AMD Phenom II 920 & 940 Full Review [UPDATED with more tests]

Quote:

---

Originally Posted by justapost

Thanks for the details overclocker_gr. Jsut to be sure, you ran two instances of orthos together with the CB10 multi core test. Correct?
Have you had C1E enabled in the bios?

---

4,nop

Guys, if you're going to talk about voltages, never ever consider what the software is telling you.
Simply put, it's wrong in most cases if not all.

Quote:

---

Originally Posted by BenchZowner

Guys, if you're going to talk about voltages, never ever consider what the software is telling you.
Simply put, it's wrong in most cases if not all.

---

That is something and others are trying to find an excuse that Phenom could consume less power
but interesting to note that PII/C2Q in Crysis uses 70+W more than CPU test while I7 uses 45W or less more on Crysis , i thought all 3D and CPU results should be linear , But i was wrong :)

excuse me?

If a mobo overvolts CPU, it consumes more power. Period. Thats a fact.
If a mobo undervolts CPU, it consumes less power. Thats another fact.

Quote:

---

Originally Posted by BenchZowner

Guys, if you're going to talk about voltages, never ever consider what the software is telling you.
Simply put, it's wrong in most cases if not all.

---

yea and really if all amd mobos are overvolting and this is not a fault then it should be taken into consideration. i find it odd that they all overvolt during load. my 9600 at stock on my m3a32 never overvolts.

Quote:

---

Originally Posted by roofsniper

yea and really if all amd mobos are overvolting and this is not a fault then it should be taken into consideration. i find it odd that they all overvolt during load. my 9600 at stock on my m3a32 never overvolts.

---

Quote:

---

Originally Posted by PetNorth

excuse me?

If a mobo overvolts CPU, it consumes more power. Period. Thats a fact.
If a mobo undervolts CPU, it consumes less power. Thats another fact.

---

guyz,read careful what benchzowner said.


Dont consider what you are seeing in windows or bios,its 99% inaccurate.

Take a DMM a measure the vcore ;)

Quote:

---

Originally Posted by PetNorth

excuse me?

If a mobo overvolts CPU, it consumes more power. Period. Thats a fact.
If a mobo undervolts CPU, it consumes less power. Thats another fact.

---

Yeap that is True , But changed Voltage in Bios then measured the Power consumption , this got nothing with the default Vcore the chip should run on, this just like i undervolt my CPU and say it consume less power on stock clocks You need to measure the real Vcore , my 3800X2 for example Stock Volt is 1.35V and i run it 1.2V all set in Bios but never measured and it consumes less power on custom settings but this got nothing to do with stock settings ,, It is similar to the fact that if you overclock your CPU it will run faster , next time we will ask power consumption test with CPUs at stock clocks and undervoltet to the max :D

Quote:

---

Originally Posted by PetNorth

excuse me?

If a mobo overvolts CPU, it consumes more power. Period. Thats a fact.
If a mobo undervolts CPU, it consumes less power. Thats another fact.

---

im assuming you are talking about current (as in amperes) being constant, then both statements are true.
edit: or current increases with voltage in the first statement and decreases in the 2nd, if i may add.

Quote:

---

Originally Posted by kemo

but interesting to note that PII/C2Q in Crysis uses 70+W more than CPU test while I7 uses 45W or less more on Crysis , i thought all 3D and CPU results should be linear , But i was wrong :)

---

Seems reasonable, the amount of work the i7 needs to do in games is much less than the amount of work it does with 8 threads of CPU intensive applications.

I was wondering what benches I can post of my chip. I am in the USA, And don't want to violate NDA.

Quote:

---

Originally Posted by charged3800z24

I was wondering what benches I can post of my chip. I am in the USA, And don't want to violate NDA.

---

if you didnt sign a NDA then you cant really violate it can you...

Quote:

---

Originally Posted by nunchucka

if you didnt sign a NDA then you cant really violate it can you...

---

HMM,, well I suppose not.. tho the person (president of my work) that gave it too me said I was alloud to post a picture or anything(noob). I though about doing some 3.0ghz bench's comparing my 9950 @ 3.0ghz. same BN HT and Ram settings. I only have the one board right now.. so I will have to swap CPUs to do so.

Quote:

---

Originally Posted by charged3800z24

I was wondering what benches I can post of my chip. I am in the USA, And don't want to violate NDA.

---

Could you see how high it'll go please, just to get an idea of what to expect thanks.

Quote:

---

Originally Posted by charged3800z24

HMM,, well I suppose not.. tho the person (president of my work) that gave it too me said I was alloud to post a picture or anything(noob). I though about doing some 3.0ghz bench's comparing my 9950 @ 3.0ghz. same BN HT and Ram settings. I only have the one board right now.. so I will have to swap CPUs to do so.

---

well if your boss signed an NDA for that chip and you post benches he could get in trouble afaik

yea id just hold off on the benchies. we only got a week until phenom II's release, show us what you got then.

