AMD's CPUs lose ground to Intel in latest Steam hardware survey, Windows 7 gains users

duckofdeath

Posts: 448   +586
Be that as it may, you can still go out and piss away $200.00 on a stinking pair of sneakers with some basketball player's name on them.

That same deuce might net you close to 4, 2 TB HDDs, a kick a** set of memory sticks, or even a mid to high range mobo.

Even at its current, "over priced state", you still get more for your money in tech than just about anything else on the planet. That's taking everything into consideration, R & D, raw materials, and especially the hyper precision equipment required to produce it.

How long does it take the average person to rip through $200.00 worth of gasoline, parking fees, and tolls? And you don't have anything to show for it when the money's gone.

Jus' sayin'.

PS: As long as you stay far away from those damned flagship smartphones, tech's a bargain.
So, that's a justification for tech to completely break Moore's Law, without an honest reason? Sneakers? I don't buy it. Literally or figuratively. 😁
 

ZedRM

Posts: 616   +387
As explained ad nauseam already, that is not true, its not as reliable as you think.
It is absolutely true. Steam has stated that surveys often have 60% to 70% participation out of all those who receive them. That is a significant portion of Steam users and MORE than enough to gain an accurate view of user stats. This grants them credibility which is why media outlets, such as this one we're discussing on, have the ability to write articles and be able to confidently cite the source. Context is important. So to is merit.

What credibility do YOU have? Hmm?
 
Last edited:

ZedRM

Posts: 616   +387
You're paying way too much for tech these days.
Another statement without merit. Technology today is much, MUCH less expensive than it once was. Sure, prices go up and right now we have an artificially inflated market(thanks to the pandemic and cryptominers), however, as a general rule, prices per capita are far less expensive than they once were. Example, in the early 1980's 1MB(that's megabyte as in 1024KB) of RAM for a given computer was around $440(I know, I was there and frequently bought RAM). For context, the average median household monthly income was $370. The late 80's and early 90's weren't much better. In the late 90's prices started coming down to the point where people could buy a PC on 2 or 3 months wages. Nowadays, you can buy a COMPELETE, very well equipped PC for less than what most people make in a month. That's a WHOLE PC, not just RAM.

Yes, context is very important.
 
Last edited:

captaincranky

Posts: 17,217   +5,972
So, that's a justification for tech to completely break Moore's Law, without an honest reason?
Reality check. "Moore's Law|". was never a physical certainty. It is just the imaginative and speculative ramblings of someone close to the technology.

The difficulty associated in shrinking the technology squares in the inverse to the transistor count. Based on my "idle speculation", the transistor count will still rise, , but in a linear, not in a logarithmic fashion.

I've been listening to this "Moore's Law' bullsh!t for the last 15 years here. You'd think god burned it into a stone tablet, and handed it to Mr. Moore in person. I can assure you, that's not how it happened, or came to be.

If you want mathematical certainty, see Pythagoras.
 
Last edited:

Lew Zealand

Posts: 1,922   +2,197
TechSpot Elite
Funny how you avoid to mention how close the midrange 5 series is to 400 bucks now. Or is that high end too? All of them high end, to justify gouging the consumer? I thought it was only apple living in a fantasy world where we don't compare contemporarily, but only with selective legacy specs? Have PC sunken this low in denial?
They saw an opportunity to cash in massively and they did. I'm not saying I wouldn't do the same, but stop kidding yourself. You're paying way too much for tech these days.

You made the claim, I was just trying to understand what processor you were referring to. And no surprise: you still gave no answer, avoiding to mention it.

FYI the R5 5600 is available for $300 now, not $400 like in your personal narrative. Sure, it was available for more in the past as constrained supply drives prices up but you seem more interested in pointing a finger at AMD saying "overpriced!"

Your argument would be decent if you stuck to the $300 list and currently available price and declare it too expensive, which IMO it still is. But you do you.
 
Last edited:

Lew Zealand

Posts: 1,922   +2,197
TechSpot Elite
It is absolutely true. Steam has stated that surveys often have 60% to 70% participation out of all those who receive them. That is a significant portion of Steam users and MORE than enough to gain an accurate view of user stats. This grants them credibility which is why media outlets, such as this one we're discussing on, have the ability to write articles and be able to confidently cite the source. Context is important. So to is merit.

What credibility do YOU have? Hmm?

What proportion of Steam users is that? You have misinterpreted something here.

Steam claims that 60 to 70% of people who are sent the survey participate, which is fine. But your comment suggests that 60-70% of total Steam users are represented in the survey, which is not the case. Based on how often I have received the survey each month on all the machines I have Steam on, that number is far less than 10%, maybe 1%, which is an insignificant portion of Steam users, making their survey not particularly accurate.

The broad strokes are probably true but most of the data is likely to be only accurate within 10-20% or so, especially at the lower ends of participation, like the video cards with ~1% representation. This becomes obvious as about once a year, the monthly data notably departs from the otherwise gradual yearlong trends, which happened this month.

So this month's data is even more crap than usual. Try again next month.
 

BSim500

Posts: 839   +1,889
Steam surveys are generally reliable. What's happened is Chinese Internet cafe's which were (and still are) using W7, stopped being counted when they uninstalled Steam to be used for mining. Now the interest in mining in China has dropped / there's a legal crackdown on it inside China, they've reinstalled Steam again and have switched back to gaming. Steam didn't "exclude" anyone nor discriminates cafe's vs private use, it's simply 24/7 miners that stop using Steam, stop being counted by Steam stats...

As for AMD, if you stop selling any CPU under £270, then you're obviously going to lose a chunk of the budget market. At one point the 5600X was triple the price of what the 1600/1600AF/2600 chips were (and is still double what the 3600 was previously). Same reason why if AMD & nVidia refuse to sell GPU's under £300, then you're going to see budget gamers hold onto 1060's, etc, for longer (which the Steam stats also show). I can't understand why people here are struggling with these two observations and seem intent on inventing bizarre conspiracies.
 

defaultname365

Posts: 34   +23
GTX 1060s still the king of GPUs. No wonder Nvidia keeps pushing for xx60 releases. The 3060 has 12Gb VRAM LMAO.... still can't get over it.
 

duckofdeath

Posts: 448   +586
Another statement without merit. Technology today is much, MUCH less expensive than it once was. Sure, prices go up and right now we have an artificially inflated market(thanks to the pandemic and cryptominers), however, as a general rule, prices per capita are far less expensive than they once were. Example, in the early 1980's 1MB(that's megabyte as in 1024KB) of RAM for a given computer was around $440(I know, I was there and frequently bought RAM). For context, the average median household monthly income was $370. The late 80's and early 90's weren't much better. In the late 90's prices started coming down to the point where people could buy a PC on 2 or 3 months wages. Nowadays, you can buy a COMPELETE, very well equipped PC for less than what most people make in a month. That's a WHOLE PC, not just RAM.

Yes, context is very important.
Context is indeed exceptionally important. You deciding to selectively pick one wind back of moore's law and fix the other is outright laughable.
 

captaincranky

Posts: 17,217   +5,972
Context is indeed exceptionally important. You deciding to selectively pick one wind back of moore's law and fix the other is outright laughable.
When, if ever, are you going to get it through your head that, "Moore's Law", isn't a law, but conjecture?

Now the "inverse square law", is an actual physical law. If you can manage figure out how it works, you should begin to understand why, "Moore's Law", can't hold up forever.
 

captaincranky

Posts: 17,217   +5,972
. I can't understand why people here are struggling with these two observations and seem intent on inventing bizarre conspiracies.
I like to imagine myself as being, "ambidextrous". Of which, I'm euphemistically referring to as, "the ability to talk out either side of my a**, with equal skill". I don't "brag" about it very often, since I'm certain others will try to belittle me by claiming, "oh that's nothing, I can do that standing on my head". :rolleyes:
 
Last edited:

ZedRM

Posts: 616   +387
What proportion of Steam users is that? You have misinterpreted something here.

