Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Nvidia’s stronger, faster Pascal architecture expected to debut Friday // GTX 1080 Benchmarked

13»

Comments

  • MalaboogaMalabooga Member UncommonPosts: 2,977
    edited May 2016
    SEANMCAD said:
    Malabooga said:
    Agreed, 950 and 370 (and previous gen equivalents) as entry level gaming cards. But then, when you see how many players LoL and CS:GO and such have and they only really need an APU....what is really "gaming". AAA games only? Latest games only?
    to be clear I am not really saying 'entry level'

    so in video cards there is a 'sweet spot' its generally the price point in which you are getting the best performance for your dollar. It also often represents the highest volume of sales AMONG the 'gaming card sector' are you familar with this general 'ruleism'?
    to this is what I am refering

    its usually in the middle between the most extreeme in high performance and the 'entry level for gaming'
    Not to continue in other topic:

    There are many "gaming standards":

    720p playable/720p 60FPS/720p max detail 60 FPS
    1080 playable/..../....
    1440p.... ....
    4k.... ....

    970/390 would be entry level to high end - 1080p GPUs/capable 1440p GPUs (though 970 3,5GB memory starts to be an issue here so generally not recommended for 1440p)

    "Sweet spot" would be 100-200$ GPU segment because those GPUs offer decent performance (and often best perf/$) and perf/$ ratio tends to fall off as you go up in price (all the more quickly the more you go up in price and often when you go down in price)

    But also there is no "ultimate" sweet spot, if you have 1440p monitor your "sweet spot" isnt the same as someone with 1080p monitor. But vast majority still has 720/1080p so that "sweet spot" could be considered "majority sweet spot"
  • SEANMCADSEANMCAD Member EpicPosts: 16,775
    Malabooga said:
    SEANMCAD said:
    Malabooga said:
    Agreed, 950 and 370 (and previous gen equivalents) as entry level gaming cards. But then, when you see how many players LoL and CS:GO and such have and they only really need an APU....what is really "gaming". AAA games only? Latest games only?
    to be clear I am not really saying 'entry level'

    so in video cards there is a 'sweet spot' its generally the price point in which you are getting the best performance for your dollar. It also often represents the highest volume of sales AMONG the 'gaming card sector' are you familar with this general 'ruleism'?
    to this is what I am refering

    its usually in the middle between the most extreeme in high performance and the 'entry level for gaming'
    Not to continue in other topic:

    There are many "gaming standards":

    720p playable/720p 60FPS/720p max detail 60 FPS
    1080 playable/..../....
    1440p.... ....
    4k.... ....

    970/390 would be entry level to high end - 1080p GPUs/capable 1440p GPUs (though 970 3,5GB memory starts to be an issue here so generally not recommended for 1440p)

    "Sweet spot" would be 100-200$ GPU segment because those GPUs offer decent performance (and often best perf/$) and perf/$ ratio tends to fall off as you go up in price (all the more quickly the more you go up in price and often when you go down in price)
    so gxt950 would be that sweet spot?

    fair enough I have no issue with that i just wouldnt consider 'entry level' and the sweet spot to be the same but its likely just a matter of symantics choice on what is 'entry level' so no worries there

    Please do not respond to me, even if I ask you a question, its rhetorical.

    Please do not respond to me

  • RidelynnRidelynn Member EpicPosts: 7,383
    Well, Malbooga pointed to a price segment - which is a good way of doing it. A $100 card today and a $100 card from 5 years ago... they both cost $100. And that will appeal to about the same group of potential buyers today, and 5 years ago. But the cards offered today versus those from 5 years ago will be different in terms of available power. 

    If you go and say a specific card, that gets invalidated the instant the next generation comes out, or the price structure shifts. You've then pegged it on a performance standard, and not a budget standard.

    The price structures/tiers tend to not really change over time, but the power available in each of those tiers goes up as the generations of cards move along Moore's Law. There is the $100-120 range, A ~$150 range, a $180-220 range, a $300-350 range, a $400+ range, etc. and there's been those price ranges of cards for over a decade now. I postulate that it will remain about the same, at least until integrated GPU becomes powerful enough to make the discrete market redundant (much like how the discrete sound card market pretty much dried up once on-board sound become more or less ubiquitous).
  • SEANMCADSEANMCAD Member EpicPosts: 16,775
    Ridelynn said:
    Well, Malbooga pointed to a price segment - which is a good way of doing it. A $100 card today and a $100 card from 5 years ago... 
    Actually I dont think that is true. When a card gets lower down the 'food chain' of the current production cycle the cost of getting them actually goes up. I am not 100% sure on that but I think that is the case

    Please do not respond to me, even if I ask you a question, its rhetorical.

    Please do not respond to me

  • RidelynnRidelynn Member EpicPosts: 7,383
    edited May 2016
    SEANMCAD said:
    Ridelynn said:
    Well, Malbooga pointed to a price segment - which is a good way of doing it. A $100 card today and a $100 card from 5 years ago... 
    Actually I dont think that is true. When a card gets lower down the 'food chain' of the current production cycle the cost of getting them actually goes up. I am not 100% sure on that but I think that is the case
    Only if you are looking at it as the same card from 5 years ago versus the same card today. Which isn't a common way of looking at it, as older SKUs get phased out for more efficient production methods (same performance for less price) or better production methods (more performance for same price). Unless you are looking at Long-term availability, like some server enterprise-grade or government contract parts get, but that isn't really in the scope of this discussion.

    If I look at a $100 card from 2011 - just picking one out of a hat - AMD's 6670 was released in 2011, for right around $100.

    Now I look today - Newegg has several R7 360's for right around $100.

    6670 - 480 SPUs at 800Mhz, up to 1G GDDR5 VRAM, around 768 GFLOPs
    R7 260 - 768 SPUs up to 1Ghz, up to 2G GDDR5 VRAM, around 1971 GFLOPs

    That's around a 2-3x increase in performance, for the same budget. The price hasn't moved, but the price versus performance metric has changed dramatically.

    But if you tried to buy the 6670 today, which I don't know why you would, but that seems to be what you are proposing, that 6670 today, assuming you can find one NITB, yeah that would still be around $100. Not exactly the same bargain it was 5 years ago.
  • MalaboogaMalabooga Member UncommonPosts: 2,977
    edited May 2016
    SEANMCAD said:
    Ridelynn said:
    Well, Malbooga pointed to a price segment - which is a good way of doing it. A $100 card today and a $100 card from 5 years ago... 
    Actually I dont think that is true. When a card gets lower down the 'food chain' of the current production cycle the cost of getting them actually goes up. I am not 100% sure on that but I think that is the case
    Old chip production gets discontinued before new chip hits the market. Chip production during the lifetime of the product tends to generally get cheaper as process on which they are produced matures and yields get better (you get more good chips out of same amount of materials and as you fulfill planned production scale). But its also dependant on maetrial costs.

    NVidia isnt producing GM204 and probably GM200 chips any more (Titan X, 980ti, 980, 970) for consumer parts.

    AMD isnt producing Hawaii chips any more for consumer GPUs (as that chip is also used in Pro/HPC GPUs.)
  • SEANMCADSEANMCAD Member EpicPosts: 16,775
    edited May 2016
    Malabooga said:
    SEANMCAD said:
    Ridelynn said:
    Well, Malbooga pointed to a price segment - which is a good way of doing it. A $100 card today and a $100 card from 5 years ago... 
    Actually I dont think that is true. When a card gets lower down the 'food chain' of the current production cycle the cost of getting them actually goes up. I am not 100% sure on that but I think that is the case
    Old chip production gets discontinued before new chip hits the market. Chip production during the lifetime of the product tends to generally get cheaper as process on which they are produced matures and yields get better (you get more good chips out of same amount of materials and as you fulfill planned production scale). But its also dependant on matrial costs.
    well to be fair to everyone including

    Ridelynn

    its a minor point that I thought was true but was unsure

    Please do not respond to me, even if I ask you a question, its rhetorical.

    Please do not respond to me

  • RidelynnRidelynn Member EpicPosts: 7,383
    SEANMCAD said:
    Malabooga said:
    SEANMCAD said:
    Ridelynn said:
    Well, Malbooga pointed to a price segment - which is a good way of doing it. A $100 card today and a $100 card from 5 years ago... 
    Actually I dont think that is true. When a card gets lower down the 'food chain' of the current production cycle the cost of getting them actually goes up. I am not 100% sure on that but I think that is the case
    Old chip production gets discontinued before new chip hits the market. Chip production during the lifetime of the product tends to generally get cheaper as process on which they are produced matures and yields get better (you get more good chips out of same amount of materials and as you fulfill planned production scale). But its also dependant on matrial costs.
    well to be fair to everyone including

