AMD CPUs had 92% Market Share at German PC Hardware Retailer in January

[_roman_]

Not only MFactory, and not only to DE. "Lieferung in weitere Länder auf Anfrage."

That is totally unrelated. I'm sorry

This is german text, last changes were done 7 days ago: https://www.preisjaeger.at/deals/info-deal-mit-guteschein-allespost-deutschland-kommt-326987

I will translate it in my words.

Germany does not sell goods to Austria? Pay me cash so you get a german shipment address. We repack the order and send it to joo, for a fabulous fee of 20 burgers.

Note: There are more tricks and services. Preisjaeger is kinda trustworthy on those shaddy tricks.

Note: Just that these services exists - that there are pages and such how to deal with such stuff - is proof enough.

Mindfactory is an instantly do not buy shop. Same with alternate.de // Check the net about that topic.

edit: just saw #37 explain it very well also. - how that shipment works

 

[mkppo]

Intels terribly regressive architecture beats or matches amds top end in everything, gaming includes. The review from this very site shows the 285k being a better product (especially when it comes to efficiency) than the 9950x. And yet, marketing wins once again and amd sells more. What can you do.

It's slower than the 9950x in both games and productivity but close enough in the latter. In games, increase that difference because 24H2 wasn't tested as 285K had issues with it during launch. Subsequent testing at other sites revealed that the 285K didn't get any performance boost from it but AMD certainly did and it was pretty significant, more so than the patches and updates intel brought to the 285K.

The other problem is it regressed in gaming performance to the prior generation quite significantly as 7950x3d/14900k are easily faster in gaming while being close enough in productivity. So instead of moving the needle forward, they took a step forward and a step back. That's reflected in the sales numbers, the DIY numbers are pretty bad. They're doing everything they can to push it to SI's though at heavily discounted prices.

If by marketing you mean Intel's bad press for the 285K and all, they brought it upon themselves.

 

[The Shield]

This is a real BLESSING for the users. Intel has been caught in shenanigans for decades, it is actually a miracle that AMD survives after all the anti-competitve programs directed against its operations.

In general terms, I agree with you - in fact, I've been purchasing AMD for 20 years - both CPU and GPU (starting with Athlon 64 and ATI Radeon). But is now crystal clear that what we THINK we were seeing with Intel (14 nm process production over and over and over, no competition whatsoever, etc.), is nothing compared to what AMD and Nvidia are doing.
I mean, during the old intel-days, have you ever seen a "gaming CPU" (like 3770K, 4770k, 7700k, etc.) sold with the marketing BS of the "demand-larger-than-supply"? Because the 9800X3D - a simple 8 core - il sold at scalper-price, IF you can find IT.
Intel NEVER did something like that, even when was dominating the market.

 

[Tek-Check]

I find it weird that reports are always from this one store and always give 90%+ sales for AMD chips while I have contacts in distribution, and it doesn't look anything close in many other places.

Sure. Give us numbers from other places.

 

[JustBenching]

It's slower than the 9950x in both games and productivity but close enough in the latter. In games, increase that difference because 24H2 wasn't tested as 285K had issues with it during launch. Subsequent testing at other sites revealed that the 285K didn't get any performance boost from it but AMD certainly did and it was pretty significant, more so than the patches and updates intel brought to the 285K.

The other problem is it regressed in gaming performance to the prior generation quite significantly as 7950x3d/14900k are easily faster in gaming while being close enough in productivity. So instead of moving the needle forward, they took a step forward and a step back. That's reflected in the sales numbers, the DIY numbers are pretty bad. They're doing everything they can to push it to SI's though at heavily discounted prices.

If by marketing you mean Intel's bad press for the 285K and all, they brought it upon themselves.

Feels like you and @Vayra86 are not looking at the reviews, im using the data provided by TPU. The 285k is faster and more efficient (a lot) in ST workloads, as fast as the 9950x in games (2% difference) while being a lot more efficient and matching the 9950x in MT performance and efficiency. There is nothing that the 9950x is better at.

Anyways, as i''ve said in my previous post, the same people that are buying 90% nvidia gpus are the same people buying 90% amd cpus.

 

[mkppo]

Feels like you and @Vayra86 are not looking at the reviews, im using the data provided by TPU. The 285k is faster and more efficient (a lot) in ST workloads, as fast as the 9950x in games (2% difference) while being a lot more efficient and matching the 9950x in MT performance and efficiency. There is nothing that the 9950x is better at.

Anyways, as i''ve said in my previous post, the same people that are buying 90% nvidia gpus are the same people buying 90% amd cpus.

I did. TPU's review:

9950X is faster in both and as I said, you have to increase the gap in games because of 24H2. So it's clearly better at both, more so in games, so I have no idea what you mean by 9950X is not better at anything. 285K is more efficient and faster in ST workloads? Sure, not a new thing either. But your initial statement wasn't that.

Edit: Where are you getting the ' A lot more efficient in MT' from? That same review has the 9950X more efficient in MT workloads.

 

[Tek-Check]

If I wanted one I'd definitely buy somewhere else.

If... exactly.
It does not look like many want it from anywhere else either.

 

[Vayra86]

Feels like you and @Vayra86 are not looking at the reviews, im using the data provided by TPU. The 285k is faster and more efficient (a lot) in ST workloads, as fast as the 9950x in games (2% difference) while being a lot more efficient and matching the 9950x in MT performance and efficiency. There is nothing that the 9950x is better at.

Anyways, as i''ve said in my previous post, the same people that are buying 90% nvidia gpus are the same people buying 90% amd cpus.

I am looking at the reviews, that's why your statements are so puzzling

MT perf

Application average

Only in ST, the 285k uses 10W less, or some 30% less which I agree is substantial. Now I'm curious what ST workloads you have that warrant that to be an argument here, using an i9 with 24 threads. Overall, these CPUs are quite the same in terms of power consumption. Even if you would certainly run more MT workloads where you'll easily lose the 10W you might have gained momentarily.

But the 9950x is not just the equal of the 285K (I think the only real differentiator here, is price of the whole platform/build), it also has 32 threads instead of 24, and 16 full fat cores instead of just 8. Intel is not faster, does not offer more to work with, and is not notably more efficient. So I can understand why people land at the 9950x for a build that needs to have lots of threads. After all, a review is not putting a varied continuous load on the CPU, but tests scenarios, and its clear the 285K will find its limits faster than a CPU that has 8 additional threads AND 8 additional cores that are fit for anything; an E core still isn't a fully functional core.

When it comes to a system for primarily just gaming, the 285K should not even be on your shopping list though.

And last, the incessant need to keep pulling every debate into your shitty little fanbase camp mentality is getting very tiring. We get it, you disagree with the way people look at things. Welcome to the world.

 

[JustBenching]

Only in ST, the 285k uses 10W less, or some 30% less which I agree is substantial. Now I'm curious what ST workloads you have that warrant that to be an argument here, using an i9 with 24 threads. Overall, these CPUs are quite the same in terms of power consumption. Even if you would certainly run more MT workloads where you'll easily lose the 10W you might have gained momentarily.

But the 9950x is not just the equal of the 285K (I think the only real differentiator here, is price of the whole platform/build), it also has 32 threads instead of 24, and 16 full fat cores instead of just 8. Intel is not faster, does not offer more to work with, and is not notably more efficient. So I can understand why people land at the 9950x for a build that needs to have lots of threads. After all, a review is not putting a varied continuous load on the CPU, but tests scenarios, and its clear the 285K will find its limits faster than a CPU that has 8 additional threads AND 8 additional cores that are fit for anything; an E core still isn't a fully functional core.

When it comes to a system for primarily just gaming, the 285K should not even be on your shopping list though.

And last, the incessant need to keep pulling every debate into your shitty little fanbase camp mentality is getting very tiring. We get it, you disagree with the way people look at things. Welcome to the world.

None of what you said is different to what I said - although you forgot - besides the ST and idle efficiency, the 285 is also way more efficient in gaming (while performance is within margin of error).

 

[Neo_Morpheus]

I find it weird that reports are always from this one store and always give 90%+ sales for AMD chips while I have contacts in distribution, and it doesn't look anything close in many other places.

I think they are the only one in the world that publishes such results.

The rest doesn’t.

Because this is the reality, 80-90% of all CPU sales are for AMD, Mindfactory, Amazon, etc. everywhere where the sales are reported, but this is DIY. Still Intel have 75% of the x86 CPU market share, thanks to its schemes and anti-competition practices forcing their trashes across all OEMs forcing them to not sell AMD and there we are, when the people have choice they choose AMD, only the most desperate fanboys or completely clueless fooled by the Intel's fanboys buy Intel.

Look at CES 2025, 90% of all gaming laptops are equipped with Arrow Lake, imagine how deep is the Intel's cartel to force the worst gaming CPU that is on par with 2022 Alder Lake in all these gaming laptops

Previous company I worked only bought Dell and by consequence intel. I tried to diversify the fleet with the introduction of some Lenovos with AMD and quickly was placed in a meeting with several vips who clearly were receiving some nice kickbacks from dell to keep all other vendors away.

And yes, intel is still bribing dell to ignore AMD. Even though, it looks like they ran out of money, since they apparently will be switching to AMD, per the CES keynote.

Yes, thats how bad AMD has it and why Intel is still alive.

But this still applies

 

[john_]

In general terms, I agree with you - in fact, I've been purchasing AMD for 20 years - both CPU and GPU (starting with Athlon 64 and ATI Radeon). But is now crystal clear that what we THINK we were seeing with Intel (14 nm process production over and over and over, no competition whatsoever, etc.), is nothing compared to what AMD and Nvidia are doing.
I mean, during the old intel-days, have you ever seen a "gaming CPU" (like 3770K, 4770k, 7700k, etc.) sold with the marketing BS of the "demand-larger-than-supply"? Because the 9800X3D - a simple 8 core - il sold at scalper-price, IF you can find IT.
Intel NEVER did something like that, even when was dominating the market.

Intel enjoys a huge production capacity, thanks to it's fabs. That's why you never seen "demand larger than supply" BS for it's CPUs. You can see it with the new graphics cards that are NOT produced in Intel's fabs, so Intel needs to wait in the line.
On the other hand AMD seems to be extremely conservative in how many units of a model it will produce. That's why they keep disappointing their investors. They don't seem to be able to take a risk and say "we expect this to sell well, let's build extra quantities and proved correct". Unfortunately this holds AMD back. AMD's financial results are usually spot on with what they announce and what most analysts expect and my belief is that is happening because as a company they are very conservative, very careful with how they will use their wafer supply. I don't believe selling a few batches of 9800X3D at $100 higher prices and losing the chance of selling double or triple those numbers, now especially that there is no Intel, makes much sense. You want people on AM5, as many as possibly, because this socket is here to stay for a few more years. They where probably saying "we expect the new Intel CPUs to sell well, so we expect to sell that many 9800X3Ds, let's not make more and end up with a huge inventory" so they probably didn't build enough to cover the extra demand. I believe that was their own excuse and it does sound like a honest reason, not like marketing BS. I don't think it is good marketing saying that your success is in part the result of someone else's failure. If you win a 100 meters race, because all of your opponents tripped, well, not much value in such win.
Also AMD has shown in the past that they can't really cover demand. They had problems with 5000 series success, they didn't manage to sell enough graphics cards, as Nvidia did, the period of the crypto mining craze, there where many rumors in the past that huge OEMs where not choosing AMD because AMD could not offer the quantities that companies like Dell for example wanted, so it wasn't always some Intel anticompetitive move, smaller manufacturers where coming out complaining that they weren't getting the full order of CPUs from AMD, that part of their order was delayed, now we see with 9800X3D supply constraints, in their previous financial they where saying that they couldn't even supply enough Instinct cards (which was the reason for the share price drop after their previous financial result), now we see them moving the 9070's closer to Nvidia's 5070 release date that looks like a suicidal move.
I don't think this is marketing BS. It's just fear of being in a situation of having a huge inventory that will have to be sold at much lower prices than expected. And unfortunately this is also the reason why AMD is not getting any bigger, it can't follow Nvidia's exploded success even by little.

 

[Visible Noise]

Of course Intel is like it's A.I. we need more A.I. AMD is like who care about A.I. speed is king.

This is the transcript of the latest AMD earnings call. I bet you aren’t able to count high enough the number of times they mentioned AI.

https://seekingalpha.com/article/4754875-advanced-micro-devices-inc-amd-q4-2024-earnings-call-transcript

You are delusional if you think they aren’t all in on AI. I mean it’s literally in the name of their new processors. Where have you been?