Steam claims that 60 to 70% of people who are sent the survey participate, which is fine. But your comment suggests that 60-70% of total Steam users are represented in the survey, which is not the case. Based on how often I have received the survey each month on all the machines I have Steam on, that number is far less than 10%, maybe 1%, which is an insignificant portion of Steam users, making their survey not particularly accurate.

The broad strokes are probably true but most of the data is likely to be only accurate within 10-20% or so, especially at the lower ends of participation, like the video cards with ~1% representation. This becomes obvious as about once a year, the monthly data notably departs from the otherwise gradual yearlong trends, which happened this month.

So this month's data is even more crap than usual. Try again next month.
Your statement is lacking basic logic. Even if only 1% participate that is still millions of systems being polled. That is hardly "insignificant" and would represent a reasonably valid sample size in ANY scientific survey. However, your assumption of numbers is as flawed as your interpretation of my statement.
 

Lew Zealand

Posts: 1,922   +2,197
TechSpot Elite
Your statement is lacking basic logic. Even if only 1% participate that is still millions of systems being polled. That is hardly "insignificant" and would represent a reasonably valid sample size in ANY scientific survey. However, your assumption of numbers is as flawed as your interpretation of my statement.
Garbage in, garbage out. Sample size doesn't matter if you don't sample the same pool of users every time.

As I said, the numbers are only generally valid, the broad strokes. However the specifics, which is what this article is all about, are completely flawed as Steam's numbers vary unusually (CPU trend), even wildly (OS trend) based on reasons that are never disclosed by Steam.

For instance: that there are 3 AMD CPUs to every 7 Intel CPUs in use is broadly accurate. That there is a trend for more people to buy/use Intel CPUs in the past month is not accurate, just like it wasn't accurate that more people bought Intel CPUs in Nov 2020 or early 2020. That even crappier than usual Steam data correct itself the following month and this will, too.
 

HardReset

Posts: 1,240   +898
Your statement is lacking basic logic. Even if only 1% participate that is still millions of systems being polled. That is hardly "insignificant" and would represent a reasonably valid sample size in ANY scientific survey. However, your assumption of numbers is as flawed as your interpretation of my statement.
Scientific survey :joy:

Now let's see https://store.steampowered.com/hwsurvey/videocard/

Intel(R) UHD Graphics
AMD Radeon(TM) Graphics
Intel Haswell
Intel Sandy Bridge

- There are gazillion Intel UHD graphics alternatives.
- Radeon graphics means what? Radeon GPU something?
- Haswell is CPU architecture, not video card.
- Sandy Bridge is CPU architecture, not video card.

More than enough proof there that Steam survey is utterly crap.
 

ZedRM

Posts: 616   +387
Last edited:

Danny101

Posts: 1,837   +790
There is often a stark contrast between feeling ripped off than actually being ripped off.

In today's world, feelings are at a high inflation rate. The return on the investment is rather low.
 

duckofdeath

Posts: 448   +586
There is often a stark contrast between feeling ripped off than actually being ripped off.

In today's world, feelings are at a high inflation rate. The return on the investment is rather low.
There is an undeniable linear relation between generation numbering by manufacturers. These generation numbers have been a solid price reference for decades.

I don't know if this comment section just decided to go full retard or if it's just liars defending shady corporations deliberately inflating prices in their reprehensible cartels. Either way, it's pathetic to see a gadget site going along with bullshit like this for the page clicks.
 

