Steam numbers for new GPUs.

[Arachnotronic]

So in other words, its a niche and irrelevant as such to the numbers.

Some perspective: 230,000 GPUs being used for mining, assume an average selling price per card of $300. That's $69 million worth of cards. That seems huge until you realize that NVIDIA's gaming GPU biz crossed over $1 billion last year.

In other words, you're probably right: it's a niche relative to overall discrete GPU sales.

 

[ShintaiDK]

Some perspective: 230,000 GPUs being used for mining, assume an average selling price per card of $300. That's $69 million worth of cards. That seems huge until you realize that NVIDIA's gaming GPU biz crossed over $1 billion last year.

In other words, you're probably right: its a niche.

Yes, not to mention the timeframe the cards have been acquired (~3 years). And how many of those cards serve a double role with gaming as well.

 

[Madpacket]

So in other words, its a niche and irrelevant as such to the numbers.

Also the hashrate went DOWN after the RX480 release.
https://etherscan.io/charts/hashrate

Hashrate went down (and goes up) depending on the price and difficulity of ETH/ETC. This is normal when you have competing coins.

More info:

https://slacknation.github.io/medium/13/13.html

230,000 dedicated GPU's is far from niche. GPU's have only recently (last 8 - 10 months) become a viable means for mining and making a profit. ASIC's took over BTC/LTC years ago due to the simplicity of the algorithm (SHA256) making GPU's irrelevant. This is currently not feasible (building ASICS) with the algorithms used for newer coins, hence the resurgence of GPU mining.

Anyways I think its become apparent to me this is a troll thread so I'll stop feeding it. Believe in the numbers that support your arguments (Conformation Bias). I'll wait for the real numbers of actual sales to be disclosed next quarter.

 

[Rvenger]

I am inclined to believe this is a troll thread with the intent to make other members start an Nvidia vs. AMD war. If I see this thread going down this path it will be locked and if I see anymore mining discussion it will be considered off topic and moderator action will take place.

Have a civil discussion and if I see any bad mouth over Nvidia vs. AMD or any trolling for either brand you will be thrown out of here for 30 days.


-Rvenger

 

[master_shake_]

If mining was so great, you think that people would be dumping older cards that gamers would buy. Yet no such thing is shown on steam.

How much of RX480 sales really goes to miners on a global scale? 1%?

is anyone really foolish enough to buy a card that was used at full throttle for months on end?

to me that means the card is on it's last legs and buying it would be a silly thing to do.

 

[Arachnotronic]

is anyone really foolish enough to buy a card that was used at full throttle for months on end?

to me that means the card is on it's last legs and buying it would be a silly thing to do.

When one buys a used card, how does one know what it had been used for previously? 😉

 

[Jaskalas]

And who are early adopters likely to be?
This is all the first month's numbers tell us. You can be certain the more expensive cards will not sell the most over time.

 

[Arachnotronic]

And who are early adopters likely to be?
This is all the first month's numbers tell us. You can be certain the more expensive cards will not sell the most over time.

To be fair, GTX 970 was a monster, doing very well over its lifetime. I think GTX 1070 and GTX 1060 on the NVIDIA side will do quite well. 1080 should do better than either 980 or 980 Ti since it will be vying for both sets of customers.

 

[Kenmitch]

When I've used a AMD cards I've never been asked to participate in the survay. Been prompted twice with Nvidia and declined both times. Not really sure why it works out that way in the end. I have zero faith in Steam survey actually portraying the real user base in the end.

 

[AtenRa]

From all those years im using Steam (5+), only once took the survey and at the time i was using an APU for benchmarking.

 

[sm625]

I'm starting to consider steam hardware survey as fully discredited and meaningless. No 290s? No 285, 280, 280x, etc. Sorry but that's just not even possible.

 

[Lepton87]

is anyone really foolish enough to buy a card that was used at full throttle for months on end?

to me that means the card is on it's last legs and buying it would be a silly thing to do.

last legs? Don't be absurd, first of all it should have warranty and it should be good for 5 years of continuous operation so having been used for months doesn't matter/

 

[boozzer]

From all those years im using Steam (5+), only once took the survey and at the time i was using an APU for benchmarking.

steam survey actually works really well, I have no doubt. sample size is big enough to represent all of pc gamers. 6 months to 1 years after, not within 1 month of hardware releases with supply problems on both sides.

this is jumping the gun even before the gun powder was invented.

 

[Arachnotronic]

I'm starting to consider steam hardware survey as fully discredited and meaningless. No 290s? No 285, 280, 280x, etc. Sorry but that's just not even possible.

I see the R9 200 series in the survey...

"

AMD Radeon R9 200 Series
1.69%
1.70%
1.74%
1.33%
0.68%
-0.65%"

Looks like a lot of R9 200 owners are jumping ship. Wonder what they're buying...

 

[jpiniero]

Steam I think split out the 300 series from the 200 series.

 

[USER8000]

The survey is biased towards people who want to answer it ,which is most likely to be hardware enthusiasts,plus it does not ask everybody consistently.

Even with me,I have had quite a few newer GPUs in the last 5 years,and Steam still thinks I have a GTX660TI.

It needs to be compulsory for it to have a better distribution of what gamers use.

Edit to post!!

Found thread now.

 

[AtenRa]

steam survey actually works really well, I have no doubt. sample size is big enough to represent all of pc gamers. 6 months to 1 years after, not within 1 month of hardware releases with supply problems on both sides.

this is jumping the gun even before the gun powder was invented.

steam doesn't disclose how many took the survey, we dont have any numbers about the sample size.

 

[Gikaseixas]

I can't believe some here think that Steam survey is accurate when in reality not every user takes the survey. It is optional and even if you install a new GPU there are no guarantees that you'll be asked to participate on it.

Troll thread indeed

 

[frozentundra123456]

last legs? Don't be absurd, first of all it should have warranty and it should be good for 5 years of continuous operation so having been used for months doesn't matter/

Personally, I would not buy *any* used dgpu unless it was from someone I knew and trusted. I just have not had much luck with longevity as far as dgpus have been concerned, even when I have bought them new, so would definitely hesitate to buy a used one.

Edit: I do not really buy top end dgpus. My limit is about 150 to 200 ish, so I dont consider that a major purchase, and the amount that I can same on a used one is just not worth the uncertainty.

 

[Arachnotronic]

I can't believe some here think that Steam survey is accurate when in reality not every user takes the survey. It is optional and even if you install a new GPU there are no guarantees that you'll be asked to participate on it.

Troll thread indeed

Steam survey seems to correlate well with the financial results of NVIDIA and AMD in their respective gaming GPU businesses. Calling this a "troll thread" isn't warranted, IMO.

 

[Gikaseixas]

Steam survey seems to correlate well with the financial results of NVIDIA and AMD in their respective gaming GPU businesses. Calling this a "troll thread" isn't warranted, IMO.

Interesting... a flood of accountants appeared in the discussion and they are using Steam as a basis for their baseless claims. About the troll thread, if it walks like a... you know how it goes

 

[USER8000]

I finally found this thread from a few years ago,when somebody looked at the HD6000 series and the Steam Hardware survey:

https://socialtechwork.com/thread...ke-for-what-people-try-to-use-it-for.2134147/