AMD Fury X

[Godlygamer23]

Gotta remember, 4gb of hbm memory is nothing like 4gb of regular gddr5. The fury X at stock is out performing the titan x at 4k in quite a few games, and even beating the 980ti with it's 6gb in about half of em as well.

If it can do that on 4gbs of memory, then shit, think of what'll come when they get 8gb models rolling out and the vram no longer becomes a bottleneck.

You're making the assumption VRAM is a bottleneck as it is. If you're not hitting the limit and the GPU is at 100% constantly, increasing the amount will do nothing.

[MAXXHEW]

You guys can never just say it's an awesome card from AMD?  Always have to compare it to other brands that many of us don't care to hear/read about... some of us don't want to buy from other brands, so why even take the time to list the competitions specs/benchmarks?!  From what I've seen nvidia has some serious issues as far as being honest when it comes to their advertising tactics, even if they make good stuff... I don't care for them as a company so I don't care about their hardwares specs/benchmarks, and I don't want the benchmarks/graphs cluttered up with other GPU's useless info.  AMD has given us on the Red Team a great card that we can all be happy with... Windows 10, DX12 and/or Vulkan will surely give us an even better experience.  Stop doing the stupid comparison charts/graphs... there's no reason for it.  If you want to do comparison vids... then make a separate video. Just my opinion... I could care less about nvidia specs.

[Dabombinable]

You guys can never just say it's an awesome card from AMD?  Always have to compare it to other brands that many of us don't care to hear/read about... some of us don't want to buy from other brands, so why even take the time to list the competitions specs/benchmarks?!  From what I've seen nvidia has some serious issues as far as being honest when it comes to their advertising tactics, even if they make good stuff... I don't care for them as a company so I don't care about their hardwares specs/benchmarks, and I don't want the benchmarks/graphs cluttered up with other GPU's useless info.  AMD has given us on the Red Team a great card that we can all be happy with... Windows 10, DX12 and/or Vulkan will surely give us an even better experience.  Stop doing the stupid comparison charts/graphs... there's no reason for it.  If you want to do comparison vids... then make a separate video. Just my opinion... I could care less about nvidia specs.

You sir appear to be ignorant and a fanboy.

[pjm0925]

You guys can never just say it's an awesome card from AMD?  Always have to compare it to other brands that many of us don't care to hear/read about... some of us don't want to buy from other brands, so why even take the time to list the competitions specs/benchmarks?!  From what I've seen nvidia has some serious issues as far as being honest when it comes to their advertising tactics, even if they make good stuff... I don't care for them as a company so I don't care about their hardwares specs/benchmarks, and I don't want the benchmarks/graphs cluttered up with other GPU's useless info.  AMD has given us on the Red Team a great card that we can all be happy with... Windows 10, DX12 and/or Vulkan will surely give us an even better experience.  Stop doing the stupid comparison charts/graphs... there's no reason for it.  If you want to do comparison vids... then make a separate video. Just my opinion... I could care less about nvidia specs.

 

You can ignore other companies if you want, but an educated consumer will look at everything that is available to get the best value for their money.  Ignoring what another company is offering is not generally the best idea if you care about how you spend your money.

 

Also, you couldn't care less about nvidia specs, please use the phrase correctly.

[GzeroD]

 

Also, you couldn't care less about nvidia specs, please use the phrase correctly.

He may have used it correctly. If people stopped bothering him with the spec comparison he would know nothing about them and would then care less than he currently does about them (because ATM he seems to be annoyed to know them)

[MAXXHEW]

You sir appear to be ignorant and a fanboy

Yeah, so? Duh... like I care?

[GzeroD]

Yeah, so? Duh... like I care?

Please don't troll in high gear it disturbs the lurkers.

[KWelz]

Well at least he is honest about being a fanboy.   Most people aren't.   

[ashtaroth]

You guys should tend to avoid some bored people. Think of the rules in a zoo. Dont feed them. Anyways against its current competition and with a horrible first impression ( i guess thats what NDA's are for) this card may discourage people looking for performance now.

[Dabombinable]

You guys should tend to avoid some bored people. Think of the rules in a zoo. Dont feed them. Anyways against its current competition and with a horrible first impression ( i guess thats what NDA's are for) this card may discourage people looking for performance now.

If your talking about the troll-they feed on the rage of others.

[MAXXHEW]

All I'm saying is that when a thread/vid is titled AMD Fury X... why do I have to look at other specs. from the competitions options?  This thread/vid should be about the AMD Fury X, if I'm gonna consider buying it... why in the hell would I want to see cluttered up graphs with useless info when all I care about is the specs. for the Fury X?  And not just that... I also ask myself, why in the hell would they even bother taking the time to make such a cluttered up graph with useless info not pertaining to the title of the thread/vid.  If we are looking for comparisons... then what would be the problem with doing a separate comparison vid?  I believe this is why some of the reviewers get the titles as "sellout" or "shills" right?  Sometimes these vids are just not about the relevant single product, why do we have to get info for the crap we don't even want?

