Phenom II info leaks out, AMD hints at something

[Hornet331]

So how does mid-range become "slowest"?

he said slowest quad.

not slowest overall.

[Sparky]

he said slowest quad.

not slowest overall.

No no. Look at the quote again:

it's very easy to match the performance of a mid range quad core Intel machine with an AM2+ one.

And accord99 then says:

It's possible to match the performance of Intel's slowest quad-core machine.

[keithlm]

That pricing is pretty much representative for europa... well, i think i forgot to factor in, the world only consists of the usa.

Oh I see. Since it was true in Europe then it is an undeniable truth even if it wasn't true everywhere.

Priceless.

[Jazzman]

Some have speculated that given the phenom underlining architecture is the same in Phenom II that you would see high current leakage as you overclocked to 4.0. On the XS AMD forum, RiverRicer found HERE that there are significant architecture changes in AMD's 45nm chip.

The transistor drive current for AMD's 45-nm devices is much lower than that of the Intel HKMG transistors. But power consumption is quickly becoming a high priority for server chips. AMD's transistors exhibit very low channel leakage. Our transistor benchmarks indicates that leakage current is less than one-third of the value measured on AMD's 65-nm process. It's also significantly lower than the Intel 45-nm HKMG process. In fact the Ion/Ioff ratio for AMD's PFET is nearly 10 times better than that for the Intel PFET.

The above would seem to imply that should AMD migrate to HKMG (High_K_Metal_Gates) you would see still more improvement. HKMG is forecast for AMD chip construction later in 2009.

In my opinion, this technology innovation would seem to suggest that later Phenom II chips could clock well above 4.0Ghz.

For those of you technically inclined, here is an explanation of the specific design innovations:

On the NFET side, stress memorization stretches the n-channel, which is enhanced later in the process flow by the addition of a nitride tensile stress liner. The liner itself is scaled down at 45 nm to ensure the required strain is adequately supplied to the transistor channel to enhance electron mobility and subsequently increase the drive current. The gate-stack design is modified as well, with a new sidewall spacer design for 45nm.

The PFET performance improvement is more dramatic with drive current now up to 660 µA/µm compared with 510 µA/µm on 65-nm transistors. Again, this increased output current is the result of optimized compressive strain for the p-channel device. The new design of the PFET moves the embedded silicon-germanium source/drain regions closer to the channel to maximize the transfer of stress, thereby increasing hole mobility. Although shorter gate lengths are not driving the improvements, it is a reduction in dimensions that allows increased channel stress to provide the performance scaling. AMD 45-nm PFET design reduces the space from embedded silicon-germanium to the channel edge by half.

[Eson]

So Stargazer, what does Phenom under performing Core 2 have to do with Phenom II? Why did you have to pull the trigger on the Intel gun? I believe the credits for yet an other ruined AMD thread goes to you. Of course Hornet331 shouldn't be forgotten either. Good work.

Haven't we discussed this matter enough?

[Katanai]

So Stargazer, what does Phenom under performing Core 2 have to do with Phenom II? Why did you have to pull the trigger on the Intel gun? I believe the credits for yet an other ruined AMD thread goes to you. .

I think that's what he does best. It's not only AMD or Intel for that matter, I'm beginning to suspect he is in it for the sport.

[STaRGaZeR]

So Stargazer, what does Phenom under performing Core 2 have to do with Phenom II? Why did you have to pull the trigger on the Intel gun? I believe the credits for yet an other ruined AMD thread goes to you. Of course Hornet331 shouldn't be forgotten either. Good work.

All Brother Esau's post and this one:

mainly because they all play SuperPi all day long brother LOL

We have a launch even today for i7 and I bet they show SuperPi there...as if in the whole scale of things it friggin matters.

volume sales are what matters, cost Vs performance...AMD proved with 4XXX GPU cards they could win at this, and I am sure they have a plan for Phenom 2 going in the same direction.

XS news is not really the place to talk about AMD anymore...every thread in the news section usually ends up going the same way

Oh wait, they are the ones talking about Intel in an AMD thread. Now some of us answer them and what? We are called thread ruiners. Good work for another call to the whaaaambulance with no reason

[Hornet331]

All Brother Esau's post and this one:



Good work for another call to the whaaaambulance with no reason.

damn you beat me.

i was going to say the same, if Brother Esau wouldn't have commented in such a dismissive way, this thread would have been already buried in the deepest of this forum.

[gallag]

This thread went off the rail's at page five and Intel fans had nothing to do with it, If you read it started with a few comments about hot nehelams and Intel's only being good for spi and fanboys thinking there epenis etc etc

As per usual its the same wee hardcore click of AMDfanboys complaining and being mega sensitive about everyone not thinking AMD is the best thing since sliced bread that has ruined this thread.

