AMD bounces back in latest Steam survey: largest-ever CPU share, RDNA 2 finally charts

cuerdc

Posts: 246   +76
Im sure nvidia is king when it comes to driver support for optimizing gameplay/benchmarks usually best to buy nvidia for that alone.
Playing with a rx6600xt currently myself no issues as yet but would of grabbed a founders edition if could source one.
My old pc still running with hd 5770 many years old held together with tape but still going strong plays a few games still mainly kids with roblox but has steam and plays a few in the library.
 

Lounds

Posts: 955   +858
The consoles use different APIs to the desktop cards. And that’s where AMD fails the hardest - software support. AMD have clearly prioritised consoles over desktop cards. Game developers also neglect Radeon users on PC they represent a minority of users. Console users get better ray tracing support than Radeon desktop users. Control has better ray tracing on XSX/PS5 than it does on a 6800XT!

Although the last time I checked the consoles weren’t exactly graphical powerhouses.
The Radeon software is pretty decent I much prefer it to Nvidia's control center which feels very dated. I'm an RTX 3060 Ti owner and it's my first Nvidia card since the GTX 460 and I'll be honest I do miss the Radeon software suite.
 

Lew Zealand

Posts: 1,973   +2,305
TechSpot Elite
The Radeon software is pretty decent I much prefer it to Nvidia's control center which feels very dated. I'm an RTX 3060 Ti owner and it's my first Nvidia card since the GTX 460 and I'll be honest I do miss the Radeon software suite.
I think it comes down to what you're familiar with. I have 2 gaming PCs set up with about as similar hardware as possible from Intel/Nvidia (8400 + 1080) and AMD (2600 + 5600XT) and I'm simply more familiar with managing the Intel/Nvidia one as I have about 10x the experience with it. (the 8400 was recently replaced with a 9700, so that one has a small leg up now).

I see where some things are better with the Radeon as it has the voltage curve SW and seemingly more single click options built in, but I'm super familiar with squeezing out the performance needed using Afterburner on the 1080. IMO Afterburner gives you finer control than the Radeon interface and that's the kind of control I like.

Still, both setups work great and every time I switch from one to another, I see flaws in what they're doing and think the first one is better. Until I switch back and see different flaws. Neither worse, all very minor, but they're there.
 

Dimitrios

Posts: 946   +758
So I logged into STEAM 1-2 nights ago it took longer than usual. It said something like, GATHERING INFORMATION or SOMETHING ABOUT SYSTEM AND HARDWARE.

First time with my new rig I built a year ago RYZEN 3600 & RX580.
 

Watzupken

Posts: 379   +351
I am struggling to understand how to use the data from Steam to determine whether AMD's market share is growing or not. But in any case, that is not important since AMD is still making money selling most of what they produce. I guess with the hype about Alder Lake that is circulating, it may slow their sales since people will adopt a wait and see approach.
 

brucek

Posts: 940   +1,360
The interesting thing is I find myself agreeing with many posters on some topics while disagreeing with them on others.
Yes. These posters are the reason I spend time in the comments here. They've convinced me in the past they are real people (vs. say a paid shrill), that they know something, and that they can make logical connections. It's worth trying to understand what they're saying and why they're saying it when it feels wrong to me at first.

(I suppose I'd also like someone who always agreed with me, if there ever was such a person, but that'd also be kind of low value add...)
 

HardReset

Posts: 1,311   +983
You tend to skew data in a way that could almost be described as, "deranged".

The fact that these older cards are in use merely indicates that their owners haven't succumbed to needing, "super duper, ultra ultimate, maximum fantastic, be all, end all, ray tracing performance", on 4K monitors.

More people likely play, "Candy Crush", on their stupid phones than use Nvidia and Radeon cards combined. But it's a "deranged" and meaningless data comparison.

If you said that 'Nvidia owners are likely way more addicted to gaming than owners of older Radeon cards", that might make some semblance of sense...or are simply poorer.
My point is that those older cards surely are NOT in use. And making arguments based on that flawed data leads into flawed arguments.

