Intel's Global CPU Market Share is on the Rise, AMD Starts the Downfall

[ExcuseMeWtf]

No surprise, hideously overpriced where I live.
And thats why I now own i3 12100F despite originally wanting to go Team Red...

 

[medi01]

AMD should have maintained their aggressive strategy on pricing.

You cannot have that and increased margins at the same time and guess what is more interesting to AMD who was dominating DIY part of the market for years by now.

OEMs is a different, shady story anyhow.

 

[birdie]

The poll doesn't make a lot of sense.

No one here has any idea how Ryzen 7000/Zen 4 will stack against Intel's 13th gen/RPL.

If the former is faster, at least in ST, it will mean a world to most gamers out there. As much as AMD fans haved been obsessed over MT performance ST performance is the king and no one cares if your desktop CPU has 1000 slow cores, 8 fast cores are better for absolute most mundane tasks, and more importantly games. Yeah, RPL will have 24 cores and it will certainly inflate some numbers.

The biggest issue for AMD is that they are fighting for TSMC's capacity and they cannot release low-range competitive parts. Intel Core 12100(F) and 12400(F) literally destroy anything AMD can offer.

 

[Makaveli]

AMD doesn't give a sht about CPU desktop. each EPYC CPU provides much more profit than CPU Desktop. One EPYC 64 cores with price $7500 is equal to 25x Ryzen 5600x at price $300 but more profit.

That is where the majority of their focus is these days and it should be. Way more money and mindshare in the enterprise space.

AMD should have maintained their aggressive strategy on pricing. Every time, Ryzen was still in 'development' getting on par with competitors or exceeding it by a notable margin on one aspect or another.

AMD is trying to remove themselves as the value brand. Intel didn't have an aggressive price strategy when it was on top, so this to me was expected from AMD. Now that ADL-S is out and abit faster than Zen 3 they could probably use a price correction but they are still selling everything they make so why bother until you really need to.

 

[john_]

For AMD winning the DIY market was a low single digit percentage change of the whole market. For Intel it could be more.

 

[Cowboystrekk]

At equal mobo + CPU price I will probably still choose 12400F though, since B660 motherboard have better feature set, sometimes better VRM, one more generation of upgrade, and pcie 4.0.

If you plan on upgrading later that is. All B550 boards have pcie 4.0 and all run 5600X and 5800X w/o issues, but not all B550 can run 5950X without problems like several B660 will struggle with 12600K and up, some even have problems with locked CPUs like 12600 (Asrock 140W long limit).

 

[Testsubject01]

AMD should have maintained their aggressive strategy on pricing. Every time, Ryzen was still in 'development' getting on par with competitors or exceeding it by a notable margin on one aspect or another.

I would agree! Supposedly their focus is on server/datacenter, why start to squeeze the Desktop market the very first gen you are slightly ahead, if you can offset it and further gather market and mind share to get another leg up.
Heard a lot of arguments, that AMD needs to establish they are a “premium brand” now, but is forcing it through pricing the way?

Given AMD's track record in CPU/APU's since 2016 and a strong comeback to high-end GPU's in 2020 and the rumors about Zen 4 / RDNA 3, putting a Headline of “Downfall” up is disgusting clickbait and TPU should be better than that, holding itself to higher standards.

 

[Valantar]

AMD is trying to remove themselves as the value brand. Intel didn't have an aggressive price strategy when it was on top, so this to me was expected from AMD. Now that ADL-S is out and abit faster than Zen 3 they could probably use a price correction but they are still selling everything they make so why bother until you really need to.

The problem with trying to do that while at the same time having made the CPU market competitive again for the first time in a decade is that in a highly competitive market with few actors, everyone needs to be the value brand. In a competitive duopoly there is no room for anyone to pretend to a premium position.

Generationally speaking it feels closer to 2 years than to 1 year (and they were announced in October but splitting hairs really). Might also be the warped perception of time because of the pandemic or the constant desire for newer and flashier hardware, but Zen3 to me is feeling a bit on the old side.

Yeah, Covid has warped everyone's perception of time - the past two years feel both excruciatingly long and as if they've just flow by to me, it just depends what I'm thinking of. But while Zen3 is ripe for replacement, it's still not all that long-lived. Longer than the previous two generations, but still nothing unreasonable. Of course perceptions are also skewed by Intel suddenly pulling off rapid-fire CPU launches. If Raptor Lake launches later this year, that'll be three generations within 18-ish months all of a sudden. Which of course mainly demonstrates how their process node troubles have held them back before this, but it also serves to make AMD look slow.