 

[Tek-Check]

That's why they keep disappointing their investors.

Some investors and analysts are aften "disappointed" by their own unrealistic expectations and predictions founded on bad math and lack of detailed insight. Successful companies do not need to behave like some investors seeking immediate gratification, like little children. Long-term, steady strategy often counts more than short-term higher gains.

Unfortunately this holds AMD back. AMD's financial results are usually spot on with what they announce and what most analysts expect and my belief is that is happening because as a company they are very conservative, very careful with how they will use their wafer supply.

You can call it conservative, but that is a prudent approach by AMD. They have showed consistently growing revenues over the years. Not meteoric growth, like Nvidia, but consistently growing flow of revenues. It's a different approach by Lisa's leadership. And rock solid so far.

You want people on AM5, as many as possibly, because this socket is here to stay for a few more years.

Around 50% of all CPUs sold are AM5 platform, and growing steadily. So, in just little over two years half of cumstomers have already adopted it over buying new or upgrading existing AM4 rigs. Considering the fact that longevity is the key feature of their platforms, 50% on AM5 CPUs is quite good. In 2027/28, they will have customers in different parts of the world using three platforms, including AM6, which will be amazing and unprescedented.

AM4 will gradually leave the scene in next couple of years, but many millions of users will still use it by 2030. Even today, one can buy new, excellent and cheap AM4 platform with a very good CPU for daily use. There is nothing wrong with it, quite opposite. So, manufacturers will simply stop producing boards and CPU for AM4 and this process will gradually reduce AM4 in use. There will still be a vibrant secondary market for a few more years.

now we see with 9800X3D supply constraints

The supply of 9800X3D has stabilized after the New Year in growing number of markets. I can see it in every major tech shop in Europe, North America and Australia.

in their previous financial they where saying that they couldn't even supply enough Instinct cards (which was the reason for the share price drop after their previous financial result)

They supplied what they could with the amount of available wafers to purchase. Obviously, it was not ideal in every segment, such as delays in laptop chips, but it is what it is. There's a long queue of customers at TSMC... and we do not have any other foundry like TMSC. Now, Zen5 CPUs are produced in new Arizona's factory. People will be soon able to see "Made in USA" on those chips.

now we see them moving the 9070's closer to Nvidia's 5070 release date that looks like a suicidal move.

It does not, really. It looks like a proper launch, with more mature software and real volume globally, so that everyone can actually buy a product, and not like catastrophic paper launch by Nvidia, scalping scandal and lack of any cards that followed.

And unfortunately this is also the reason why AMD is not getting any bigger, it can't follow Nvidia's exploded success even by little.

AMD is not getting any bigger? What kind of nonsense is this? Did you check their revenues and market capitalization in last 6 years? You have some homework to do.

Nvidia, of course, skyrocketed due to AI boom. Exploding success is not the only model of success to follow, and not the best one necessarily. Nvidia has just lost $600 billion in market cap in a few days due to DeepSeek release. It just shows how overly inflated its value is, and how easily market cap can be wiped out within days due to irrational hype in the Western investor circles.

And then you have OEMs putting those CPUs in like 50% of their PCs, keeping Intel's percentage high in the x86 market.

Sure, 50%, but not 80% anymore. Their market share is gradually melting, and consistently over the years. That is 100% true.
Don't expect any revolutionary, fast changes. This industry does not work like that.
Remember the words: gradually and consistently.

This is a real DISASTER for users.
The last thing we need is another monopoly based on demand-larger-than-supply-marketing-BS.

And what is it that you are talking about specifically?

Then why is AMD quoted to have a global market share of 28.7% for desktop PCs in Q4 2024, while 62% only in South Korea in Nov 2024? All this data is correct, but variations are too big to draw any conclusion.

Measures by different countries, market research companies, retailers and online tools will always yield different results. All those are approximations, depending on sampling data of CPU sales and PC attachment.

It looks like Mercury research has a close global approximation, which gives AMD ~30% on desktop. Steam gives it ~37% overall, PC Mark has just posted that AMD edged Intel in desktop, at ~50%, etc. Regional and geographical variations will be big from country to country, depending on historical factors, presence of tech companies, etc. It's complex.

 

[R0H1T]

the 285 is also way more efficient in gaming

There's no point comparing CPU power draw for games unless your game's running completely on the CPU. And for the last time there's no such thing as "gaming" efficiency

 

[mkppo]

the 285 is also way more efficient in gaming (while performance is within margin of error).

Way more efficient? Not really.

9950X consumes all of 10W extra in total for slightly more performance. And again, that performance differential will change with 24H2. Not sure why TPU still haven't updated their CPU charts with the new one, but most other websites show more than 5% average gains at 1080p with 24H2. That 10W is neither here nor there, but people would gladly take any extra CPU performance they can get for gaming as evident by 9800X3D.

Also something important to note: TPU's charts don't show much gains in terms of % because of the large test suite with some games really not responding to CPU's well. Take the 9800X3D as an example, it performs just 4.1% faster than 7800X3D in TPU's charts, but 11.5% faster in HUB's 9800X3D review.

So yes, with 24H2 the 9950X should at least be 3% faster than 285K in TPU's chart. That's approaching 7800X3D > 9800X3D territory.

 

[Tek-Check]

Intels terribly regressive architecture beats or matches amds top end in everything, gaming includes. The review from this very site shows the 285k being a better product (especially when it comes to efficiency) than the 9950x. And yet, marketing wins once again and amd sells more. What can you do.

And if you look into meta-review from 3D Centre? Do you still see the same?

 

[Vayra86]

None of what you said is different to what I said - although you forgot - besides the ST and idle efficiency, the 285 is also way more efficient in gaming (while performance is within margin of error).

And yet, you explain it in a way that somehow needs to be tied to 'but marketing and muh Nvidia marketshare and people logic'... mate... let it go. There's always a mindshare element in play if you know two things about hardware, and a lot of it is based on personal experiences, not some perceived camp mentality; and if you're trying to prove that people are hypocrites especially in their purchasing behaviour vs what they say about it... I don't think thats a secret

 

[JustBenching]

And yet, you explain it in a way that somehow needs to be tied to 'but marketing and muh Nvidia marketshare and people logic'... mate... let it go. There's always a mindshare element in play if you know two things about hardware, and a lot of it is based on personal experiences, not some perceived camp mentality; and if you're trying to prove that people are hypocrites especially in their purchasing behaviour vs what they say about it... I don't think thats a secret

If you agree they are hypocrites then why is it an issue that I'm exposing it...

