• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Time Spy Standard DX 12 Bench.

flopper said:
Futuremark seems paid by Nvidia?
How else to explain this?
http://www.overclock.net/t/1605674/computerbase-de-doom-vulkan-benchmarked/220#post_25351958
Click to expand...

Silent_Scone said:
Because Pascal is-quite-good at a lot of things, perhaps.
Click to expand...

So a developer of Time Spy had this to day about Async and Maxwell.

http://steamcommunity.com/app/223850/discussions/0/366298942110944664/

So while Pascal can manage some ASync, Maxwell despite NVIDIA claims simply disables it to avoid having to even do it so it doesn't take a performance hit.


It was not tailored for any specific architecture. It overlaps different rendering passes for asynchronous compute, in paraller when possible. Drivers determine how they process these - multiple paraller queues are filled by the engine.


The reason Maxwell doesn't take a hit is because NVIDIA has explictly disabled async compute in Maxwell drivers. So no matter how much we pile things to the queues, they cannot be set to run asynchronously because the driver says "no, I can't do that". Basically NV driver tells Time Spy to go "async off" for the run on that card. If NVIDIA enables Asynch Compute in the drivers, Time Spy will start using it. Performance gain or loss depends on the hardware & drivers.
Click to expand...

Ultimately some AMD cards gain quite a bit (ie. they have a lot of shader units idling while rendering and they are very good at using them for the available paraller loads). Some AMD cards gain less or not at all (either less capable at paralleriziing, less idle shader units or no idle shader units at all - for example a HD 7970 is hard pressed to have any to "spare")

Some NVIDIA cards cannot do this at all. The driver simply says "hold your horses, we'll do this nicely in order". Some NVIDIA cards can do some of it. They might use another way than AMD (more driver/software based), but the end result is the same - the card hardware is capable of doing more through some intelligent juggling of the work.
Click to expand...
 
thebennyboy said:
The guy explained why he thinks it's not right, more than your "lol it's bs" argument.
Click to expand...

You know what. I'll take what a FM developer has to say as a pretty good counter. Perhaps you can help the tinfoil hat brigade who are apparently knee deep in decompiled code and are waiting to show FM up for the paid NV shills they apparently are.

I did wonder how long it would take Mahigan to pipe up with some damage control over this test. I guess AMD pay extra for weekends. (bigwinkysmiley.gif)

Edit: You would have be be properly mental to think FM would do something so utterly blatant (if you listen to the conspiracy theorists), it would kill their business stone dead as nobody would trust them ever again.
 
Last edited:
Kaapstad said:
Wrong thread.
Click to expand...



sorry wrong link, one below is for time spy 1.0

http://www.3dmark.com/spy/49632

1070 in sli on a 1440p monitor

5930k at stock 3.5ghz

drivers 368.69

mine is:

SCORE 8985 with NVIDIA GeForce GTX 1070(2x) and Intel Core i7-5930K

Graphics Score 10036
CPU Score 5641

or

3DMark Score8985
Graphics Score10036
CPU Score5641
Graphics Test 165.75 fps
Graphics Test 257.28 fps
CPU Test18.95 fps
 
flopper said:
Futuremark seems paid by Nvidia?
How else to explain this?
http://www.overclock.net/t/1605674/computerbase-de-doom-vulkan-benchmarked/220#post_25351958
Click to expand...

That is a very interesting post in that link, and basically says that 3dMark isn't actually using Async at all, but concurrancy. Well... that's a shame, and odd that they would call if async if true.

EDIT: Just saw that other steam link, about nvidia just saying in drivers it won't support async. More interesting reading :-) Pretty good stuff! keep it up!
 
Last edited:
Panos said:
In some unbalanced systems might. Look at this



It does affect the general score but not much. Look for example the Fury block.
Some general score differences are around 1.2%, while the CPU is almost double.
Click to expand...

Systems that crash a lot? Systems that are top heavy? Systems that are on the edge of a table or is this just what you see from a Reds perspective?
 
This thread has had to be tided up once already, keep it to what it is intended for. If you want to discuss the virtues of the benchmark then please start a new thread as not to clutter this one.
 
Had some trouble with the latest v5 versions of Sapphire Trixx Corrupting the default clocks and making it so I couldn't boot into windows (even though it is set to not start or set any clocks at bootup). Took a while to work out what was going wrong. I will stick with last version of the v4 line of trixx.

http://www.3dmark.com/spy/76198
Overall Score: 4454
GFX Score 4329
CPU Score 5334

Powercolour R9 290 1170/1600 (Elpida)
Catalyst 16.7.2 Hotfix
4790k @4800 (1.33v)(4700 uncore)
2x 8gb OverclockersUK Genesis memory (DDR3 2400 11-13-13-32-1T)
 
Last edited:
I'm unsure why they're called GPU scores on the result posts as there positioned based on the overall score not the gpu's.

Just something I took note of while having an internal grumble about a 390 placing above me in the results :)

And then I noticed it wasn't 1, gpu scores it was 1 gpu as in a single gpu score, tut tut will I ever learn :rolleyes:
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom