[ComputerBase] Pascal is ahead of Vega in Wolfenstien's 2 worst case scenarios

[Muhammed]

First bench is up, Pascal is massively ahead of Vega in the worst case scenarios (where GPU power matters), though in other less problematic scenes, Vega cards are 10% faster than comparable Pascal cards

https://www.computerbase.de/2017-10...nstein-2-1920-1080-anspruchsvolle-testsequenz


Discussion continued here:
https://forums.anandtech.com/threads/various-wolfenstein-iis-benchmarks.2524410/unread

AT Moderator ElFenix

 

[Carfax83]

Damn, I was just about to post this, and you beat me to the punch! I was actually going to put benchmark pics in my thread though

And benchmarks in the area that favored AMD:

​

 

[Carfax83]

Also, PCgameshardware.de is still in the process of benchmarking, but they've already done a few GPUs.

 

[Muhammed]

New Also, PCgameshardware.de is still in the process of benchmarking, but they've already done a few GPUs.

In their sequence GTX 1080 = Vega 64, across all resolution, and this is without any game ready driver from NVIDIA, with Vega having FP16 and Async Compute as well.

 

[caswow]

oh well thread already starting good. watch amd deliver a new driver that fixes the worst case scenario

 

[Spjut]

Cool that it's Vulkan only, I thought there would have been an OpenGL path as well

 

[PhonakV30]

given by this thread , I knew Muhammed would do bash on AMD.no wonder He chose word "worst case scenario".he could write this :

Wolfenstien's 2 Review : Vega vs Pascal.

But instead we see :
Pascal is ahead of Vega in Wolfenstien's 2 worst case scenarios

Well , Prepare for War Flame by OP.I will not look into this thread.Good luck.

 

[96Firebird]

"Worst case scenario" is the type of the benchmark that ComputerBase used. Stop crying.

 

[Rifter]

given by this thread , I knew Muhammed would do bash on AMD.no wonder He chose word "worst case scenario".he could write this :

Wolfenstien's 2 Review : Vega vs Pascal.

But instead we see :
Pascal is ahead of Vega in Wolfenstien's 2 worst case scenarios

Well , Prepare for War Flame by OP.I will not look into this thread.Good luck.

way to not read the article, nice trolling job. The name of the test was worst case scenario. If you are going to troll at least read the article or you come off looking like you have no idea what you are talking about.

 

[Despoiler]

"Worst case scenario" is the type of the benchmark that ComputerBase used. Stop crying.

It's also presented second in the article and yet here we have it presented first. Vega 64 = 1080ti @ 1080p.

 

[PhonakV30]

way to not read the article, nice trolling job. The name of the test was worst case scenario. If you are going to troll at least read the article or you come off looking like you have no idea what you are talking about.

Title :
Wolfenstein 2 im Benchmark: Das Vulkan-API-Spiel mit Vega-Feature

https://www.computerbase.de/2017-10/wolfenstein-2-new-colossus-benchmark/

Troll? oh wait.... You have no idea what you are talking about ! Title should be written this :

Wolfenstein 2 in Benchmark: The Volcan API game with Vega feature

but OP didn't , He chose only part of test.

 

[96Firebird]

It's also presented second in the article and yet here we have it presented first. Vega 64 = 1080ti @ 1080p.

Did you have a point?

 

[Rifter]

Title :
Wolfenstein 2 im Benchmark: Das Vulkan-API-Spiel mit Vega-Feature

https://www.computerbase.de/2017-10/wolfenstein-2-new-colossus-benchmark/

Troll? oh wait....

"Wolfenstein 2 - 1.920 × 1.080, worst case test sequence"

nice try though. OP was obviously calling attention to a specific benchmark, and he used the same name as the article he is quoting

 

[kawi6rr]

given by this thread , I knew Muhammed would do bash on AMD.no wonder He chose word "worst case scenario".he could write this :

Wolfenstien's 2 Review : Vega vs Pascal.

But instead we see :
Pascal is ahead of Vega in Wolfenstien's 2 worst case scenarios

Well , Prepare for War Flame by OP.I will not look into this thread.Good luck.

I couldn't agree more, I figured it was going to be AMD bashing thread just based on the OP and I was correct. I'll not stop by this thread again.

 

[Rifter]

I couldn't agree more, I figured it was going to be AMD bashing thread just based on the OP and I was correct. I'll not stop by this thread again.

How exactly is it AMD bashing to show benchmarks of their performance? Weather you like the results or not they are still the results, and from cross referencing with other sites they look to be accurate.

Generally i like to know performance metric in as many situations as possible when considering GPU purchases, so this testing of worst case scenario vs standard canned benchmark runs i find very useful. I care much more about how my GPU performs in these worst case scenario situations than i do Vs a standard canned run with little happening on screen and FPS is 100+ anyways.

 

[PhonakV30]

"Wolfenstein 2 - 1.920 × 1.080, worst case test sequence"

nice try though. OP was obviously calling attention to a specific benchmark, and he used the same name as the article he is quoting

This is my question , Why he did ?! Why not choose computerbase's Title ? Why did he go that specific Test ? Just look at his history.You seriously do have problem grasp, It's not WCCFTECH's Forum, It's professional Anandtech Forum.Just look what happened to Carfax83's thread.stahlhart locked because of stupid war flame.I Think this thread will be locked.

 

[Rifter]

This is my question , Why he did ?! Why not choose computerbase's Title ? Why did he go that specific Test ? Just look at his history.You seriously do have problem grasp, It's not WCCFTECH's Forum, It's professional Anandtech Forum.Just look what happened to Carfax83's thread.stahlhart locked because of stupid war flame.I Think this thread will be locked.

He is Pro Nvidia poster for sure, im not saying he isnt.

But that doesnt change the fact that the results of this test appear to be accurate.

And as i mentioned in my last post this is information i find very useful. How your GPU handles highly stressful testing and highly stressful parts of games when FPS is dropping due to extreme amount of objects on screen and action happening is much more useful than standard canned benchmarks.

Who cares how fast your GPU is in scenes where there isnt much going on and FPS is very high already if it lets you down when it counts and the crap hits the fan

 

[Despoiler]

Did you have a point?

Do you?

The first test scene corresponds to a demanding scenario , as it occurs in the game again and again. It is a sequence inside and outside a building in combination with various particle effects. The second scene, on the other hand, is a worst-case scenario . For the first hour and a half, there has only been one sequence of this kind.

So for 99% of the game we have Vega 64 = 1080TI. Then one place the perf is off. Context is key, but conveniently ignored by many.

 

[Rifter]

Do you?



So for 99% of the game we have Vega 64 = 1080TI. Then one place the perf is off. Context is key, but conveniently ignored by many.

Obviously mores testing is needed, in other games especially. But if it becomes a trend that Vega chokes in highly stressful situations then that is something that would be good to know.

 

[Udgnim]

looks like AMD driver updates no longer help increase performance for Hawaii or Fiji

performance differences between 580 vs Fury X & 390 are significant

or maybe it's a game developer thing and they focused their Vulkan path primarily for Polaris & Vega

 

[nurturedhate]

Account created in 2009. Lays dormant for 8 years. Only becomes active right before the launch of Vega. Only posts vega flame posts and threads.

 

[Trumpstyle]

This is what you hate about pc-gaming. Old hardware ages hilariously fast. Geforce 1060 twice the performance compared to geforce 970 when they should be about equal. Fury X should be about 40% faster than radeon 580 but is 30% slower .

You always forced to buy the latest gpu architecture or u will always get a few games that runs badly.

 

[ElsayedAli]

Curious considering FP16 is used in the game, wonder how much it impacts performance.

So for 99% of the game we have Vega 64 = 1080TI. Then one place the perf is off. Context is key, but conveniently ignored by many.

Only in 1080p though, context is key again, this is probably a CPU limitation, as the 1080TI is unable to go higher than 140fps in any scenario. Also PCGamesHardware tested another scenario and Vega 64 was only GTX 1080 level, so obviously this isn't 99% of the game. More data is needed indeed.

 

[nurturedhate]

This is what you hate about pc-gaming. Old hardware ages hilariously fast. Geforce 1060 twice the performance compared to geforce 970 when they should be about equal. Fury X should be about 40% faster than radeon 580 but is 30% slower .

You always forced to buy the latest gpu architecture or u will always get a few games that runs badly.

Looks to be an issue w/ sub 6gb vram cards and some massive utilization issues for Fury X.

 

[Carfax83]

given by this thread , I knew Muhammed would do bash on AMD.no wonder He chose word "worst case scenario".he could write this :

Wolfenstien's 2 Review : Vega vs Pascal.

But instead we see :
Pascal is ahead of Vega in Wolfenstien's 2 worst case scenarios

Well , Prepare for War Flame by OP.I will not look into this thread.Good luck.

Um, did you not see those pretty graphs I posted? I uploaded best and worst case results for both AMD and NVidia.