• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Clock for Clock, Vega VS FuryX and discussion

  • Thread starter Thread starter Deleted member 50521
  • Start date Start date
Durvelle27 said:
Considering Volta is over a year away which gives time for AMD to release another series as well that may keep up
Click to expand...
way report is volta could be 4-7months if they can get mass production up by end of the year, could have it at start of q2. Its possible but nothing confirmed and just the though it could be possible could be a problem for amd
 
Batou1986 said:
Comparing GPU's clock to clock
:roll:
Click to expand...
why not?
vega is an evolutionary architecture, it's still in the gcn line of things and when it comes to amount of shaders, rops and whatnot, it's close to identical.
comparing same or similar architectures with same amount of resources (or normalized to that) clock to clock does give some insight as to how changes fared.

as far as ipc changes go, remember that nvidia also took "ipc" hit moving from maxwell to pascal. nvidia themselves said this was necessary to increase the clocks. it sounds reasonable to assume that amd had to make a similar choice.
 
arbiter said:
way report is volta could be 4-7months if they can get mass production up by end of the year, could have it at start of q2. Its possible but nothing confirmed and just the though it could be possible could be a problem for amd
Click to expand...

Well, I can tell you that there won't be any Volta till 2018 for sure. Partially because Volta is so expensive to make and partially because AMD isn't really pushing the envelope with Vega.

I was gaming pretty comfortably with GTX 980, going to GTX 1080Ti means I'll be good for minimum 2 years like I was with GTX 980. Meaning my next replacement will be just in sync with Volta and Navi if all goes as predicted.
 
BiggieShady said:
Indeed. The only way to detect architectural improvements.
Click to expand...
Sometimes performance per clock (or IPC) is lower in a more modern architecture (or process), because it makes something else possible, so in the end performance is higher. You're totally ignoring this aspect.

Generally speaking, this whole IPC fallacy is rather daft.
 
If you lower IPC by 5% to gain 30% throgh higher clock, that certainly makes sense...
 
When you make so much fuss over newly introduced features like AMD did then clock for clock is the only way to see whether they make any difference. DSBR is not making as much difference as people thought it would, if any difference at all. How do you know that ? Thanks to Fury X clock for clock comparison.
 
Last edited:
RejZoR said:
Well, I can tell you that there won't be any Volta till 2018 for sure. Partially because Volta is so expensive to make and partially because AMD isn't really pushing the envelope with Vega.
Click to expand...
NVIDIA only needs a replacement for 1080Ti. Everything else they can just shift down in the lineup thanks to good efficiency. So an expensive Volta is not a big deal.
 
Ain't that simple given how power hungry 1080Ti is. I mean it's not Vega level but a 600mm2 Pascal would be chugging a lot unless clocked low, which defeats the point of making it. My bet is there will be no Pascal refresh but Volta will come in Q1 2018, as soon as GDDR6 is ready. They don't even have to release a xx70 SKU along with the flagship, just 2080 with GDDR6 and 2060 with DDR5X.
 
I think they can stick few more shader units into Pascal, clock it to 2GHz and make it even more beefier that way. At expense of power consumption. But realistically, if they bump Pascal to 250W and give it extra 20-30% boost, that would still work imo. 250W is imo area where people are still willing to use. Anything past that is questionable as it brings larger cooling issues.
 
I'm kind of surprised they don't respond to V56 at all, but if the launch price scam is real then they don't really have to, 1070 is $100 cheaper and almost as fast, 1080 is about the same price or slightly more expensive while it's definitely faster.
 
Yeah they don't really need to do anything, their margins are gonna be far better with the 314 mm² GP104, especially compared to a 486 mm² HBM2 chip.
 
Last edited:
So many people make the wrong assumption that real life performance IPC gains are only linked to the hardware implementation. Which is simply not true , you can have a million extensions or changes that increase performance by 1000% , if the software does not account for all that it's all for nothing. It is dumb/superficial to compare 2 different architectures clock for clock without knowing that aspect.

Intel and AMD have struggled with this issue ( and still are ) for ages. Every time they introduced new SMID extensions for example , there was no software that would use it or no compiler that would efficiently make of them for long periods of time. They launched architectures that clearly had much improved IPC but that gain was no where to be seen in real life performance.
 
Last edited:
xkm1948 said:
https://www.computerbase.de/2017-08...ts-escalation-rx-vega-56-vs-vega-64-vs-fury-x

View attachment 91166

So basically the performance per clock cycle, which in theory should be purely from improved design, only gives a tiny amount of performance boost. The most performance boost comes from the bumped clock rate.
http://www.anandtech.com/show/11717/the-amd-radeon-rx-vega-64-and-56-review/2


This feels like Pentium4 and Faildozer to me. RTG tries to maximize clock hoping it will give them some performance boost. What they ended up getting is a power hungry monster. I am sure @W1zzard would do some in depth analysis of Vega design in the future.
Click to expand...
Pascal (vs Maxwell) says hi o_O

I have a feeling that purely gaming related improvements aren;t going to be that high on Volta, or Navi, either. Pascal also had more compute (performance) uplift as compared to gaming, the only silver lining for Vega should be that it's compute heavy design might come in handy in the future just like the original GCN.
Vya Domus said:
So many people make the wrong assumption that real life performance IPC gains are only linked to the hardware implementation. Which is simply not true , you can have a million extensions or changes that increase performance by 1000% , if the software does not account for all that it's all for nothing. It is dumb/superficial to compare 2 different architectures clock for clock without knowing that side of things.

