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

AMD RX Vega First Pricing Information Leaked in Sweden - "Feels Wrong"

Cut the crap. Do you a quote from AMD saying "just wait a year and we'll give you what Nvidia gives you today"? For 2017, Vega's performance is low, period.

Electronics output is not measured in years, you know.
 
Cut the crap. Do you a quote from AMD saying "just wait a year and we'll give you what Nvidia gives you today"? For 2017, Vega's performance is low, period.
So are you saying that 1080s performance is low too? or it's a low segment of video cards? so what's 1070? ridiculously low?
Just pointing out. It's not out yet. but it's nothing you already know it's crap as 1080 GTX?
Question.
So what's your 1060 if 1080 and Vega are crap?
 
So are you saying that 1080s performance is low too? or it's a low segment of video cards? so what's 1070? ridiculously low?
Just pointing out. It's not out yet. but it's nothing you already know it's crap as 1080 GTX?
Question.
So what's your 1060 if 1080 and Vega are crap?
Well, I can't explain it any better to you, if you can't grasp the concept of "context".
Vega is a (very) high end part. Vega needs more power than GTX 1080. Judging by what we know, Vega does not conclusively beat GTX 1080. Vega is coming out a year after GTX 1080. Maybe this is all you were waiting for, but I don't think that's what most users were waiting for.
 
Just a few days left, I hope all the drama goes away after launch (it won't). On topic, if the price is really that high it better deliver something to justify it :D
 
Well, I can't explain it any better to you, if you can't grasp the concept of "context".
Vega is a (very) high end part. Vega needs more power than GTX 1080. Judging by what we know, Vega does not conclusively beat GTX 1080. Vega is coming out a year after GTX 1080. Maybe this is all you were waiting for, but I don't think that's what most users were waiting for.

Vega is competing(and beating) with a GTX 1080 overclocked @~1900MHZ on games and engines that were practically wtritten for the harware.

So, where is the context here?

We are talking a piece of hardware that is, performance wise, the near top of current performance and we don't know how much it will cost.
 
Well, I can't explain it any better to you, if you can't grasp the concept of "context".
Vega is a (very) high end part. Vega needs more power than GTX 1080. Judging by what we know, Vega does not conclusively beat GTX 1080. Vega is coming out a year after GTX 1080. Maybe this is all you were waiting for, but I don't think that's what most users were waiting for.
Can't explain? Well you could if you choose words carefully. Then people will get your point but objective not this. Now it's more clearer. all you gotta do is try a bit. Maybe it will be all I've been waiting for mister. It's not out yet. :)
 
Vega is competing(and beating) with a GTX 1080 overclocked @~1900MHZ on games and engines that were practically wtritten for the harware.

So, where is the context here?

We are talking a piece of hardware that is, performance wise, the near top of current performance and we don't know how much it will cost.

A. Vega is competing(and beating) with an OCed 1080 while sucking more power than the Hoover Dam can produce.

B. Stop the crap with "written for the hardware" unless you're ok with the console ports that keep popping up every few months which means squat. The gap between AMD and nvidia's dx12 has been closed down well enough.

C. 1.5 Years wroth of development for Fiji v2.0? Yet some people called out Pascal being Maxwell on a smaller node. The only catch is Pascal works as intended. Vega on the other hand points out to the next space-heater your room will want for the winter. Also a fun fact nVidia had dominance for a full gen (almost) in the top tier with 0 competition. The market tends to saturate at some point thus making this a very though sell.

D. A piece of hardware that will be obsolete once nVidia releases Volta. Adding insult to the injury they can't match the current gen's high end after that "nice" PR campaign they put out.
 
Vega is competing(and beating) with a GTX 1080 overclocked @~1900MHZ on games and engines that were practically wtritten for the harware.

1900mhz is not an oc'd card, thats a boost clock, and a low one at that.
 
1900mhz is not an oc'd card, thats a boost clock, and a low one at that.

This is correct. A large majority of 1080s will boost themselves to 1900+ without any overclocking at all, thanks to GPU Boost 3.0. I put a 1080 FTW2 in a client's system, and without any changes it was running at 2000 while in Superposition.
 
1900mhz is not an oc'd card, thats a boost clock, and a low one at that.

Which GTX 1080 exactly? reference?
Because you can't compare 3rd party cards with reference ones.
The GTX 1080 founder edition boost is 1733 per specification.

Let's try to play "how to be an adult" and compare apples to apples please.
 
Last edited:
everyone assumes rx vega to be slow.. and thinks 800 dollars is to high, we do have to remember it was nvidia that was pricing gpus up the wazzzoo for the last years

also i sense amd has been doing alot of sandbagging and most of the leaks of their products has been eerm wierd lately, in the past most of them has been spot on, but now they are all over the place

setting the preliminary prices to 800 basically gives them a way to mess with nvidia


its amusing that everyone gets up in arms over a not set in stone, unreleased product price on a unreleased product that we dont know shit about yet

Wut?
 

Ok, I read that entire excerpt just to be sure it is saying the same things we all supposedly know and it is:
at 83°C a stock reference 1080 clocks at ~1705MHz under load.

I'll show you one better
index.php


1759.

What are you trying to demonstrate exactly?
 
1733 is the MINIMUM boost. And because the FE is such a hunk of HEYNANUNANU, the cooler, its sitting down there one bin up over minimum. You put a reference clocked card with any decent cooler on it, or turn up the damn fan on the FE, those clocks will shoot right back up as you can see from that review. It won't hit 1900Mhz, but it is damn close.
 
1733 is the MINIMUM boost. And because the FE is such a hunk of HEYNANUNANU, the cooler, its sitting down there one bin up over minimum. You put a reference clocked card with any decent cooler on it, or turn up the damn fan on the FE, those clocks will shoot right back up as you can see from that review. It won't hit 1900Mhz, but it is damn close.

I agree 100%

but when i say
Vega is competing(and beating) with a GTX 1080 overclocked @~1900MHZ on games and engines that were practically wtritten for the harware.

and someone responds
1900mhz is not an oc'd card, thats a boost clock, and a low one at that.
A large majority of 1080s will boost themselves to 1900+ without any overclocking at all, thanks to GPU Boost 3.0. I put a 1080 FTW2 in a client's system, and without any changes it was running at 2000 while in Superposition.

I say "WTF is the culprit in comparing apples to oranges"?
We all know custom boards perform better than references, what is the point of using that as a counter argument?
 
A. Vega is competing(and beating) with an OCed 1080 while sucking more power than the Hoover Dam can produce.

B. Stop the crap with "written for the hardware" unless you're ok with the console ports that keep popping up every few months which means squat. The gap between AMD and nvidia's dx12 has been closed down well enough.

C. 1.5 Years wroth of development for Fiji v2.0? Yet some people called out Pascal being Maxwell on a smaller node. The only catch is Pascal works as intended. Vega on the other hand points out to the next space-heater your room will want for the winter. Also a fun fact nVidia had dominance for a full gen (almost) in the top tier with 0 competition. The market tends to saturate at some point thus making this a very though sell.

D. A piece of hardware that will be obsolete once nVidia releases Volta. Adding insult to the injury they can't match the current gen's high end after that "nice" PR campaign they put out.

A. maybe
B. I'd never use this argument myself, but I'm the one that calls a tie when the FPS breaks 120 in a comparison or the difference from one board to the other is 2 or 3. When you port something from or to a console you are capping at 30 or 60, all those extra 90 or 60 frames are lost. Last gen consoles are capable of rendering most games over 30, set aside the cpu taxing ones.
C. Completely biased statement I find useless to respond to.
D. Your vision of what is obsolete is childish. Let's see how much it will cost when Volta delivers and compare performance to performance.
 
This is correct. A large majority of 1080s will boost themselves to 1900+ without any overclocking at all, thanks to GPU Boost 3.0. I put a 1080 FTW2 in a client's system, and without any changes it was running at 2000 while in Superposition.

You have to understand that the reference 1080 is capped at 82°C with a blower and it is a situation where you cannot "run at 2000".
This is higly dependant on custom cooler, case airflow, heat dissipation of other components.
Why in the world would you compare this specific situation of a custom board in a custom case with the reference, yet to be seen in real world situations, Vega?
 
1900mhz is not an oc'd card, thats a boost clock, and a low one at that.
The boost is an overclock. It increases voltage and core frequency. The difference is that it is not being done manually by the user but software which is called boost 2.0 or 3.0 whatever you wanna call it. same happens with CPUs. You can call it controlled overclock if you want where the temperature is being controlled by software preventing the hardware from overheating and damage.
 
You have to understand that the reference 1080 is capped at 82°C with a blower and it is a situation where you cannot "run at 2000".
This is higly dependant on custom cooler, case airflow, heat dissipation of other components.
Why in the world would you compare this specific situation of a custom board in a custom case with the reference, yet to be seen in real world situations, Vega?

The same will be true of the Vega RX on air. For a firestrike run the 1080 card will stay near it's top boost clock until the temps reach 82. With a tweak of the fan curve the 1080 will stay high with some audible fan noise. Base model Pascals operate this way - they'll ramp up to higher clocks without any overclock if temps allow. It's why a wise 1080/1080ti owner buys the cheapest card and pops it under water - instant 2GHz+ clocks without any need for a fancy custom card. I dare say the Vega RX card will be the same - needs water to shine.
 
The boost is an overclock. It increases voltage and core frequency. The difference is that it is not being done manually by the user but software which is called boost 2.0 or 3.0 whatever you wanna call it. same happens with CPUs. You can call it controlled overclock if you want where the temperature is being controlled by software preventing the hardware from overheating and damage.
Boost isn't an overclock...its running within specifications set by the factory. Overclocking is going outside of those parameters it's set with out of the box. by usThis automatic, pre-programmed functionality.

Same is true with XMP... the majority of the ICs are rated for 2400Mhz and all.................but its not overclocked/overclocking when its labeled to make XXXX Mhz.
 
Boost isn't an overclock...its running within specifications set by the factory. Overclocking is going outside of those parameters it's set with out of the box. by usThis automatic, pre-programmed functionality.

Same is true with XMP... the majority of the ICs are rated for 2400Mhz and all.................but its not overclocked/overclocking when its labeled to make XXXX Mhz.
They aren't apples to apples comparisons by you.
The sold spec on the 1080 garunteed it will boost upto 1733 ,if you didn't get that Rma , any extra boost 3 gives is still an overclock, its just a system and environment driven one. And not a warrantied one so not base spec.
I get that an overclock to you takes a guy and some time but that was how it used to be, in this day and age chip companies try to use all available resources.

Im yet to hear of any memory that automatically overclocks itself , and if you buy a speed of memory for its speed, that speed is warrantied not more.
 
I do not deem this automated process whos maximums and minumums are set by the mfg, named boost 3.0, to be overclocking. You also arent breaking the warranty when the card goes past its 'rated' boost automatically .... they dont warranty overclocking, so, boost 3.0 cant be overclocking, right (a bit of a joke there, the last part).

There is a small contigent that believes xmp is overclocking, lol... i sure dont... but the ics are rated for 2400 mhz and binned up. Apples to oranges... fine. :)
 
Last edited:
I do not deem this automated process whos maximums and minumums are set by the mfg, named boost 3.0, to be overclocking. You also arent breaking the warranty when the card goes past its 'rated' boost automatically .... they dont warranty overclocking, so, boost 3.0 cant be overclocking, right (a bit of a joke there, the last part).

There is a small contigent that believes xmp is overclocking, lol... i sure dont... but the ics are rated for 2400 mhz and binned up. Apples to oranges... fine. :)
I think the fact that the top clock it boosts to is not set and truly is based on its environment prooves my case over yours.
Sitting your pc in the middle of antartic with a geni the same card could theoretically clock higher in the same pc then it would sat in a bedroom , if it crashes at 2ghz in every game is that warrantied , could i trade for a new possibly better one because I heard these boost to 2ghz on water but if i tried it and it didn't i am pretty sure I couldn't just rma trade it for that.
 
Back
Top