[VC]AMD Fiji XT spotted at Zauba

Page 10 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Feb 19, 2009
10,457
10
76
+1

I would say Russian's statement is true. 4hrs a day every day for a year is someone with a lot more time on their hands than the average gamer. Even though I'm considered a "gamer" to all my friends, I generally only average 3-4hrs a week.

Power efficiency certainly has it's place but $$ savings isn't really one of the reasons for most people.

Back OT - It sounds like the 390X would be pretty short-lived on 20nm if it debuts this spring and 16nm FF is supposed to be ready at the end of 2015.

When I was in my last job, I was gaming about 1hr a weekday average, due to wife & kids after work time.

On weekends, I probably squeeze in about 8 hrs of gaming for both sat & sun, but it depends if we have family outings planned or not.

Perf/w matters, but not to a great extent where I would have to folk out a big $ premium for it. Saying that, AMD absolutely needs to match NV on perf/w because they need to fight for the notebook market.

@f1sherman
My CF stock R290 clocks with -50mV takes 200W ea under mining load and ~170W in Valley. Efficiency is good with a slight under volt. :)
 
Last edited:

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
4 hours a day? That's it?

I thought we were talking about no life, no friends, gaming addict (MMO?!).
And it looked like you wanted to get your message across (kWh being BS) by taking almost the worst case scenario.
Now be generous and give a man his 12 hours/day. He has no life, remember?



But wait, what's this ^^ Canada??
Out of all places, you chose Canada, which has the cheapest electric energy on the planet relative to purchasing power.
You are not going to get your message across if you continue like this ;)

Lets not take cheapo Canada OK? Nor 8 - 37 cents USA(much variance).
Lets take middle of the road 30 cents (Australia, Belgium, Netherlands, Italy , Ireland, Sweden) which accidentally is the number you started your calculation with.

$0.30 x 0.135 kWh x 12h x 365 x 100 / 80 (PSU efficiency) = $221 /year (~per 300W card)

So it looks like during the period of 2 years, our nolifer is going to pay for yet another nice GPU($450) on electricity alone.
Or $300 per 2 years, if he's no-lifer with a job, playing only 8h/day.




Right, imagine the marketing slogans:

  • Buy 2 Radeons, get only 1 - GAMING EVOLVED 2 year plan.
  • Radeon inside - get a job or GTFO
  • AMD Gaming on a high budget.

And this is just for our no-lifer. It gets multiplied many times over in HPC environment.
AMD won't get a single high value supercomputer/data center contract if you need a freaking water just to use their consumer GPU.

You must have forgotten the /sarc tags.
 

escrow4

Diamond Member
Feb 4, 2013
3,339
122
106
+1

I would say Russian's statement is true. 4hrs a day every day for a year is someone with a lot more time on their hands than the average gamer. Even though I'm considered a "gamer" to all my friends, I generally only average 3-4hrs a week.

Power efficiency certainly has it's place but $$ savings isn't really one of the reasons for most people.

Back OT - It sounds like the 390X would be pretty short-lived on 20nm if it debuts this spring and 16nm FF is supposed to be ready at the end of 2015.

4hrs is nothing here. I devote at least 30hrs a week if there is a decent AAA game out. Minimum. And I have a $1K to spend on a GPU but it will most likely be going to Nvidia.
 

geoxile

Senior member
Sep 23, 2014
327
25
91
4hrs is nothing here. I devote at least 30hrs a week if there is a decent AAA game out. Minimum. And I have a $1K to spend on a GPU but it will most likely be going to Nvidia.

You realize 30 hours in a week is an average of approximately 4 hours 20 minutes a day? It's certainly more than 4 hours but not by much
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
146
106
no healthy adult plays 4 hours a day every day unless that's his job or he is a millionaire / disabled person who uses gaming like reading/listening to music, etc.

Your attitude...

So besides the usual rubbish about GPUs. You also want to determine peoples life now to force your wishes for GPUs on the rest of the people so its defendable.

And talking about job, how much time do you spend on that? If its more than 6-7½ hours a day, 5 days a week and your commute is more than 15mins each way we can put that into your box as well of unhealthy. Or if you watch more than 4 hours of TV a day for that matter. (An average american watches 5 hours of TV per day.)
 
Last edited:

AtenRa

Lifer
Feb 2, 2009
14,003
3,362
136
http://en.wikipedia.org/wiki/Video_gaming_in_the_United_States

The average American gamer spends an average of 13 hours per week playing video games
http://www.theaveragegamer.com/averagegamers/

UK Gamers
Based on a study commissioned by gaming social network Pixwoo in 2013.

  • Play computer games five days a week, for 2 hours and 32 minutes each day.
Also, more Gamer stats

http://www.bigfishgames.com/blog/2014-global-gaming-stats-whos-playing-what-and-why/

http://www.theesa.com/facts/pdfs/ESA_EF_2014.pdf
 

Gikaseixas

Platinum Member
Jul 1, 2004
2,836
218
106
I play 2 hours max per day. Gotta play with kids, pretend listening to her rumblings, work around the house, listen some music, etc

4 hours a day is for a younger gamer or someone who has lots of free time after a full day of work
 

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
I don't understand: AIB's and third party vendors offer water or liquid cooling if a gamer desires it.

1. Those cost too much $, usually come with $150-200 premiums.
2. Perception is key for the avg uninformed gamer. If you release a hot and loud reference card like 7970Ghz, 290X, the avg. gamer will think that all 7970s and 290Xs run hot and loud. You have seen it on our forums. You have also seen how people continue to link power usage, temperatures and noise levels of reference 7970/290X cards.

All that non-sense would end with a WC card like 295X2. As Silverforce noted, AIO LC isn't that expensive.

The craziest part of all is just how successful Corsair, Asetek, CM, Enermax have gotten with AIO LC systems for CPUs. The difference is WC benefits GPUs far more than CPUs. Where we need WC the most are GPUs, not CPUs. A high end ocerclocked GPU uses as much power as a max overclocked 5820K, while i7 4790K is barely going to hit 160-180W max overclocked.

I want to see your rebuttal against WC for the failed throttled mess that was Titan Z. I guess you would boycott a WC flagship Maxwell card too because you are SOOO against WC?

It's interesting how nearly all opponents of WC are NV users on our board. Do you guys fear that much that 390X WC will smoke the 980 that badly? Don't worry, you will have your air cooled GM200 and perf/watt to talk about. :)
 

f1sherman

Platinum Member
Apr 5, 2011
2,243
1
0

Wow we really went all the way back to "average" and married with kids.
Remember it started with the worst case scenario and forget-about-power-difference attitude:

Let's assume some worst case scenario: R9 390X uses 300W of power and is only 20% faster than 980 for $550. That's a delta of 135W against a reference 980.

Let's take some unhealthy gamer who has no life, no kids, no friends, no real world job, and plays 4 hours a day every day all year.

As if Canada and 4h/day is really the worst case.
Thats like presenting average alcohol consumption statistics when discussing the health issues of a heavy drinker with the liver well beyond repair.

And the starting approach by RS was a good one.
Similarly to TDP when constructing cooler, take the worst case scenario and you got the entire base covered.
But now that we don't like the results, we are back to fixing the starting idea. Lets nots consider the worst case, lets take the average instead :thumbsdown:
 

SirPauly

Diamond Member
Apr 28, 2009
5,187
1
0
1. Those cost too much $, usually come with $150-200 premiums.

There is a premium -- as there was with the 295x2.


2. Perception is key for the avg uninformed gamer. If you release a hot and loud reference card like 7970Ghz, 290X, the avg. gamer will think that all 7970s and 290Xs run hot and loud. You have seen it on our forums. You have also seen how people continue to link power usage, temperatures and noise levels of reference 7970/290X cards.

All that non-sense would end with a WC card like 295X2. As Silverforce noted, AIO LC isn't that expensive.

So your answer is reference 7970 and 290x should of been water cooled?



I want to see your rebuttal against WC for the failed throttled mess that was Titan Z. I guess you would boycott a WC flagship Maxwell card too because you are SOOO against WC?

It's interesting how nearly all opponents of WC are NV users on our board. Do you guys fear that much that 390X WC will smoke the 980 that badly? Don't worry, you will have your air cooled GM200 and perf/watt to talk about. :)


I'm not going to respond to this but do desire to discuss over-all strategies, architectures, engineering aspects, features, immersion, gaming experience, software, etc. No rebuttal, because I am not against water cooling choice, but certainly not a strong vocal advocate for it for reference designs - subjective view.

