AMD is competitive!

upload_2018-1-23_7-54-16.jpeg

upload_2018-1-23_7-54-32.jpeg

images

images

images


The good ol Celeron, may be able to do something from 2005 in 2018.
 
Intel is in a heap of trouble and it looks like a long term affair as well. Intel has nothing basically to compete against Nvidia, as more processing is being done by GPUs as time goes on, exponentially. Every market AMD has placed product with Ryzen, they penetrated significantly. Zen + will probably make Coffee Lake looked like Kaby Lake of last year. I can only imagine what Threadripper 2 will do. Intel better release 10nm processors fast! Wait - they delayed it again. At this rate Zen 2 with 7 nm will be well established before Intel gets their 10 nm processors down. Biggest issue I see with AMD is GPUs - they are way more important for future processing then CPUs and they are behind the 8 ball - at least 7nm Vega's have been revealed. I just see Nvidia exploding if they can make the GPU's needed/wanted. AMD I see doing well in continue capturing Intel market even though it maybe shrinking - I don't think AMD can afford to ignore GPU's from the top end to the low end.
 
They do not post sales numbers, just most popular demand for the cycle, and for the first time in 60-80 days Coffee lake has stock, but again Paul Alcorn has covered the trivialities of relying on bad data and making an opinion on it.

Celerons and Pentiums sold so much they can't even make the list.

Paul Alcorn article confirms AMD got 3% desktop marketshare in 2017. But that was before CoffeeLake took the market back for Intel.

The guy that collects the Mindofactory data leaves out lots of CPUs. If you check Amazon best-seller list you will see there are Pentiums and Celerons among the top ten.
 
But amazon provides no data, lets call that legit. Like I know with 100% certainty that nobody where I am from buys off amazon as it is expensive and our distributors and resellers are more dependable in dealing with RMA's without data, it is just a monthly popularity log with no numbrs of units sold. Again mindfactory gives the most legitimate number of units sold and Ryzen has sold a boat load in 2017, ie ~40% of their total CPU's sold.
 
Thats what we get reading your posts over and over, where does it tell you how many units have sold? one can feasibly have a top seller if you sell 1 unit and the others don't sell. Mindshare showed units sold
 
After watching Newegg top selling CPU's since the launch of Ryzen, Ryzen always seems to have a presence in the top 10. The chips in these spots change all the time, currently the #1 selling CPU is the R5 1600 followed by the 7700K. Interestingly, the 8700K (reportedly the best selling processor to have ever existed or will exist until Intel releases the 9700K) is ceremoniously holding it's #1 spot at #6 position.
 
After watching Newegg top selling CPU's since the launch of Ryzen, Ryzen always seems to have a presence in the top 10. The chips in these spots change all the time, currently the #1 selling CPU is the R5 1600 followed by the 7700K. Interestingly, the 8700K (reportedly the best selling processor to have ever existed or will exist until Intel releases the 9700K) is ceremoniously holding it's #1 spot at #6 position.
lol, I wonder how that works out, #6 = #1. Only in some minds that would ever make sense.
 
After watching Newegg top selling CPU's since the launch of Ryzen, Ryzen always seems to have a presence in the top 10. The chips in these spots change all the time, currently the #1 selling CPU is the R5 1600 followed by the 7700K. Interestingly, the 8700K (reportedly the best selling processor to have ever existed or will exist until Intel releases the 9700K) is ceremoniously holding it's #1 spot at #6 position.

What has to do the position that a chip has on Newegg sales with the position in Amazon sales?
 
Last edited:
What has to do the position that a chip has on Newegg sales with the position in Amazon sales?

Not much really, I guess about as much as any other individual sales data point, be it from Amazon or Newegg or whoever else. I guess since you agree it actually doesn't represent total sales in any way, I don't see how being #1 in Amazon translates to anything more than being #1 on Amazon. If it does, could you explain how it does represent more than what the data point actually represents.
 
Not much really, I guess about as much as any other individual sales data point, be it from Amazon or Newegg or whoever else. I guess since you agree it actually doesn't represent total sales in any way, I don't see how being #1 in Amazon translates to anything more than being #1 on Amazon. If it does, could you explain how it does represent more than what the data point actually represents.

Amazon represents worldwide sales. Newegg is basically a local store. Also we did spend some time in another thread discussing the issues with Newegg, including the lack of stock for CFL chips and the abusive pricing. Both issues affected sales without any doubt.
 
Amazon represents worldwide sales. Newegg is basically a local store. Also we did spend some time in another thread discussing the issues with Newegg, including the lack of stock for CFL chips and the abusive pricing. Both issues affected sales without any doubt.
Wait... What? So are you implying Amazon doesn't inflate their prices or have stock issues as well? And by the by Newegg is not even close to a local store, Microcenter maybe. Amazon total sales are larger but so is the scope of there catalog, Newegg is primarily a computer/electronics store. In comparison Newegg had $2.63B in sales to Amazons $135B in 2016. Take into account catalog size differences and world wide impact and I dare say Newegg would not be too far off as a great indicator of sales, more so because that is the bulk and near entirety of their business. And as far as stock Amazon is well known to sell way beyond stock levels so I wouldn't credit them with having better stock than anyone else, most cases making their rating on such circumstances worse (look at Ryzen launch).
 
The one thing I can imagine is that AMD's market share could have potentially been greater had the ram prices not been so stupid expensive. It is beautiful to see a company come back from their greatly exaggerated death and be successful once again. Now AMD still needs to make themselves a name that will be easily recognizable to those outside of the computing circles but, at least it is a start. :)
 
I'm checking Amazon's best-seller lists. USA, France, Germany, the UK, the 8700k/7700k is #1.

Ryzen shows up #3, #8, #4, and #3 respectfully on those lists. Within the Top 20, all Ryzen 3, 5, 7 show up as well. (Except 7 in France, #22).

Strangely, a FX-8350 is also showing up at #13 (USA), #12 (GER), #10 (UK), the FX-8300 at #18 (GER), the FX-4300 at #17 (UK), and FX-6300 #19 (UK).
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
upload_2018-1-25_8-54-49.jpeg


since getting actual market data is hard, I have appealed to local distributors and resellers for their unit sold but not all are willing to disclose it.

From what I have it looks much like mindshare.de the desktop trend shows AMD have taken around ~40% of the total sales of CPU's from Intel in 2017 with this to trend up in 2018 with a new release of both performance desktop and APU's.

total addressed market which is currently in the 9-12% range will likely trend upwards insofar as:

1) Eypc servers will be adopted more.
2) AMD will feature on mobility parts which is a bigger market than desktop.
3) Growing mindshare, AMD will have the market believing in alternatives again, something they hadn't done for 10 years. There is reason to buy AMD again.
4) APU's address 30% market space for users not needing dGPU's, currently AMD has no noteworthy iGPU part on the market but that will change soon.

closer to 50-50 share in performance PC marketshare is possible still and the targeted 20% overall market is a reasonable projection for AMD in 2018
 
Last edited:
Amazon represents worldwide sales. Newegg is basically a local store. Also we did spend some time in another thread discussing the issues with Newegg, including the lack of stock for CFL chips and the abusive pricing. Both issues affected sales without any doubt.

Amazon represents Amazon sales, that is all. Newegg represents Newegg sales, that is all. Neither is an indication on total CPU sales, they only represent themselves in their numbers. Any implication that one set of numbers from an individual retailer is indicative of all sales at all retailers of a specific item is questionable at best. I dare say that the information that we both gave is unconnected B.S., the only thing sad about that is I seem to be the only one of us that knows it.
 
Wait... What? So are you implying Amazon doesn't inflate their prices or have stock issues as well?

No.

Amazon represents Amazon sales, that is all. Newegg represents Newegg sales, that is all.

Amazon sales can be taken as an indicator of worldwide sales. Newegg and Mindfactory cannot.

Moreover, Amazon best selling list is not being mentioned in a vaccum of information. Amazon sales are being used in conjunction with orher parameters and all the parameters together point in the same direction: AMD got momentum in 2017 with RyZen launch and then lost almost all of it when its competitor launched CoffeLake.
 
View attachment 51554

since getting actual market data is hard, I have appealed to local distributors and resellers for their unit sold but not all are willing to disclose it.

From what I have it looks much like mindshare.de the desktop trend shows AMD have taken around ~40% of the total sales of CPU's from Intel in 2017 with this to trend up in 2018 with a new release of both performance desktop and APU's.

total addressed market which is currently in the 9-12% range will likely trend upwards insofar as:

1) Eypc servers will be adopted more.
2) AMD will feature on mobility parts which is a bigger market than desktop.
3) Growing mindshare, AMD will have the market believing in alternatives again, something they hadn't done for 10 years. There is reason to buy AMD again.
4) APU's address 30% market space for users not needing dGPU's, currently AMD has no noteworthy iGPU part on the market but that will change soon.

closer to 50-50 share in performance PC marketshare is possible still and the targeted 20% overall market is a reasonable projection for AMD in 2018

Mercury Research, Roland analysis,... show that AMD got about 3% marketshare in desktops with Ryzen

AMD Desktop Unit Share
4Q16: 9.9%
4Q17: 12.9%

But this data doesn't count the latter effect of CoffeeLake launch. CoffeeLake puts Intel back to #1 in sales (as Amazon, Mindfactory,... show). Consequently AMD loose marketshare in 2018 and it is back to pre-zen times now, as confirms the evolution of Passmark marketshare graph.

EPYC is having harder times than Ryzen. The marketshare gained in servers is much smaller than the 3% measured in desktops the past year.

Ryzen mobile also performed worse than desktop Ryzen: "However, Intel still gained a bit of share in laptops as Ryzen Mobile’s limited ramp underwhelms thus far". It remains to be seen the impact of Ryzen mobile this year, with full production of chips. I expect mobile Ryzen to gain marketshare in the single digit percent.
 
And then Coffee Lake will wain when Ryzen refresh is released, then the Ryzen refresh will wain because the Intel 9000 series is released, then the intel 9000 series will wain because Ryzen 2 is released and so on and so on. This fluctuation is normal and it would be abnormal if it didn't fluctuate back and forth.

As for Amazon's numbers being anything more than what Amazon has sold over a time period is ludicrous. Implying that those numbers should be venerated more than others is simply short sided.
 
And then Coffee Lake will wain when Ryzen refresh is released, then the Ryzen refresh will wain because the Intel 9000 series is released, then the intel 9000 series will wain because Ryzen 2 is released and so on and so on. This fluctuation is normal and it would be abnormal if it didn't fluctuate back and forth.

I expect RyZen refresh to bring some marketshare back to AMD this 2018. But how much? I expect low sales for this RyZen refresh.

It seems evident that most people that upgraded to first gen RyZen from a previous Bulldozer/Piledriver build is not going to upgrade again this year. It also seems evident that most owners of Coffelake will not migrate to RyZen refresh.

I will share my thoughts about Intel 9th series and Zen2 processors another day.
 
We can put together the data of Amazon, Mindfactory, and Newegg together, and in my opinion, use that to compare and hash out trends. Looking at everything, I would agree with Juanrga that Intel is back on top with the 8700k, the charts are basically in agreement with that, but unlike Juanrga, I don't think it's a slam dunk, AMD's place in the market is lessened but the charts state that AMD continues to sell Ryzen at a decent rate and continues to be competitive on the sale charts. Intel will be reporting earnings today anyhow, but I won't be surprised if they match or exceed expectations regarding Coffee Lake which imo are good. AMD's earnings will be next week Tuesday.
 
I expect RyZen refresh to bring some marketshare back to AMD this 2018. But how much? I expect low sales for this RyZen refresh.

It seems evident that most people that upgraded to first gen RyZen from a previous Bulldozer/Piledriver build is not going to upgrade again this year. It also seems evident that most owners of Coffelake will not migrate to RyZen refresh.

I will share my thoughts about Intel 9th series and Zen2 processors another day.
Hell no! I see huge gains for AMD as Intel's security fixes cause HUGE performance degradation and a general displeasure among current owners so much so they flock to AMD to express said displeasure. And assuming AMD gets Epyc availability up worldwide, then the gains may be nothing short of astronomical.
 
Hell no! I see huge gains for AMD as Intel's security fixes cause HUGE performance degradation and a general displeasure among current owners so much so they flock to AMD to express said displeasure. And assuming AMD gets Epyc availability up worldwide, then the gains may be nothing short of astronomical.

Most workloads take a 0--5% performance impact on Intel hardware with both Meltdown+Spectre patches. The impact of Spectre on AMD hardware has to be measured still.

I am convinced the current predictions of huge marketshare gains for AMD this year will follow the same route that the older claims of huge gains made the past year.
 
Most workloads take a 0--5% performance impact on Intel hardware with both Meltdown+Spectre patches. The impact of Spectre on AMD hardware has to be measured still.

I am convinced the current predictions of huge marketshare gains for AMD this year will follow the same route that the older claims of huge gains made the past year.
On the cherry picked ones of course. ;)
 
Most workloads take a 0--5% performance impact on Intel hardware with both Meltdown+Spectre patches. The impact of Spectre on AMD hardware has to be measured still.

I am convinced the current predictions of huge marketshare gains for AMD this year will follow the same route that the older claims of huge gains made the past year.