Radeon HD 8800 series was OEM only cards, rebranded 7000-series. Also series contained 15 models, of those only two were 8800-series. Sounds great eh? Popular on Steam survey, Rebranded, OEM only card, 2 models from 15, almost 10 years old. Sounds realistic 🤔
 
I'm not sure it was me who called Radeon cards "crap", but if you want to isolate me saying that, I guess I'll have to cop to it

IIRC, Radeon GPUs tended to dim the room lights when they kicked in. Why some extremely rude individuals even referred to them derisively as "space heaters".

Nvidia had their space heater moment as well with the Fermi line up which you failed to mention. Its all and well making comments such as these however if you fail to mention both companies bad products then its seen as favouritism on which ever side you are.

https://www.techspot.com/review/263-nvidia-geforce-gtx-480/page13.html

As for the Intel their anti-competitive practices amounted to astronomical sums thrown at the top tier PC makers to stop selling AMD servers.

The figures are actually eye watering but you seem to give them a pass on them for it.

For example Dell was getting kick backs over a billion dollars in one year alone not to launch AMD products, and over a prolonged period. The payments were 38 per cent of Dell's operating profit in the year 2006, and amazingly 76 per cent (or $720m) in one quarter, Q1 2007. Total payments came to some 4.3 Billion dollars, Fraud on a grand scale

https://www.theregister.com/2010/07/26/after_the_dell_settlement/?page=1

https://fortune.com/2007/02/15/suit-intel-paid-dell-up-to-1-billion-a-year-not-to-use-amd-chips/

https://money.cnn.com/2010/07/23/technology/dell_intel/index.htm

And those were the payments just to Dell. Add in IBM, HP etc and the scale of monies just get bigger and more ridiculous

The amount of deprived revenue to AMD was huge and certainly contributed to their bulldozer line up due lack of R&D funds. One directly leads to another. (That's not to dispute that AMD made some obvious wrong turns in developments at times).

The worst thing is that Intel got away with it. AMD got a settlement of just over a billion and and Michael Dell a 4 million personal fine, basically a slap on the wrist considering the monies involved.

Intel ? They carried on as normal and tried to do same thing in their attack on ARM and the mobile market. They threw billions in 'Contra-Revenue' payments (Whatever you want to call it, rebate, Contra Revenue, or old fashioned kickback, it's just paying the customer to use your part that they would not have otherwise used) to top tier mobile makers in effort to secure a foothold which fortunately this time around didn't work.
 

Maxiking

Posts: 151   +169
If you are clueless when comes to statistics, do not comment.

When doing a large survey like this, you pick random pcs and you do return them back into the basket, so the surveyed pcs don't have to be unique. The possibility of skewed results is minimal and you can measure it by n/N.

n - the number of pcs randomly chosen
N - the entire sample

n/N is usually <0.00xxxxx

I know, I know, you know better, that's why you are giving away your worthless comments online.
 

HardReset

Posts: 1,311   +983
If you are clueless when comes to statistics, do not comment.

When doing a large survey like this, you pick random pcs and you do return them back into the basket, so the surveyed pcs don't have to be unique. The possibility of skewed results is minimal and you can measure it by n/N.

n - the number of pcs randomly chosen
N - the entire sample

n/N is usually <0.00xxxxx

I know, I know, you know better, that's why you are giving away your worthless comments online.
Three bolded words tells it all. First, this survey is not necessarily "large". Since all numbers are percentages, it's only guess. Secondly, there is no single proof that picked PC's are random. Thirdly, even if something is usually something, it cannot be generalized to apply all cases.

I don't know what you were trying to prove tbh. I have studied statistics if that is problem.
 

Maxiking

Posts: 151   +169
:D

If you really studied statistics, my sides.

Valve's results align pretty much with the Nvidia customer distribution they made public a long time ago, for example, 1060GTX, as the most popular GPU, we knew this fact first from the Valve survey and then from the Nvidia.

And of course, Mr. conspiracy theorist, there is also no proof, that you studied statistics. But there is plenty of evidence you seriously lack the knowledge and instead of arguing constructively, you are making up conspiracy theories about how Valve is either incompetent or malign.

The survey pretty much aligns with everything with know about pc gamers. As a 4 core being most popular during a certain period of time, 1080p the most popular resolution, 1080gtx and 1080ti having also relatively high market share due to their cheap price. When AMD introduced 6 and 8 core Ryzen, the trend in the survey was clear, their share was increasing. When Intel introduced 8700k, 8400k.. the share of 6 core cpu jumped significantly.

So by judging these factors, I CAN SAY.

The survey is large enough to cover the trends accurately including detecting all the ancient hw, and their methodology is right.

I don't have time for this, so bye bye.
 

HardReset

Posts: 1,311   +983
:D

If you really studied statistics, my sides.

Valve's results align pretty much with the Nvidia customer distribution they made public a long time ago, for example, 1060GTX, as the most popular GPU, we knew this fact first from the Valve survey and then from the Nvidia.
Or then Nvidia just looked what is most popular on Steam survey to solidify their own claims. I know these tricks.
And of course, Mr. conspiracy theorist, there is also no proof, that you studied statistics. But there is plenty of evidence you seriously lack the knowledge and instead of arguing constructively, you are making up conspiracy theories about how Valve is either incompetent or malign.
Steam survey has many undeniable flaws. I have listed some of them many times and that clearly proves Valve is incompetent. Having simple errors for years tells more than enough about competitiveness.
The survey pretty much aligns with everything with know about pc gamers. As a 4 core being most popular during a certain period of time, 1080p the most popular resolution, 1080gtx and 1080ti having also relatively high market share due to their cheap price. When AMD introduced 6 and 8 core Ryzen, the trend in the survey was clear, their share was increasing. When Intel introduced 8700k, 8400k.. the share of 6 core cpu jumped significantly.
You basically take some random data you Think is right and then use Steam survey data to "prove" it.

In case you missed news, there was huge demand for 5-core CPU's last month. However demand suddenly dropped this month. Steam survey agrees with me https://store.steampowered.com/hwsurvey/cpus/

5 cpus 0.01% 0.02% 0.01% 0.24% 0.02% -0.22%

So by judging these factors, I CAN SAY.

The survey is large enough to cover the trends accurately including detecting all the ancient hw, and their methodology is right.

I don't have time for this, so bye bye.
Mr. statistics don't have single proof for "large enough" and methodology" claims. Just like I predicted. You're just another anti-AMD troll.
 

captaincranky

Posts: 17,646   +6,438
Nvidia had their space heater moment as well with the Fermi line up which you failed to mention. Its all and well making comments such as these however if you fail to mention both companies bad products then its seen as favouritism on which ever side you are.
Look, I still have an EVGA 9500 GT in service. I use in in an XP box, which is dedicated to the sole purpose of seeking out, and enjoying "erotic art". I certainly don't game on it, nor do I do my banking with it. Wouldn't it be comical if I joined Steam, tried to game with it, and then wound up in the survey?.

Since these type threads routinely devolve into "AMD vs. Intel" & Radeon vs Nvidia, I don't feel any obligation whatsoever to go through a line by line, code name by code name comparison of VGA power draw, just to humor you.

I buy Intel CPUs. This is not because I'm a fan of Intel. It's the reason that each one I have bought, booted up. first try, in a system I built, and is still working superbly, as much as a decade later.

I buy Nvidia's card, because they, (usually) work. My "most potent" is a GTX-1050 ti. I didn't even buy that, "because I needed it " (I don't), I bought it so that I would have a build, that I at least could imagine, "I was a real boy", who owned, "a real computer"..

The 1050 cost me $140.00. Newegg has them in stock now. They are asking for $280.00 them. I'd like to have another, since they're the lowest card in Nvidia's line that are able to have a special driver installed, which will enable 10 bit color depth, on a compatible "one billion color monitor", of which I have 2, 2K versions, still sitting in their boxes. Hell, I don't even know if they work.

So, the next 1050 will have to wait, since I'm not stupid, desperate, affluent, or addicted enough, to pay double what I did for the last one.Period.
 
Last edited: