News AMD 4Q25 Earnings

[poke01]

Just the opposite. They really cut down the clock speed in portable mode. Would have been nice with some Samsung node later than 8 nm.

Does it matter for a $450 product? No

Still sold 17 million units. These are not premium devices
Nintendo likes to cut BoM a lot to reach a specific price point.

 

[fastandfurious6]

Handhelds don’t need more than 8nm for the mass market. Also Nintendo has ip

GAMING handhelds don't need more than 8nm and apple iphones for instagram need 2nm !!!!!!???

!!!!

 

[poke01]

GAMING handhelds don't need more than 8nm and apple iphones for instagram need 2nm !!!!!!???

!!!!

yep. Unlike handhelds, iPhones are a much bigger market. Makes even Nintendo look irrelevant in terms of volume shipped.

News flash, gaming hardware is a very small market

 

[adroc_thurston]

News flash, gaming hardware is a very small market

Oh it's not small at all.

 

[fastandfurious6]

[GAMING handhelds don't need more than 8nm and apple iphones for instagram need 2nm !!!!!!???]

yep. Unlike handhelds, iPhones are a much bigger market. Makes even Nintendo look irrelevant in terms of volume shipped.

News flash, gaming hardware is a very small market

News flash, you keep admitting very weird pro-apple bias even when completely meaningless

Forcing nearly every game studio to start long projects just to port their game on switch 2 abysmal 8nm Tegra RTX2050 equivalent, only because it's Nintendo and it will sell because it's Nintendo... is that normal for you?

 

[Joe NYC]

$AMD currently down 14%, Micron dipping below $400, looks like the smart money is starting to hedge their AI bets.

There is this crazy arrangement for NVidia to invest $100 billion in OpenAI.

OpenAI started to buy stuff (including big order with AMD) like they have the $100 billion in their pocket. And now Jensen is seen as backing out of this commitment.

The biggest opportunity for AMD is the deal with OpenAI, and the biggest cloud is if OpenAI can pay for this big order with AMD.

It seems this situation has as much or more to do with the stock action of AMD stock today than the earnings report yesterday.

 

[adroc_thurston]

There is this crazy arrangement for NVidia to invest $100 billion in OpenAI.

OpenAI started to buy stuff (including big order with AMD) like they have the $100 billion in their pocket. And now Jensen is seen as backing out of this commitment.

The biggest opportunity for AMD is the deal with OpenAI, and the biggest cloud is if OpenAI can pay for this big order with AMD.

It seems this situation has as much or more to do with the stock action of AMD stock today than the earnings report yesterday.

no the stock sell-off is AMD not posting a massive Q1 guide despite all the ML capex bumps

 

[poke01]

ews flash, you keep admitting very weird pro-apple bias even when completely meaningless

Forcing nearly every game studio to start long projects just to port their game on switch 2 abysmal 8nm Tegra RTX2050 equivalent, only because it's Nintendo and it will sell because it's Nintendo... is that normal for you?

im just giving a taste of what you sound like with your AMD bias. Annoying isn’t?

You have no idea how game development works… yes game devs ship on volume hardware and Switch 2 is essentially 2020 hardware sold at a price Nintendo thinks the market can bare.

If you don’t like that Sony has a plan in 2027 and is happy to offer you an AMD N3 SoC handheld but it’s going to be more expensive.

 

[fastandfurious6]

You have no idea how game development works

You have no idea how impossible it is to make AAA games from 2022 run on nintendo switch 2 (2025) with fluid framerate at just 30FPS. Tegra is very weak, even upgraded at 8nm / Ampere.

I expect to see more headlines like this:

Elden Ring Switch 2 Gamescom performance called a "disaster", raising concerns about third-party games

Demanding third-party Switch 2 games are arriving on the console, with sometimes mixed results. Gamescom 2025 attendees worry about how Elden Ring Tarnished Edition will perform on the handheld. In open areas, impressions highlight frame rates that plummet well below 30 fps.

www.notebookcheck.net

Bare minimum would be a steam deck equivalent - AMD Zen2 APU which is already old, we're at Zen6 this year - but Tegra is like 3 times slower than Zen2... Nintendo execs picked it for "compatibility" with Switch 1... and obviously profit $$$$

 

[tajoh111]

AMD guidance wasn't good enough as the price of the stock already included them beating expectations. The beat was not strong enough. AMD stock price was double what it was a year ago. Even this quarters revenue beats vs from the previous guidance was not that strong. The previous guidance did not expect any Chinese revenue because of trade restrictions. That was lifted which was built into the price during the previous months. China added about 400 million during Q4. Considering their previous outlook was 9.6 billion, add in Chinese revenue of 400 million and AMD was still in their guidance of 9.6 billion plus or minus 300 million. This is not beating expectations and for stocks to rise, they need to beat it by more than this because the market has already priced AMD beating expectations which has lead to the stock gaining value.

Nvidia earnings generally beat their guidance which generally seem optimistic because they show much more growth quarter to quarter. And they still beat it. E.g they projected 54 billion but had 57 billion in revenue. Even with this beat Nvidia lost 5% or so. People expected better from AMD considering how much the stock has risen over the year and being the smaller player, AMD has more room for growth as Nvidia has saturated the market(own the market). It is expected AMD can take Nvidia's marketshare and grow their own modest sales. This is not shown with such so modest outlook. Nvidia is expecting 65 billion which they will likely still beat, which has lead to Nvidia meteoric rise.

AMD guidance was weak and has them not even beating their current quarter which now has Chinese revenue coming in from data center. Along with the current quarter not really beating their guidance, it should be no surprise this resulted in a big stock drop(the biggest single day loss since Lisa Su became CEO) when the driving force for stock value is growth.

Take off your AMD glasses and you will see AMD financial have done little to justify it's stock price jumping from 100 dollars 1 year ago to its peak of 252 dollars.

 

[gdansk]

Take off your AMD glasses and you will see AMD financial have done little to justify it's stock price jumping from 100 dollars 1 year ago to its peak of 252 dollars.

If all you care about is stock price, it's hard to justify the current prices of:
Seagate,
Western Digital,
Micron,
AMD,
Qualcomm

Some of them don't have huge sell offs following the recent record quarters but most did. And I suspect AMD will still have far better margins than Seagate and WD in 2030, but we'll see.

 

[Josh128]

Take off your AMD glasses and you will see AMD financial have done little to justify it's stock price jumping from 100 dollars 1 year ago to its peak of 252 dollars.

AMD reached much higher than $100 more than a year ago. It hit $150ish in the COVID era. They've more than doubled their quarterly revenue from then to now, I'd say thats more than "little", lol.

 

[Doug S]

There is this crazy arrangement for NVidia to invest $100 billion in OpenAI.

OpenAI started to buy stuff (including big order with AMD) like they have the $100 billion in their pocket. And now Jensen is seen as backing out of this commitment.

The biggest opportunity for AMD is the deal with OpenAI, and the biggest cloud is if OpenAI can pay for this big order with AMD.

It seems this situation has as much or more to do with the stock action of AMD stock today than the earnings report yesterday.

Look at the action on Microsoft, Meta, Amazon and Google in recent days. Investors are starting to question the crazy amounts they're looking at spending on AI capacity. That's bad for companies selling the AI chips. Especially the "second choice" AI chip seller after Nvidia, which is AMD.

I'm sure OpenAI stock would have been pummeled in the past week for the same reasons even more than those other names, had they been publicly traded.

 

[branch_suggestion]

AMD guidance wasn't good enough as the price of the stock already included them beating expectations. The beat was not strong enough. AMD stock price was double what it was a year ago. Even this quarters revenue beats vs from the previous guidance was not that strong. The previous guidance did not expect any Chinese revenue because of trade restrictions. That was lifted which was built into the price during the previous months. China added about 400 million during Q4. Considering their previous outlook was 9.6 billion, add in Chinese revenue of 400 million and AMD was still in their guidance of 9.6 billion plus or minus 300 million. This is not beating expectations and for stocks to rise, they need to beat it by more than this because the market has already priced AMD beating expectations which has lead to the stock gaining value.

Nvidia earnings generally beat their guidance which generally seem optimistic because they show much more growth quarter to quarter. And they still beat it. E.g they projected 54 billion but had 57 billion in revenue. Even with this beat Nvidia lost 5% or so. People expected better from AMD considering how much the stock has risen over the year and being the smaller player, AMD has more room for growth as Nvidia has saturated the market(own the market). It is expected AMD can take Nvidia's marketshare and grow their own modest sales. This is not shown with such so modest outlook. Nvidia is expecting 65 billion which they will likely still beat, which has lead to Nvidia meteoric rise.