Not sure who is claiming huge marketshare gains for AMD, I would even venture to agree with you on this point. I see steady growth and increased marketshare of Ryzen systems and while not toppling the entrenched Intel juggernaut in any market segment, AMD should steadily increase it's marketshare in all segments they currently inhabit. I would consider this positive for the company and the industry especially since AMD was supposed to go bankrupt in 2013 or 2014 or 2015 or 2016, according to the multitude of random forum Nostradamus's 'absolutely going to happen' predictions.
 
On the cherry picked ones of course. ;)

Are Microsoft, Google, Red-Hat, Computerbase, Hardewaredeluxx,... cherry picking workloads? Sure they aren't.

Not sure who is claiming huge marketshare gains for AMD

I replied to a poster that "sees huge gains for AMD as Intel's security fixes cause HUGE performance degradation and a general displeasure among current owners so much so they flock to AMD to express said displeasure"

He expects a massive migration of users from Intel to AMD. That is not happening.
 
Are Microsoft, Google, Red-Hat, Computerbase, Hardewaredeluxx,... cherry picking workloads? Sure they aren't.



I replied to a poster that "sees huge gains for AMD as Intel's security fixes cause HUGE performance degradation and a general displeasure among current owners so much so they flock to AMD to express said displeasure"

He expects a massive migration of users from Intel to AMD. That is not happening.

Yes but Intel could be burning to the ground and you would just insist they are doing thermal testing on their next processors and all is well.
 
Yes but Intel could be burning to the ground and you would just insist they are doing thermal testing on their next processors and all is well.

Thermal testing is apt, given the 26% power spike from 4.7 to 4.8+ and how only high end cooling is used to bench coffee lake, where are the stock fans? even the 8400 tested with high end cooling that is almost as much as the chip itself. Also delid benches are horse crap, 80% of buyers would not even consider that especially the premium put on pre delidded parts. If I were intel I would mark those CPU's and void warranties entirely, even use inferior materials to ensure failures and more money off their fanboys who think buying a CPU to overclock just to get the performance you expect is a good deal.
 
Are Microsoft, Google, Red-Hat, Computerbase, Hardewaredeluxx,... cherry picking workloads? Sure they aren't.



I replied to a poster that "sees huge gains for AMD as Intel's security fixes cause HUGE performance degradation and a general displeasure among current owners so much so they flock to AMD to express said displeasure"

He expects a massive migration of users from Intel to AMD. That is not happening.
Actually I was imitating you by just blowing crap out of my behind using just a little truth of the current situation and adding a HEAP LOAD of bias. I think I emulated you post habits almost exactly.

Anyway for you since I have yet to see you post once in any of the Intel Meltdown/Spectre patch threads:

https://hardforum.com/threads/linus-torvalds-rips-into-intel.1953015/page-2#post-1043450600
Thank you!! inspectre confirmed Meltdown was patched and reporting as "SLOW". I ran some benchmarks for Shogun2 and then disabled Meltdown patch and rebooted and re-ran the benchmarks. Massive improvement!

  • DX11, ultra settings: 22.85 fps -> 38.5 fps
  • DX9 benchmark: 17.05 fps -> 54 fps
No idea why DX9 was so gimped.. but ignoring that, looks like disabling Meltdown patch showed a 68% improvement (or rather, the patch introduced a 40% performance degradation for DX11 ultra settings).

The cynic in me thinks Intel is trying to get folks to buy new processors.
That thus far looks like WAY MORE than 1-5%. Granted not really scientific as only one data point but it definitely proves you are full of manure.
 
  • Like
Reactions: Zuul
like this
That thus far looks like WAY MORE than 1-5%. Granted not really scientific as only one data point but it definitely proves you are full of manure.

My claim was that most workloads take a 0--5% performance impact. I didn't claim that all workloads are bound to that.

In the same way, when Microsoft, Google, Red-Hat,... measure performance under different workloads and report us the impact is "negligible" in most cases, they don't claim you cant find few workload where it is not.
 
Last edited:
We can put together the data of Amazon, Mindfactory, and Newegg together, and in my opinion, use that to compare and hash out trends. Looking at everything, I would agree with Juanrga that Intel is back on top with the 8700k, the charts are basically in agreement with that, [...] Intel will be reporting earnings today anyhow, but I won't be surprised if they match or exceed expectations regarding Coffee Lake which imo are good. AMD's earnings will be next week Tuesday.

