AMD Phenom II X4 Model 940 @ [H]

[Mr. Bluntman]

A thorought albeit blunt review. Good to know I switched to a Core 2 Duo when I did because Phenom II is certainly looking like a let-down clock-for-clock.

 

[gtg465x]

A) I'm idling right now at 2.63GB/4GB "used".

Damn son, you must have a virus or something . Vista idles at about 1.1 / 4GB on my system, and around 400MB of that gets freed up when I launch a memory intensive application.

Also, might want to take a look at THIS website. It was the first one that popped up when I searched Google, but you'll see pretty much the same results everywhere you look.

Like Kyle said, 2GB normally gives you slightly better performance in games due to the lower latencies. Get out n00b.

 

[MrWizard6600]

using a drive with a single 80 gig platter. nice.

 

[cannondale06]

Damn son, you must have a virus or something . Vista idles at about 1.1 / 4GB on my system, and around 400MB of that gets freed up when I launch a memory intensive application.

Also, might want to take a look at THIS website. It was the first one that popped up when I searched Google, but you'll see pretty much the same results everywhere you look.

Like Kyle said, 2GB normally gives you slightly better performance in games due to the lower latencies. Get out n00b.

Stalker Clear Sky and Warhead will hitch with just 2 gigs of ram. I had to add another gig on my previous pc because 2 wasnt cutting it.
this site even confirms what i was saying for Warhead. "As our benchmarks reveal, a smooth gameplay with 2 GiByte memory is possible on XP only, on Vista the screen freezes up to one second in some situations."http://www.pcgameshardware.com/aid,...ecial_and_general_info_about_the_game/?page=2

 

[Slash3]

I have a decent rig and some spare sticks of DDR2, I might do a quick Far Cry 2 2GB/4GB/8GB comparison later today. Kind of curious, myself. I know that my memory usage peaks quite high at various times, depending on what I'm doing, but that's on a system that's left on 24/7. On a fresh boot that jumps straight into a benchmark session, the 2GB might go slightly further. Although right now in Vista x64 I'm using an indicated 2.44GB and have little other than Firefox, AV and some IM clients running. It's my understanding that Vista grabs available memory to perform caching and reports it differently than XP did - leading to a higher indicated memory use on task manager vs XP, when the reality is that the usage is similar. I typically hover at about 1.1-1.2GB of used memory at a recently booted desktop.

In any event, I'll go get some popcorn.

Edit: Just to create more discussion, look at the interesting result PCGH gets with GTA4 and the Phenom II X4..

http://www.pcgameshardware.com/aid,672218/Reviews/AMD_Phenom_II_X4-_Deneb_reviewed/?page=4

 

[MrWizard6600]

I have a decent rig and some spare sticks of DDR2, I might do a quick Far Cry 2 2GB/4GB/8GB comparison later today. Kind of curious, myself. I know that my memory usage peaks quite high at various times, depending on what I'm doing, but that's on a system that's left on 24/7. On a fresh boot that jumps straight into a benchmark session, the 2GB might go slightly further. Although right now in Vista x64 I'm using an indicated 2.44GB and have little other than Firefox, AV and some IM clients running. It's my understanding that Vista grabs available memory to perform caching and reports it differently than XP did - leading to a higher indicated memory use on task manager vs XP, when the reality is that the usage is similar. I typically hover at about 1.1-1.2GB of used memory at a recently booted desktop.

In any event, I'll go get some popcorn.

Vista Prechaches the crap out of everything. Its constantly moving stuff in and out of memory, aggressivly, but that doesnt mean that for some reason its not actually using that memory.

 

[Slash3]

Vista Prechaches the crap out of everything. Its constantly moving stuff in and out of memory, aggressivly, but that doesnt mean that for some reason its not actually using that memory.

Right. What I meant was that in XP memory usage appears to be lower than it actually is, because not all allocated cache is presented in task manager, whereas in Vista it is (with few exceptions).

It makes it look like Vista is hogging much more memory, and while it is true that the actual working footprint is higher, it's also reported more accurately.

Doesn't necessarily have much to do with the current discussion though, so I'll stop talking about XP vs Vista in that context.

 

[marzzes]

Hahahaha………. Jesses Kyle, brutal.

 

[ALL4AMD]

Poor AMD. and here i built a P1 to help them out



P.S. I too think the amount of ram needs to be kept the same.

 

[Atech]

Poor AMD. and here i built a P1 to help them out



P.S. I too think the amount of ram needs to be kept the same.

That was a stupid choice, always go for you OWN needs.

Nice read, no surprizes for me...not even the whining of the Phaliure2® fannny girls

 

[jonneymendoza]

"the damn Arabs" is that some sort of racist remark?

This is one of the worst reviews I have ever read on [H] and one of the worst reviews I have read in general. And you arguing that different amounts of memory on each platform don't matter, wow. The difference might not matter, but it may. As pointed out, a quality review removes as many variables as possible. It's common sense and good practice.

But I expect you to pull a fit, probably ban me, use several 4 letter words and claim how you know better than the "payed off" sites that are showing Phenom II is a much more favourable light.

LOL thats not a racist remark. come on man lighten up! his saying that in a kind of good sarcastic way

 

[MrWizard6600]

nice to know someone agrees with me...

even if he is all4AMD

Anyways, pulled this off guru3d, its not insignificant:

normally when you see these CPU "gaming" benchmarks, they put out identical numbers at higher resolutions when the load is (by percentage) shifted to the gpu, but in COD4's case (and thus COD5's case? same engine), theres significant gap between the two architectures:

*note the red boxed 940, they mention before that its OC'd to 3.8GHz, but I'm not sure if the 940 used in that specific test is OC'd or non-OC'd.

 

[Valset]

I hate to add to this but the ram should be kept the same if for no other reason then to prevent the argument before it started. and the same for the cpu choice, if H had included a Q9400/9300 and over clocked them there would not be the debate. all the same I am disappointed, I was hopping for better. I may still buy this but then again I may just dump my current setup entirely and go with a good C2Q system. it seems that it can't quite match the core two quads.

 

[kllrnohj]

First, just to get this out of the way, there is a typo on the test system page. Either that, or Corsair can work some crazy black magic "2GB Corsair DDR2-8888 RAM"

Damn son, you must have a virus or something . Vista idles at about 1.1 / 4GB on my system, and around 400MB of that gets freed up when I launch a memory intensive application.

Also, might want to take a look at THIS website. It was the first one that popped up when I searched Google, but you'll see pretty much the same results everywhere you look.

Like Kyle said, 2GB normally gives you slightly better performance in games due to the lower latencies. Get out n00b.

Vista easily eats up 2GB of memory, and even if it will give it up easily, freeing RAM is *VERY SLOW*. So for the tests dealing with 500mb or so of data, the Phenom rig is going to get stuck waiting for Vista to free up enough RAM from the death grip of SuperFetch. I'm fairly positive that that WILL impact the results. SuperFetch has a tougher time filling up 4GB, so the Core 2 system won't have to wait for nearly as much RAM to be freed, and of course the i7 rig won't have to wait at all.

That website only shows the results for Far Cry 2, where it doesn't really matter how long it loads. I'm more concerned with the application benchmarks, where waiting for RAM to be freed is going to impact the results.

I must admit, this is the first [H] eval I've been disappointed in. It just feels so sloppy, rushed, and half-assed. Even if the Phenom II is a disappointment, it should at least get a fair, proper eval.

 

[AtkinS]

.....

Have youre test methods changed alot, or ????????????????????????


All other reviews i see, there isnt this big diffrence between phenom 940 and qx9770. and the qx9770 and core 7 in games.

 

[jonneymendoza]

.....



All other reviews i see, there isnt this big diffrence between phenom 940 and qx9770. and the qx9770 and core 7 in games.

thats because GAMES ARE GPU DEPENDANT!!!! when will people learn that?

 

[cannondale06]

thats because GAMES ARE GPU DEPENDANT!!!! when will people learn that?

if you already have a good cpu then yes there wont be much difference. throw in an X2, Pentium D, or even a low end Core 2 like E2180 cpu and the cpu becomes pretty important in most newer games.

