Steam players are reverting back to GTX 1660 graphics cards, despite Nvidia's best efforts

NightReaver said:
Honestly doesn't mean much because things get knocked around on that list all the time. If we look at just number of reviews or thread activity it's really not super hot. Overall sales for a certain period of time would be more telling.
Click to expand...
I could imagine any mid to big budget new game of the week to always score really high that week true.

But anyway, the premise is just strange as it start by looking at the steam hardware survey, see the top 20 most popular skus are cards that were sold above $200
 
Brackle said:
No one, because its hard to know what people are playing with at home. If you got 200-300 people logging into steam at a single Internet Cafe on 1 day. Times that by 365 days, then you can easily see how the charts are skewed.

So, lets say 250 people login to steam at 1 single internet cafe and multiply that by 365 days (just an example). That would be 91,350 for 1 single internet cafe. Now lets say there are a 1000 internet cafe's in the world. That would be over 91,250,000 people using a 1660.....When they aren't really that many.

See how the chart can be easily skewed?
Click to expand...
You trying to say Steam doesn’t know it’s the same computer for all those users at gaming cafes?——because it absolutely does know.
 
Brackle said:
Do you have any evidence to back up your claim?
Click to expand...
Steam takes hardware snapshots on sign in, how else do you think they know that you are signing in from a new device blah blah blah that prompts their 2 factor?
Those hardware ID's are unique, unless the cafe is doing something to randomize those on every sign in then they would be the same every time, and any basic report filter would handle that.
 
chameleoneel said:
Haha. Do you have any to back that it doesn’t?

Steam takes a lot of info about your computer, with the standard user agreement. They know.
Click to expand...
I dont, and neither do you. So again, it is possible that it counts the same PC over and over.

I never once stated I knew that it does. I just said it was possible.
 
Lakados said:
Steam takes hardware snapshots on sign in, how else do you think they know that you are signing in from a new device blah blah blah that prompts their 2 factor?
Those hardware ID's are unique, unless the cafe is doing something to randomize those on every sign in then they would be the same every time, and any basic report filter would handle that.
Click to expand...
and why people that go game on an hardware coffee (and I imagine pay by hours-minutes) would accept when asked to go through the hardware survey to start with ? I would think they are more likely than the average user to say no thank you. Or why would those users get randomly selected more often than others ?

I am not sure if I get the premise at all, it seem to be based on the wrong assumption that steam use every account machine info in a systematic way and not a small subset of random users that accept to fill it when prompted.

Lakados said:
Steam takes hardware snapshots on sign in, how else do you think they know that you are signing in from a new device blah blah blah that prompts their 2 factor?
Click to expand...
Survey do send an hash of the pc mac addresses I think.
 
It also seems we're assuming they only count one PC per account, but I don't see why that would have to be the case, either. Would make more sense to me if they took one data point as one data point, segregate/aggregate them based on the year/date it was logged, since it is very likely some users game on more than one PC, and there is no reason to throw out one or the other(s).
 
LukeTbk said:
and why people that go game on an hardware coffee (and I imagine pay by hours-minutes) would accept when asked to go through the hardware survey to start with ? I would think they are more likely than the average user to say no thank you. Or why would those users get randomly selected more often than others ?

I am not sure if I get the premise at all, it seem to be based on the wrong assumption that steam use every account machine info in a systematic way and not a small subset of random users that accept to fill it when prompted.


Survey do send an hash of the pc mac addresses I think.
Click to expand...
So on your account you have your authorized devices and when you sign into steam it checks the authorized list and if the device isn’t there it adds it after you do all your 2 factor stuff.

Valve collects hardware hashes regardless of any opt-ins.

The survey opt in would just include that data in the survey and gives Valve permission to scan installed software and such.

So people who use Internet cafes would probably have a pretty long list of authorized computers and numerous users would have matching hardware hashes which valve would know to filter out.
Not any different from households where one PC might have multiple steam accounts.
 