ZedRM

Posts: 616   +387
Me: Who cares? It's not like Intel CPUs are good, they're just AVAILABLE.
Oh that is a silly statement. Of course Intel's CPU's are good. Just because AMD has a line of CPU's that's better doesn't suddenly make Intel's offerings crap. CONTEXT is important!
 

hwertz

Posts: 65   +28
Welp, I just got my AMD system a week or two back (Ryzen 5 3450U) and it's pretty sweet. I lucked out, got a "returned and checked out" system that for reasons unknown to me, the person who returned it yanked the 12GB of RAM and put 32GB in it. Sweet!

I could totally see AMD's steam share dropping. First, focusing on high-end chips, I wouldn't spend the $$$$ on a Threadripper to play games, probably 20 out of 24 cores would be idle (obviously depends on the game but still.) Also, with the price of NVidia cards going back to reality, some who would have bought a Ryzen (for the pretty good integrated graphics) will probably go back to Intel + NVidia card.

Windows 7? I have no idea... (I find BSim500's explanation by far the most likely, Chinese gaming cafes went back to gaming instead of cryptomining.) I can say Steam proton (and plain wine for stuff outside steam) have very close to 100% compatibility these days, with the bonus (if you're into this style of retrogaming) that the 90's-era games that typically won't run in Win7+ will run in Wine (my friend likes to play Tiger Woods '99 for instance.)
 

Avro Arrow

Posts: 1,727   +2,047
TechSpot Elite
Oh that is a silly statement. Of course Intel's CPU's are good. Just because AMD has a line of CPU's that's better doesn't suddenly make Intel's offerings crap. CONTEXT is important!
Your level of reading comprehension isn't strong enough to realise that the word "good" is being used in a relative manner in this case? Well, it looks like everyone else understood my words just fine.

BTW, ALL Intel offerings are crap because Intel is crap. Now try using it in the relative manner in which it must be meant because obviously I'm not comparing Intel to Bosch, am I?
 
Last edited:

Gezzer

Posts: 153   +79
Because 10th gen Intel were sold at large discount

Yup, pull the trigger on an upgrade and decided to grab a i5 10600KF even though I'd been waiting to grab a 5600X for ages. What made up my mind for me? AMD $370CN, Intel $255CN for pretty much the same thing or debatably even better. I want AMD to do well, competition is good for everyone, even bought a first gen Ryzen system when they came out (sold it to a work buddy). But I just can't justify paying a $115CN premium for the privilege of supporting AMD.
 

Avro Arrow

Posts: 1,727   +2,047
TechSpot Elite
Yup, pull the trigger on an upgrade and decided to grab a i5 10600KF even though I'd been waiting to grab a 5600X for ages. What made up my mind for me? AMD $370CN, Intel $255CN for pretty much the same thing or debatably even better. I want AMD to do well, competition is good for everyone, even bought a first gen Ryzen system when they came out (sold it to a work buddy). But I just can't justify paying a $115CN premium for the privilege of supporting AMD.
Nor should you. AMD's pricing has been stupid lately because of the demand and lack of availability. There's no point in paying a crap-tonne more because of the brand name if the performance of the less-expensive option is good enough. That's one of the reasons that the last Intel CPU I owned was a Core2Duo. The value preposition of Phenom II and FX-8300 was just too good. Having a good hate-on for Intel helped too. LOL
 

Danny101

Posts: 1,837   +790
Nor should you. AMD's pricing has been stupid lately because of the demand and lack of availability. There's no point in paying a crap-tonne more because of the brand name if the performance of the less-expensive option is good enough. That's one of the reasons that the last Intel CPU I owned was a Core2Duo. The value preposition of Phenom II and FX-8300 was just too good. Having a good hate-on for Intel helped too. LOL
That's why I say that the motherboard you choose is immensely the more important choice in the build. It sets the tone of the your upgrade path. I use to cheap out on this part, but no longer. Got to pay attention to the VRM here or you will find yourself with a dead board. My last upgrade included a quality board and an entry level CPU. It's also why I prefer AMD. More upgrade options. Intel is to Apple like AMD is to Microsoft.