News AMD Reaches Highest Overall x86 Chip Market Share Since 2013

watzupken

Reputable
Mar 16, 2020
1,058
547
6,070
I feel AMD's gained market share at an accelerated pace over the last few years because Intel stumbled and isn't nimble enough to pick themselves up. I think they grossly underestimated AMD's come back, and gave themselves too much time to rectify their 10nm issues. The next few years will likely see a lot of competition between the 2 as Intel claimed that their 10nm issue is fixed and a flurry of 10nm products to be released starting 2021. Though I still feel they will continue to bleed market share because Intel tend not to compete on price, which AMD easily takes advantage of. Let's see.
 
It makes me happy to see AMD doing well for themselves again, competition is good and prevents the market from getting stale (again).
Now if only AMD had the revenue to manufacture in house like intel does, imagine the progress.
 

PapaCrazy

Distinguished
Dec 28, 2011
311
95
18,890
I feel AMD's gained market share at an accelerated pace over the last few years because Intel stumbled and isn't nimble enough to pick themselves up. I think they grossly underestimated AMD's come back, and gave themselves too much time to rectify their 10nm issues. The next few years will likely see a lot of competition between the 2 as Intel claimed that their 10nm issue is fixed and a flurry of 10nm products to be released starting 2021. Though I still feel they will continue to bleed market share because Intel tend not to compete on price, which AMD easily takes advantage of. Let's see.

Intel's failure defines the current market. Hard to tell who is running to AMD, and who is running away from Intel. Unfortunately, Intel's failures also affected x86 market share as a whole. With MS experimenting with Windows ARM and Apple completely shifting over, Intel created an x86 crisis that will impact AMD too. The next few years will be interesting to say the least.
 

escksu

Reputable
BANNED
Aug 8, 2019
878
354
5,260
This is the reality. Intel still has 80% of the end user market and almost 95% of the server market.

Amd needs to continue its momentum and chip away intel's share. A few years of superior product alone isnt enough to overtake intel. Its a marathon.
 

Jimbojan

Distinguished
May 17, 2017
83
37
18,560
AMD's share gain is down in the noise. Just wait until Intel's 10 nm Tiger Lake CPU and Xe graphic come in volume, and see how much market share AMD will gain. AMD is actually competing with its 7nm chip with Intel's 14 nm chip so far. Intel is still owns the market, clearly it cannot hold over forever, now that it declared 10 nm has arrived both in CPU and graphic, let's see what the battle field is looking like. Intel said it is increasing its capex to increase its 10 nm production in light of the 7nm delay. Just wait for a few more days, Intel will say something about it.
 
Intel's failure defines the current market. Hard to tell who is running to AMD, and who is running away from Intel. Unfortunately, Intel's failures also affected x86 market share as a whole. With MS experimenting with Windows ARM and Apple completely shifting over, Intel created an x86 crisis that will impact AMD too. The next few years will be interesting to say the least.

To be fair AMD caused it by pushing AMD64 instead of us going to IA64.

Either way though it wont be as bad as people think. Apple is still a marginal market share and has very little bearing on the overall PC market. I also wouldn't be surprised if a lot of people get upset since now all their software will be useless.

I am not sure we will see major changes even with MS doing ARM. They are doing it because its smart to do but people hate and resist change, its why AMD64 won so easily over IA64, so it may be interesting but we may not see any real change to the consumer desktop market apart from AMD having more market share.
 

Jimbojan

Distinguished
May 17, 2017
83
37
18,560
It is not clear AMD has gained any market share from NVDA or not, however, Intel is very likely to gain graphic market share from both AMD and NVDA when its Xe graphic is out. That was the reason why NVDA was trying to acquire other companies to increase its revenue to cover the market share loss in graphics.
 
Intel's failure defines the current market. Hard to tell who is running to AMD, and who is running away from Intel. Unfortunately, Intel's failures also affected x86 market share as a whole. With MS experimenting with Windows ARM and Apple completely shifting over, Intel created an x86 crisis that will impact AMD too. The next few years will be interesting to say the least.
Intel increased their market share by 20% last years third quarter compared to the second and has remained at that level ever since.
People are running towards intel in such numbers that intel can't keep up causing people to buy CPUs from wherever possible, even from AMD.
AMD made 1.9bil compared to intels 19.7bil so roughly 10% the sales but intel made 5.1bil net income (money in the pocket) while AMD only made 157mil so less than a third of 10% of what intel made.
Even if AMD would be gaining market share doing so by undercutting prices is the most terribad idea ever, any market share they win will be immediately lost if they ever increase prices again, this tactic would only make sense if you could flood the market and completely take over so people would be forced to pay through the nose for the next gen.

https://www.intc.com/investor-relat...d-Quarter-2020-Financial-Results/default.aspx
https://ir.amd.com/news-releases/ne...reports-second-quarter-2020-financial-results
 
I don't think their 10nm is quite as "Fixed" as they proclaim it to be.