[othertomperson]

Gotta remember, 4gb of hbm memory is nothing like 4gb of regular gddr5. The fury X at stock is out performing the titan x at 4k in quite a few games, and even beating the 980ti with it's 6gb in about half of em as well. 

 

If it can do that on 4gbs of memory, then shit, think of what'll come when they get 8gb models rolling out and the vram no longer becomes a bottleneck.

 

Here's a fun experiment. Get two 980s (not Tis), overclock the arse out of them, play a tonne of games at 4K.

 

Spoiler: the cases where 4GB of GDDR5 actually bottlenecks performance are so few and far between you'd have difficulty proving this as a limitation with 980s, let alone a Fury X. If 4GB (HBM or otherwise) doesn't bottleneck you to start with then adding another 4GB will do precisely nothing.

 

And actually all of the benchmarks I've seen have put the Fury X behind the 980 Ti. In fact one or two have put it only a couple of fps ahead of an overclocked 980 in a couple of games.

[GzeroD]

All I'm saying is that when a thread/vid is titled AMD Fury X... why do I have to look at other specs. from the competitions options?  This thread/vid should be about the AMD Fury X, if I'm gonna consider buying it... why in the hell would I want to see cluttered up graphs with useless info when all I care about is the specs. for the Fury X?  And not just that... I also ask myself, why in the hell would they even bother taking the time to make such a cluttered up graph with useless info not pertaining to the title of the thread/vid.  If we are looking for comparisons... then what would be the problem with doing a separate comparison vid?  I believe this is why some of the reviewers get the titles as "sellout" or "shills" right?  Sometimes these vids are just not about the relevant single product, why do we have to get info for the crap we don't even want?

Nope. They need to include that information to help their viewers make an informed decision. It's that simple. sure you might just have clicked the video too learn about the new and shiny toy from your favorite vendor. Some people are actively debating getting one of them (top teir GPU's) and don't have a brand preference. For those people they need to include all of that information that seems to be getting beaten into your head so violently so they can make an informed decision.

[Hunter7263]

Here's a fun experiment. Get two 980s (not Tis), overclock the arse out of them, play a tonne of games at 4K.

 

Spoiler: the cases where 4GB of GDDR5 actually bottlenecks performance are so few and far between you'd have difficulty proving this as a limitation with 980s, let alone a Fury X. If 4GB (HBM or otherwise) doesn't bottleneck you to start with then adding another 4GB will do precisely nothing.

 

And actually all of the benchmarks I've seen have put the Fury X behind the 980 Ti. In fact one or two have put it only a couple of fps ahead of an overclocked 980 in a couple of games.

To be fair memory on GDDR5 does bottleneck at least a little even on Titan X's. In terms of speed anyway. This is why you can still get measurable performance gains by overclocking just the memory on both the Titan X and the 980ti.

 

Also based on how the 300 series reacted to the increase in memory speed over the 200 series it seems that AMD has a lot of cards responding very well to higher bandwidths. Its particularly noticeable in the 390. The 390 which is the same cut down gpu core as the 290 actually outperforms the fully enabled core of the 290x because of the memory bandwidth increases. Since the Fuji core does borrow architecture of some of the previous chips its reasonable to assume that the higher bandwidth was necessary to compete with Nvidia based on the hardware.

 

But of course that's just a theory... A GAME TH.... nvm....

[Hunter7263]

All I'm saying is that when a thread/vid is titled AMD Fury X... why do I have to look at other specs. from the competitions options?  This thread/vid should be about the AMD Fury X, if I'm gonna consider buying it... why in the hell would I want to see cluttered up graphs with useless info when all I care about is the specs. for the Fury X?  And not just that... I also ask myself, why in the hell would they even bother taking the time to make such a cluttered up graph with useless info not pertaining to the title of the thread/vid.  If we are looking for comparisons... then what would be the problem with doing a separate comparison vid?  I believe this is why some of the reviewers get the titles as "sellout" or "shills" right?  Sometimes these vids are just not about the relevant single product, why do we have to get info for the crap we don't even want?

Ya but specs don't tell the whole story. Take HBM for the perfect example. In the fury its a 4096 bit width operating at 500mhz. Seriously though how do you look at that and say well GDDR5 running a 384 bit width and at 1200mhz and say that it has to be better. I mean in this case ya... but how do you compare those stats?

 

Take a 390 for example. The specs show a clearly cut down version of a 290x gpu core with faster memory. How could you possibly know that the 390 is faster at a cheaper price by just knowing the specs?

 

If they have similar stats but are based on different cores how do you tell then? Lets say a Hawaii based gpu vs a Tonga based GPU with the rest of the stats the same. That doesn't tell you anything about the card and unfortunately you cant just base performance on stats anymore because there are things that it will never tell you.