Not to mention that this imaginary Internet cafe is going to have basically the same customer base logging in day after day, week after week. It’s not a whole new clientele every week.
 
cjcox said:
At 1080p, do we care?
Click to expand...
Could be a nice plus, a 1660-1660 Super users will probably use FSR upscaling in the Last of us type of title at 1080p, even 2060 I can see use DLSS quality.

Lakados said:
The survey opt in would just include that data in the survey and gives Valve permission to scan installed software and such.
Click to expand...

Which is what we were talking about, I thought.
 
To add what Lakados said in post #51.

The way it should work (I've done some analytics work) is the client app will collect basic telemetry data (probably OS, CPU, GPU) or hash the data to provide a generic metric that doesn't count as telemetry data. Though the client agreement probably allows for basic telemetry data under the guise "we need to check you're not a bot or cheating".

From that overall internal(non public) metric, they'll determine who to send a survey to get more detail info (versions, drivers, monitors, etc.) until they get enough responses that provides statistical accuracy for every category they're interested in. While the public (surveyed data) maybe a small subset of internal metrics, it probably matches percentage wise on the base categories they have internally (OS, CPU, GPU, etc).
 
Last edited:
Eymar said:
To add what Lakados said in post #51.

The way it should work (I've done some analytics work) is the client app will collect basic telemetry data (probably OS, CPU, GPU) or hash the data to provide a generic metric that doesn't count as telemetry data. Though the client agreement probably allows for basic telemetry data under the guise "we need to check you're not a bot or cheating".

From that overall internal(non public) metric, they'll determine who to send a survey to get more detail info (versions, drivers, monitors, etc.) until they get enough responses that provides statistical accuracy for every category they're interested in. While the public (surveyed data) maybe a small subset of internal metrics, it probably matches percentage wise on the base categories they have internally (OS, CPU, GPU, etc).
Click to expand...
It also collects IP addresses, for the sake of location login identification, and maybe system NetBIOS name, looking at my authorized devices I don't know if it pulled that automatically or if I was that anal about naming the entry when it asked, if it asked, I can't remember it has been a hot minute since I did a reinstall.
 
Lakados said:
It also collects IP addresses, for the sake of location login identification, and maybe system NetBIOS name, looking at my authorized devices I don't know if it pulled that automatically or if I was that anal about naming the entry when it asked, if it asked, I can't remember it has been a hot minute since I did a reinstall.
Click to expand...
It used to ask you to name your device when you tell it to remember/trust your device. Dunno if it still does.
 
Nobu said:
It used to ask you to name your device when you tell it to remember/trust your device. Dunno if it still does.
Click to expand...
That would track, pre medicated me was sorta anal about naming conventions and such.
 
kac77 said:
Steam players are reverting back to GTX 1660 graphics cards, despite Nvidia's best efforts

The Nvidia GTX 1660 released back in 2019, but Valve’s latest Steam Hardware Survey figures suggest people are still buying the graphics card. Rather than necessarily jumping from lower spec favorites like the GTX 1650 to the new RTX 4060, players appear to be sticking with the GTX line of GPUs for the most part.
Click to expand...
despite Nvidia's efforts, to do what? Sell video cards?

The 1660 is still being marketed, and it fills in the price points below the 4060 cards. And people are still buying them apparently. 1xxx was a great generation of GPU.

1691105439579.png
 
GoodBoy said:
despite Nvidia's efforts, to do what? Sell video cards?

The 1660 is still being marketed, and it fills in the price points below the 4060 cards. And people are still buying them apparently. 1xxx was a great generation of GPU.

View attachment 587887
Click to expand...

If people aren't gonna buy Nvidia cards, they're just gonna turn around and buy another Nvidia card

That's the truth of it
 
This is the console scarcity effect. People are realizing that they dont need a fancy card, just a card.

Im still gaming with a gt950 4gb. Sure it’s at low settings on a premium title, but most of what I play runs fine at medium or high settings.

The market discovered this as well. If a lower end viable card existed, they’d buy it instead.
 
mustang_steve said:
This is the console scarcity effect. People are realizing that they dont need a fancy card, just a card.

