AMD Phenom II Review Thread

[Shintai]

Its already available at newegg, 920 & 940...I cant understand why they are labelled at 125W because all denebs consume less power than high-end yorkies, which are rated 95W....very strange

Not really...



Maybe if you pair the yorkfields with an nVidia chipset. Or if just using idle numbers?

[ryboto]

Not really...



Maybe if you pair the yorkfields with an nVidia chipset. Or if just using idle numbers?

Look to Lost-Circuits review, and the table that was already posted in this thread. They do consume around what yorkfields do, but it can also depend on the motherboard.

[Shintai]

Look to Lost-Circuits review, and the table that was already posted in this thread. They do consume around what yorkfields do, but it can also depend on the motherboard.

In that chart I see a QX9650 use less power than a PH2 940 (Lowest of the 2 PH 940 numbers). All the rest of the yorkfields are missing besides the QX9770. Plus everything is old steppings I bet on the rest. There aint even a Q6600 there.

So no?

Its a THG wannabee site it seems.

[Manabu]

It seems like AM3 processors will be backwards compatible with AM2+ (and AM2 too?). But that those AM2+ phenom II will not be compatible with AM3.

http://images.bit-tech.net/content_i...ew/stars-8.jpg

[Kaldor]

I have to say, I am disappointed. I wasnt expecting i7 levels of performance out the P2, but was hoping that the 940 would about = the Q9650, which it clearly does not. I think some trickle down will happen as other have stated, it should drop the price on upper C2Q's. Oh well, I guess its i7 on my next machine.

[Shintai]

It seems like AM3 processors will be backwards compatible with AM2+ (and AM2 too?). But that those AM2+ phenom II will not be compatible with AM3.

http://images.bit-tech.net/content_i...ew/stars-8.jpg

Yep. Its abit like buying the cat in the sack with these AM2+, unless you want a complete upgrade of everything next time.

925 should come next month. 945 in April.

[Morais]

Not really...


Maybe if you pair the yorkfields with an nVidia chipset. Or if just using idle numbers?

http://www.bit-tech.net/hardware/200...-920-review/13

Well, doesnt change my point of view in rating denebs from 125W to 95W

[ryboto]

In that chart I see a QX9650 use less power than a PH2 940 (Lowest of the 2 PH 940 numbers). All the rest of the yorkfields are missing besides the QX9770. Plus everything is old steppings I bet on the rest. There aint even a Q6600 there.

So no?

Its a THG wannabee site it seems.

wow, you really pick and choose what you read in people's posts. I said "around".
at idle:


and at load:


Even if it's an old stepping, how much has power consumption improved on those chips? Do you know? 68W vs ~65W, so they are "around" the same, or approximately similar. I remember seeing measurements of a Q6600, putting it at ~70-80W at 2.4ghz.

[Glow9]

http://www.bit-tech.net/hardware/200...nd-920-review/

In this review when they pair Yorkfields with ddr2 ram the gap between deneb and yorkfield is very little...I think that with am3 mobos and dd3 ram...the step behind i7 will be deneb

You mean a step behind yorkfield on a ddr3 board.

Which Quad sells the best, that's the real question?
http://anandtech.com/cpuchipsets/sho...px?i=3492&p=20

I wait Intel price cut to buy Drwho?! Make us happy

Might happen their sales are down 23% and not moving expensive mobos is not going to make their partners happy I'm sure.

Anyone else think www.bjorn3d.com review sucked? looking at the resolutions they used, then with them Using a Phenom II thats OCed against a stock Intel just seems stupid.

Since most reviewers have been sitting on demo chips for a while I think all of these should state if they are retail or not since I'm guessing most aren't. I think when I see someone on here take pics of their own opened box that will have more credibility.

[Donnie27]

Yes, and it was the same for 65nm and I bet it'll be the same for 32nm; the fact is that until Intel seriously blunders, AMD is going to have a hard task retaking the performance crown. Personally, I see a very focused Intel beating AMD to the core shrink and maintaining their edge for the foreseeable future.

@Donnie: Where are the usual suspects? It's time to defend their honor.

Don't know but still waiting for the Damage Control

[Rammsteiner]

Argh, I still dont get it.

WHAT'S THE DEAL WITH PICKING ON POWER CONSUMPTION!?. Seriously, if you care about environment, then commit suicide, that's the most green thing you can do. If you care about your wallet, dont look at high end parts and cry they use so much.

Just like a low power CPU wouldnt be able to run your 4 GPU's and Far Cry 2, a high-end part cant be low power. I thought this was basics.

Although, it's quite dissapointing, on the otherhand it's still better than Agena so whatever

[Zucker2k]

Don't know but still waiting for the Damage Control

Tough task.

For me, this visualizes what some have argued a lot here. It's about power consumption, and while PII may have better idling power consumption than Ci7, if you're a Ci7 owner, the graph below is why you love your system.

Tech Report:

We can quantify efficiency even better by considering specifically the amount of energy used to render the scene. Since the different systems completed the render at different speeds, we've isolated the render period for each system. We've then computed the amount of energy used by each system to render the scene. This method should account for both power use and, to some degree, performance, because shorter render times may lead to less energy consumption.

[Jacky]

AMD could skip 32nm push a 16 bulldozer out by researching more on 22nm and design and development.

I didn't know they were involved in the pancake business where moving to a new node is as simple as getting a new frying pan ?

Oh, this made me laugh so hard. Pure genius.

Other than that.. mhm PII doesn't look that bad. Is there someone going to crunch the numbers and give us the average changes? e.g. clock for clock vs 65nm intel, cache castrated 45nm, 45nm, nehalem, etc
If no, I'll do it (I will probably do it anyway).

[bingo13]

you always find strange comments, the real strange part is how reviewers actually OC, in core2 times you see oc's with the best air coolers like ultra120 extreme, its already known phenom2 needs cool down to clock higher. Now they test on stock cooler. Where are the NB oc's????

Even though I got our chip stable at 3.9, it more or less sucked when trying to clock HTT or NB speeds up. The chip would lock at 211 HTT and NB set over 2200 resulted in a no-POST situation on six different boards, so the chip was not that good in my opinion. We have three retail chips arriving today to test and we will push HTT/NB as mentioned in the review. AMD told us that week 50 and up parts are seeing significant improvements in both core clock and HTT clock capabilities compared to our ES samples.

[Eastcoasthandle]

AMD Phenom II Review
Anandtech
Bit-tech
Computer Base
Driver Heaven
EB
Guru3d
H
Hexus
Hot Hardware
Lab 501 (Romanian)
Legion Hardware
Legit Review
Lost Circuit
Over Clockers Club
Tech Report
Xbit Labs

[[XC] gomeler]

I am certain that the reason so many overclocks by reviewers were so low was the sheer lack of experience with the systems. I was afraid of blowing up my sample before the launch date so I stuck with a conservative 1.45vCPU which yielded a paltry 3800MHz. Now that the article is live, I'm blasting 1.55vCPU through this chip at 4200MHz and climbing

The biggest issue I have though is the lack of temperature readings, I'm currently using a thermalprobe wedged up against the IHS of the actual CPU to get a ballpark figure :-\ It's reading around 55C at idle while the thermal diodes on the chip still read 19C

[Zucker2k]

I am certain that the reason so many overclocks by reviewers were so low was the sheer lack of experience with the systems. I was afraid of blowing up my sample before the launch date so I stuck with a conservative 1.45vCPU which yielded a paltry 3800MHz. Now that the article is live, I'm blasting 1.55vCPU through this chip at 4200MHz and climbing

The biggest issue I have though is the lack of temperature readings, I'm currently using a thermalprobe wedged up against the IHS of the actual CPU to get a ballpark figure :-\ It's reading around 55C at idle while the thermal diodes on the chip still read 19C

That's a bit toasty; is 4.2Ghz stable, benchable? Nice clocks.

[Eternalightwith]

NB at 3.4 ?!?!

if you consider NB clocks up to 3.4GHz from 2.6-2.7GHz and...

[[XC] gomeler]

That's a bit toasty; is 4.2Ghz stable, benchable? Nice clocks.

This chip won't last long Next week it gets 2.0vCPU and -196 It is stable and benchable for everything but Cinebench x64 SMP. For some reason Cinebench x64 just kills this CPU, everything else including Cinebench x32 SMP runs fine

[Lightman]

Even though I got our chip stable at 3.9, it more or less sucked when trying to clock HTT or NB speeds up. The chip would lock at 211 HTT and NB set over 2200 resulted in a no-POST situation on six different boards, so the chip was not that good in my opinion. We have three retail chips arriving today to test and we will push HTT/NB as mentioned in the review. AMD told us that week 50 and up parts are seeing significant improvements in both core clock and HTT clock capabilities compared to our ES samples.

Exactly what I've said in earlier in this thread.
You seem to have oldest samples of revision RB-C2 I've seen on internet. Anything from 0839 to 0843.
At least in Europe retail stock is 0850 and onwards. Tomorrow I should find out what we have in UK.
Granted it came from Macci, but he already demonstrated NB clocks in the range of 3.3GHz on a 0850 retail chip. Also HondaGuy, member of XS, has shown close to 3GHz NB.

I wonder how much further they can improve it in RB-C3 revision.

PS. Have you already received AM3 platforms for evaluation? I hope the answer for this question is not under NDA

[[XC] gomeler]

Is there a listing of the various weeks? Here's my chip, 0843. Don't know anybody keeping track, 0850 retail silicon globally? Was told mine was a retail sample from AMD :shrugs:

[Shintai]

The biggest issue I have though is the lack of temperature readings, I'm currently using a thermalprobe wedged up against the IHS of the actual CPU to get a ballpark figure :-\ It's reading around 55C at idle while the thermal diodes on the chip still read 19C

Pretty much confirms what alot have been saying while people reported ultra low temps. They simply read wrong data or just bugged diodes.

For gaming Phenom 2 looks horrible. With basicly Q9400 area for the 940.







Thats a LOOOONG road to 2011 for Phenom 2. 3.2 or 3.4Ghz I guess with a 2ghz NB (AM3 is 2Ghz NB) might be the fastest on stock in its lifetime. 3.0Ghz seems to be the top for 2009 according to AMD roadmaps.

i5 in summer 2009. 32nm i7 in winter 2009. 32nm i5 in 2010. 32nm i8? in late 2010....

[[XC] gomeler]

Diodes are broken. BIOS reads 17C to 19C, AOD reads 19C no matter what.

[demonkevy666]

Pretty much confirms what alot have been saying while people reported ultra low temps. They simply read wrong data or just bugged diodes.

no his chip senor is probably broken or never worked at all.

hondaguys temps changed when OC.

[gOJDO]

Let's not forget that the first review was made by the hwbox.gr! Their results were inline with the results we see from today's reviews. So some hardcore AMD fanboys who were bi*ching at the results are owing an apology to the greek gurus.

So to sum it up:
- Deneb is great only as an upgrade for those who have AM2/AM2+ platforms.
- It offers no advantage over what is available on the market for quite some time.