 

[Vayra86]

If you agree they are hypocrites then why is it an issue that I'm exposing it...

Because its aimed at people and not tech, and leads to nothing.

 

[john_]

Some investors and analysts are aften "disappointed" by their own unrealistic expectations and predictions founded on bad math and lack of detailed insight. Successful companies do not need to behave like some investors seeking immediate gratification, like little children. Long-term, steady strategy often counts more than short-term higher gains.

Look. AMD's time is not infinite. Intel could fix it's manufacturing tomorrow. That could mean OEMs running back to Intel, Xeons starting taking back market share. Or AMD could lose a console or something. If AMD doesn't take advantage of the time it is winning, increase market share and get as much money as they can, they wouldn't increase in size enough to be able to defend their position. It's not about investors and analysts and expectations and predictions. Look where Nvidia is today. Thanks to it's size it gets priority and probably much better prices at TSMC. How can AMD fight a company that gets priority and better prices from TSMC? How are they going to fight Intel if Intel recovers? If you laugh at the possibility of Intel recovering, maybe 10 years ago you had an i7 and was laughing at the possibility of AMD becoming competitive. I am sorry, if you can't see past today, I just can't help with that.

You can call it conservative, but that is a prudent approach by AMD. They have showed consistently growing revenues over the years. Not meteoric growth, like Nvidia, but consistently growing flow of revenues. It's a different approach by Lisa's leadership. And rock solid so far.

In a market that moves rapidly, that opportunities are presented today and they are history tomorrow, this is not prudent. It's just a series of lost opportunities. Opportunities that made Nvidia a monster and put AMD in a tough position having to fight that monster. Nvidia could probably be coming to desktops and laptops with SOCs this year or the next. With or without Mediatek. With it's size and it's deep pockets full of money, not to mention the influence it exerts in the PC market, Nvidia can move a part of the laptop/desktop market to the ARM platform. That will be bad for both AMD and Intel.

Around 50% of all CPUs sold are AM5 platform, and growing steadily. So, in just little over two years half of cumstomers have already adopted it over buying new or upgrading existing AM4 rigs. Considering the fact that longevity is the key feature of their platforms, 50% on AM5 CPUs is quite good. In 2027/28, they will have customers in different parts of the world using three platforms, including AM6, which will be amazing and unprescedented.

AMD almost shoot it's feet with the AM5 platform. Twice. First when it released AM5 for the first time and Intel's LGA 1700 was looking a better or/and a cheaper(with DDR4) choice. Then with the stupid pricing on 9000 Ryzen CPUs compared to the 7000 series CPUs, when in fact the 9000 series CPUs weren't really offering anything to the typical consumer. What saved the AM5 platform was X3D chips and the incompetence of Intel's engineers to take advantage of TSMC's 3nm node. AM5 holding only 50% of the AMD platforms after so many years, shows that it didn't convinced many AMD fans to pay the money to upgrade or chose AM5 over AM4. Is like saying that Windows 11 is doing great when many insist on staying on WIndows 10. That 50% wouldn't have happened if AMD wasn't selling X3D. Fortunately they sell X3D chips and that's what saved AM5, that's also what helped AM4 to continue being an option for many.

AM4 will gradually leave the scene in next couple of years, but many millions of users will still use it by 2030. Even today, one can buy new, excellent and cheap AM4 platform with a very good CPU for daily use. There is nothing wrong with it, quite opposite. So, manufacturers will simply stop producing boards and CPU for AM4 and this process will gradually reduce AM4 in use. There will still be a vibrant secondary market for a few more years.

Nothing new here. Even a 10 years old platform is fine for typical office usage if an SSD is used.
That being said, an AM5 platform that could be targeting every user, by offering CPUs and motherboards from the $50 price point, would have almost replaced the AM4 platform by now. AM5's percentage would have being closer to 80%-90% than 50%.

The supply of 9800X3D has stabilized after the New Year in growing number of markets. I can see it in every major tech shop in Europe, North America and Australia.

That was going to happen eventually. But they could avoid the shortage and the higher prices.

They supplied what they could with the amount of available wafers to purchase. Obviously, it was not ideal in every segment, such as delays in laptop chips, but it is what it is. There's a long queue of customers at TSMC... and we do not have any other foundry like TMSC. Now, Zen5 CPUs are produced in new Arizona's factory. People will be soon able to see "Made in USA" on those chips.

That's why you don't lose opportunities to grow and fill your pockets with cash. The prices at the American TSMC fab will be higher and tariffs will make chips build in Taiwan also pricier. That could lead the whole industry in recession because most people will chose to keep their electronic equipment longer. That will affect the financial of all companies that sell to consumers.

It does not, really. It looks like a proper launch, with more mature software and real volume globally, so that everyone can actually buy a product, and not like catastrophic paper launch by Nvidia, scalping scandal and lack of any cards that followed.

Either AMD decided to drop prices after seeing the price on the 5070, so they need higher quantities to meet the expected demand, or they don't allocate enough wafers from TSMC for the gaming cards, so they get quantities at a lower rate, meaning they had to wait to build inventory. Or they are trying to implement Multi Frame Generation in their FSR 4.0 to be able to compete with Nvidia's cards. In my opinion the first and third theory are closer to the truth with the second one probably not being the case considering they where getting ready for a January release.
Waiting to see if it will be a proper launch and at prices that will trigger demand. Obviously I wish to see plenty of cards available at $500 or less. I might even upgrade my RX 6600.

AMD is not getting any bigger? What kind of nonsense is this? Did you check their revenues and market capitalization in last 6 years? You have some homework to do.

It's not me that has to do homework. They don't get any bigger. They get more money and better profit margins from Instinct and EPYC CPUs, thanks to X3D chips and Intel's incompetence they started seeing AM5 beating Intel's platforms, finally, they also get about a billion each year from Xilinx, but the AMD we talk about here, the combination of CPUs and GPUs without Xilinx is not getting any bigger. In fact the gaming division is a shadow of it's past. An AMD that was really growing would have been hitting over 10 billions per quarter by now and server chips would not be killing gaming GPUs, because of priorities with wafers and lack of funding to keep Nvidia in check.