Intel report is out: "Intel's Client Computing Group (CCG) shipped a record volume of Intel® Core i7 processors".

As I said Amazon was really representing the trend in worldwide sales. AMD numbers are expected for next week.
 
My claim was that most workloads take a 0--5% performance impact. I didn't claim that all workloads are bound to that.

In the same way, when Microsoft, Google, Red-Hat,... measure performance under different workloads and report us the impact is "negligible" in most cases, they don't claim you cant find few workload where it is not.
What did you or others do to test most?
 
Would you trust his tests if he did share them?

Why not pick apart his references, that should be easy, right?
I don't think quite a few people trust what comes out of his mouth as a whole, much less testing. Easy to cherry pick reference graphs and such.

I don't get paid by AMD or Intel, so I don't really take sides.
 
Remember: denying the facts will not make they disappear.

People denying the conclusions draw from Passmark marketshare graphs, Amazon best selling list, Mindfactory sales reports,... didn't make AMD to sell more processors. Same happens with the performance impact of Meltdown and Spectre. Denying the data given by Microsoft, Google, Red-Hat, Computerbase, Hardwaredeluxx,... will not make AMD to sell a huge amount of processors.

I will remark again that I agree that AMD is more competitive with Zen. I disagree with all the hype and bias surrounding it. And when I write "hype" I mean all those fantasies like RyZen getting 40% marketshare and EPYC getting 16% marketshare the past year.

Will AMD be more competitive this year? Yes, but again by small quantities. E.g. Pinnacle Ridge is a minor refresh of Summit Ridge, so don't expect more than some few % gain over current marketshare.
 
Last edited:
Remember: denying the facts will not make they disappear.

People denying the conclusions draw from Passmark marketshare graphs, Amazon best selling list, Mindfactory sales reports,... didn't make AMD to sell more processors. Same happens with the performance impact of Meltdown and Spectre. Denying the data given by Microsoft, Google, Red-Hat, Computerbase, Hardwaredeluxx,... will not make AMD to sell a huge amount of processors.

I will remark again that I agree that AMD is more competitive with Zen. I disagree with all the hype and bias surrounding it. And when I write "hype" I mean all those fantasies like RyZen getting 40% marketshare and EPYC getting 16% marketshare the past year.

Will AMD be more competitive this year? Yes, but again by small quantities. E.g. Pinnacle Ridge is a minor refresh of Summit Ridge, so don't expect more than some few % gain over current marketshare.

With the concession of now that AMD will increase market share in the coming year, he must also mean the inverse is true, Intel will also lose market share in the coming year. Considering that would be 2 years on a row of increased market share, that would also mean that Intel will have lost market share in those same 2 years. Thanks, Juan for clearing that up.
 
How is Mindfactory.de not technically the same as Newegg.com? In your words they should be considered local retailers and their results don't matter compared to Amazon numbers because the Amazon numbers have clairvoyant properties that predict world trends. Next time you are around these magical numbers, could you ask them if they know the lotto numbers next week, that would be great.
 
With the concession of now that AMD will increase market share in the coming year, he must also mean the inverse is true, Intel will also lose market share in the coming year. Considering that would be 2 years on a row of increased market share, that would also mean that Intel will have lost market share in those same 2 years. Thanks, Juan for clearing that up.

Concession of now? You don't read my posts. The past year I predicted that AMD would increase market share with Zen products in 2017, with my low-single-digit expectations being finally confirmed. I am confident that my expectations for this year will be confirmed as well.
 
Last edited:
Even though Intel 8th Gen has a CPU on top of the Amazon list, there is still a huge disparity in total number of verified reviews, which should give some idea of total cpus sold. Intel 8th gen has 111 COMBINED reviews. Meanwhile, the R5 1600 has 259, the 1700 has 345, the 1300x has 56, the 1800x has 169, the 1200 has 78, the 1600x has 141, the 1500x has 96, the 1700x has 173, and the 1400 has 58.

So, 1385 combined for Ryzen and just 111 combined for Intel 8th Gen. Yes, Ryzen has been out longer, but I do no see the 8th Gen catching Ryzen anytime soon on US Amazon, especially with Ryzen+ around the corner.

On the world stage Intel is still unchallenged, especially laptops. Some regions most likely don't even have Ryzen. Hell, in areas like China and India, they probably cell more 7th gen Celerons than ALL Ryzens combined. The two countries combined with other developing and undeveloped Asian countries cover about half of the World's population alone. Because of these factors, it could still be a while before we really see the effect of Ryzen to the CPU market share.
 
Back
Top