There were some interviews around CES where they talked about the dynamics of DDR5 supply and how AM5 would be enthusiast first (I remember some more explicit quotes about APUs only later but that's the one I can find now)


source: https://www.tomshardware.com/news/a...desktop-pcs-ddr5-pricing-impacts-release-date

Hm, interesting. I guess that's understandable, though I'd be surprised if DDR5 supplies weren't fine come late Q3 or early Q4 this year. Still more expensive than DDR4, sure, but it shouldn't be too bad a year into consumer availability.

 

[Makaveli]

The problem with trying to do that while at the same time having made the CPU market competitive again for the first time in a decade is that in a highly competitive market with few actors, everyone needs to be the value brand. In a competitive duopoly there is no room for anyone to pretend to a premium position.

Its only been a competitive duopoly since Zen came out and more so at Zen 2. Intel has been in the premium position for over a decade charging $500 for Quad core cpu's. Even with the market the way is it now why is AMD not allowed to do the same? They have shareholders to please that want to see an increase in ASP. I think too many people are use to AMD being the budget brand and think they should stay there. They are trying to hold market leading positions and will price accordingly. No business ever has said I want to stay in 2nd place to keep pricing low for consumers that will not grow your business.

 

[mechtech]

AMD doesn't really have anything competitive at the lower-end.
Hopefully they've managed to work out a better top to bottom range of processors for AM5.

This.

Intel with their own fabs can pump these out and set whatever price they want without paying a middle man.
So they have
Quantity
Selection/variety

something AMD is struggling with.

other factors too like oooo new and shiny I want one!!

Edit lots of good comments

As for amd being aggressive on low end and price have to remember mainly using TSMC only now and they have raised prices several times and are supply constricted. If AMD put out a chip 2x performance of alder lake tomorrow for same price as 5k series it’s probable that they would still lose market share just due to supply constraints. Maybe if they kept low end and used GF to fab them it would provide more variety at the cost of efficiency from 12nm node?
This flip flop cycle happens when something new and better comes out just like Ryzen and then Ryzen 5k series. May happen again with AM5 who knows.

the GPU shortage probably isn’t helping new pc building either.

with rumors of covid rules going lax in the near future I’m just going to wait and see if anything happens with pricing.

 

[Chrispy_]

Laptops are where the fight is at and AMD's offerings just look better than Alder Lake for the moment. Better performance/Watt, more attractive TDPs for thin&light designs, vastly superior IGPs....

AL > Zen3 on desktop, but that's to be expected until Zen4 arrives. Hopefully we'll see fierce competition across all price points on desktop moving forwards into the DDR5 generation.

 

[TheinsanegamerN]

This "downfall" take also ignores that the latest processors from AMD (ignoring APUs) are almost 2 years old (1 and a half, close enough). New stuff sells better, which just makes it more surprising that Zen 3 is still maintaining more or less the same launch prices.

Your take ignores that AMD isnt gearing up to release anything new. This isnt AMD being in the middle of launchign a new product, with the 5000 series APUs they made it very clear they just....dont care about the low end. AMD decided to not bother releasing budget chips. AMD decided that their "budget" chip the 5600g would retail for the price of an i5($259, effectively $100 more then the 3600) and only feature PCIe 3.0 and a cut down cache, screw you plebs, stop being poor! Anyone complaining that intel low end parts are being compared to 2 year old AMD stock needs to realize that the 2 year old stock is all AMD offers, and that is AMD's fault alone.

 

[mechtech]

AMD starts a downfall is certainly over dramatic click bait. But Intel is definetly back in the game, i say this outside tech, mainly in investment terms and get all types of backlash. Everyone is blindly on board the AMD hype train.

But i notice this even in less informed consumers. At this point the tables have reversed, even if Intel has a better, cheaper CPU, people don't even consider it, all they want is Ryzen, period. I blame a lot of the media, influencers for this.

Another thing that's hillarious is common people and even some pc builders, when they hear Ryzen 5, great cpu's. In the Intel side, if it isn't I7 it isn't that good. A relic of the past

Interesting. And in enterprise I would say it’s opposite. 3 years ago we needed a server and I suggested AMD and response was only Intel. Regardless or performance price or power consumption. Oh well not my money. But if I was the bean counter I wouldn’t be impressed. Should always go for max product for least money. Where max is whatever suits that specific requirement regardless of brand.

 

[Dave65]

AMD starts a downfall is certainly over dramatic click bait. But Intel is definetly back in the game, i say this outside tech, mainly in investment terms and get all types of backlash. Everyone is blindly on board the AMD hype train.