AMD guidance was weak and has them not even beating their current quarter which now has Chinese revenue coming in from data center. Along with the current quarter not really beating their guidance, it should be no surprise this resulted in a big stock drop(the biggest single day loss since Lisa Su became CEO) when the driving force for stock value is growth.

Take off your AMD glasses and you will see AMD financial have done little to justify it's stock price jumping from 100 dollars 1 year ago to its peak of 252 dollars.

That is a lot of words to miss the point, explosive growth ain't happening until Helios ramp is in swing and MI500 ramp will supercharge this.
Trader is an anagram of, well you can figure that one out. AMD forever has been a long hold stock, trading it any other way is not conducive to success, things will look quite different in 1-2 years.
Oh, and CN revenue next Q is basically flat, seasonality is the reason for the dip somewhat offset by Instinct growth.

 

[Markfw]

All of this talk of stock price, but here is my take. AMD was one a small player in the CPU world (less than 10% or so) and now is almost 50/50 with Intel, and may surpass them shortly. This alone makes any talk of stock price most likely just guessing until they actually take the lead, but regardless if they do, they have proven for the last several years, they are more than just a "player" in this CPU world. They are a serious leader.

 

[DrMrLordX]

All of this talk of stock price, but here is my take. AMD was one a small player in the CPU world (less than 10% or so) and now is almost 50/50 with Intel, and may surpass them shortly. This alone makes any talk of stock price most likely just guessing until they actually take the lead, but regardless if they do, they have proven for the last several years, they are more than just a "player" in this CPU world. They are a serious leader.

AMD makes things people need first and things people want second. Their stock price is all over the place due to the oftentimes bizarre calculations of investors that are caught up in hype cycles and/or their own hubris more than they are in corporate fundamentals.

I dunno if being an AMD shareholder is a good thing or a bad thing. Actually being AMD is (apparently) a good thing, if you like money.

 

[itsmydamnation]

i also think Lisa has GCN truma , never wants to even have the slightest chance to have so much unsellable inventoy again.

 

[Joe NYC]

That is a lot of words to miss the point, explosive growth ain't happening until Helios ramp is in swing and MI500 ramp will supercharge this.
Trader is an anagram of, well you can figure that one out. AMD forever has been a long hold stock, trading it any other way is not conducive to success, things will look quite different in 1-2 years.
Oh, and CN revenue next Q is basically flat, seasonality is the reason for the dip somewhat offset by Instinct growth.

My take on the China sales:
Q4 guide: $0
Q4 actual revenue $390m
Q4 inventory reserves used up $360m, $440m remaining

Q1 guide Mi308: $100m
Q1 guide Mi325: $0

What will actually happen in Q1 - I think a lot more than $100m of Mi308 will be sold and > $0 of Mi325 will be sold. Probably all of the Mi308 will be sold by the end of Q2, at whatever price AMD can get for them.

 

[adroc_thurston]

Look at the action on Microsoft, Meta, Amazon and Google in recent days. Investors are starting to question the crazy amounts they're looking at spending on AI capacity

The market is weird.
Insane opex guides but the component supply chain (NVDA/AMD/AVGO/friends) is flat or down?

 

[mikegg]

Look at the action on Microsoft, Meta, Amazon and Google in recent days. Investors are starting to question the crazy amounts they're looking at spending on AI capacity. That's bad for companies selling the AI chips. Especially the "second choice" AI chip seller after Nvidia, which is AMD.

I'm sure OpenAI stock would have been pummeled in the past week for the same reasons even more than those other names, had they been publicly traded.

The cap ex increase by big tech is justified by AI demand. Wallstreet just decided to derisk the last few days.

 

[adroc_thurston]

The cap ex increase by big tech is justified by AI demand. Wallstreet just decided to derisk the last few days.

where's da ROIC be at for all that GEMM slop

 

[jpiniero]

Think I saw some speculation that some Data Center buildouts are in trouble because they can't get the bonds sold or something like that.

Maybe at least some on Wall Street have finally realized that blowing money on a Circlejerk is a bad idea.

 

[Josh128]

where's da ROIC be at for all that GEMM slop

Right here.

 

[adroc_thurston]

Right here.

Good lord yes, to afford 200B capex they'll have to lay off at least 30k heads on top of that.

 

[adroc_thurston]

Think I saw some speculation that some Data Center buildouts are in trouble because they can't get the bonds sold or something like that.

It's an Oracle issue.