    Ridelynn

    its a minor point that I thought was true but was unsure
    Your statement wasn't entirely incorrect, it's just that you had it in the wrong context.
  • SEANMCADSEANMCAD Member EpicPosts: 16,775
    Ridelynn said:
    SEANMCAD said:
    Malabooga said:
    SEANMCAD said:
    Ridelynn said:
    Well, Malbooga pointed to a price segment - which is a good way of doing it. A $100 card today and a $100 card from 5 years ago... 
    Actually I dont think that is true. When a card gets lower down the 'food chain' of the current production cycle the cost of getting them actually goes up. I am not 100% sure on that but I think that is the case
    Old chip production gets discontinued before new chip hits the market. Chip production during the lifetime of the product tends to generally get cheaper as process on which they are produced matures and yields get better (you get more good chips out of same amount of materials and as you fulfill planned production scale). But its also dependant on matrial costs.
    well to be fair to everyone including

    Ridelynn

    its a minor point that I thought was true but was unsure
    Your statement wasn't entirely incorrect, it's just that you had it in the wrong context.
    ok thanks

    Please do not respond to me, even if I ask you a question, its rhetorical.

    Please do not respond to me

  • [Deleted User][Deleted User] Posts: 0
    edited May 2016
    The user and all related content has been deleted.

    image

    Somebody, somewhere has better skills as you have, more experience as you have, is smarter than you, has more friends as you do and can stay online longer. Just pray he's not out to get you.
  • HrimnirHrimnir Member RarePosts: 2,415

    So mtibbs, when the NDA releases on the 17th and it's not a 15-20% gain over the previous generation, will you admit you were wrong?  BTW I will admit I am wrong if it *does* end up only being 15-20%.

    "The surest way to corrupt a youth is to instruct him to hold in higher esteem those who think alike than those who think differently."

    - Friedrich Nietzsche

  • RidelynnRidelynn Member EpicPosts: 7,383
    I think you'll need to clarify that wager a bit Mrimnir.

    It likely will be a good deal faster than that over the 980, but compared to the 980Ti, yeah, probably only 10-15% (if that)

    All depends on what you are trying to compare it to. nVidia will tell you to compare it to the 980, obviously. The high-end gamer crowd are going to want to compare it to the Ti, since they are always looking for some excuse to upgrade.
  • filmoretfilmoret Member EpicPosts: 4,906
    So what was the thing this thread was all about?  I see it is past Friday and does anyone have an update?
    Are you onto something or just on something?
  • filmoretfilmoret Member EpicPosts: 4,906
    Come on really?
    Are you onto something or just on something?
  • RidelynnRidelynn Member EpicPosts: 7,383
    edited May 2016
    No real updates possible until May 27, when we see if the 1080 actually ships or is supply constrained.

    That target date was the big reveal of the conference, and is really what this thread was about.
  • BrutiiBrutii Member UncommonPosts: 42
    You must gather your party before venturing forth.
  • MalaboogaMalabooga Member UncommonPosts: 2,977
    edited May 2016
    1070 news, only 1920 cuda cores (only 65% of 1080), low boost clocks. But as expected they dump all crappy chips on 1070.

    http://wccftech.com/gtx-1070-1920-cuda-cores/

    Im pretty sure that they will do 1070ti with GDDRX5 later on and screw all those who bought 1070 as 1070 has VERY bad bandwidth with only GDDR5 and just 1920 cores...that also explains why 1070 is so "cheap" compared to 1080.

    If you compare that to 980 vs 970, 980 has 2048 cores, 970 has 1664 cores and thats ONLY 23% less than 980 and they have SAME banwidth.

    Post edited by Malabooga on
  • vadio123vadio123 Member UncommonPosts: 593
    Shipped one 1080 to me right now! 
  • RidelynnRidelynn Member EpicPosts: 7,383
    Malabooga said:
    1070 news, only 1920 cuda cores (only 65% of 1080), low boost clocks. But as expected they dump all crappy chips on 1070.

    http://wccftech.com/gtx-1070-1920-cuda-cores/

    Im pretty sure that they will do 1070ti with GDDRX5 later on and screw all those who bought 1070 as 1070 has VERY bad bandwidth with only GDDR5 and just 1920 cores...that also explains why 1070 is so "cheap" compared to 1080.

    If you compare that to 980 vs 970, 980 has 2048 cores, 970 has 1664 cores and thats ONLY 23% less than 980 and they have SAME banwidth.

    Will be interesting to see what the real world performance delta is.
Sign In or Register to comment.