• 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.

Interesting BF 4 CPU usage [GameGPU.Ru]

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
i5 strikes back, so much for HT!

😀

Almost 20% (18%?) increase in IPC from SB to Haswell, or am I diluted/personally bias?
 
Last edited:
beginner99 said:
All in all this really makes the 780 look a lot more interesting than a 290x especially if you do not have an OCed HW.
Click to expand...

The cake is a lie:
bf4_gpu_high2_1920.png
 
You're in single player though 🙁

Compare the 780 cpu results to the R290X results in multiplayer...

bf4_cpu_radeon.png


Every cpu takes a big hit, the 8350 at 5GHz only manages 51 fps with a R290X.
 
BallaTheFeared said:
i5 strikes back, so much for HT!

😀

Almost 20% increase in IPC from SB to Haswell, or am I diluted/personally bias?
Click to expand...

17%ish, which is okay I guess.

HT averages about 5% in this test which is pretty minimal.


I'm curious at to which map and how many players they're running though. I'd imagine 64 player on a levolution map with lots of changes could use the extra threads.
 
ChronoReverse said:
17%ish, which is okay I guess.

HT averages about 5% in this test which is pretty minimal.


I'm curious at to which map and how many players they're running though. I'd imagine 64 player on a levolution map with lots of changes could use the extra threads.
Click to expand...

By your logic they >>removed<< some of their "optimizations for HT" going from beta build to release build.
 
Last edited:
its good showing for the 8350 but even then its basically right there with the slower clocked 4 core 4670k. oc them both and the 4670k leaves it behind while using quite a bit less power. heck even the 2500k would also pass up the 8350 with both oced. and again this is about as good as the 8350 can look.
 
Last edited:
Communism said:
Living in a bubble must be fun.
Click to expand...

Okay, let's take the results from the SAME site running the SAME test comparing the Beta and Retail. They weren't good enough to use the same GPUs each time but let's look at the performance deltas instead:

http://pclab.pl/zdjecia/artykuly/chaostheory/2013/10/bf4_beta/charts/bf4_cpu_geforce.png
Haswell no-HT to HT: 3.9%

http://pclab.pl/zdjecia/artykuly/chaostheory/2013/10/bf4/charts/bf4_cpu_geforce.png
Haswell no-HT to HT: 4.2%


Now WHO'S the one in denial? Are you going to try to explain this away with "optimizations"?

Face it, they just happened to have a workload that wasn't suited for it and between beta and retail, the difference is negligible.
 
Last edited:
Communism said:
Beta:
91.4/88.0 = 1.04195804196
Release:
91.4/88.0 = 1.03863636364
Click to expand...

You copy and pasted wrong but at least you managed to confirm my math? (ah you fixed the typo)

4.2% vs 3.9%

That's well within the margin of error.


Oh btw:
Beta Ivy Bridge: 2.9%
Release Ivy Bridge: 6.1%
Beta Sandy Bridge: 2.3%
Release Sandy Bridge: 4.7%



I could trumpet this as some conclusive evidence but I understand that the difference is actually not that huge.
 
Last edited:
ChronoReverse said:
You copy and pasted wrong but at least you managed to confirm my math? (ah you fixed the typo)

4.2% vs 3.9%

That's well within the margin of error.


Oh btw:
Beta Ivy Bridge: 2.9%
Release Ivy Bridge: 6.1%
Beta Sandy Bridge: 2.3%
Release Sandy Bridge: 4.7%



I could trumpet this as some conclusive evidence but I understand that the difference is actually not that huge.
Click to expand...

It's the comparison you chose. I didn't even notice that they aren't the same GPU, making your choice of that comparison pointless.

Also, the multiquote system on this forum is horribad, I'm definitly not wrestling with it next time, couldn't even get my response out in time.
 
The more important question. Will my core i5-2500 survive playing this game about 5-10hrs a week until sky lake comes out, or will this heavy use melt it before then? 🙂
 
The game is fantastically optimized, no question. Final build that is. Waaayyy better than bf3 was.

If you don't have the best CPU/gpu, aka 4960k and sli, lower your f'ing settings and accept your 4670k is not top of the line.
 
Communism said:
It's the comparison you chose.
Click to expand...

And it shows that your idea that improved optimizations going from beta to retail is reducing HT benefits significantly is bunk. So you have no case for your accusation that I'm living in a bubble for saying that this site's test may just happen to be one that doesn't use HT well.


Meanwhile, the first page still features another test on retail BF4 that shows HT having quite a bit of benefit. So there are some scenarios at least where HT is an advantage.
 
