Medal Of Honor Warfighter CPU Performance

humbug · 2 Nov 2012 at 12:45

http://www.tomshardware.co.uk/medal-of-honor-warfighter-performance-benchmark,review-32557-7.html

Well i never......

Is this AMD working with EA Game developers?

Speed · 2 Nov 2012 at 12:49

Seems to like more true cores, although I don't see how this is a proper test without an i5 in the mix somewhere. It goes from low end to extreme high end with the Intel processors.

humbug · 2 Nov 2012 at 12:51

Speed said:
Seems to like more true cores, although I don't see how this is a proper test without an i5 in the mix somewhere. It goes from low end to extreme high end with the Intel processors.

I know and agree, but; there is a hexa-core SB-E in there.... :eek:

And the old Athlon beats the SB Pentium.

Martini1991 · 2 Nov 2012 at 12:55

Those results don't make any sense.

GPU limited from the Pentium to the 3960k, but the FX8350 is ahead outside of acceptable margin of error?

Speed · 2 Nov 2012 at 12:56

The FX-8350, isn't that a true 8 core processor though? Where as the i7-3960X is 6 cores with hyper threading. It would appear that the game engine prefers true cores, not hyper-threading.

humbug · 2 Nov 2012 at 12:59

Martini1991 said:
Those results don't make any sense.

GPU limited from the Pentium to the 3960k, but the FX8350 is ahead outside of acceptable margin of error?

Is it GPU limited? people used to think that about BF3 until they found 2500K's bottle-necking SLI / CF setup's.

Martini1991 · 2 Nov 2012 at 12:59

Speed said:
The FX-8350, isn't that a true 8 core processor though? Where as the i7-3960X is 6 cores with hyper threading. It would appear that the game engine prefers true cores, not hyper-threading.

Wouldn't matter.
You're GPU limited from a Pentium, and then the 3960k with its 4 more cores each clocked 1GHZ + higher with an extra 6 threads from HT is pretty much the same?

The 6 cores of the 3960k at 4.25GHZ would smash the 8 cores at 4GHZ in the FX8350.

Price aside obviously, the FX8 is much better value/money.

humbug said:
Is it GPU limited? people used to think that about BF3 until they found 2500K's bottle-necking SLI / CF setup's.

Well, a Pentium and 3960k giving similar results? Yeah it's GPU limited.
Unless of course, you can't read graphs

Unless of course you think an FX8350 is better than a higher clocked 3960k?
Both have pretty much identical instruction sets.

humbug · 2 Nov 2012 at 13:02

Speed said:
The FX-8350, isn't that a true 8 core processor though? Where as the i7-3960X is 6 cores with hyper threading. It would appear that the game engine prefers true cores, not hyper-threading.

True 8 core? pass....

it has 8 integer cores, it only has 4 L2 cache threads (2 per core) so while it has 8 cores those cores have to share 4 threads.

Martini1991 · 2 Nov 2012 at 13:03

http://www.anandtech.com/bench/Product/443?vs=697

When the 3960 has a 700MHZ clock disadvantage, in the graph the 3960 has a 250MHZ lead.

The FX8350 doesn't win a single benchmark.

My only rational explanation is it's an outlier/wrong result, not because I'm bashing AMD, but it's a result that doesn't make any sense.

humbug · 2 Nov 2012 at 13:04

Martini1991 said:
Well, a Pentium and 3960k giving similar results? Yeah it's GPU limited.
Unless of course, you can't read graphs

Unless of course you think an FX8350 is better than a higher clocked 3960k?
Both have pretty much identical instruction sets.

Avg FPS are 30% lower, i'd be willing to bet thats low Min FPS which is usually where the CPU is the weak link.

Martini1991 · 2 Nov 2012 at 13:06

The average frame rate difference is 1FPS, that isn't 30%.
The minimum FPS is different by a fair amount (Probably is 30%)

Also, I don't see the Athlon II besting the pentium (In minimum FPS it does)

I think you're reading this graph wrong.
At any rate, you've made this thread to be pro AMD, but the result doesn't make any sense, unless of course you disregard all reasoning

humbug · 2 Nov 2012 at 13:09

Martini1991 said:
The average frame rate difference is 1FPS, that isn't 30%.
The minimum FPS is different by a fair amount (Probably is 30%)

Also, I don't see the Athlon II besting the pentium (In minimum FPS it does)

Pentium G860 = Avg 41FPS

i7 3960X = Avg 59FPS

I think you're reading this graph wrong.
At any rate, you've made this thread to be pro AMD, but the result doesn't make any sense, unless of course you disregard all reasoning

Don't start getting ****y again because you don't like the thread :rolleyes:

Martini1991 · 2 Nov 2012 at 13:09

That's minimum FPS.
L2 Legend

Not sure with the swearing, I wasn't saying anything, it's an outlier result, anyone can see that.
I was also saying the "Disregard as tongue in cheek, hence the

.

Don't be losing your cool when rationale thinking comes along.

humbug · 2 Nov 2012 at 13:12

Martini1991 said:
That's minimum FPS.
L2 Legend

That's what i said, Minimum FPS (Usually CPU bound)

Martini1991 · 2 Nov 2012 at 13:13

humbug said:
That's what i said, Minimum FPS (Usually CPU bound)

No you didn't.

Pentium G860 = Avg 41FPS

i7 3960X = Avg 59FPS

Avg FPS are 30% lower

The average FPS difference is 1 FPS, ergo GPU bound (The minimums are CPU bound, but again, you have the Athlon II with a lesser instruction set and quite a fair bit lower average FPS with a fair higher minimum, the pentium will always be the better CPU out of the two, doesn't make sense.)

humbug · 2 Nov 2012 at 13:15

Martini1991 said:
No you didn't.

humbug said:
Avg FPS are 30% lower, i'd be willing to bet thats low Min FPS which is usually where the CPU is the weak link.

Martini1991 · 2 Nov 2012 at 13:15

The average FPS isn't 30% lower, it's 1 FPS.

humbug · 2 Nov 2012 at 13:20

Martini1991 said:
The average FPS isn't 30% lower, it's 1 FPS.

So it is, Min FPS is 30% lower. Not Avg, your right.

If anything it just proves my point, i speculated the Min FPS would be lower; and they are, just because its only 1FPS lower on the Avg does not mean its wrong, like with BF3 that Min FPS could be quite occasional thus not making much difference to the Avg.

I'm also not saying its right, but your analysis is no prof its wrong.

Martini1991 · 2 Nov 2012 at 13:24

My analysis isn't wrong.
Unless of course you find completely inferior CPU's getting higher frames logical and rational (Outside of the margin of error), which would be disregarding reasoning.

Also, you were wrong with what you were saying, I already know and agree minimums are CPU bound, but you kept blabbing on about averages being 30% lower and they weren't.
In post 11 I understood where your 30% came from.

The minimums you were right with, but it's a bit of a stretch to say you speculated on something with the figures staring you in the face, with an already accepted thesis for many years

humbug · 2 Nov 2012 at 13:31

Martini1991 said:
My analysis isn't wrong.
Unless of course you find completely inferior CPU's getting higher frames logical and rational (Outside of the margin of error), which would be disregarding reasoning.

Thats just a blanket statement based on your eye's only.

The Min FPS in a benchmark does not necessarily result in significantly lower Avg FPS, it depends on how often it dips that low and in a 2 minute bench it unlikely to do that a lot.

I'm not saying this result is false or correct. its a surprising result to me but thats about all until i see actual evidence it is wrong, not just your opinion.