 

[Soulstice]

Kyle,

[H]ardOCP's reviews are generally unbiased and informative, especially the new PSU reviews based on a near-perfect rating system. Kudos!

However, over the years of being a [H]ardOCP forum member and supporter, I've noticed that every now and then you guys will review a piece of equipment and use the same methodology that you vehemently criticize, especially when other hardware sites do it.

In this particular case, you compared processors of a different nature, and (God knows why?) decided to install a completely different amount of memory into each test-bed system. Seriously man... ...what were you thinking?

And, please, spare me the rational that "paging to disk was not an issue". Who cares if it was or not? The point is that you failed to compare apples to apples. You took an already defunct (as in "dead"... no chance of competing) technology, and disabled it further by reducing its memory by half. Other than being lazy, I can't think of any suitable excuse as to why you simply didn't swap the two single GB sticks with 2 x 2 GB modules. Furthermore, it would have been more accurate to use a recent board (as in just released!), as opposed to the MSI model chosen for this review. Don't get me wrong, I realize the MSI board is known for it's quality and configurable options, but after admitting the the board failed to operate with the memory running at 1066 MHz, it's fully possible that the board's latest BIOS may have some compatibility issues with the latest Phenom II units, resulting in poorer performance overall (not just reduced memory speeds). You could have just as easily ported the system over to another newer board based on the same chipset for the purpose of the review.

Think about this for a second. All you had to do to remain unbiased, prove your point, and eliminate any resentful response on the forum was to keep the test-boxes as closely aligned (in terms of hardware configuration) as possible. Not doing so jeopardizes all of the article's results, and calls into question [H]ardOCP's integrity. I believe you, in that you're probably correct in your hypothesis that, for the purpose of your benchmark tests, the additional RAM would have made a minimal impact. However, until you show us the actual results of an apples-to-apples comparison, the current result-set is worth as much as not having performed the any of the tests in the first place.

In fact, just for laughs, I would like to see the results in reverse! Throw in 3 GB of tripple-channel into the I7, 2 GB into the Core 2's and 4 GB (2 x 2 GB) into the Phenom II. Still not apples-to-apples, but at least it will shut people like me up!

Seriously though, just like a science experiment, the goal should always be to eliminate as many variables as possible... ...and, unfortunately, you've failed in this review to adhere to an unbiased, non-debatable testing methodology.

 

[Slash3]

Briefly back on the 2GB vs 4GB thing, it looks like Far Cry 2 does not exceed 2GB when the built-in benchmarker is used, regardless of settings. But only just... the engine itself uses about ~700-800MB and a bare vista boot sits just under a gig of allocated memory. No difference for the canned runs at 640x480/DX9/low or 1680x1050/DX10/8XAA/Ultra (max of my monitor) for any built-in run going from 2GB to 4GB. This may not apply to the other application suite tests or game tests but FC2 is fine. (Crysis: Warhead is well recognized as using more than 2GB right off of the bat, and it is reflected in the minimum FPS scores in benchmarks)

And of course, right after I swap out some sticks and reboot a few times, I find a similar comparison on CDRinfo.com: http://www.cdrinfo.com/Sections/Reviews/Specific.aspx?ArticleId=24373&PageId=6

Still, I'd be pretty uncomfortable benchmarking a system with less RAM than video memory!

 

[Brackle]

Just wish the review wasnt so biased we know it sucks compared to the i7...shrug

 

[stormy1]

Enjoyed the review .. colorful and funny..
I did not enjoy the junk that AMD has put on the market.
My 5000BE is still going strong so I think im going to sit out this round, I hope AMD has something competitive to drive prices down by the time I am ready to upgrade.

