.

 

[Doug S]

When will they reveal their foundry customers?

Those customers may be hedging their bets by designing for both Intel and TSMC, in case Intel can't deliver on its commitment to them. That's something they wouldn't want TSMC to know, as they could get pushed back in the line in favor of more "committed" customers if demand exceeds supply on the leading edge like it did a year or two ago.

It is probably safe to guess those customers are Qualcomm and Nvidia since this is talking about leading edge capacity right out of the gate. They would dual source, and if one foundry outperforms the other then they have a higher end version and lower end version of their design to sell.

 

[jpiniero]

QC and nVidia are like the least likely to bother with 18A. It's probally the usual suspects like Erickson or perhaps some DoD stuff.

 

[gdansk]

Unless it's weird research DoD projects they prefer old, proven nodes with expanded operating temperatures and radiation tolerance. And research projects are unlikely to be considered major.

 

[itsmydamnation]

QC and nVidia are like the least likely to bother with 18A. It's probally the usual suspects like Erickson or perhaps some DoD stuff.

Could also see , Cisco , Palo , HP/aruba. Cisco were first on board the last time Intel tried this before the real first iteration of 10nm did 10nm things.

 

[Doug S]

QC and nVidia are like the least likely to bother with 18A. It's probally the usual suspects like Erickson or perhaps some DoD stuff.

Why wouldn't Nvidia want it? They can sell all the AI stuff they can make so a second supplier - especially if it is was the best performing node at the time it starts mass production - would be something I'd think they'd want if Intel is able to pull off the claimed schedule.

And Qualcomm was on Intel's original list of "partners" and if nothing else they'd like a second source even at modest volumes just to keep TSMC honest.

 

[dullard]

When will they reveal their foundry customers?

They already have revealed some of them.

Department of Defense, Boeing, Northrop Grumman: https://www.intel.com/content/www/u...ense-industrial-base-customers.html#gs.7hza5v

Ericsson: https://www.intel.com/content/www/u...ollaboration-5g-infrastructure.html#gs.7hzb0x

Arm (although technically it would be Arm's customers): https://www.intel.com/content/www/u...on-collaboration-leading-edge-soc-design.html

Synopsys: https://www.intel.com/content/www/us/en/newsroom/news/intel-partnership-news-aug-2023.html#gs.7hzfva

Tower Semiconductor (after Intel's purchase of them fell through): https://www.intc.com/news-events/pr...services-and-tower-semiconductor-announce-new

Nvidia is a maybe: https://www.tomshardware.com/news/nvidia-ceo-intel-test-chip-results-for-next-gen-process-look-good