Nvidia, of course, skyrocketed due to AI boom. Exploding success is not the only model of success to follow, and not the best one necessarily. Nvidia has just lost $600 billion in market cap in a few days due to DeepSeek release. It just shows how overly inflated its value is, and how easily market cap can be wiped out within days due to irrational hype in the Western investor circles.

Nvidia is making money from AI that will use tomorrow into expanding in other markets if AI bubble is burst. Nvidia's $600 billion loss was just a hiccup with it's share price going back up again. On the other hand AMD lost over half of it's valuation in the last year. How is AMD getting bigger by losing half it's valuation and how is Nvidia inflated when it's share price is going up again?

Sure, 50%, but not 80% anymore. Their market share is gradually melting, and consistently over the years. That is 100% true.
Don't expect any revolutionary, fast changes. This industry does not work like that.
Remember the words: gradually and consistently.

OEMs will jump back to Intel if Intel can warranty them a strong stream of 18A chips. We have seen it in a number of cases in the past, how difficult it is for OEMs to increase their AMD options and how easily they can go back to Intel. Because Intel owns fabs and if those fabs start working as they should, they can warranty the quantities huge companies like Dell need. AMD not getting bigger and not securing more wafers at TSMC will be always a dissadvantage. Ans as I said in the beggining of my comment, 10 years ago people would be advising to not expect revolutionary, fast changes with Ryzen CPUs.
And yes, I also can play that game.
"Remember my words. Slow and too late".

 

[Tek-Check]

Look. AMD's time is not infinite. Intel could fix it's manufacturing tomorrow. That could mean OEMs running back to Intel, Xeons starting taking back market share. Or AMD could lose a console or something.

- read Dr Ian Cutress' report on Intel's Q4 revenues; it will take them a few good years to try to come back
- https://morethanmoore.substack.com/p/intel-2024-q4-financials
- AMD has just gained new console contract, so they are not going to be losing it any time soon, at least for a decadeIf AMD doesn't take advantage of the time it is winning, increase market share and get as much money as they can, they wouldn't increase in size enough to be able to defend their position. Look where Nvidia is today. Thanks to it's size it gets priority and probably much better prices at TSMC. How can AMD fight a company that gets priority and better prices from TSMC?

f AMD doesn't take advantage of the time it is winning, increase market share and get as much money as they can, they wouldn't increase in size enough to be able to defend their position. It's not about investors and analysts and expectations and predictions. Look where Nvidia is today. Thanks to it's size it gets priority and probably much better prices at TSMC.

- AMD is exactly taking advantage and little-by-little building market share across several segments; their main strength here is consistency
- read Dr Ian Cutress' report on AMD's Q4 revenues
- https://morethanmoore.substack.com/p/amd-q4-2024-and-fy-financials
- as I said, Nvidia is a phenomenon on its own. It's neither the only way to be a successful company, nor the best model to follow by entire world
- if everybody followed blindly what Nvidia is doing, our global climate change would significantly accelerate... but that's another topic that investors are completely blind and irresponsible about, as they seek monetary profits while the planet is heating up...
- joint installation base of large data centers now consume more energy and emit more heat than entire countries, such as Netherlands or Argentina
- just pose and think about it for a second...
- soon, many investors will need twice as many AC devices in their offices, to cool themselves during adrenaline-fueled investment sessions in unbearable summers

How can AMD fight a company that gets priority and better prices from TSMC? How are they going to fight Intel if Intel recovers? If you laugh at the possibility of Intel recovering, maybe 10 years ago you had an i7 and was laughing at the possibility of AMD becoming competitive. I am sorry, if you can't see past today, I just can't help with that.

- as seen in Q4 report, AMD has posted record numbers ever across segments, so I am not sure what you are banging about here
- they are coping pretty well and predictions for this year are positive
- if Intel recovers? They have a long way to go. They lost almost $19 billion last year in comparison to 2023. Arrow Lake and Xeons do not look good either.

In a market that moves rapidly, that opportunities are presented today and they are history tomorrow, this is not prudent. It's just a series of lost opportunities. Opportunities that made Nvidia a monster and put AMD in a tough position having to fight that monster. Nvidia could probably be coming to desktops and laptops with SOCs this year or the next. With or without Mediatek. With it's size and it's deep pockets full of money, not to mention the influence it exerts in the PC market, Nvidia can move a part of the laptop/desktop market to the ARM platform. That will be bad for both AMD and Intel.

- make no mistake. With new tarrifs, wafers will become way more expensive, which TSMC will announce soon, and Nvidia will not think twice where to allocate vast majority of money earning chips. Spoiler alert - it's not going to be consumer market
- sure, they will have some presence, but mega iGPUs from Apple and AMD will slowly eat out their discrete graphics in mobility sector. Watch out in 2030.
- any bringing of ARM into desktop will need to be with modular DIY designs and Nvidia would need to make a huge investment to introduce such competitor to x86. Such adventure is very unlikely to happen any time soon. They prefer their beloved data center. Much easier.
- Nvidia is not capable of making such an entrance into desktop space. They are not capable of stocking enough of 5080 and 5090 cards in the first place, let alone anything more serious. It was a joke of paper launch. Entire consumer industry is laughing how badly they handled 5000 series launch. Nonsense.

AM5 holding only 50% of the AMD platforms after so many years, shows that it didn't convinced many AMD fans to pay the money to upgrade or chose AM5 over AM4. Is like saying that Windows 11 is doing great when many insist on staying on WIndows 10. That 50% wouldn't have happened if AMD wasn't selling X3D. Fortunately they sell X3D chips and that's what saved AM5, that's also what helped AM4 to continue being an option for many.

