AMD Phenom II 920 & 940 Full Review

[Jochenp]

Don't even try playing that game with Gosh.
He's just an arrogant troll, I'm still wondering why he is still here.
You can start talking to your walls, that kind of conversation equals to a conversation with the "knowledgeable" gosh.

Ignore, and STAY ON TOPIC.

Too bad, I hoped to actually learn something from this thread
I'll take your advice

[moiraesfate]

roofsniper, stop it. There is no reason to behave the way you are.

I'm with qurious, please don't let this thread be closed. I suggest deleting the rude posts and letting the thread itself continue with a pm warning to those continuing the misbehavior.

overclocker, may I suggest creating a new thread with the current results when they're done? Thanks for your efforts. I can't wait to see the numbers.

[roofsniper]

To the Mods. It appears that some people are not liking the Deneb results that are being presented on this thread, so they are doing everything possible in getting this thread closed again. It would be a shame that they get their way, please don't let them succeed

it appears that some people are trying to make rational decisions yet the same people keep popping up every single time saying that they are trolls. many people have said that the q6600 is better than the phenom II but you have never had a problem with that but when people say that phenom II doesn't suck and its alright thats when all hell breaks lose.

[roofsniper]

roofsniper, stop it. There is no reason to behave the way you are.

I'm with qurious, please don't let this thread be closed. I suggest deleting the rude posts and letting the thread itself continue with a pm warning to those continuing the misbehavior.

overclocker, may I suggest creating a new thread with the current results when they're done? Thanks for your efforts. I can't wait to see the numbers.

yes i really want the thread to go in a good direction but the fact is that many people are saying things that are not true and whenver someone says something good about phenom II they get flamed or someone disagrees with them. as i recall this is a phenom II review thread so i think it would only be fair if we could talk about phenom II instead of talking about how amd users are crying because they don't like deneb results when in fact they are just trying to say some simple things about deneb and how some results look good. now if i can actually make one post about phenom II without having to get flamed for it then all will be good. but i think its total bs that everytime i try to say something rational that a load of people come up and have nothing to say but intel is better. i don't want to start anything or act like a troll but it seems that somehow just commenting about the review and the cpu in a positive way can create an uproar.

[TL1000S]

Windows X64

This is an important aspect.

In the review Vista SP1 is listed as OS for *all* rigs.
As this may be 32-bit (and I usually take that as default) you *must* document this.
And preferably show actual screenshots from the benches.

In the norwegian hardware forum (diskusjon.no) someone doubts the credibility of the entire set of tests mainly due to the very bad results for PII in Cinebench10.
Care to elaborate?

[qurious63ss]

it appears that some people are trying to make rational decisions yet the same people keep popping up every single time saying that they are trolls. many people have said that the q6600 is better than the phenom II but you have never had a problem with that but when people say that phenom II doesn't suck and its alright thats when all hell breaks lose.

See that's the problem. You have to show proof to backup those rational decisions or whatever else your trying to prove. I asked Overclocker a few pages back why his results for Cinebench were so low on Deneb as compared to what Anand got when he tested Agena. I had proof and showed him the numbers and he politely answered that it was because different OSs were used.

[roofsniper]

See that's the problem. You have to show proof to backup those rational decisions or whatever else your trying to prove. I asked Overclocker a few pages back why his results for Cinebench were so low on Deneb as compared to what Anand got when he tested Agena. I had proof and showed him the numbers and he politely answered that it was because different OSs were used.

ok and results from other people who have reviewed it are not good? plus i believe no where in this thread there was proof of a q6600 being better than a phenom II yet that went on for pages. mostly the problem is that so many people are coming to conclusions so quickly without seeing many important things like more real world results, power consumption and pricing. without knowing everything you can't just say something is going to be horrible. if someone has proof that deneb sucks and that it is worse than the core 2 quads then ok. but from the results i have seen that mattered, that people will actually use, deneb isn't looking to shabby with more results to come. so before the other results come out and before there are more reviews it just doesn't make sense to come to conclusions so fast. we haven't even seen oced results yet although many people have said that the competing intel quads are better based on the fact that they can oc to phenom II speeds. they still have more results coming and they are some things that they won't do that other reviewers will and we need those before making such quick judgments on things.