Ultimately the market decides, rightly or wrongly and will refrain posting in this thread based on too much off topic.
 

james1701

Golden Member
Sep 14, 2007
1,791
34
91
Why are we not see more Asus Poseidon type cards being released. Everyone wins, air and water alike.
 

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
I will respond later to the ludicrous claims of gaming for 5-6 hours a day 5 days a week as representative of a normal/avg PC gamer, and insinuations that my 4 hours a day x 365 days representation is "too conservative", but for now I will leave this article as it's entirely on topic for 390X:

Wednesday, 12 November 2014 09:46
20nm node broken for GPUs

"It looks like we might never see 20nm GPUs from either Nvidia or AMD. From what we know, both companies spent a lot of time looking into the new 20nm manufacturing process and they have decided that it is simply not viable for GPUs.

Yields are not where they are supposed to be and from a business perspective it doesn’t make sense to design and produce chips that would end up with very low yields. At this point we do not expect to see any high-end chips in 20nm, as there are obvious manufacturing obstacles and both companies might even skip the 20nm process altogether and move directly to 16nm FinFET."

http://www.fudzilla.com/home/item/36284-20nm-node-broken-for-gpus

As I said before, there are many positive reasons why WC is the way forward and should be implemented in stock form because if done right, it is superior to air cooled blower solution. If cards like 7970Ghz and 290X came with stock WC, they would have smashed launch reviews, and overclocking would have been a dream with extemely comfortable noise levels and sub-80C maxed out.

I don't ever recall the hate and opposition against Intel bundling WC with their flagship CPUs, or anyone making fun of someone with an overclocked i5/i7 buying a Corsair H60-80-90-100-110 for improved cooling over CM212 or stock Intel fan.

If AMD still decides to be dumb enough to use a loud reference blower, instead of WC or a powerful open air solution, they deserve all the negative press coverage of their reference card. They didn't listen to us gamers with 5870/6970/7970/290X so I am confident finally they get serious enough and make a giant leap in improving cooling. 295X2 was a major step in proving that WC > Air cooling. If AMD implements water in their flagship cards, this will be a big competitive advantage for anyone with poor case airflow where open air cooling fails, and it will be a MASSIVE win for us SLI/CF users who will instantly solve the biggest gripe of open air dumping all the heat into your case. The naysayers can continue buying more expensive water blocks or $120+ Hybrid Accelero cooler. Never have I seen an enthusiast forum so stuck in their ways. I guess you all forgot how progress happened with CPU cooling, moving from horizontal coolers, to heatpipes, to towers and now to full blown 280MM AIO LC kits. But when it comes to GPUs, it's like it's forbidden somehow to introduce the far superior Heat dissipation properties of water to PC components that actually benefit THE most from water.
 

monstercameron

Diamond Member
Feb 12, 2013
3,818
1
0
I will respond later to the ludicrous claims of gaming for 5-6 hours a day 5 days a week as representative of a normal/avg PC gamer, and insinuations that my 4 hours a day x 365 days representation is "too conservative", but for now I will leave this article as it's entirely on topic for 390X:

Wednesday, 12 November 2014 09:46
20nm node broken for GPUs

"It looks like we might never see 20nm GPUs from either Nvidia or AMD. From what we know, both companies spent a lot of time looking into the new 20nm manufacturing process and they have decided that it is simply not viable for GPUs.

Yields are not where they are supposed to be and from a business perspective it doesn’t make sense to design and produce chips that would end up with very low yields. At this point we do not expect to see any high-end chips in 20nm, as there are obvious manufacturing obstacles and both companies might even skip the 20nm process altogether and move directly to 16nm FinFET."

http://www.fudzilla.com/home/item/36284-20nm-node-broken-for-gpus

As I said before, there are many positive reasons why WC is the way forward and should be implemented in stock form because if done right, it is superior to air cooled blower solution. If cards like 7970Ghz and 290X came with stock WC, they would have smashed launch reviews, and overclocking would have been a dream with extemely comfortable noise levels and sub-80C maxed out.

I don't ever recall the hate and opposition against Intel bundling WC with their flagship CPUs, or anyone making fun of someone with an overclocked i5/i7 buying a Corsair H60-80-90-100-110 for improved cooling over CM212 or stock Intel fan.

If AMD still decides to be dumb enough to use a loud reference blower, instead of WC or a powerful open air solution, they deserve all the negative press coverage of their reference card. They didn't listen to us gamers with 5870/6970/7970/290X so I am confident finally they get serious enough and make a giant leap in improving cooling. 295X2 was a major step in proving that WC > Air cooling. If AMD implements water in their flagship cards, this will be a big competitive advantage for anyone with poor case airflow where open air cooling fails, and it will be a MASSIVE win for us SLI/CF users who will instantly solve the biggest gripe of open air dumping all the heat into your case. The naysayers can continue buying more expensive water blocks or $120+ Hybrid Accelero cooler. Never have I seen an enthusiast forum so stuck in their ways. I guess you all forgot how progress happened with CPU cooling, moving from horizontal coolers, to heatpipes, to towers and now to full blown 280MM AIO LC kits. But when it comes to GPUs, it's like it's forbidden somehow to introduce the far superior Heat dissipation properties of water to PC components that actually benefit THE most from water.

Lol, you are being taken for a ride, WC is only bad on gpus because AMD did it first [being a reference cooler]. Were it nvidia then the conversation would be different, as usual.
 

SirPauly

Diamond Member
Apr 28, 2009
5,187
1
0
If AMD still decides to be dumb enough to use a loud reference blower, instead of WC or a powerful open air solution, they deserve all the negative press coverage of their reference card.

What if AMD's new architecture is very efficient?

How would one x-fire, 2, 3, 4 if each single GPU card has a similar set up like the 295x2?
 

Keysplayr

Elite Member
Jan 16, 2003
21,219
56
91
Lol, you are being taken for a ride, WC is only bad on gpus because AMD did it first [being a reference cooler]. Were it nvidia then the conversation would be different, as usual.

There isn't anything wrong with water cooling a GPU. The perception of the "need" for it however might be an issue among some. I don't care personally one way or the other.
 

III-V

Senior member
Oct 12, 2014
678
1
41
Lol, you are being taken for a ride, WC is only bad on gpus because AMD did it first [being a reference cooler]. Were it nvidia then the conversation would be different, as usual.
I strongly disagree. But anyway, I've got nothing per se against water cooling, as long as failure rates are tolerably low, as are costs.
 

f1sherman

Platinum Member
Apr 5, 2011
2,243
1
0
I will respond later to the ludicrous claims of gaming for 5-6 hours a day 5 days a week as representative of a normal/avg PC gamer, and insinuations that my 4 hours a day x 365 days representation is "too conservative", but for now I will leave this article as it's entirely on topic for 390X:


Please don't.
I won't respond to a wall of text that has wrong claim as the premise.
I'll do it here and now:

RS you are backpedalling with your worst case scenario.
From nolife, no kids, no friends, no job, very heavy gamer to this clean cut guy that's representative of normal/avg PC gamer. Wow really?

I find that misleading and dishonest.

Matter of fact I thought electricity does not matter before considering your little calculus exercise. You showed me otherwise.

PS
Bold part in your quote: you are making this up. And also you forgot to factor in PSU efficiency ;)


##
 
Last edited:

crisium

Platinum Member
Aug 19, 2001
2,643
615
136
How would one x-fire, 2, 3, 4 if each single GPU card has a similar set up like the 295x2?

A legitimate concern, but x-fire is for enthusists who rarely buy reference. There will still be non reference open air. Though it would be awkward if we have a large gap before non reference like with 290x.
 

AtenRa

Lifer
Feb 2, 2009
14,003
3,362
136
Wow we really went all the way back to "average" and married with kids.
Remember it started with the worst case scenario and forget-about-power-difference attitude:

There are more than 80M PC gamers in the world, playing 8 Hours 24/7 will not even be 1% of them.
Even when i was in Uni playing Ultima Online, I wasnt able to play every day for 8 hours. I could do it some weekends but not 24/7 365.

also, I just posted some stats about Gamers, i wasnt responding to you directly ;)
 

alcoholbob

Diamond Member
May 24, 2005
6,390
469
126
lol at that article, how is 20nm broken for GPUs?

Don't all TSMC nodes have a low performance and a high performance version? Just because TSMC decided to call their 20nm High Performance node "16nm FinFET" when it's obviously a 20nm node based on size and performance, and you can't build a GPU on their 20nm low performance node doesn't mean 20nm is broken, it just means its like how it always has been.