And inside leaks state 7nm is a even worse debacle than 10nm.

I agree with you. It's not fixed. While yields are up, it's speed and power usage is well below what they had hoped for. No matter how Intel tries to paint 10nm a solved problem, it's not. The only thing saving it is new chip architecture designs, not the node itself.
 
AMD's share gain is down in the noise. Just wait until Intel's 10 nm Tiger Lake CPU and Xe graphic come in volume, and see how much market share AMD will gain. AMD is actually competing with its 7nm chip with Intel's 14 nm chip so far. Intel is still owns the market, clearly it cannot hold over forever, now that it declared 10 nm has arrived both in CPU and graphic, let's see what the battle field is looking like. Intel said it is increasing its capex to increase its 10 nm production in light of the 7nm delay. Just wait for a few more days, Intel will say something about it.

All early previews of Xe are looking bad, like Really REALLY REALLY bad. It needed a really effective and efficient 10nm. That node is a failure. There's rumors that Intel will dump it as it's costing them a ton of money, and it's not producing anything really competitive. Raja is a nice guy, but if this happens, his days of being an upper level leader are over. He may be brilliant, but that would mark two times he failed in a huge way. (Running RTG, and now Intel with near unlimited resources) I can't fault Raja for doing so poorly with AMD as they were so strapped for cash at the time. But with Intel, he really doesn't have any excuses.

Intel needs to maintain P2E to keep their stock up. And right now their CapEx on non successful divisions is insane. Intel has a very solid history of starting projects and gobbling up startup companies, and then just dropping them like hot potatoes. This includes their modem business, Atom for mobile devices, 1st gen dGPU's, Knights Landing/Xeon Phi series, their autonomous drone series, and so many other ventures and startups I can't even list them all.

The fact is intel buying so much capacity from TSMC says a lot about how they view their own internal prospects.

BTW: 10nm is generally regarded as a failure by the majority of tech press.
 
Last edited:
To be fair AMD caused it by pushing AMD64 instead of us going to IA64.

Either way though it wont be as bad as people think. Apple is still a marginal market share and has very little bearing on the overall PC market. I also wouldn't be surprised if a lot of people get upset since now all their software will be useless.

I am not sure we will see major changes even with MS doing ARM. They are doing it because its smart to do but people hate and resist change, its why AMD64 won so easily over IA64, so it may be interesting but we may not see any real change to the consumer desktop market apart from AMD having more market share.

IA64 had a very large instruction word that depended on huge amounts of software optimization by the compiler that never materialized. It pushed even further into what made CISC CISC. While ARM has become more CISC like they are still a much simpler architecture beneath. One of the things that makes CISC chips so complex is circuitry for legacy modes, including memory access. Do you know how many modes there are to support memory access? I think it was 8+, last I counted. Local 64K Paged jumps, Base, System, Extended, Expanded, Virtual, contiguous 32 Bit, 64 bit. It's insane.

However I think Apple moving to ARM for desktops is just going to kill off their desktop brand. Overall this is a bad decision. While they MIGHT get better performance for recompiled applications, Apples desktop and laptop sales have been going into the sewer. Prices have sky rocketed, while quality has actually gone down. The real effort here is to control margins as Apple gets a bigger revenue slice on new hardware if they control all the chips. Plus it locks out Hackintosh systems.

The only saving grace for Apple was, IMHO, the dual boot mode so users could access windows when they needed to run applications that were commonly available only on Windows. Windows adoption for ARM has been painfully slow and app compatibility slow and unstable.

If you remove this capability to run Windows in CISC efficient mode, then you will remove the ability of Apple users to dynamically switch between environments. Recompiled CISC emulation is never efficient and never guaranteed compatible. For example, I can dynamically issue a flag to set the number of bits the x86 FPU uses for computations. I can use 80 bit, 64bit, 32bit and 16bit. It is an assembly instruction. You can't do that with ARM. 80bit IEEE internal is an x86 thing. It would have to be emulated, or you risk getting different results back. A lot of MIMD and SIMD instructions would have to be emulated to ensure compatibility. And MIMD and SIMD instructions are used extensively for things like video editors/encoders,decoders/CAD/3D.

We have a few testers here that love their MAC desktops. However they haven't updated their hardware in years. And whenever they log into work, they have to go to Windows boot to do work. I would love to see the dual boot mode statistics. I bet Apple has them.
 
Last edited:
I think is fair to said that AMD is where it is now, because they did a good job makig it posible to have a nice product for a very good price. A product that can now compete to intels finest in the desktop, and even kick intel ass really bad on the HEDT segment, not to mention the nice server CPUs.

Having intel issues or not with its 10nm technology does not takes away the fact of what AMD has been able to pull-off coming from the FX era..
Also keep in mind that we are already on the 14nm++++. Intel had a lot of time, and they did in fact an amazing job refining the same tech over and over again, to a point that maybe, just maybe, they managed to get this latest 14nm++++ iteration to be really close in performance to a first gen 10nm CPU would have been a year ago (power wise thats a diferent story).
 
Last edited:
  • Like
Reactions: prtskg

Soaptrail

Distinguished
Jan 12, 2015
302
96
19,420
To be fair AMD caused it by pushing AMD64 instead of us going to IA64.

Either way though it wont be as bad as people think. Apple is still a marginal market share and has very little bearing on the overall PC market. I also wouldn't be surprised if a lot of people get upset since now all their software will be useless.

I am not sure we will see major changes even with MS doing ARM. They are doing it because its smart to do but people hate and resist change, its why AMD64 won so easily over IA64, so it may be interesting but we may not see any real change to the consumer desktop market apart from AMD having more market share.

Well you could also argue that Intel's Itanium paper announcement that caused the entire PC industry to believe the hype caused DEC to stop making Alpha chips which lead to their designers and workers going to AMD which allowed AMD to create their first Althon 64 bit CPU's. I know John C. Dvorak wrote about it but i cannot find that article now.
 

TJ Hooker

Titan
Ambassador
Intel increased their market share by 20% last years third quarter compared to the second and has remained at that level ever since.
Their revenue increased by 20%, not market share.

Even if AMD would be gaining market share [...]
I don't know why you're stating this as a hypothetical; AMD is gaining market share, as shown quite clearly by the data in this article.

You're really giving the impression you don't understand what market share is.
 
  • Like
Reactions: prtskg and King_V
If you remove this capability to run Windows in CISC efficient mode, then you will remove the ability of Apple users to dynamically switch between environments.
I mean, technically they could have two computers, one for their Mac work, and one for their Windows work. Maybe continuing to use their old Mac as a Windows system, while requiring the purchase of a new one in order to keep running the latest Mac-exclusive software. Apple seems to have a tendency of forcing their old hardware and software into obsolescence, and this will be their third time switching to a different processor architecture over the years. They kept the Motorola 68K architecture for around 10 years, then switched to PowerPC for around 10 years, and then to x86 for around 15 years. Now, they'll be switching to ARM, but it wouldn't surprise me if they switched again to something else another decade or so down the line, assuming they are even making traditional desktops and laptops at that point.
 
Their revenue increased by 20%, not market share.
While their gross margin dropped by 6.6% and their taxes increased by 2.4% ,so they made more revenue although they had higher costs=made less per unit,so explain to me like I'm 5, how is that possible without increasing your market share?
Desktop platform average selling prices was also down 3%
I don't know why you're stating this as a hypothetical; AMD is gaining market share, as shown quite clearly by the data in this article.
They do have tables with numbers but without being able to cross check them with the source they are referencing they are just tables with numbers.
Also if AMDs 18% of market share comes down to less than a third of 10% of what intel is making then my point is still valid,as soon as AMD increases prices people will drop them like a hot potato.
 
While their gross margin dropped by 6.6% and their taxes increased by 2.4% ,so they made more revenue although they had higher costs=made less per unit,so explain to me like I'm 5, how is that possible without increasing your market share?
Desktop platform average selling prices was also down 3%

They do have tables with numbers but without being able to cross check them with the source they are referencing they are just tables with numbers.
Also if AMDs 18% of market share comes down to less than a third of 10% of what intel is making then my point is still valid,as soon as AMD increases prices people will drop them like a hot potato.

Could easily be other sources than just CPUs. Intel has a hand in a lot of things so even with some market share loss their revenue can continue to be high or even grow if the markets they are in are emerging markets or they have a decent enough product.

Consumer markets are not high margin markets compared to HPC, AI or other markets. Compare that a top end consumer chip is $1K-$2K while top end server chips can reach near $10K.
 
  • Like
Reactions: TJ Hooker

spongiemaster

Admirable
Dec 12, 2019
2,289
1,290
7,560
While their gross margin dropped by 6.6% and their taxes increased by 2.4% ,so they made more revenue although they had higher costs=made less per unit,so explain to me like I'm 5, how is that possible without increasing your market share?
Desktop platform average selling prices was also down 3%

Ok. Billy and Sally are both selling lemonade. They both sell 50 lemonades each so Billy has 50٪ marketshare. The following week, Billy sells 60 lemonades and makes more money, but lemons now cost more so his profits drop. Not only that, but Sally is now wearing a bathing suit and sells 100 lemonades by herself. Total market has increased from 100 to 160. So now Billy's market share has dropped from 50% to 37.5% despite generating more revenue.
 
  • Like
Reactions: TJ Hooker

TJ Hooker

Titan
Ambassador
They do have tables with numbers but without being able to cross check them with the source they are referencing they are just tables with numbers.
The results from Mercury Research are widely reported by other tech publications, and have been for years. AMD themselves have referenced Mercury Research results on their investor relations page in the past, which would open them to a lawsuit if the results were bogus.

spongiemaster has explained how it's possible to generate more revenue while not gaining market share.