AMD Running DOOM Vulkan Benchmarks On Intel Hardware

H

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
I wanted to ask you guys what your thoughts on this were. According to AMD, Radeon graphics takes DOOM to the "next level with Vulkan implementation," but they are using Intel CPUs during their tests. More than a few of you feel that it is a bit odd to see AMD using Intel hardware when it has CPUs / APUs of its own. So, is this a big deal to you? Not a big deal? I enlarged the text from the link above so you could read it.
 
They did it with their Fury X mini PCs already so it's not a new thing really.
 
Yeah they have been using X99 platform system for all of their benchmarks recently, all of the ones from Fury era, Polaris era, etc. I think they are using it because they know its faster and if they used an AMD CPU and then showed benchmarks everyone would say how they wouldn't be comparable, etc.
 
To show off graphics capabilities they need to be sure they're not CPU-limited. They're smart enough to know that the only way to get there right now is on Intel's platform.
 
Well I consistently score about 10% lower than my Intel brethren depending on how multithreaded the game is. Not enough to upgrade to an Intel platform of course. If I were selling video cards, I would want that 10% score in my benchmarks of course. So I don't see much wrong with what they are doing.
 
Sounds like you are simply bending over backwards to create drama where there is none to distract from AMD's 30%+ performance gains using Khronos Mantle, err Vulkan. I mean if AMD cpus matter so much to you all of a sudden that you had to write this post then why don't we see your reviews include systems built using AMD cpus?
 
Lorien said:
Sounds like you are simply bending over backwards to create drama where there is none to distract from AMD's 30%+ performance gains using Khronos Mantle, err Vulkan. I mean if AMD cpus matter so much to you all of a sudden that you had to write this post then why don't we see your reviews include systems built using AMD cpus?
Click to expand...
Because he doesn't work for AMD and isn't trying to sell AMD products?
 
Well, if they are going to compare their cards to nvidia (which I assume have been demonstrated on Intel hardware) then it makes sense....all else being equal. I don't think I've seen an nvidia benchmark for Doom on an AMD platform yet. Storm in a teacup, its just for ease of comparison.....nothing to see here.
 
trick0502 said:
until zen comes out, they dont have a high performance cpu. i really dont see anythng wrong with it.
Click to expand...
Fixed that for you.

Believing zen has a good chance of upsetting the balance is wishful thinking. I mean, i hope it really is a leap forward and that it's successful, but AMD really doesn't have history of living up to expectations.
 
The real sad part is that [H] hasn't kept up with the rest and doesn't have performance tests of their own so they are left talking about the small print on an AMD article. Others already beat them to the punch and have found truly interesting things like Nvidia gaining NOTHING from Vulkan which is by far more interesting to discuss.

index.php
 
Lorien said:
Sounds like you are simply bending over backwards to create drama where there is none to distract from AMD's 30%+ performance gains using Khronos Mantle, err Vulkan. I mean if AMD cpus matter so much to you all of a sudden that you had to write this post then why don't we see your reviews include systems built using AMD cpus?
Click to expand...

So asking you what your thoughts are on a subject that it is repeatedly brought up by our readers is "bending over backwards to create drama?" Really? Since when has giving our readers the opportunity to voice their own opinions on a subject "creating drama?"
 
Lorien said:
The real sad part is that [H] hasn't kept up with the rest and doesn't have performance tests of their own so they are left talking about the small print on an AMD article. Others already beat them to the punch and have found truly interesting things like Nvidia gaining NOTHING from Vulkan which is by far more interesting to discuss.
Click to expand...

I'd say this is a bit more interesting...

Part of me thinks that AMD didn't do these tests themselves, maybe an ad agency or some other third party, and that explains the CPU choice. Even if it's not, I don't see the big deal.

They probably know most people have Intel CPUs anyway since they have their own sales numbers to look at, lol.
 
I'm still curious as to what the performance would be on a 9590. yes it will/may be slower then a 5960x but I want to see it!

edit: speeling
 
Last edited:
I agree with others here, AMD CPU's just cant keep up with Intel, so if I was wanting to demonstrate GPU performance, I would use Intel as well.
 
I'm surprised by this. I thought one of the points for Vulkan/Mantle was that it helped lower end CPUs (read AMD's) compete with higher end CPUs by using more of the GPU. I had thought they'd use these early benchmarks slides to show the performance gains on their own hardware (CPU and GPU).
 
Lorien said:
The real sad part is that [H] hasn't kept up with the rest and doesn't have performance tests of their own so they are left talking about the small print on an AMD article. Others already beat them to the punch and have found truly interesting things like Nvidia gaining NOTHING from Vulkan which is by far more interesting to discuss.

index.php
Click to expand...
Sonicks said:
I'd say this is a bit more interesting...

Part of me thinks that AMD didn't do these tests themselves, maybe an ad agency or some other third party, and that explains the CPU choice. Even if it's not, I don't see the big deal.

They probably know most people have Intel CPUs anyway since they have their own sales numbers to look at, lol.
Click to expand...

Not a big deal. AMD's "current" CPUs would definitely be holding back their GPUs. Of course you want to show the best possible scores.
 
Hypergreatthing said:
Fixed that for you.

Believing zen has a good chance of upsetting the balance is wishful thinking. I mean, i hope it really is a leap forward and that it's successful, but AMD really doesn't have history of living up to expectations.
Click to expand...

i dont think it will kill 1151 cpus, but maybe within 5-10%.
 
Lorien said:
The real sad part is that [H] hasn't kept up with the rest and doesn't have performance tests of their own so they are left talking about the small print on an AMD article. Others already beat them to the punch and have found truly interesting things like Nvidia gaining NOTHING from Vulkan which is by far more interesting to discuss.

SNIP
Click to expand...

I've seen multiple posts that it's faster on NV hardware as well.

I mean, I'll know in a bit. I have an 1080 (and 6700k)

Those numbers also don't mean too much when for all we know it lopped several milliseconds off the CPU time per frame and it's just completely and utterly GPU limited and the framerate held still as a result.
 
I see absolutely nothing wrong with this. When you are testing something you want to only test 1 variable while keeping everything else the same to keep an accurate comparison.

Intel CPU w/ AMD GPU vs. Intel CPU w/ Nvidia GPU

-or-

AMD CPU w/ AMD GPU vs. AMD CPU w/ Nvidia GPU

Changing the CPU type when doing a GPU comparison can lead to invalid results. It's sad that AMD can't do the comparison with AMD CPU's, but I suspect the need is there to use Intel to insure there is zero possibility of a CPU bottleneck affecting the results and to present benchmarks consistent with the rest of the industry. If AMD had the CPU performance crown, they could use their CPU's for the comparison instead since their CPU's would then insure zero CPU bottleneck.

Just because they used Intel CPU's does not have any bearing on how well their CPU's work for gaming. You need a proper CPU benchmark comparison test for that. In that case you would reverse the above and keep the GPU constant while changing the CPU.
 
Last edited:
There isn't anything wrong with it. It is however a sad state for them to be in. I do hope things pick up for them eventually. Not getting my hopes up too high though.
 
Has anyone put those AMD graphs to shame yet?
 
True Story - Using AMD got me kicked out of my WoW Guild.

I was running an AMD cpu, a new build, I think it was the 1090t with a video card I cannot remember. This was around 2010 or 2011. Anyways, I was in a high-end guild doing heroics ( now mythic ) and my frame rates was causing all kinds of shit during boss pulls. They basically had to pull me from raid and then I was moved to inactive and so I basically was forced to quit the guild.

I had to sell the system and rebuild using intel. A 2600 or a 920 or 940 .. I can't remember. Long story short. I saw my frame rates double and triple.

AMD is a horrible gaming CPU. I don't care what people say.
 
Last edited:
Vulkan (API) - Wikipedia, the free encyclopedia

Vulkan is intended to provide a variety of advantages over other APIs as well as its spiritual predecessor, OpenGL. Vulkan offers lower overhead, more direct control over the GPU, and lower CPU usage.[11] Intended advantages include:

  • Reduced driver overhead, reducing CPU workloads.[19]
  • Reduced load on CPUs through the use of batching, leaving the CPU free to do additional computation or rendering than otherwise.[20]
  • Intelligent and even CPU scaling for multi-core CPUs, which are by far the majority type of CPU on the market. Previous APIs like DirectX 11 and OpenGL 4 were initially designed for use with single-core CPUs and could not easily use multiple or scale the workload evenly across them, leaving some CPU cores underworked and some not even utilized at all, resulting in performance issues and frequent CPU bottlenecks.[21]
  • OpenGL uses the high-level language GLSL for writing shaders which forces each OpenGL driver to implement its own compiler for GLSL that executes at application runtime to translate the program's shaders into executable code for the target platform. Vulkan instead provides an intermediate binary format called SPIR-V (Standard Portable Intermediate Representation), analogous to the binary format that HLSL shaders are compiled into in DirectX. By allowing shader pre-compilation, application initialization speed improved and a larger variety of shaders can be used per scene. This also removed the need for GPU vendors to implement GLSL compilers, resulting in easier driver maintenance, and smaller driver packages in theory (as GPU vendors still have to include OpenGL/CL).[22]
  • Unified management of compute kernels and graphical shaders, eliminating the need to use a separate compute API in conjunction with a graphics API.
  • Cross-platform API supported on both mobile devices and high-end graphics cards.
  • OS agnostic to improve the portability of applications created using the API.
 
SixFootDuo said:
True Story - Using AMD got me kicked out of my WoW Guild.
I was running an AMD cpu, a new build, I think it was the 1090t with a video card I cannot remember. This was around 2010 or 2011. Anyways, I was in a high-end guild doing heroics ( now mythic ) and my frame rates was causing all kinds of shit during boss pulls. They basically had to pull me from raid and then I was moved to inactive and so I basically was forced to quit the guild.
I had to sell the system and rebuild using intil. A 2600 or a 920 or 940 .. I can't remember. Long story short. I saw my frame rates double and triple.
AMD is a horrible gaming CPU. I don't care what people say.
Click to expand...

Just don't forget to download a little less smut over torrent while you are raiding. Maybe remove some mallware will work good as well ;)
 
Lorien said:
The real sad part is that [H] hasn't kept up with the rest and doesn't have performance tests of their own so they are left talking about the small print on an AMD article. Others already beat them to the punch and have found truly interesting things like Nvidia gaining NOTHING from Vulkan which is by far more interesting to discuss.

index.php
Click to expand...

maybe because async is not on on Nvidia hardware, also it's mostly older cards that would get better performance according to id soft.
 
Nvidia has been optimizing all three api's, DX, OpenGL and Vulcan. That fact that Vulcan puts it right on par with DX now should tell you its doing very well. A-sync or not. When's the last time that happened?
 
Barton anyone? That's the last I remember that was a better option than Intel... I liked my 2500+
 
SixFootDuo said:
True Story - Using AMD got me kicked out of my WoW Guild.

I was running an AMD cpu, a new build, I think it was the 1090t with a video card I cannot remember. This was around 2010 or 2011. Anyways, I was in a high-end guild doing heroics ( now mythic ) and my frame rates was causing all kinds of shit during boss pulls. They basically had to pull me from raid and then I was moved to inactive and so I basically was forced to quit the guild.

I had to sell the system and rebuild using intil. A 2600 or a 920 or 940 .. I can't remember. Long story short. I saw my frame rates double and triple.

AMD is a horrible gaming CPU. I don't care what people say.
Click to expand...
Your issues had ABSOLUTELY ZERO to do with the CPU. Motherboard, drivers, GPU, network drivers, maybe even virus or malware.

My fiance can max WoW at 100+ fps with Rig#2 in my sig. It too is a Thuban 6 core like yours. Yes an Intel chip would run it faster, but the AMD chip should still have run it well enough.
 
CaptNumbNutz said:
Your issues had ABSOLUTELY ZERO to do with the CPU. Motherboard, drivers, GPU, network drivers, maybe even virus or malware.

My fiance can max WoW at 100+ fps with Rig#2 in my sig. It too is a Thuban 6 core like yours.
Click to expand...

I imagine it had more to do with the netcode and the way it handled the "very new" 6 core cpu's. To blame AMD is kind of a cop out but understandable at the time.
 
CaptNumbNutz said:
My fiance can max WoW at 100+ fps with Rig#2 in my sig. It too is a Thuban 6 core like yours. Yes an Intel chip would run it faster, but the AMD chip should still have run it well enough.
Click to expand...

Yeah WoW still works best on high IPC cpu that Thuban prolly is very hard to beat if it is above(ish) 4.5 ghz for most if not all Piledriver cpu out there ....
 
CaptNumbNutz said:
Your issues had ABSOLUTELY ZERO to do with the CPU. Motherboard, drivers, GPU, network drivers, maybe even virus or malware.

My fiance can max WoW at 100+ fps with Rig#2 in my sig. It too is a Thuban 6 core like yours. Yes an Intel chip would run it faster, but the AMD chip should still have run it well enough.
Click to expand...

Bullshit.

I kept the same hardware. The same video card, ssd, memory. Only thing I changed was the CPU. Anyone here will tell you Intel can double your frame rates in World of Warcraft over AMD at least as far as the 1090t I had.

Even the settings in WoW were the same. In fact, if you go back and look at posts I made here on HardOCP back in 2010 or 2011 you can see where I talked about the ordeal.

So if the CPU didn't have anything to do with it, you're saying the chipset and motherboard doubled my frame rates?

lol
 
You must log in or register to reply here.
Back
Top