But i notice this even in less informed consumers. At this point the tables have reversed, even if Intel has a better, cheaper CPU, people don't even consider it, all they want is Ryzen, period. I blame a lot of the media, influencers for this.

Another thing that's hillarious is common people and even some pc builders, when they hear Ryzen 5, great cpu's. In the Intel side, if it isn't I7 it isn't that good. A relic of the past

Total click bait. They seem giddy about the prospect of a "downfall" of AMD..

 

[Makaveli]

Interesting. And in enterprise I would say it’s opposite. 3 years ago we needed a server and I suggested AMD and response was only Intel. Regardless or performance price or power consumption. Oh well not my money. But if I was the bean counter I wouldn’t be impressed. Should always go for max product for least money. Where max is whatever suits that specific requirement regardless of brand.

I've met alot of IT directors like this. And I think the reasons are becoming less relevant as AMD continues to grow.

TCO is obviously important in a business and I know you can't just swap AMD for Intel with certain products even if they are both x86 due to compatibilities issues with software. Then you have IT support contracts which usually cover a few years, and vendor discounts tied into it. With how well EPYC is doing now i've seem some of these people now coming around but there are alot that are still Intel only...

 

[catulitechup]

mmmmmmm

 

[zlobby]

The dip was back in 2016

$1.54 per share! What a time it was... Kaaaaa-ching!

 

[Valantar]

Its only been a competitive duopoly since Zen came out and more so at Zen 2. Intel has been in the premium position for over a decade charging $500 for Quad core cpu's. Even with the market the way is it now why is AMD not allowed to do the same? They have shareholders to please that want to see an increase in ASP. I think too many people are use to AMD being the budget brand and think they should stay there. They are trying to hold market leading positions and will price accordingly. No business ever has said I want to stay in 2nd place to keep pricing low for consumers that will not grow your business.

You're completely misunderstanding me here. 1) I completely understand why AMD is doing this. I'm just saying it's poor judgement. And 2) Nobody is saying AMD "is not allowed to do the same." Again: it's just a poor choice. Why? Well, let's look at what you're saying:

Intel could do this, why? Because they were a monopolist. They had no real competition. That the CPU market is now a competitive duopoly is precisely why this doesn't work for AMD. Intel got their butts handed to them by AMD precisely because they stuck to this approach as the world changed around them. AMD can't turn around and try to act if they're an uncontested market leader when they have a competitor breathing down their neck. And remember, it's always more difficult to be an up-and-comer than an incumbent, so AMD establishing themselves as a premium brand is always going to be harder than it was for Intel to maintain their position as a premium brand. So, AMD acting as if they're in a secure lead and can charge a premium over Intel just doesn't make logical sense, even if they currently have a (very much deserved) mindshare advantage. The main outcomes of their current strategies in the CPU market will be as follows: short-term shareholder appeasement (due to high ASPs), and long-term loss of mindshare due to turning away all but the most well off customers, abandoning the midrange and budget markets entirely. This undermines any goodwill they might have gained from serving these markets previously, which is just bad PR in the long term.

And that's why I brought up that this is a problem inherent to late-stage capitalist economics and the profit-first mode of thinking: it's a short-term mode of thinking that is inherently self-destructive through undermining the basis on which it stands. Pleasing shareholders through being profitable is a possible outcome of producing good (enough) products, marketing them properly, adapting to the market, and being lucky/sleazy in the right ways at the right times. You can only plan on it as a secondary effect with complex and interwoven causes. Attempting to plan first for profits almost entirely guarantees that in the mid-to-long term, those profits will disappear, as you will have alienated your customer base and likely paved the way for a competitor to step up. (This is also why all large corporations with incumbent market leader positions buy out any and all potential threats: that's a lot quicker and easier than competing, and is thus cheaper in the short run, which pleases shareholders.)

I'm not at all arguing for AMD to remain a budget brand. If you had read what I wrote, that should be obvious - just like there's no room for a 'premium only' brand in a competitive duopoly, there isn't room for a 'budget only' brand either. They need to compete as broadly as possible, but most importantly: they need to compete. I'm arguing that they've gone from a fantastic combination of delivering great products at great prices and competing fiercely, to suddenly abandoning the vast majority of their previous customer base. They've gained inroads into previously inaccessible markets (laptops, servers) which gain them volume and continues their growth to some degree, but their pivot to "premium only brand" seems premature, especially in light of Intel suddenly beating their IPC again while also delivering significantly better value. I don't mind AMD's Zen3 pricing overall - at launch, they were still decent value. But the lack of adjustments as Intel has responded, and the lack of midrange and budget options? That's borderline inexcusable this late after launch.