 

[Atech]

Briefly back on the 2GB vs 4GB thing, it looks like Far Cry 2 does not exceed 2GB when the built-in benchmarker is used, regardless of settings. But only just... the engine itself uses about ~700-800MB and a bare vista boot sits just under a gig of allocated memory. No difference for the canned runs at 640x480/DX9/low or 1680x1050/DX10/8XAA/Ultra (max of my monitor) for any built-in run going from 2GB to 4GB. This may not apply to the other application suite tests or game tests but FC2 is fine. (Crysis: Warhead is well recognized as using more than 2GB right off of the bat, and it is reflected in the minimum FPS scores in benchmarks)

And of course, right after I swap out some sticks and reboot a few times, I find a similar comparison on CDRinfo.com: http://www.cdrinfo.com/Sections/Reviews/Specific.aspx?ArticleId=24373&PageId=6

Still, I'd be pretty uncomfortable benchmarking a system with less RAM than video memory!

This should make the whiners shut up...or post evindence to the opposite...but wanna bet all they do is whine and produce no data?

 

[Diesel_Power]

What a shame AMD. What are they going to do a year or two from now when i7 prices go down and this is still the best they've got? Dead meat!

 

[Rage_3k_Moiz]

The Phenom II X4 940 should have been compared to the i7 920, since they sit at similar price points. The review is clearly biased.

tbreak's review here rightfully compares the Phenom II X4 940 to the Core i7 920. Anandtech's review here also compares the X4 920 and X4 940 against all the major CPUs from Intel.

 

[eric66]

problem here is seems ppl really had high expectation from old tech (mainly kyle ) this looks like r600 to rv670 thing ati did earlier nothing special have done to old arch

 

[Silus]

Which is strange. TechReport gives a much more favorable review, and places the 940 has being roughly equal to the Q9400:

http://www.techreport.com/articles.x/16147

--In the Phenom II, AMD has produced a chip that comes strikingly close to duplicating the performance of Intel's mid-range Core 2 Quad processors, the Q9300 and Q9400.--

Guru3D pretty much comes to the same conclusion:

http://www.guru3d.com/article/amd-phenom-ii-x4-920-and-940-review-test/1

--See, when you place things in perspective, the $278 Phenom II X4 940 is here and there getting in close vicinity of that Core i7 920 processor. But let's compare to the Core 2 series for a moment. One thing we can definitely state is that Phenom II's performance is now comparable to the mid- to more high-end Core 2 Quad performance levels.--

So now I'm wondering whether something is up with [H]'s testing platform or not.

You seem to have missed the rest of the Tech-Report's conclusion, which reflects what Kyle also reproduced in his article:

Some tough realities remain for AMD, but I'd say they're now tempered by a little more hope. Although the Phenom II is a marked improvement over the original 65nm Phenom, AMD still can't match the fastest Core 2 Quads in clock-for-clock or outright performance. And obviously, the Core i7 is yet another step beyond the Core 2.

 

[Atech]

You seem to have missed the rest of the Tech-Report's conclusion, which reflects what Kyle also reproduced in his article:

Xbitlabs also agrees:
http://xbitlabs.com/articles/cpu/display/amd-phenom-ii-x4_15.html#sect0

They missed the window of opportunity and the launch of Phenom II X4 doesn't have the desired effect on the market. The new Phenom II X4 doesn’t look too impressive against the background of contemporary Core 2 Quad and especially Core i7 CPUs. The results of our tests show that the top Phenom II X4 processors can only be worthy rivals to the Core 2 Quad CPUs from the “junior” Q8000 series. Unfortunately, Phenom II X4 cannot yet do better than that.

 

[Casper312]

wow this thread is turning into a love hate one! seems like some people dont like the review here

Their just mad hard does not follow the crowd and reviews shit how its supposed to

Im actually not surprised how it performs, and im loving peoples price arguments, Cause plugging a lower cost c2q at the same clocks will no perform the same? Never knew the Qx chips performed 50% better per clock than reg ones

 

[AtkinS]

you whoever you where who said its gpu dependant.

the benches are done on samy hardware in both tests, the qx9770 bested the core 7, and then suddenly it dont ?

uhm, both tests are done by hardocp, which makes this uhm ?

even though it doesnt beat core 2, its priced way lower, which makes it value. for now.

Legit Bottom Line: The AMD Phenom II processor series processor may not be the performance winner in all the benchmarks, but it might not need to be competitive in this day and age when the price tag matters most.

bit tech: AMD is getting back in the saddle, and we can't wait to see what 2009 will bring them, and us. Good stuff, very stable, nice , affordable and fast. What's not to like?

the whole core 7 platform is quite expensive. im using an cheap AM2+ board, in fact its half the price of core 7's cheapest board.
my memory cost me 40 bucks and i got 4 gigs.
the cpu will cost me abit, but still, its half the price.

 

[Michaelius]

From my point of view and budget which don't allow triple sli this review was pointless.

 

[Rossi~]

Im not suprised by this, but still disappointed by the results as i was a bit of an AMD fan boy. But i am alos slightly glad as this makes my i7 a bit of a king and less of a waste of money
Would be interested to read some more reviews

 

[jonneymendoza]

you whoever you where who said its gpu dependant.

the benches are done on samy hardware in both tests, the qx9770 bested the core 7, and then suddenly it dont ?

uhm, both tests are done by hardocp, which makes this uhm ?

even though it doesnt beat core 2, its priced way lower, which makes it value. for now.

Legit Bottom Line: The AMD Phenom II processor series processor may not be the performance winner in all the benchmarks, but it might not need to be competitive in this day and age when the price tag matters most.

bit tech: AMD is getting back in the saddle, and we can't wait to see what 2009 will bring them, and us. Good stuff, very stable, nice , affordable and fast. What's not to like?

the whole core 7 platform is quite expensive. im using an cheap AM2+ board, in fact its half the price of core 7's cheapest board.
my memory cost me 40 bucks and i got 4 gigs.
the cpu will cost me abit, but still, its half the price.

but core 2 uses ddr2 ram as well and core 2 boards are just as cheap as your beloved AMD board. can you price up a typical phenom 2 setup along with a typical quad core 2 setup please?

i would like to see where this so called cheaper build for a AMD build exists. Yes i7 costs an arm and a leg but newsflash, your definitely getting your money's worth for one as it wont be beaten for at least a year. its only the mobo and ram that costs alot. cpu is cheap to mine and everyones surprise for a new tech really as i was expecting the low end i7 to be priced around £400+

what have we learned from this new AMD, that they provide a cheaper budget solution for people? newsflash again, the current core 2 fits that bill quite nicely as it stands with the option to drop its price even further if they want to.

why cant AMD compete with the i7 i wonder? its taken AMD THIS LONG to compete with INTELS LAST GEN platform. thats pathetic really.

 

[PhantomAu]

Well as an AMD fan, I'm very disappointed.

When AMD says they are not competing with the i7 they are only fooling themselves.

I fully expect intel to dump prices on their Yorkfield line in preparation for the release of the i5 mainstream part. Enthusiasts will buy the new i7 and system integrators will most likely be able to move their existing inventory of mainstream intel hardware compared to new stock of AMD. Which pretty much leaves only people who have an AM2+ board to grab a PII 940 and overclock the crap out of it to make the most of their investment. That said I do admire the power savings on the PII. I hope all traces of the P1 are quickly remove as I cannot see them being a product anyone would want to buy now.

(edit) there is one market which I missed & I think AMD will have success here, but not solely based on the strength of its CPU and that is for HTPC's. The cheap & fully integrated mobo's which can decode bluerays in hardware are a winner, but people don'r need AMD's latest and greatest quad core for that.

Kyle I think your review was fair & unbiased. I know as an enthusiast site, overclocking is the order of the day, but given your no BS, no fudge reviews I and others would love to see some straight out benchmarks of the PII 920 & vs the Q6600, Q9400 & Q9550 purely for value consideration when buying new systems from these now "mainstream" chips. I know we can get these elsewhere, but I simply trust your results more.

After reading the review I can only hope your problems with DDR2 1066 on the Phenom board are not an indication of a fault with your hardware as other sites were able to run at that speed without error. Bring on another round of BIOS updates.

Thanks again for the review, I'll start specing out my i7 upgrade now.

Danny

 

[AtkinS]

my standing here.


New build, and youre not in need of having the ultimate at whatever the cost might be.
Why not phenom 2 ? if you need quad.

Core 2 represent a good value aswell, limited overclocking, and less fps per watt.


Do you have a core 2 quad or something like that, i dont see any reason to change.
Got phenom 1 and running 3 ghz, you will be able to run whatever you like.