Last edited:
ChronoReverse said:
And it shows that the idea that optimizations going from beta to retail is reducing HT benefits significantly is bunk. So you have no case for your accusation that I'm living in a bubble for saying that this site's test may just happen to be one that doesn't use HT well.


Meanwhile, the first page still features another test on retail BF4 that shows HT having quite a bit of benefit. So there are some scenarios at least where HT is an advantage.
Click to expand...

So that's 5 times the goal posts have been moved, I'm interested to see how many more times they do.

The onus of proof has been on you the whole time bro, I'm just having fun with you.
 
Last edited:
Okay, name the goal posts.

You claimed that optimizations reduce benefits of HT. Proven wrong with the charts comparing beta to retail.

You claimed that it takes less optimization to use HT. I demonstrated it's wrong simply explaining what SMT is.

You claimed that Java can somehow have 100% boost with HT. This doesn't even make sense!
 
Last edited:
ChronoReverse said:
Okay, name the goal posts.

You claimed that optimizations reduce benefits of HT. Proven wrong with the charts comparing beta to retail.

You claimed that it takes less optimization to use HT. I demonstrated it's wrong simply explaining what SMT is.

You claimed that Java can somehow have 100% boost with HT. This doesn't even make sense!
Click to expand...

Lol, claiming charts aren't significant and then claiming them as proof. You be Heisenberging me bro.

Btw, onus is still on you bro.
 
Communism said:
Lol, claiming charts aren't significant and then claiming them as proof. You be Heisenberging me bro.

Btw, onus is still on you bro.
Click to expand...

No, I said the DIFFERENCES aren't significant. Reading comprehension.

3.9% versus 4.2% is a small difference. Hence insignificant. Therefore going from beta to retail did not change the HT performance increase significantly. Optimizing code did not reduce the HT boost in the case of BF4.


So I have backed up my refutations because you keep saying the onus is on me. You however have shown no proof and have studiously ignored my explanations instead of actually demonstrating that I've made a mistake.
 
Last edited:
Communism said:
The fact that the software even uses 4770k hyperthreading cores as real cores shows how terribly optimized the game is for Intel CPUs.

Hyperthreading is just putting two threads on the same resource pile (core).

However this makes sense since the entire Frostbite 3 engine is optimized for jaguar cores.

It is a console game running on DirectX 11 after all.
Click to expand...

Communism said:
Read my post again.

If the software was optimized for Intel, in a well threaded program (one that actually uses 5+ balanced load threads) the Hyperthreads would give negligible performance.

The fact that the Hyperthreads are giving such a significant significant performance boost means that the game is absolutely horribly optimized for Intel processors.
Click to expand...

ChronoReverse said:
Wait what. This isn't like back in the days of Pentium 4 where HT doesn't have real benefits. It's pretty hard to completely load the execution units of CPU cores like Haswell with a single thread 100% the time when running something as dynamic as gameplay.

Besides, we're seeing BF4 using 6 full cores to the max and 2 more lightly. That's pretty much a best case scenario for a HT-capable CPU.
Click to expand...



^^This is where you "lost the thread" of this thread^^

You deliberately used a different definition of "optimization" here, which is why I left in the first place.

Now you forgot that you purposefully used a different definition of "optimization" and completely lost your train of thought. I was having fun with it with my comments after that.

I left because arguing with anyone who's going to use grammatical tricks is just non-productive
 
Last edited:
Okay, let me confirm this.

Are you saying that with the optimization you're talking about, a program optimized for Intel will be faster if it's 4 threads instead of 8 threads, ASSUMING that the workload is parallizeable to 8 threads?


Think very carefully before you answer that.

[edit]On a 4 core, HT enabled chip of course
 
Last edited:
ChronoReverse said:
Okay, let me confirm this.

Are you saying that with the optimization you're talking about, a program optimized for Intel will be faster if it's 4 threads instead of 8 threads, ASSUMING that the workload is parallizeable to 8 threads?


Think very carefully before you answer that.

[edit]On a 4 core, HT enabled chip of course
Click to expand...

No need to repeat myself, I'll just point you to the same conversation I've been having with people since forever.

http://forums.anandtech.com/showthread.php?t=2329338&highlight=hyper

For Hyperthreading to show improved results over just the cores without hyperthreading, the thread would need to be optimized around a target with less compute resources [or just not optimized well at all]. Hyperthreading is just a second set of dispatches and cache (In ivy bridge and haswell, the cache will be scavenged and used for the primary thread if hyperthreading is turned off)

Hyperthreading doesn't have more compute resources of non-Hyperthreading, it merely allows a second thread to share the available compute resources.
 
Last edited:
You must log in or register to reply here.

TRENDING THREADS

Back
Top