[OverClocker_gr]

This is an important aspect.

In the review Vista SP1 is listed as OS for *all* rigs.
As this may be 32-bit (and I usually take that as default) you *must* document this.
And preferably show actual screenshots from the benches.

In the norwegian hardware forum (diskusjon.no) someone doubts the credibility of the entire set of tests mainly due to the very bad results for PII in Cinebench10.
Care to elaborate?

Yes,the os is Vista x86 32bit.


Show actual screenshots?



Show to your friend these pics.(although i have uploaded few pages back, Phenom II 940 @ x64 + 9550 @ x64)

pii 940 + 9550(actual is a qx9650 @ 9550 clocks so there is no difference )

[qurious63ss]

ok and results from other people who have reviewed it are not good? plus i believe no where in this thread there was proof of a q6600 being better than a phenom II yet that went on for pages. mostly the problem is that so many people are coming to conclusions so quickly without seeing many important things like more real world results, power consumption and pricing. without knowing everything you can't just say something is going to be horrible. if someone has proof that deneb sucks and that it is worse than the core 2 quads then ok. but from the results i have seen that mattered, that people will actually use, deneb isn't looking to shabby with more results to come. so before the other results come out and before there are more reviews it just doesn't make sense to come to conclusions so fast. we haven't even seen oced results yet although many people have said that the competing intel quads are better based on the fact that they can oc to phenom II speeds. they still have more results coming and they are some things that they won't do that other reviewers will and we need those before making such quick judgments on things.

Agreed, so when more results come out and more importantly people get there hands on this chip then we will see if Overclockers numbers were correct or not, but until then let's give him the benefit of the doubt and if you dispute his numbers then call him on it but only if you have data to back it up.

[roofsniper]

Agreed, so when more results come out and more importantly people get there hands on this chip then we will see if Overclockers numbers were correct or not, but until then let's give him the benefit of the doubt and if you dispute his numbers then call him on it but only if you have data to back it up.

oh no im not disputing his numbers. i know for a fact that most of his numbers are correct. i have seen some better results at stock from others but im sure that was because of a different configuration. the results here are good but theres just not many that matter to real world performance and theres more that i want to see which he is working on now. i just think its good to have more reviews because different reviews have different configurations and different chips. if i can see power consumption, x64 results, and how well it can oc plus a few more real world benchmarks then in my books its a perfect review.

[OverClocker_gr]

for those that they have doubts about this review,i am just quoting what gorillakos said here

Are you serious with the things you are saying? ? ?

What it is most possible? A small "credible" site or a huge "credible" site?

Think before you post, and if you wish i am paying you the tickets to come to Greece and review with us, and let us know again if we are making real reviews or bs .

[qurious63ss]

oh no im not disputing his numbers. i know for a fact that most of his numbers are correct. i have seen some better results at stock from others but im sure that was because of a different configuration. the results here are good but theres just not many that matter to real world performance and theres more that i want to see which he is working on now. i just think its good to have more reviews because different reviews have different configurations and different chips. if i can see power consumption, x64 results, and how well it can oc plus a few more real world benchmarks then in my books its a perfect review.

So then if your problem is with the posters then ignore them, or let the Mods handle them.

[Drwho?]

SSE is Intel's icing on the cake.

Well, actually, let s readjust this one ... Since the beginning, SSE is 128bits, but since AMD had 64 bits, they got the /flavor:AMD to compile with 64 bits loads and stores with the Athlon64 ... when they moved to Phenom, they were victime of their short vision and the 128bits execution unit is used only half.(because their code path detection) AMD is now using Intel flavor.
On few test, you can expect the Phenom II performance to increase a little when AMD is done fixing the mess they did build for themselve.
Using movhps and movlps instead of movaps was not very smart, it was a nice way to slow down the Pentium 4, but that 's about it.

Give them the credit to lose few % with Phenom II because of this, it is not going to change the overall picture, but know that the processor Phenom I and II is not responsible for the short coming on SSE, it is AMD software enabling who messed up.

next time, when an instruction set is 128bits, don t use 64

This is my personal opinion.

[roofsniper]

Well, actually, let s readjust this one ... Since the beginning, SSE is 128bits, but since AMd had 64 bits, they got the /flavor:AMD to compile with 64 bits loads and stores with the Athlon64 ... when they moved to Phenom, they were victime of their short vision and the 128bits execution unit is used only used half.(because their code path detection) AMD is now using Intel flavor.
On few test, you can expect the Phenom II performance to increase a little when AMD is done fixing the mess they did build for themselve.
Using movhps and movlps instead of movaps was not very smart, it was a nice way to slow down the Pentium 4, but that 's about it.

Give them the credit to lose few % with Phenom II because of this, it is not going to change the overall picture, but know that the processor Phenom I and II is not responsible for the short coming on SSE, it is AMD software enabling who messed up.

next time, when an instruction set is 128bits, don t use 64

This is my personal opinion.

hopefully now that amd has changed around all the positions around the company and got a new ceo that they will actually think things through. i think they got very arrogant with k8 and though that they could do w/e they wanted. thats why they went on this quest for a true quad core and ended up screwing themselves. the problem was that they acted as if no matter what they did they could still come out over intel. they knew that someday they would need a true quad so they thought that hey k8 is doing good lets go for a true quad first instead of jumping into it slowly. they screwed themselves and could of made k10 amazing. hopefully with bulldozer they will actually think it through instead of going for the gold.

[Drwho?]

Don't get me wrong, i don t want to start a bashing of AMD here, i am just saying that I saw in few case, the Phenom to use a Code 64bit path (MMX) instead of the SSE code path, because it is detected as a latest generation of Athlon64, but does not have the upgraded code path for 128 bits, and I think and it deserve to be known. It is not going to improve much, and I am sure AMD is working diligently to fix this.

[demonkevy666]

Don't get me wrong, i don t want to start a bashing of AMD here, i am just saying that I saw in few case, the Phenom to use a Code 64bit path (MMX) instead of the SSE code path, because it is detected as a latest generation of Athlon64, but does not have the upgraded code path for 128 bits, and I think and it deserve to be known. It is not going to improve much, and I am sure AMD is working diligently to fix this.

I want to know what you mean by intel flavor ? lol

I thought that was because it go's two 64 bit ones for each 128 bit one.

[G.Foyle]

Yes,the os is Vista x86 32bit.
Show to your friend these pics.(although i have uploaded few pages back, Phenom II 940 @ x64 + 9550 @ x64)

pii 940 + 9550(actual is a qx9650 @ 9550 clocks so there is no difference )

This is an unfortunate choice of testing environment. Under Vista x64 Phenom II is marginally slower in Cinebench R10 x64 than Yorkfield (clock for clock; buck for buck it's much faster).

[OverClocker_gr]

This is an unfortunate choice of testing environment. Under Vista x64 Phenom II is marginally slower in Cinebench R10 x64 than Yorkfield (clock for clock; buck for buck it's much faster).

Only on native x64 applications phenom gains.And there are not many.

[Zucker2k]

hopefully now that amd has changed around all the positions around the company and got a new ceo that they will actually think things through. i think they got very arrogant with k8 and though that they could do w/e they wanted. thats why they went on this quest for a true quad core and ended up screwing themselves. the problem was that they acted as if no matter what they did they could still come out over intel. they knew that someday they would need a true quad so they thought that hey k8 is doing good lets go for a true quad first instead of jumping into it slowly. they screwed themselves and could of made k10 amazing. hopefully with bulldozer they will actually think it through instead of going for the gold.

You're only saying all these things with the benefit of hindsight. Assuming Intel didn't exist, you wouldn't be making that statement. What most fail to see is that Intel jumped leaps and bounds ahead with core/core2; it was a true revolution and we're still living it. Its been three years since Intel turned around the clock/clock deficit, and AMD still seems confounded by Intel's momentum. I mean that "fake" quadcore, Q6600, is still very strong and trades blows with AMD's best two years after the fact. That alone should give any doubters an idea about how revolutionary core2 was/is.

[TL1000S]

Only on native x64 applications phenom gains.And there are not many.

So true (and I also posted the same in diskusjon.no).
And thanks for clarifying the platform

Edit@Zucker2k:
No, Core/Core2 was not that of a revolution.
It was just Intel leaving a wrong "trail" (Netburst) and introduced revorked "Dothan-architecture" to desktop/server.

[roofsniper]

You're only saying all these things with the benefit of hindsight. Assuming Intel didn't exist, you wouldn't be making that statement. What most fail to see is that Intel jumped leaps and bounds ahead with core/core2; it was a true revolution and we're still living it. Its been three years since Intel turned around the clock/clock deficit, and AMD still seems confounded by Intel's momentum. I mean that "fake" quadcore, Q6600, is still very strong and trades blows with AMD's best two years after the fact. That alone should give any doubters an idea about how revolutionary core2 was/is.

/off topic
as calmatory has said before the reason why companies jump each other is because one fails to make their latest product. amd failed with k10 and thats why the core 2 quads overtook them. the core 2 idea is pretty amazing. who would of thought that just slapping two of your cpus that failed in the past together would of worked. if i remember correctly the first core 2 duo was made from two pentium M laptop cpus. and the core 2 quads were made from two of the core two duos. who would of thought that some crappy laptop cpus would work so well. it was a simple design that somehow worked and didn't take much time at all to do. i doubt it will ever work again but this set amd back. as we all know k10 can't take the performance crown right now but bulldozer is on its way and its a complete new design so anything is possible. the only thing that is pissing me off here is that even tho phenom II can't overtake the performance crown both phenom I and phenom II are nice cpus and anyone that has used one before can't doubt that.

[Chad Boga]

the core 2 idea is pretty amazing. who would of thought that just slapping two of your cpus that failed in the past together would of worked. if i remember correctly the first core 2 duo was made from two pentium M laptop cpus.

That is a bizarre interpretation of how C2D came about.

How was the Pentium M laptop chip a failure?

[gallag]

/off topic
as calmatory has said before the reason why companies jump each other is because one fails to make their latest product. amd failed with k10 and thats why the core 2 quads overtook them. the core 2 idea is pretty amazing. who would of thought that just slapping two of your cpus that failed in the past together would of worked. if i remember correctly the first core 2 duo was made from two pentium M laptop cpus. and the core 2 quads were made from two of the core two duos. who would of thought that some crappy laptop cpus would work so well. it was a simple design that somehow worked and didn't take much time at all to do. i doubt it will ever work again but this set amd back. as we all know k10 can't take the performance crown right now but bulldozer is on its way and its a complete new design so anything is possible. the only thing that is pissing me off here is that even tho phenom II can't overtake the performance crown both phenom I and phenom II are nice cpus and anyone that has used one before can't doubt that.

Make your 1000th post better than your 999th

[roofsniper]

Make your 1000th post better than your 999th

wow how about you grow up. ever single post i have ever made intel fanboys have flamed even if i have given intel praise or given amd praise. its almost as if i am being looked at differently just because my cpu is different. i think this is known as racism.good thing theres an ignore button so i can just skip over the posts of the people that have an iq below 100.

That is a bizarre interpretation of how C2D came about.

How was the Pentium M laptop chip a failure?

it was more of the pentium 4 as a whole.

[TL1000S]

OK. Back to results & facts.

Next relevant question about Cinebench10 (32-bits btw).

Why did "Overcklock gr" get so much better numbers for i920/i940 than for instance Xbitlabs?