Is the gameplay experience like, really better on core 7, i dunno, i cannot say i can notice much diffrence from platform to platform, i really cant notice, ohh im on a 2500 usd comp, or an 1000.

at this level there isnt about best performing if you ask me, but what gives you the possibility to run whatever game you throw at it, at a low cost.

again look at previous site, when im posting the images of farcry 2 benches.

how can the scores be that diffrent, look at HW tested, its the same hardware, so......

First test of core 7.
GTX280OC
TT TP 1200W.
2560x1600 0xaa very high same scene.
core7 47fps 3.2 ghz
qx9770 50fps 3.2ghz
e8500 50fps 3.16ghz

Phenom 2 review
GTX280OC
TT TP 1200W.
2560x1600 0xaa very high same scene.
core7 111fps 3.2ghz
qx9770 84fps 3.2ghz


Doesnt benefit of multicores ?... core 7 8 threads doesnt matter ?

how can the scores be that diffrent is my question .

***edit.
You can then say easily overclock that to 3.4GHz or so. I can fully see going this route if the hardware backbone is there to already make this happen

this is easy on any board that support 140 W cpu, all new that is.

 

[conscript]

not your best work Kyle. I don't think your conclusion is wrong, but how you got there leaves something to be desired.

 

[jonneymendoza]

Well as an AMD fan, I'm very disappointed.

When AMD says they are not competing with the i7 they are only fooling themselves.

I fully expect intel to dump prices on their Yorkfield line in preparation for the release of the i5 mainstream part. Enthusiasts will buy the new i7 and system integrators will most likely be able to move their existing inventory of mainstream intel hardware compared to new stock of AMD. Which pretty much leaves only people who have an AM2+ board to grab a PII 940 and overclock the crap out of it to make the most of their investment. That said I do admire the power savings on the PII. I hope all traces of the P1 are quickly remove as I cannot see them being a product anyone would want to buy now.

Kyle I think your review was fair & unbiased. I know as an enthusiast site, overclocking is the order of the day, but given your no BS, no fudge reviews I and others would love to see some straight out benchmarks of the PII 920 & vs the Q6600, Q9400 & Q9550 purely for value consideration when buying new systems from these now "mainstream" chips. I know we can get these elsewhere, but I simply trust your results more.

After reading the review I can only hope your problems with DDR2 1066 on the Phenom board are not an indication of a fault with your hardware as other sites were able to run at that speed without error. Bring on another round of BIOS updates.

Thanks again for the review, I'll start specing out my i7 upgrade now.

Danny

im about to press the buy button on a i7 rig within the next few days i just wished the P2 would at least make intel slash their prices abit more for the i7's

 

[AciDeX]

Great job Kyle, pretty much what I expected. This line made me laugh my ass off! ...How true, how true.

” Please for the love of someone’s god, put down the bong and get sober before you mail this crap out."

 

[Casper312]

not your best work Kyle. I don't think your conclusion is wrong, but how you got there leaves something to be desired.

It gets to the point and shows the performance of the chip, how is that bad work? Maybe if he would of said some good things about the chip instead of getting straight to the point?

I love the review, for the hype the chip had, and performance it has, im surprised they even published the review after doing testing.

 

[marzzes]

I’m not one to shoot the messenger here but how can I refer any of my clients to your website when you insist on checking your intellect at the door and before you give me your “Howard Stern of hardware” speech, I’ll leave you with one word: “Educate”……
In general I get a kick out of your reviews but man- o- man, just wish you could find a better way of expressing yourself, especially when it comes down to belittling and or assuming everybody on your forms is beneath you!

A concerned reader,

Mario

 

[panfist]

In this particular case, you compared processors of a different nature, and (God knows why?) decided to install a completely different amount of memory into each test-bed system. Seriously man... ...what were you thinking?

Furthermore, it would have been more accurate to use a recent board (as in just released!),..the the board failed to operate with the memory running at 1066 MHz

I agree with these two points.

I'd also like to see your assessment of power consumption and performance per watt. I'm guessing that this is another area where Intel is in the lead.