Intel and AMD have struggled with this issue ( and still are ) for ages. Evey time they introduced new SMID extensions , there was no software that would use it or no compiler that would efficiently make of them for a long period of time. They launched architectures that clearly had much improved IPC but that gain was no where to be seen in real life performance.
Click to expand...
So what you;re saying is that Vega will age better, then say Pascal or even Polaris? I agree however that is the only silver lining for this release.
With the launch of AMD's newest series of high-end graphics cards, RX Vega, we have also learned that some of the new technologies could also lead to massive performance boosts on the PlayStation 4 Pro since the console comes with a chip manufactured by AMD based on its Polaris architecture.

The GPU architecture is said to feature the Rapid Packed Math (RPM) technology that is basically hyper-threading for GPUs, but although the Xbox One X uses AMD GPU, Microsoft opted not to use the RPM.

The RPM technology allows for double the amount of operations to be conducted when less precision is required, meaning if 32-bits of precision are not required, the GPU could instead run two 16-bit operations in parallel and doubling processing throughput, from 4.2 teraflops to 8.4 teraflops. For reference, the Xbox One X has 6 teraflops of graphical performance. The only problem is that developers will need to code for the new technology in order to take advantage of it.
Click to expand...
http://www.nextpowerup.com/news/37826/playstation-4-pro-could-end-up-being-faster-than-xbox-one-x/
 
R0H1T said:
So what you;re saying is that Vega will age better, then say Pascal or even Polaris? I agree however that is the only silver lining for this
Click to expand...



Funny how every time I bring up aspects that are very real with regards to the way hardware&software implementations go along in general , everyone just assumes I am implying Vega will turn into a Titan Xp killer in the near future.

No , I am not saying Vega will most certainly age better. What I am saying is that this "clock for clock" comparison is misleading/useless without more in-depth information at hand.
 
Vya Domus said:
Funny how every time I bring up aspects that are very real with regards to the way hardware&software implementations go along in general , everyone just assumes I am implying Vega will turn into a Titan Xp killer in the near future.

No , I am not saying Vega will most certainly age better. What I am saying is that this "clock for clock" comparison is misleading/useless without more in-depth information at hand.
Click to expand...
Who said anything about Titan X or P killer?

It's not when you're comparing IPC in case of GPU, however most of the untapped potential for GPU's takes a long time to be fully tapped into.
That's mainly because the consoles aren;t updated as quickly & game studios (& publishers) generally concentrate there first.
 
Durvelle27 said:
Considering Volta is over a year away which gives time for AMD to release another series as well that may keep up
Click to expand...
It's almost September 2017, and more than a year is at least 13 months. Are you aware of those ? Or maybe you mean Volta is more than one year away from your current location if you walked all the way through to Santa Clara, hence "volta is a more than a year away" like "I live 2 hours from here" ? I say next amd generation is probably gonna be more than a year away from volta, not in the meantime.
 
Last edited:
cucker tarlson said:
It's almost September 2017, and more than a year is at least 13 months. Are you aware of those ? Or maybe you mean Volta is more than one year away from your current location if you walked all the way through to Santa Clara, hence "volta is a more than a year away" like "I live 2 hours from here" ? I say next amd generation is probably gonna be more than a year away from volta, not in the meantime.
Click to expand...
Q3 2018 is a year
 
BiggieShady said:
Indeed. The only way to detect architectural improvements.
Click to expand...
I don't see how, yes you can run crysis on both at same clocks and compare how they both run in effect legacy software .
But most of the architecture changes were for non functioning in that test ,features ,and i dont buy gpus for games i can already run fine.
I buy them for games due to be released.
But it Is an easy way to troll Amds software team who admittedly have not yet enabled CBR or rpm etc etc.
 
Durvelle27 said:
Q3 2018 is a year
Click to expand...
Are you teaching me the calendar ?
We all have no idea when Volta is coming. But more than a year from now ? Seriosly, do you believe nvidia is just gonna say "ahhh screw it let amd catch up with the drivers, we'll give them a year or more" ? Isn't V100 in production already and GDDR6 due for production in early 2018 ?
 
Last edited:
cucker tarlson said:
Are you teaching me the calendar ?
We all have no idea when Volta is coming. But more than a year from now ? Seriosly, do you believe nvidia is just gonna say "ahhh screw it let amd catch up with the drivers, we'll give them a year or more" ? Isn't V100 in production already and GDDR6 due for production in early 2018 ?
Click to expand...
It's Nvidia they do stupid things and it's never to benefit the consumer

They just love sitting on what they have and twiddling their fingers
 
cucker tarlson said:
Are you teaching me the calendar ?
We all have no idea when Volta is coming. But more than a year from now ? Seriosly, do you believe nvidia is just gonna say "ahhh screw it let amd catch up with the drivers, we'll give them a year or more" ? Isn't V100 in production already and GDDR6 due for production in early 2018 ?
Click to expand...

Who cares that V100 is in production , it is hardly representative of whatever Volta will be for the gaming variant , if there's even going to be one.
 
I'm not saying it's representative of what it will be, if you read that correctly, I'm saying that v100 was paper launched in may 2017 and available in q3 same year, suggesting that we have to wait 1.5 year for the gaming variant is a stretch. Also, unless he means that amd will just do a straight rebrand of the cards that literally just launched, expecting team red to launch another series before volta is not something I can really logically comprehend given how much time it usually takes them to release a high end card.
 
Last edited:
You must log in or register to reply here.
Back
Top