Quote:

---

Originally Posted by kemo

That is something and others are trying to find an excuse that Phenom could consume less power
but interesting to note that PII/C2Q in Crysis uses 70+W more than CPU test while I7 uses 45W or less more on Crysis , i thought all 3D and CPU results should be linear , But i was wrong :)

---

the reason for that is simple, artificial cpu tests are stressing the cpu to the max (eg. prime with 8 threads).

Games on the other hand dont stress all cores on Ci7 so the PCU shuts down not needed parts. :yepp:

I hope AMDs next cpu has something similar to intel PCU.

Quote:

---

Originally Posted by BrowncoatGR

well if your boss signed an NDA for that chip and you post benches he could get in trouble afaik

---

He was being pretty picky about me posting info..even a screenshot or 2. He won't even tell me how he got it..I guess I could block the info like everyone else..and post a screen shot.

Edit: I am loading Vista 7 and want to see how it is. Here is all I did. I only ran a 1M Pi..So I can't say for sure how stable it is...

http://3800z24.info/Phenom/Deneb/Pi_1M_4ghz_small.JPG

Quote:

---

Originally Posted by OverClocker_gr

So,this is the follow-up of the first thread regarding the new Phenom II processor series review that we did @ hwbox.gr.
here is again the link
http://www.hwbox.gr/showthread.php?t=3189

---

All are fake !:ROTF: You can see again ! special , benchmark in game !
http://i279.photobucket.com/albums/k...md_ph2_940.png
example with Crysis Warhead.
benchmark in http://www.hwbox.gr/showthread.php?t=3189&garpg=22
1280x1024 : FPS>125FPS
1680x1050 : FPS~~24FPS
benchmark in http://guru3d.com/article/top-10-gam...ore-216-test/7

Quote:

---

system
Mainboard
ASUS X58 ROG edition Rampage II Extreme
Processor
Core i7 965 @ 3750 MHz
Graphics Cards
GeForce GTX 260 core 216 896 MB
Radeon HD 4870 1024 MB
Memory
3072 MB (3x1024MB) DDR3 1800 MHz OCZ
Power Supply Unit
1200 Watt
Monitor
Dell 3007WFP - up to 2560x1600

---

1280x1024 : FPS<50FPS
1600x1200 : FPS ~~32FPS
-----------------------------------------------------------
Why ? who will explain ? thanks !:rofl:

Quote:

---

Originally Posted by vietthanhpro

All are fake !:ROTF: You can see again !

---

Cause you are a little bit a ****.


We run crysis with the lower details.

At guru 3d they run it @ 1280*1024 + 2xAA + Gamer Details.
@1600*1200 which is a higher resolution than 1680*1050 , they run it with Gamer Details and 2xAA
We run it @ 1280*1024 without AA + Performance Details.
@1680*1050 we run it at Enthusiast without AA

Ok now?



write my answer here too

http://vozforums.com/showthread.php?t=146108

Quote:

---

Originally Posted by OverClocker_gr

Cause you are a little bit a ****.


We run crysis with the lower details.

At guru 3d they run it @ 1280*1024 + 2xAA + Gamer Details.
@1600*1200 which is a higher resolution than 1680*1050 , they run it with Gamer Details and 2xAA
We run it @ 1280*1024 without AA + Performance Details.
@1680*1050 we run it at Enthusiast without AA

Ok now?

---

what kind of VGA do you use ?:shrug:
AND you can bring somes pictures !:( about games benchmark:yepp:
OK now ?

don't worry about him i swear he is related to lo squartatore.

When do clock-per-clock comparisons arrive? I don't mean to be impatient, I'm just excited! :D

Quote:

---

Originally Posted by vietthanhpro

what kind of VGA do you use ?:shrug:
AND you can bring somes pictures !:( about games benchmark:yepp:
OK now ?

---

oh yes.

deneb
http://img229.imageshack.us/img229/11/shi11ub0.th.jpg

i7 @ 965 @ 940 turbo on speeds
http://img122.imageshack.us/img122/4489/sh1t2kv1.th.jpg

Quote:

---

Originally Posted by YukonTrooper

When do clock-per-clock comparisons arrive? I don't mean to be impatient, I'm just excited! :D

---

Maybe today(for us in greece :p:)

Quote:

---

Originally Posted by OverClocker_gr

oh yes.

deneb
http://img229.imageshack.us/img229/11/shi11ub0.th.jpg

i7 @ 965 @ 940 turbo on speeds
http://img122.imageshack.us/img122/4489/sh1t2kv1.th.jpg

---

oh ! in DX9c, i7 run faster:clap:
But I don't see Deneb's score in resolution 1680*1050 ! NDA?:(http://img229.imageshack.us/img229/11/shi11ub0.jpg
in i7 benchmark, you can reduce QPI speeds.