- "only 50%"? Hahaha. You are starting to sound like a heavily biased copium who is not willing to look into numbers and stats. Only 7 years ago, they did not have more than 20%. Building market share to parity on desktop in just 7 years is a very commendable achievement. Your problem is that you are not able to appreciate such seismic change that took place rather quietly and not over night.
- X3D is the right product for the right market. That's all you need, like Nvidia in data center. A jackpot product, which is exactly what they have on x86
- the key strength of AMD's desktop platforms are their longevity. In 2027/28, they will have AM6 and will be able to support not two, but three long-distance platforms. No one, not even Intel will be able to do this if they do not change their ways.
- as you probably know, different parts of the world run different hardware and have different upgrade cycles. AM4 will still be popular in 2030 in many developing parts of the world and AMD will still be releasing firmware updates for millions who enjoy those systems. Majority of the world does not live in developed economies. Never forget that. While folks in rich countries will enjoy new shinny AM6 PCs and mature AM5 systems, millions upon millions of users elsewhere will still enjoy AM4 systems. Is your imagination able to cope with this, namely that AMD will be running globally three long-distance platforms? Cheaper AM4 and AM5 server boards and CPUs will still be popular even in 2030.

AM5 platform that could be targeting every user, by offering CPUs and motherboards from the $50 price point, would have almost replaced the AM4 platform by now. AM5's percentage would have being closer to 80%-90% than 50%.

- $50? Hahaha. You really live up there, in your imagination
- there is no reason to replace AM4 platform if users enjoy it; no reason to create more e-waste if a system can run for 7-10 years
- this is what you are missing. AMD is neither Intel nor Nvidia, they will never be and they should never be similar to them. They have their own path.
- every platform needs time to mature and get higher adoption rate. There is nowhere to rush.

But they could avoid the shortage and the higher prices

- there is no evidence of being able to aovid shortages; don't make up stuff, please; stock was being sold as quickly as it arrived; everywhere
- this is clearly visible in their Q4 client revenues; a record sales never seen before
- they did not anticipate such storm on 9800X3D; and yes, you can partially blame Intel users who massively migrated to AM5 platforms
- AMD quickly ramped up production and almost every major retailer had pre-order system, so everyone who ordered eventually got their CPU within a few weeks; unlike Nvidia's top cards that do not have a globally efficient pre-order system and buyers will need to wait...and wait...

That's why you don't lose opportunities to grow and fill your pockets with cash. The prices at the American TSMC fab will be higher and tariffs will make chips build in Taiwan also pricier. That could lead the whole industry in recession because most people will chose to keep their electronic equipment longer. That will affect the financial of all companies that sell to consumers.

- pockets with cash? Gosh... what a dreadful outlook
- well, you can't just shove down the throats silicon at all times and force people to buy more and often; the world does not work like that
- there's a saturation point at the other end; and customers live on one planet only
- you can swallow maybe three bananas, but one more will cause you nausea and unplesant side effects; it's called human nature
- people should keep PCs for longer and not create unnecessary e-waste; that's why Intel recently announced a new, modular PC initiative
- recession is a part of life; it comes and it goes; most of time, ordinary folks lose out... another topic

Obviously I wish to see plenty of cards available at $500 or less. I might even upgrade my RX 6600.

- 9070 yes, but not 9070XT for $500

the combination of CPUs and GPUs without Xilinx is not getting any bigger

- this is blatantly false. Check Q4 report. Client CPU have posted record revenues
- GPU is lagging behind, that is true, but they will be releasing RDNA4 with consoles are almost at OEL

Nvidia is making money from AI that will use tomorrow into expanding in other markets if AI bubble is burst. Nvidia's $600 billion loss was just a hiccup with it's share price going back up again. On the other hand AMD lost over half of it's valuation in the last year. How is AMD getting bigger by losing half it's valuation and how is Nvidia inflated when it's share price is going up again?

- what you call a 'hiccup' could be the first symptom of deeper changes that market will need to adjust to, as open source AI models flood the market
- AMD has not lost "half", please stop making stuff up and look into numbers
- you would be the worst possible long-term investor adviser with such insights

"Remember my words. Slow and too late"

- unfortunately, it is Intel who has lost two out of three major trains in semiconductor industry in last 15 years
- read Ian Cutress' articles that I linked here and you will find out more

 

[john_]

- read Dr Ian Cutress' report on Intel's Q4 revenues; it will take them a few good years to try to come back
- https://morethanmoore.substack.com/p/intel-2024-q4-financials
- AMD has just gained new console contract, so they are not going to be losing it any time soon, at least for a decade

As I told you before, Intel only needs one good product from their fabs. OEMs are still favoring them. As long as Intel is going from one disaster to the next, AMD will keep gaining market share. But that could change as easily as it changed 20 years ago with Core2Duo and a few years ago with 12th gen. Unfortunately OEMs are biased towards Intel. They even shown some love for Qualcomm that I haven't seen the last 25 years for AMD. As for consoles, is Microsoft onboard or just SONY?

- AMD is exactly taking advantage and little-by-little building market share across several segments; their main strength here is consistency
- read Dr Ian Cutress' report on AMD's Q4 revenues
- https://morethanmoore.substack.com/p/amd-q4-2024-and-fy-financials
- as I said, Nvidia is a phenomenon on its own. It's neither the only way to be a successful company, nor the best model to follow by entire world
- if everybody followed blindly what Nvidia is doing, our global climate change would significantly accelerate... but that's another topic that investors are completely blind and irresponsible about, as they seek monetary profits while the planet is heating up...
- joint installation base of large data centers now consume more energy and emit more heat than entire countries, such as Netherlands or Argentina
- just pose and think about it for a second...
- soon, many investors will need twice as many AC devices in their offices, to cool themselves during adrenaline-fueled investment sessions in unbearable summers

They gain slowly in some markets, they lose in other markets. If AMD was holding a 40% in gaming GPUs and was at 15% in AI, we would be agreeing in everything. What I see is a slow gain in servers thanks to Intel's failures, slow gains in desktop and mobile(finally) CPUs thanks to Intel's failures, little gains to AI, total disaster in gaming GPUs and Xilinx remaining at the same levels as when it was a separate company. You see fields full of flowers and rainbows, I see reality.

I really can't consider as arguments what you post about global climate. You desperately try to invent arguments here that have nothing to do with the subject.

- as seen in Q4 report, AMD has posted record numbers ever across segments, so I am not sure what you are banging about here
- they are coping pretty well and predictions for this year are positive
- if Intel recovers? They have a long way to go. They lost almost $19 billion last year in comparison to 2023. Arrow Lake and Xeons do not look good either.

