News GTX 1650 Still Most Popular GPU According to Newest Steam Survey

[Admin]

The latest Steam hardware survey shows minor changes in hardware usage compared to May 2023.

GTX 1650 Still Most Popular GPU According to Newest Steam Survey : Read more

 

[LolaGT]

Not any shock. The average gamer doesn't spend $400-500 on a discrete card.
Let alone the ones significantly more than that, those are for people with way more money than sense.

We are still in a 1080p world for the majority.

 

[Amdlova]

I have one here... but get a 1660 super. I have waited two year to Replace the 1650 but with these new cards can't do. 1650
On rdr2 got 42fps full hd with the 1660 super finally will get the 60 fps. Great card for the price.

 

[zecoeco]

A superb and extremely power-efficient GPU that doesn't require extra power, additional ports , or newer PCIe Gen4. Gets the job done for 99% of people who just want to "game" and have "fun".
As powerful as an Xbox Series-S, and sold for no more than $100 (Non-Super Variant).
At 720p its very impressive, and is great at 1080p resolution. Speaking from experience as I have owned the card for about 1 year.

 

[PlaneInTheSky]

People are too busy paying their grocery bills to care about stupid stuff like a raytraced puddle of water, or the latest "AAA" game full of microtransactions.

A 1650 is good enough to play indie games and 99% of games that came out before 2020.

 

[Math Geek]

got a 1650 super that has no problems at 1080p.

so long as you don't suffer from FOMO, it's more than ample for enjoying most any game.

 

[Alvar "Miles" Udell]

No shock as GPU prices are at the levels they are, even for entry level GPUs, and the up to $380 RTX 4060 isn't going to make a dent in it.

The GPU market is highly competitive

No it's not and I disagree most strongly.

The 7900XTX and RTX 4080, for example, are essentially equal in performance according to TH, but while the 4080 is $1150-$1200, the 7900XTX is $1000-$1100 (current prices on Newegg). Even if you don't weight vendor specific features (like DLSS) all and just go on the numbers, $1175 vs $1050 isn't very "competitive", they're both still incredibly expensive.

Then at the more mainstream/entry level gaming with the 4060 vs the current AMD competitor, the 6700XT, the 6700XT is $340-$400 and the 4060 is $300-$390. There is absolutely ZERO competition, the prices are identical.

And this is not unexpected considering nVidia (especially) and AMD are making money hand over fist in other market segments so there is zero reason for them to actually COMPETE in consumer graphics. When AMD released the 4970, they priced it half of what nVidia priced their card and it shocked the market, they cut prices quickly across the board as AMD released their lower tiers. Same thing happened with the 5970. When AMD released the RX 480 it was priced extremely competitively, and the rebrand RX 580 did as well.

No, until Intel gets their GPUs in order there effectively is no competition between AMD and nVidia in the consumer space (and in the professional space vendor specific features and software performance mean more). No matter if you go Team Red or Team Green, you're going to pay about the same price for the same level of performance, especially in the more entry level and mainstream tiers.

 

[hotaru.hino]

It's important to note that this data should only really be taken as that: data.

Because if you look at the history (since Valve also provides the last four months), the RTX 3060 was in 10.67% of computers that were surveyed in March. It dropped down to 4.66% in April, though the GTX 1650 went up about 2 percentage points. So unless a significant amount of people sold their 3060s and got a GTX 1650, this doesn't really say much about actual ownership of video cards across Steam's landscape.

I mean, I haven't submitted my hardware profile to Steam in several months. Either because I wasn't asked or I just wasn't logging in as much.

 

[IamNotChatGpt]

Can people stop talking about Steam surveys? (and the damn submarine<-nobody cares, the news cycle is REALLY dead dry for people to still talk about that garbage init?).
These surveys don't mean anything.

It doesn't survey properly. Especially if you are on linux. If everyone would take the survey, the numbers would look very different, They make it seem like that there are only a handful number of 2000, 3000 and 4000 owners out there....

 

[baboma]

>the RTX 3060 was in 10.67% of computers that were surveyed in March. It dropped down to 4.66% in April, though the GTX 1650 went up about 2 percentage points. So unless a significant amount of people sold their 3060s and got a GTX 1650, this doesn't really say much about actual ownership of video cards across Steam's landscape.

I agree in part. The data collection is too uneven to definitively or even tentatively state whether X is more popular than Y at any given time. But IMO it's good enough to divine broad findings, such as Nvidia GPUs have a dominant share over the other brands.

Likewise, you can divine broad trends over time, ie with successive monthly snapshots, such as whether a new generation is making inroads, and how quickly.

>These surveys don't mean anything.

It's not perfect, but it's the only viable indicator of GPU ownership available to the public. Amazon's best selling GPUs is another indicator, but it only cover new sales, not what's being used (ownership), and it has no quantitative data, making it much less useful. But it is a good snapshot of the hot sellers.

Amazon Best Sellers: Best Computer Graphics Cards

Discover the best Computer Graphics Cards in Best Sellers. Find the top 100 most popular items in Amazon Computers & Accessories Best Sellers.

amazon.com

Right now, the top two best-sellers are 3060 12GB, followed by a 3060 8GB at over $300. Despite the sturm und drang over the 4060 from the hardcore crowd, it's reasonable to expect the 4060 slotting into the 3060's spots in the coming months. This is how to tell whether the sentiments expressed on these forums are representative of the overall public. Read: They aren't.

Of course, there are various sales/use trackers available with subscription. If you really want to know, you can. It just costs money.

 

[hotaru.hino]

These surveys don't mean anything.

It doesn't survey properly. Especially if you are on linux. If everyone would take the survey, the numbers would look very different, They make it seem like that there are only a handful number of 2000, 3000 and 4000 owners out there....

I think the surveys are meaningful, it's just people need to understand how to interpret the data.

 

[in_the_loop]

Not any shock. The average gamer doesn't spend $400-500 on a discrete card.
Let alone the ones significantly more than that, those are for people with way more money than sense.

We are still in a 1080p world for the majority.

"The Average gamer" isn't using a GTX 1650 though. Just because it is the top card among very many other cards doen't mean that it is the "median" card that a gamer uses.
Somebody did a calculation for the "better cards" where he included all the Nvidia cards from the 20,30 and 40 series that were at least a xx60 card, ie that excluded cards like the 3050, 3050t etc. He also included the still fast 1080ti. But not the GTX-1650
In other words mostly raytracing cards (besides the 1080ti) that are a lot faster than a GTX 1650
The amount was 35.08%!
And that was for only the Nvidia cards. If we would have included both AMD and Intel the number would probably have come closer to 50%, aka the avereage user with cards at least being a xx60 card or better.

So the market is not limited by the average user having a crap card.
Only including Nvidia, the 35% "Good cards" out of the 132 million user on Steam in 2021 means a potential 46 million gamers with these cards. (Could be higher now since the 132 million users was 2 years ago)
Add AMD to that. Also that games exist outside of Steam and the market for producing games that use the latest features is simply big.
In addition to this, the gamers with the better cards are probably also more avid gamers buyng many more games.
For the developers, the number of games sold is more important than the number of users of a specific card.
So the average should be the average card for all the games sold, which would mean an even higher number since the gamers with better cards probably buys much more games.

 

[PlaneInTheSky]

"The Average gamer" isn't using a GTX 1650 though. Just because it is the top card among very many other cards doen't mean that it is the "median" card that a gamer uses.

It is though.

The average gamer has a 1650.

In addition to this, the gamers with the better cards are probably also more avid gamers buyng many more games.

Or they're playing Fortnite/CoD 24/7, and bought a fancy card to get "200fps". These players are often just playing 1 game.

 

[in_the_loop]

It is though.

The average gamer has a 1650.




Or they're playing Fortnite/CoD 24/7, and bought a fancy card to get "200fps". These players are often just playing 1 game.

No, the statistics only show that the GTX 1650 is the most used card.
It says nothing about averages. As I showed a large amount of people use much stronger cards and the average card, in performance would probably land somewhere with a 3060-card or so, even though that specific card isn't the most sold.
We are talking about averages here, not the most used card at a specific moment.