Im still gaming with a gt950 4gb. Sure it’s at low settings on a premium title, but most of what I play runs fine at medium or high settings.

The market discovered this as well. If a lower end viable card existed, they’d buy it instead.
Click to expand...
I have friends who until this morning around 8:30 a.m. were happy with their GTX 970s and 1050s, and now they are very bummed.

They are likely to order up some RX6600s and be amazed, and I am more than a little envious they can still find joy in these things.
 
Lakados said:
I have friends who until this morning around 8:30 a.m. were happy with their GTX 970s and 1050s, and now they are very bummed.

They are likely to order up some RX6600s and be amazed, and I am more than a little envious they can still find joy in these things.
Click to expand...
They should get a free $60 game with the RX6600, that is kind of the point, the rising cost of games matters more than what low end budget video cards people around the world like to use.
 
crazycrave said:
the rising cost of games
Click to expand...
? Games cost was quite more when I was young, some of the cartridge then box in a brick and mortar store to bits to download rebate was passed down to the consumer (and I imagine volumes of sales), you can build a collection from free epyc-gog-steam games over time now and the price of the top new games has not rises since forever.

game-prices.001.jpg
sam_naji_pricing_2.jpg
 
Last edited:
It's hard to compare with something made that every game released was only made for one platform like Atari 2600.
 
Last edited:
My overclocked 2070S is running fine, it actually pumps the frames quite well. Realistically speaking, the main reason you'd upgrade (vram aside) is to gain access to newer technologies that Nvidia blacklist on holder hardware that's more than capable of running such technologies - Oh how I'd love to be able to take advantage of resizable BAR and DLSS3.
 
Cheap GPUs sell very well. My first GPU when I was a kid was a Geforce 2 MX400. It was $70 brand new. Is there a brand new GPU currently on the market that can get even remotely close to that price?
 
madpistol said:
Is there a brand new GPU currently on the market that can get even remotely close to that price?
Click to expand...
Sometime RX580 are close to that price ($120 2023 dollars):
https://www.bestbuy.com/site/xfx-am...=198&refdomain=pcpartpicker.com&skuId=6136515

6400 not that far.

Back in those days I would imagine it was around the performance of the best top end GPU of the previous generation and having any 3d card made the pc much better than the average one still running without one, now you need to significantly beat a 770 HD/what come in a 5700G.
 
I think the 1660 also offers one of the newer versions of NVENC encoding for the lowest price. Good for a dedicated stream PC at a low cost, until the new intel gpus came out with better options
 
I had a RX 5500 XT 8Gb once, with AM5 i would love to see what it could do now, I believe the card supports SAM and PCIe 4 and it cost $189 new back then for the small single fan model by VisionTek before Bit Coin at the Dell store, never had anything faster than a Ryzen 3 3100 back then, Ryzen 5 3600 was no faster at pushing the video card than 4 / 8 core 3100 http://www.3dmark.com/fs/23525080

I had just bought my B550 board and was getting it bios flashed back then.
 
Last edited:
July stats. With the 1050, 1050ti, and 1650 making up over 10%, I'm thinking low profile is not so niche after all. Also, what the heck is a 1650ti? Chinese only product?
Screenshot_20230807_224909_Samsung Internet.jpg
 
Last edited:
I know I would have never got a 1050ti if I didn't need LP. Even ignoring AMD, the 3 GB 1060 was a better buy at about the same price and likely would have flexed just a bit more foe the 1650 Super, which is considerably faster.

Now that the 4060 has an LP offering, it will be interesting to see if it climbs up the charts.
 
OKC Yeakey Trentadue said:
July stats. With the 1050, 1050ti, and 1650 making up over 10%, I'm thinking low profile is not so niche after all. Also, what the heck is a 1650ti? Chinese only product?
View attachment 588838
Click to expand...
1650ti is a Mobile chip, lots of budget “gaming” laptops came with that guy in 2020, mostly Acer and the like, it’s about half the performance of a 1660 Super.
 
You must log in or register to reply here.
Back
Top