Yep, the gaming division is posting record numbers. I guess you know what "you are banging about here", right? Please...
Also, I didn't said that AMD is not having some success, I am saying that they are losing huge opportunities to have great success. They are also wasting time that could cost them in a year or two.
Intel is investing in fabs. They didn't lost 19 billions. They are investing those billions to build fabs. They didn't got robbed, that money didn't gone in a dust bin. You can say "lost" if those fabs end up useless.
And I will repeat again. Well. I will not. I will ASK you. 10 years ago did you had an Intel i7 or an Intel i5? When someone was saying to you 10 years ago, that Ryzen will make AMD competitive again, was your reply "If AMD recovers? They have a long way to go."?

- make no mistake. With new tarrifs, wafers will become way more expensive, which TSMC will announce soon, and Nvidia will not think twice where to allocate vast majority of money earning chips. Spoiler alert - it's not going to be consumer market
- sure, they will have some presence, but mega iGPUs from Apple and AMD will slowly eat out their discrete graphics in mobility sector. Watch out in 2030.
- any bringing of ARM into desktop will need to be with modular DIY designs and Nvidia would need to make a huge investment to introduce such competitor to x86. Such adventure is very unlikely to happen any time soon. They prefer their beloved data center. Much easier.
- Nvidia is not capable of making such an entrance into desktop space. They are not capable of stocking enough of 5080 and 5090 cards in the first place, let alone anything more serious. It was a joke of paper launch. Entire consumer industry is laughing how badly they handled 5000 series launch. Nonsense.

You throw arguments about Nvidia, that apply even more on AMD, because AMD is a less important customer for TSMC. I don't know if you realize it, or just don't care. If you just think that throwing "more arguments" will make you more convincing. As for mega iGPUs, they are fine and great and we have been waiting for those over a decade, but guess what. They are not exactly cheap. Why do you think Nvidia is pushing discrete GPU pricing higher the last 12 years (with the first Titan card)? Because they know that iGPUs will replace lower end cards. But guess what. If Nvidia decides to ender the laptop market, they also have the capability to build mega iGPUs. And no, they don't necessarily need DIY. DIY is probably 5-10% of the market. Probably closer to 5%. The rest is mostly laptops and pre build systems from big OEMs. And with the money Nvidia has, they can move into laptop AND desktop space. With Intel having problems and AMD playing it safe, this is a huge opportunity for Nvidia. If Qualcomm managed to sell X number of Snapdragon laptops, Nvidia can sell 5 times that number, if not more. That's why I keep saying to you that AMD "DOES NOT HAVE INFINITE TIME". They have to grab all opportunities when those are presented.

- "only 50%"? Hahaha. You are starting to sound like a heavily biased copium who is not willing to look into numbers and stats. Only 7 years ago, they did not have more than 20%. Building market share to parity on desktop in just 7 years is a very commendable achievement. Your problem is that you are not able to appreciate such seismic change that took place rather quietly and not over night.
- X3D is the right product for the right market. That's all you need, like Nvidia in data center. A jackpot product, which is exactly what they have on x86
- the key strength of AMD's desktop platforms are their longevity. In 2027/28, they will have AM6 and will be able to support not two, but three long-distance platforms. No one, not even Intel will be able to do this if they do not change their ways.
- as you probably know, different parts of the world run different hardware and have different upgrade cycles. AM4 will still be popular in 2030 in many developing parts of the world and AMD will still be releasing firmware updates for millions who enjoy those systems. Majority of the world does not live in developed economies. Never forget that. While folks in rich countries will enjoy new shinny AM6 PCs and mature AM5 systems, millions upon millions of users elsewhere will still enjoy AM4 systems. Is your imagination able to cope with this, namely that AMD will be running globally three long-distance platforms? Cheaper AM4 and AM5 server boards and CPUs will still be popular even in 2030.

Oh, nice. Copium. The ultimate.... argument that someone uses when in despair realises that the other person probably does have a point. In all your excitement you probably missed the fact that I am talking about percentage in AMD platforms, not all platforms. Re read my post. As for seismic change, that happened in 2006 with Core2Duo when AMD's market share collapsed in a single year, not after 7 years.
X3D is saving AMD in mainstream CPUs. No one deny this. Probably only trolls deny this.
Longevity was something that AMD always was using as an advantage. From the AM3 era. Unfortunately most people where keep buying Intel, because most people either didn't knew enough about computers to knew what longevity meant, where buying pre build PCs, so that longevity wasn't really there because OEM hardware and BIOS is limited, or where enthusiasts or wannabe enthusiasts who where selling their whole platform every year or two to go to a new platform.
2030? Oh come on, this isn't a argument. If someone doesn't need the latest and greatest, even today can use an older platform. As long as an SSD is installed in that platform. Until recently one of my systems was a 6 core Thuban. Unfortunately the motherboard died. Well, it served me greatly for over 12 years. A 10 years old Haswell system can be used nicely even for gaming today. So, please try to avoid talking about my copium and my imagination. And AMD wouldn't keep three platforms alive. It's costly. The first thing Su did when she took over AMD, was to kill AM1, because having and more importantly, supporting too many platforms in the market is a bad thing.

- $50? Hahaha. You really live up there, in your imagination
- there is no reason to replace AM4 platform if users enjoy it; no reason to create more e-waste if a system can run for 7-10 years
- this is what you are missing. AMD is neither Intel nor Nvidia, they will never be and they should never be similar to them. They have their own path.
- every platform needs time to mature and get higher adoption rate. There is nowhere to rush.

More HAHA arguments. You are definitely going in my ignore list after that. When you pass juvenile, I might consider losing my time again answering you.

But let's finish my mistake of answering you. If you had the option to buy an AM5 motherboard and AM5 CPU for the same price of an AM4 motherboard and an AM4 CPU, would you go for the AM4, just because of DDR5 costs? E waste is buying the old platform that it's upgrade options are limited to what is already in the market instead of investing in a new platform that will keep going for years. AMD should follow Intel's path of getting OEMs onboard and Nvidia path of grabbing opportunities, of creating new features that others would want to have, would be willing to pay extra to get them. As for platform maturity, 1-2 years is more than enough for maturity. Support, that's something else. Support yes, 5+ years would be great.