To be clear, I still think Intel deserves to be in the doghouse for their consistent bribery and monopolist tactics throughout the decades, but AMD is doing a pretty poor job currently at being a better option.

 

[TheEndIsNear]

Yes, yes! I need people to dump their shares, so I can buy the dip.

I bought TMSC, Applied materials, and AMD. I got AMD at 29. I had a 200% return total on all stocks. Until this centuries napoleon decided to start a war. And we have a pansy of a president. Now it's 100%. The stock market is a slot machine they attach to your retirement in this country. I think market share will swing back and forth as new products are released. My 2 cents worth a halt a cent due to inflation.

 

[Max(IT)]

Say thank you to Lisa Su and her pricing policy with Zen 3. AMD lost contact with reality and started being... Intel.
Zen 3 prices were ridiculous since the beginning, and no 5300X was another big mistake.

AMD should have maintained their aggressive strategy on pricing. Every time, Ryzen was still in 'development' getting on par with competitors or exceeding it by a notable margin on one aspect or another.

And now that Zen is on point, its too expensive. Not smart, they haven't consolidated their lead yet and want to cash in too early. All the while its pretty silly for Ryzen MSDT chips to cost a lot, they're the weaker derivatives of the exact same stack top to bottom.

Once again I'm left with that nagging feeling AMD never truly wants to win the war, every time they win a few battles, they drop the ball somehow. I do hope I'm wrong.

exactly. Because they did the same mistake with RDNA 2...

 

[Operandi]

I'm not at all arguing for AMD to remain a budget brand. If you had read what I wrote, that should be obvious - just like there's no room for a 'premium only' brand in a competitive duopoly, there isn't room for a 'budget only' brand either. They need to compete as broadly as possible, but most importantly: they need to compete. I'm arguing that they've gone from a fantastic combination of delivering great products at great prices and competing fiercely, to suddenly abandoning the vast majority of their previous customer base. They've gained inroads into previously inaccessible markets (laptops, servers) which gain them volume and continues their growth to some degree, but their pivot to "premium only brand" seems premature, especially in light of Intel suddenly beating their IPC again while also delivering significantly better value. I don't mind AMD's Zen3 pricing overall - at launch, they were still decent value. But the lack of adjustments as Intel has responded, and the lack of midrange and budget options? That's borderline inexcusable this late after launch.

They are competing and winning where it counts, and thats the datacenter. Thats what Zen was set out to do from the start, the only reason it was released into the consumer DIY desktop space is because of of how insanely hard it is to break (back) into that market, Zen1 and Zen2 desktop was essentially nothing more than a good PR campaign with enthusiasts to win back the datacetner. AMD is still nowhere near the size of Intel when it comes to volume of CPUs moved into the market, nor do they have all the long term contracts and established designs with established partners to justify several different CPUs to target every market segment let alone the excess of available engineering hours to make it happen like Intel does. If they ever get close to parity in terms of engineering teams then we'll see dedicated smaller CPUs to better fit mid-range and low-end segments but currently all focus is on what will work for Epyc.

AMD only has so much capacity to use at TSMC and it would be utter insanity to sacrifice datacenter market share to satisfy anything in the mobile notebook market or puny DIY market let alone the low-end budget segments of either.

 

[kapone32]

Its only been a competitive duopoly since Zen came out and more so at Zen 2. Intel has been in the premium position for over a decade charging $500 for Quad core cpu's. Even with the market the way is it now why is AMD not allowed to do the same? They have shareholders to please that want to see an increase in ASP. I think too many people are use to AMD being the budget brand and think they should stay there. They are trying to hold market leading positions and will price accordingly. No business ever has said I want to stay in 2nd place to keep pricing low for consumers that will not grow your business.

Especially without choosing desultory moves to try to maintain their position. I am not worried though because if the 6000 APUs are anything to go by the next release from AMD should be interesting (not 5800X3D). I heard 5 GHZ all cores and that would be insane.

They are competing and winning where it counts, and thats the datacenter. Thats what Zen was set out to do from the start, the only reason it was released into the consumer DIY desktop space is because of of how insanely hard it is to break (back) into that market, Zen1 and Zen2 desktop was essentially nothing more than a good PR campaign with enthusiasts to win back the datacetner. AMD is still nowhere near the size of Intel when it comes to volume of CPUs moved into the market, nor do they have all the long term contracts and established designs with established partners to justify several different CPUs to target every market segment let alone the excess of available engineering hours to make it happen like Intel does. If they ever get close to parity in terms of engineering teams then we'll see dedicated smaller CPUs to better fit mid-range and low-end segments but currently all focus is on what will work for Epyc.

AMD only has so much capacity to use at TSMC and it would be utter insanity to sacrifice datacenter market share to satisfy anything in the mobile notebook market or puny DIY market let alone the low-end budget segments of either.

You forget that Gaming (147 billion) is in the DIY space. I do agree though that the Datacenter was a Endgame but they had to get public sentiment first and Youtube, Twitch and Reddit are all about Gaming, enthusiasts are for the nose bleed products like 5950X and 12900K. The fact that having a free Epic account allows you to have a nice free library of compelling Games is anecdotal but still a fact. Even though all GPUs are overpiced they still fly off the shelves. All you have to do is look at the lineups for any brick and mortar "notification" of stock. I do believe that if they could AMD would release their non X CPUs and APUs to DIY instead of OEM they would but you are spot on with the supply constraints and indeed meddling from Intel. There is also the fact that you can actually buy real AMD laptops that eat into the "budget" supply for wafers.

 

[seth1911]

Sometimes people say to me i should make them a PC for maybe 400€ etc. ok why not, its possible to play many games on medium in 1080p

I3 10100F 75€
H510 64€ (If they want a really faster cpu in future maybe a i5 11400f)
16GB RAM 56€
Case + PSU 100€
Second Hand GTX 960 80€
--------------------------------
375€ with shippment on a straight 400€


No Way to make this with an AMD
200€ for a 3600 Tray + 15€ for a Cooler, there are no 3100 or 3300x



Its like 2006 Amd fly high and falling down, now its the same they though they can take more money than intel with similar performance and upside down again
History repeats

AMD dont geddit, they main customers for them are Price/Performance buyers in the entry and midrange segment even in OEM too.

 

[Valantar]

They are competing and winning where it counts, and thats the datacenter. Thats what Zen was set out to do from the start, the only reason it was released into the consumer DIY desktop space is because of of how insanely hard it is to break (back) into that market, Zen1 and Zen2 desktop was essentially nothing more than a good PR campaign with enthusiasts to win back the datacetner. AMD is still nowhere near the size of Intel when it comes to volume of CPUs moved into the market, nor do they have all the long term contracts and established designs with established partners to justify several different CPUs to target every market segment let alone the excess of available engineering hours to make it happen like Intel does. If they ever get close to parity in terms of engineering teams then we'll see dedicated smaller CPUs to better fit mid-range and low-end segments but currently all focus is on what will work for Epyc.

AMD only has so much capacity to use at TSMC and it would be utter insanity to sacrifice datacenter market share to satisfy anything in the mobile notebook market or puny DIY market let alone the low-end budget segments of either.

Sadly your argument is undermined by the fact that AMD's Computing and Graphics division has higher revenue than its Enterprise, Embedded and Semi-Custom division (which includes tens of millions of chips for consoles in the last year, representing a significant portion of its revenue) (~$2.6B in Q4 '21 vs. ~$2.2B), and for Intel, the Client Computing Group similarly has much higher revenue (and margins!) than their Data Center Group ($10.1B/$3.5B gross margin (~30.1%) vs. $7.3B and $1.7B margin (~23.3%)). The advantage of getting into the datacenter markets are in a larger total addressable market than just the consumer markets, larger purchases, and long-term contracts, not necessarily in overall profitability. Server chips have massive list price margins, but are sold at heavy rebates in any kind of volume, and those service contracts come with a lot of labor costs - they still earn them a lot of money, but they aren't cheap markets to compete in.

Also, if AMD needed their consumer fans for PR for their enterprise parts, do you think it's reasonable to abandon that strategy after one generation in the lead (two if we're a bit generous)? That still seems incredibly short-sighted.

As for limited supply, that is absolutely a factor. But the priorities AMD has made with that supply is precisely why we have grounds to criticize them. There are plenty of things they could have done to improve low-end and mid-range chip supplies, including making a tiny, 4-core, half L3 chiplet for low end parts, for which they would have gotten massive numbers off of each wafer (and of course they could also have used those for low core count, PCIe/RAM bandwidth oriented EPYC chips). They've chosen to aim narrowly at the premium market with no visible concern for anything else. And with that, they are shooting themselves in the foot.

 

[Cutechri]

Can't say I'm surprised. Hope Zen 4 is as monstrous as it's rumored to be and that Raptor Lake can throw hands with it.