 

Last example. The FuryX can process like 8.9 TFLOPS and the Titan X can only do 6.6 TFLOPS. Does that clearly show that the Fury X is superior in games? I mean come on superior bandwidths more stream processors yada yada yada.....

 

Like they guy above me said stats aren't as useful as comparisons because you have so many different architectures. (Now if you compare say the 980ti vs the Titan X you can use numbers because its the same gm200 gpu)

[Godlygamer23]

Ya but specs don't tell the whole story. Take HBM for the perfect example. In the fury its a 4096 bit width operating at 500mhz. Seriously though how do you look at that and say well GDDR5 running a 384 bit width and at 1200mhz and say that it has to be better. I mean in this case ya... but how do you compare those stats?

Frequency is a meaningless metric because it doesn't tell you what the cycles consist of - in the case of DDR, each cycle is effectively doubled due to the Double Data Rate. It is able to complete two data transfers per cycle. Because of this, we could say the effective frequency of DDR RAM running at 600MHz is 1200MHz or 1.2GHz.

[gcubed]

That's kinda like saying you can't trust any benchmarks because you don't know how it will change in the future.  The results are accurate for now with current resources, there might be improvements in the future or there might not.  What we do know is that the current results are accurate and can be trusted. 

 

I am not saying you can't trust benchmarks, but the benchmarks are likely to improve over the next two or three driver releases.

 

Yes, the results are accurate now and can be trusted. All I am saying is that there will be improvements unless AMD has done something that nobody else has ever done in the history of driver software and shipped a perfect driver.

[gcubed]

All I'm saying is that when a thread/vid is titled AMD Fury X... why do I have to look at other specs. from the competitions options?  This thread/vid should be about the AMD Fury X, if I'm gonna consider buying it... why in the hell would I want to see cluttered up graphs with useless info when all I care about is the specs. for the Fury X?  And not just that... I also ask myself, why in the hell would they even bother taking the time to make such a cluttered up graph with useless info not pertaining to the title of the thread/vid.  If we are looking for comparisons... then what would be the problem with doing a separate comparison vid?  I believe this is why some of the reviewers get the titles as "sellout" or "shills" right?  Sometimes these vids are just not about the relevant single product, why do we have to get info for the crap we don't even want?

 

The only reason to run a benchmark is for comparison. If you don't want comparisons, just go to the store and read the box.

[mr moose]

I am not saying you can't trust benchmarks, but the benchmarks are likely to improve over the next two or three driver releases.

 

Yes, the results are accurate now and can be trusted. All I am saying is that there will be improvements unless AMD has done something that nobody else has ever done in the history of driver software and shipped a perfect driver.

 

When you say to take something with a grain of salt you are literally insinuating a large variable in the truth of the subject. That is what the expression means.  So when you say to take these benchmarks with a grain of salt what you are actually communicating is that the benchmarks have a high probability of being erroneous in some manner.  This is why we say to take first party benchmarks (from AMD or Nvidia) with a grain of salt, but unless an independent tester gives us reason to doubt the results and said results are within margin of error of other independent benchmarks then there is no need to insinuate some degree of inaccuracy.

 

Typically all cards improve with driver updates,  there is no guarantee the FuryX will be anything different.  This makes driver updates largely moot.

[GzeroD]

The only reason to run a benchmark is for comparison. If you don't want comparisons, just go to the store and read the box.

Well there is also the "To check to see if it will work/be playable for your intended uses."

Arguably that's a big part of benchmarks especially the in game benchmarks.

[mr moose]

Well there is also the "To check to see if it will work/be playable for your intended uses."

Arguably that's a big part of benchmarks especially the in game benchmarks.

 

Not arguably, but actually , You only have to look at many of the threads here to see people asking what card to get listing specific games they want to play. 

How many times do you see,  "How will card XX perform in BF4 on Ultra" or "I want to play game XX, which cards should Ii get for $XX"?

[thirtyfpsonpcisgood]

I'm still waiting on a single card that will give me 60 frames per second solid, maximum settings at 4K at an affordable price and so this card and others after it I will not be buying.  2018 maybe is when I buy a new video card!

[Godlygamer23]

When you say to take something with a grain of salt you are literally insinuating a large variable in the truth of the subject. That is what the expression means. 

When you take something with a grain of salt, it means you're skeptical.

http://dictionary.reference.com/help/faq/language/e21.html

https://en.wikipedia.org/wiki/Grain_of_salt

[mr moose]

When you take something with a grain of salt, it means you're skeptical.

http://dictionary.reference.com/help/faq/language/e21.html

https://en.wikipedia.org/wiki/Grain_of_salt

 

Are you insinuating that skepticism is not questioning the truth? or are you insinuating that questioning the variability of the truth in a subject is not the same as taking something with a grain of salt?

 

Either way I have no idea what you are trying to say sorry.

[Hunter7263]

It doesn't mean that there is a large variable in the truth. It means you should be skeptical that it is even the truth at all.