- there is no evidence of being able to aovid shortages; don't make up stuff, please; stock was being sold as quickly as it arrived; everywhere
- this is clearly visible in their Q4 client revenues; a record sales never seen before
- they did not anticipate such storm on 9800X3D; and yes, you can partially blame Intel users who massively migrated to AM5 platforms
- AMD quickly ramped up production and almost every major retailer had pre-order system, so everyone who ordered eventually got their CPU within a few weeks; unlike Nvidia's top cards that do not have a globally efficient pre-order system and buyers will need to wait...and wait...

No evidence? AMD THEMSELVES SAID THAT THEY WHERE NOT EXPECTING SUCH DEMAND BECAUSE THEY WHERE NOT EXPECTING INTEL TO F UP THAT MUCH. My God...... You make up stuff thinking that making up stuff will prove the other person wrong. Well it doesn't work that way.
Record sales because Intel F UP. Twice in a year!
Blame Intel users.....So, now you are making up the stuff that you accuse me of making up. Is there any consistency in your thought or are they random?
Nvidia is doeing what it knows best. Faking MSRP and inflating prices. They know that they will get away with this. I am expecting some high IQ trolls to keep accusing AMD for Nvidia's pricing, while paying what Nvidia wants with a smile.

- pockets with cash? Gosh... what a dreadful outlook
- well, you can't just shove down the throats silicon at all times and force people to buy more and often; the world does not work like that
- there's a saturation point at the other end; and customers live on one planet only
- you can swallow maybe three bananas, but one more will cause you nausea and unplesant side effects; it's called human nature
- people should keep PCs for longer and not create unnecessary e-waste; that's why Intel recently announced a new, modular PC initiative
- recession is a part of life; it comes and it goes; most of time, ordinary folks lose out... another topic

The world is going that direction. More and more silicon in our houses. One console or one PC, turned to one PC and one console, turned to one(or more) console and one(or more) PC and one(or more) laptop and one(or more) smartphone and one(or more) tablet and more smart devices and more electronics and so on. We might see people keeping their equipment for more time, not because they care about environment, but because of tariffs making things more expensive. And that will affect also AMD.

- 9070 yes, but not 9070XT for $500

9070 will probably still be fine at that price. I say probably because we have to see benchmarks first.

- this is blatantly false. Check Q4 report. Client CPU have posted record revenues
- GPU is lagging behind, that is true, but they will be releasing RDNA4 with consoles are almost at OEL

GPU is not lagging behind, it's getting obliterated year after year. AMD's management probably feels that iGPUs are enough for now, enough to warranty them sales in laptops, handhelds and consoles. With people willing to pay more to get less from Nvidia, I don't blame them.
And yes, GPUs and CPUs are not getting much bigger. What CPUs are winning, GPUs are losing.

- what you call a 'hiccup' could be the first symptom of deeper changes that market will need to adjust to, as open source AI models flood the market

Nvidia's share price is recovering. For now it's just a hiccup.

In 6 months, 12 months, 24 months, 100 months, things could change and you end up right. But for now, it look like just a hiccup.

- AMD has not lost "half", please stop making stuff up and look into numbers
- you would be the worst possible long-term investor adviser with such insights

View attachment 383565

Look, I am not making stuff up. You just missing the picture. AMD's share price was at $227 and not is retreating to under $107. That's more than half market capitalization lost.

If you think that marrying a stock is a good strategy, well, maybe you are a better long term investor than me. The stock market disagrees with you.

- unfortunately, it is Intel who has lost two out of three major trains in semiconductor industry in last 15 years
- read Ian Cutress' articles that I linked here and you will find out more

Intel's management was a disaster. That was really a blessing for the market, because it gave AMD the chance and time to get market share, create a good name in CPUs, fix the damage that Bulldozer did. We are "enjoying" a duopoly instead of suffering a monopoly as it is the case in GPUs today. As for Ian, I was always a fan of his articles, but I am not reading those like I am reading the Bible or something.

 

[Tek-Check]

Microsoft onboard or just SONY?

Microsoft does not have much choice at the moment, do they? For a new console to be released in 2027, they need to decide now.

They didn't lost 19 billions.

They literally lost $19 billion in revenues in comparison to previous year, which means they did not earn that money to invest into anything. Sacking CEO is only one symptom of current internal business culture. It should get better in a few years, but it does not look good at all now. Also, dropped by Dow Jones and replaced by Nvidia.

You see fields full of flowers and rainbows, I see reality.

Yes, you see reality that suits your filters.

When someone was saying to you 10 years ago, that Ryzen will make AMD competitive again, was your reply "If AMD recovers? They have a long way to go."?

Yep, I said that they had a long way to go, because they did, and they recovered, indeed.

Look, I am not making stuff up. You just missing the picture. AMD's share price was at $227 and not is retreating to under $107. That's more than half market capitalization lost.

Market cap is not the same as share price. Nonsense. I gave you the numbers for market cap.

No evidence? AMD THEMSELVES SAID THAT THEY WHERE NOT EXPECTING SUCH DEMAND BECAUSE THEY WHERE NOT EXPECTING INTEL TO F UP THAT MUCH. My God...... You make up stuff thinking that making up stuff will prove the other person wrong.

I literally said in the text you quoted that they did not anticipate. Read more carefully before replying

GPU is not lagging behind, it's getting obliterated year after year. AMD's management probably feels that iGPUs are enough for now, enough to warranty them sales in laptops, handhelds and consoles.

Ian Cutress explained this nicely on his YT channel, as he attended the call. Have a watch.

ntel's management was a disaster. That was really a blessing for the market, because it gave AMD the chance and time to get market share, create a good name in CPUs, fix the damage that Bulldozer did. We are "enjoying" a duopoly instead of suffering a monopoly as it is the case in GPUs today. As for Ian, I was always a fan of his articles, but I am not reading those like I am reading the Bible or something.

I agree. As for Ian, no one expect him to be read like a "Bible", but a very insighful analysis of what's happening in tech world.

 

[john_]

Market cap is not the same as share price. Nonsense. I gave you the numbers for market cap.

Good day.

 

[Tek-Check]

Good day.

I don't do the numbers. The website clearly states how they measure historical capitalization, at the end of year, which makes sense, as it eliminates momentary spikes and capricious anomalies.

AMD (AMD) - Market capitalization

As of February 2025 AMD has a market cap of £140.13 Billion. This makes AMD the world's 80th most valuable company according to our data.

companiesmarketcap.com