• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

【CPU Tests】Warhammer 40,000 Dawn of War III

csbin

Senior member
http://www.gamegpu.com/rts-/-стратегии/warhammer-40-000-dawn-of-war-iii-test-gpu

wd2_proz.png



wd2_intel.png



wd2_amd.png
 
Great results from Ryzen. Crazy progress over Piledriver. That's an 80% improvement per clock (assuming 3.8 vs 4.9 average turbo using the FX-9590).

And yet, barely faster than a more than 200.00 cheaper non-hyperthreaded i5 , and not even a k model at that. But yea, not that hard to improve over piledriver.
 
And yet, barely faster than a more than 200.00 cheaper non-hyperthreaded i5 , and not even a k model at that. But yea, not that hard to improve over piledriver.
You are just asking for us to turn that around on you. It's obvious, but I'll point it out for you anyway.

Of course, the same criticism can be leveled at the $320 6700, and I guess you'd rather have the $1000 5960X that gets 16% lower minimums than the same i5? 5 times the price, less consistent performance.
 
You are just asking for us to turn that around on you. It's obvious, but I'll point it out for you anyway.

Of course, the same criticism can be leveled at the $320 6700, and I guess you'd rather have the $1000 5960X that gets 16% lower minimums than the same i5? 5 times the price, less consistent performance.
It's all about perspective - too bad some people refuse to broaden theirs.
 
Sure would be nice to see a 7700k and a G4560 thrown into the mix. I have a feeling the G4560 would embarrass every CPU on the list for price/performance. I just built one for a client and I'm surprised at how snappy it is.
 
Sure would be nice to see a 7700k and a G4560 thrown into the mix. I have a feeling the G4560 would embarrass every CPU on the list for price/performance. I just built one for a client and I'm surprised at how snappy it is.

Just add a few percent to the i3-4330. It's the same clock speed.
 
Now this is interesting .. Ryzen giving a run of his money to Intel?


Doesn't look that good for Ryzen. There is no i5-7600k, no 4790k, no 6700k, no 7700k, no Broadwell-E.....only on par with an old Haswell i5-4670k. AMD is in a difficult situation later this year when Skylake-E and Coffee Lake come out, at least when it comes to gaming.
 
Doesn't look that good for Ryzen. There is no i5-7600k, no 4790k, no 6700k, no 7700k, no Broadwell-E.....only on par with an old Haswell i5-4670k. AMD is in a difficult situation later this year when Skylake-E and Coffee Lake come out, at least when it comes to gaming.

Somehow 1 FPS above a 2013 i7-4770K on min and 10% slower average FPS with twice the cores/threads, higher clocks by the top Ryzen SKU (much more expensive, lower OC headroom as well) is being celebrated as a win here. Hopefully GameGPU finds out its not (pre) 2015 anymore and updates their Intel CPUs. Not to mention the misleading numbers for this particular game, no way Skylake is behind Haswell.
 
Last edited:
Somehow 1 FPS above a 2013 i7-4770K on min and 10% slower average FPS with twice the cores/threads, higher clocks by the top Ryzen SKU (much more expensive, lower OC headroom as well) is being celebrated as a win here. Hopefully GameGPU finds out its not (pre) 2015 anymore and updates their Intel CPUs.
Hopefully Intel gives GameGPU a compelling reason to upgrade since Haswell and Haswell-E.🙄

I also hear people celebrate a transistor power-frequency curve at 0.7V as proof that CFL will smash records at high frequencies, so there's that.🙄
 
Somehow 1 FPS above a 2013 i7-4770K on min and 10% slower average FPS with twice the cores/threads, higher clocks by the top Ryzen SKU (much more expensive, lower OC headroom as well) is being celebrated as a win here. Hopefully GameGPU finds out its not (pre) 2015 anymore and updates their Intel CPUs.
Pretty embarrassing that the 6700 has no per clock advantage against that 2013 CPU (it even has a higher boost clock too). In fact, the 4770K is 5.7% ahead. What a win! Intel sure has come a long way since 2013.

But hey, while GameGPU is busy updating its Intel CPUs, maybe we could see some Ryzen 5s thrown in there as well. I wonder how a $250 1600X might match up to that $310 6700...
 
Last edited:

So this graph basically shows that (aside from some very poor IPC performers, ie. AMD FX) the higher the clock frequency, the better it performs with this game. On my second pass of these stats, I wondered why there aren't any Intel ~4GHz processors in there. For example, the 4790k. Initially I thought this might be because they were showing certain price bracket CPUs, but then we have a few i7s in the list, including more than one former flagship, and the 5960X is currently selling for over £1k UKP.
 
The game does seem to scale with HT or more cores very well so the Ryzen 5 1500X and 1600X should have similar performance to the Ryzen 7 1800X.
 
The game does seem to scale with HT or more cores very well so the Ryzen 5 1500X and 1600X should have similar performance to the Ryzen 7 1800X.

It doesn't seem to be using much more than 4 threads though; look at the 2500K vs 2600K. So you would have to overclock to get near the 1800X's score but certainly doable.

The 5960X is pretty funny; it's clearly running near 3 Ghz. Not because it can't without violating the TDP but because Intel. If you look at the GPU scores at 4.6 Ghz the 5960X gets 82 fps min, so that's pretty close to straight scaling.
 
How relevant is this CPU test? It isn't specific on what the actual test scenario is.

We tested the CPU dependency on 16 models of basic configurations that are relevant for today. The test was conducted in those places where the value of the video card for the game is minimal and its load was less than 99%, this time at a resolution of 1920x1080.

Test is pretty meaningless unless you test with high unit counts in an engagement during a late game situation. Those situations would also be GPU demanding so it doesn't seem like they tested there.

People who primarily play single player skirmish would also find tests with the AI relevant although those aren't directly enabled in the open beta (you could force them on by having the other players leave the game).
 
The game does seem to scale with HT or more cores very well so the Ryzen 5 1500X and 1600X should have similar performance to the Ryzen 7 1800X.
Actually if a game scales well with cores/threads, then one would expect the 1800x to be much faster than the 1500 or 1600. This particular game seems to use a lot of cores, but does not actually scale well with cores at all, as evidenced by the i5 6600 being very close to the 1800x, and also the low results for the 6950x. Just look at the Ryzen core usage. Two cores are at nearly 60%, while the other cores/threads are at about half that. Overall, it seems, as another poster said, all that is needed is four real cores with decent ipc and clockspeed. So as to your point about the 1500 and 1600, I would agree that if they can be overclocked to the level of the 1800x, they would probably perform similarly.
 
Ryzen is doing "OK", but obviously Intel still has an advantage here. With that said I'm happy with what AMD has managed to do to this point considering the behind the 8-ball condition they are in. MUCH better than the crap they called the FX and Zen + will hopefully be what they are saying it's supposed to be. If not, I'll just have to jump back on the Intel HEDT bandwagon for my next upgrade. I'm pulling for AMD, but not holding my breath.
 
AMD will rise from the ashes and destroy the terminators!!!

16 thread, 8 cores, at 65 watt tdp and for $320 bucks, what more can you ask for?
 
Back
Top