Techspot does BattleField 4 CPU and GPU Comparison

[AliG]

Hey guys,

Since there seems to be some weird BF4 numbers floating around, I thought I'd post a review from a reputable site.

http://www.techspot.com/review/734-b...-4-benchmarks/

Unfortunately XS doesn't recognize the image types, but here are the key results:
1080p DX11 Performance
1440p DX11 Performance

Feel free to make your own conclusions, but I think it looks like AMD got their money's worth from their BF4 collaborations

[Greg83]

and more then 20% better then titan, with mantle to come. wow
what a beast

[Tim]

Phenomenal performance for Hawaii.

CPU wise, it's shocking to think it's nearly 2014 and reviewers still don't understand CPUs need to be reviewed showing minimum frame rates.

[kromosto]

what i get from this results is if you want more performance buy another 7970ghz.

[radier]

Who cares about Single Player performance ?

For now AMD sux in Multi real bad.

[Tim]

Who cares about Single Player performance ?

For now AMD sux in Multi real bad.

First I've heard of that. Why would that be? Seems odd, same engine, same settings right?

[radier]

SP vs MP performance is different story.

MP needs top of the notch CPU while SP can be handeled with any modern CPU.
Also pclab tests already shown how bad AMD cards work in MP.

[LordEC911]

SP vs MP performance is different story.

MP needs top of the notch CPU while SP can be handeled with any modern CPU.
Also pclab tests already shown how bad AMD cards work in MP.

Just gloss over the fact that they did all those test in one day.

[Tim]

That shouldn't make any difference, surely. Same engine, same settings. I don't see why multiplayer would do worse. Maybe a different map, but I highly doubt that would make such a performance decrease. Review seems flawed tbh, look at the performance in the techspot review.

[radier]

@LordEC911
With two benching setups they had it's not a big problem.

@Tim
No minimum FPS? Techspot review is a joke

This is how proper CPU test result should look like:

[vario]

Who cares about Single Player performance ?

For now AMD sux in Multi real bad.

Uhm, i do.Most people i know cares.That type of generalization is kinda silly.
On the topic however, seems like AMD had a bug in drivers, just released drivers to testers for 290 reviews (which is why they pushed the NDA back) solves some problems like multi in BF4.
There are already other results than those from pclab.pl which have WEIRD disparity between single and multi results.

[radier]

This "WEIRD disparity" was already present in BF3.
Did you play any of this games? I gues not.

[LordEC911]

@LordEC911
With two benching setups they had it's not a big problem.

I see a lot more than 2 setups...

[tbone8ty]

@LordEC911
With two benching setups they had it's not a big problem.

@Tim
No minimum FPS? Techspot review is a joke

This is how proper CPU test result should look like:

Link for that review please thanks

Edit

Found it
http://www.bf4blog.com/battlefield-4...pu-benchmarks/

[informal]

I've never really bothered to read "pclab.pl" reviews. Always different results from other well known review websites, seemed to me very fishy.

[tbone8ty]

Also we should consolidate bf4 performance reviews and maybe move to PC gaming.

[radier]

@informal
Maybe because focus from pclab knows how to choose CPU and GPU intensive places in different games.

[nossy23]

Much better benchmarks can be found here:


http://www.sweclockers.com/artikel/1...-battlefield-4


Use Google translate if you don't understand the results.

[tajoh111]

I am just thinking out a loud here but, doesn't techspot only test for single player results and if we match PClabs single player results, they mirror techspots.

And if I was going to do graphics card testing, particularly if I wanted to do it quickly, I would test single player and multiplayer testing in the same session so I wouldn't need to remove the graphics cards twice as many times and twice as many driver installations.

Thus if there were something going wrong with the testing methodology for PClabs, wouldn't multiplayer problems manifest in single player problems?

But because the single players results are okay in the PClab review and AMD performs better just like the techspot review, people accept them. However even those it is from the same benching suite, same testing session, the multiplayer results aren't as favorable for AMD, people discard them.

[vario]

This "WEIRD disparity" was already present in BF3.
Did you play any of this games? I gues not.

Well you guessed bad.And what it has to do with anything if i had or had not ...

But its pretty clear you just one of those with attitude that ALWAYS knows better.
Im not saying pclab.pl results are WRONG, far from it, ive read multiple times their different reviews.BUT im saying different people that made such benchmarks got different results, which means probably disparity with driver versions.And or different setup that punishes for some reason ati cards, which isnt necessarily good on average.

[filmbot]

CPU wise, it's shocking to think it's nearly 2014 and reviewers still don't understand CPUs need to be reviewed showing minimum frame rates.

Just curious, why would minimum frame rates need to be included? Trying to come up with a reason but just can't at the moment.

[FlanK3r]

SP vs MP performance is different story.

MP needs top of the notch CPU while SP can be handeled with any modern CPU.
Also pclab tests already shown how bad AMD cards work in MP.

I read few reviews here. Its many work, but still there are some mistakes.

[blindbox]

Just to clear things up:

- pclab.pl benchmarked the multiplayer, showing the 780 beating R9 290x.
- sweclockers.com benchmarked the multiplayer, showing R9 290x beating the 780 (and the Titan).

Gonna need more reviewers to test the game out before we can draw any conclusions to who's right and who's wrong.

EDIT: Also, notice the graphics quality seems to affect who's the winner. Drawing the above conclusions based on the highest settings they benched at.

EDIT2: Checked, and checked again. See the difference in settings between these two.

http://www.sweclockers.com/artikel/1...d-4/3#pagehead
http://pclab.pl/art55318-4.html

sweclockers had r9 290x ahead of GTX 780 by a little bit at 1920x1080 ultra 4x MSAA
pclab.pl had r9 290x a little bit behind gtx 780 at 1920x1080 ultra (no msaa probably, lower slides indicate they'd tell if 4x MSAA is enabled)

Pclab.pl does not have any 2560x1440/1600 results.
sweclockers showed r9 290x beating titan at 2560x1440 ultra 4x msaa

I think both sites are correct.

[kromosto]

and i benched 7970 ghz it beats neither titan nor 290x .

i think we have to wait more benchmarks to see what is going on. jumping to conclusion with one benchmark may also lead to wrong results.

[blindbox]

and i benched 7970 ghz it beats neither titan nor 290x .

i think we have to wait more benchmarks to see what is going on. jumping to conclusion with one benchmark may also lead to wrong results.

Hope you checked my above edit. I got the habit of jumping the gun, only to re-edit it later after finding out more. Do we question why 1080p + MSAA or 1440p/1440p + MSAA are not benched in pclab.pl? You're right though, need more benches.