Can you not all just get along? Read from page five and honestly say where you thought it started.

[Sparky]

It takes two to make a fight. You could have just ignored him. Typical "he started it" defense. Doesn't mean you have to finish it.

So how about everyone just drop it then and get back on the original topic

[Eson]

I must have missed Brother Esau's Core 2 vs. Phenom comment. I thought he was talking about how tired he was of Intel fanboys destroying AMD threads and that there are nothing wrong with the Phenom. And Tony mentioned the name i7. I am very sorry for being stupid, i must have overreacted.

[STaRGaZeR]

There is no reason to be sorry It's always the same. The funny thing is that the "Intel fanboys" including myself have already said several times that Phenom 2 can be a bomb, and probably will, before the crapping.

[gallag]

There is no needreason to be sorry It's always the same. The funny thing is that the "Intel fanboys" including myself have already said several times that Phenom 2 can be a bomb, and probably will, before the crapping.

I see a lot of "Intel fanboys" also saying that if deneb pans out like it looks like it will it might be there next system. Strange that.

[Hornet331]

There is no reason to be sorry It's always the same. The funny thing is that the "Intel fanboys" including myself have already said several times that Phenom 2 can be a bomb, and probably will, before the crapping.

It doesn't matter how many good things you say, one "bad" word and its like in left4dead when a boomer pukes on you, the hord is going to rape you.

eidt:
btw. if thers gona be a quad core with ~65W TDP -> super cheap cruncher here i come.

right now, I could build a phenom cruncher for 250€ but the power consumption for 24/7 is a nogo for me right now. (Yorkfield system costs me 60€ more but saves that in less then a year)

[gosh]

After the horrible failure of Phenom, they absolutely have to bring something competitive to the table in order to sustain themselves, at the least.

Yesterday I did a report for a customer having a large database. The report took about 40 minutes to complete in debug mode. When you do this kind of work you don't sit and wait for the report to be ready, you plan it and do other work while the report is being created. On the AMD you will not notice that the computer is working but I on the Intel, the computer slows down considerably.

[Zucker2k]

Yesterday I did a report for a customer having a large database. The report took about 40 minutes to complete in debug mode. When you do this kind of work you don't sit and wait for the report to be ready, you plan it and do other work while the report is being created. On the AMD you will not notice that the computer is working but I on the Intel, the computer slows down considerably.

Come on man, you can do better than that. I don't know any systems, or any system configurations, for that matter, known by "on the AMD" and "on the Intel." Maybe next time list some system specs and Ocs IF they apply. Thank you.

[gallag]

Yesterday I did a report for a customer having a large database. The report took about 40 minutes to complete in debug mode. When you do this kind of work you don't sit and wait for the report to be ready, you plan it and do other work while the report is being created. On the AMD you will not notice that the computer is working but I on the Intel, the computer slows down considerably.

I think you miss the point, If you remember when the first gossip of phenom started and then with AMD pimping it themselves expectations were high, When phenom launched it was not a bad cpu, It was just worse that the competition. It was slower clock for clock, Max stock clocks and oc potentiol AND used more power. This made it a failure in many of the enthusiasts eyes just like the pentium dual cores were compared to the x2 (i.e x2 3800 beating a extreme edition with hyper threading pentium lol)

So don't get upset when someone calls the first phenom a failure, Sure it is a better option in some cases and with later revisions and new chipsets etc it made its self a valid option in more and more scenarios but at the end of the day extreme systems is 99% guys with a hardware hobby and to them the phenom was a flop. So when Zytek_Fan says that phenom was a failure he is not talking about severs vmware etc, He is just giving a opinion on what phenom was to him.

Deneb will be what phenom1 was meant to be anyway and we can all just forget about phenom1 like we did the p4d.

[Donnie27]

I think you miss the point, If you remember when the first gossip of phenom started and then with AMD pimping it themselves expectations were high, When phenom launched it was not a bad cpu, It was just worse that the competition. It was slower clock for clock, Max stock clocks and oc potentiol AND used more power. This made it a failure in many of the enthusiasts eyes just like the pentium dual cores were compared to the x2 (i.e x2 3800 beating a extreme edition with hyper threading pentium lol)

So don't get upset when someone calls the first phenom a failure, Sure it is a better option in some cases and with later revisions and new chipsets etc it made its self a valid option in more and more scenarios but at the end of the day extreme systems is 99% guys with a hardware hobby and to them the phenom was a flop. So when Zytek_Fan says that phenom was a failure he is not talking about severs vmware etc, He is just giving a opinion on what phenom was to him.

Deneb will be what phenom1 was meant to be anyway and we can all just forget about phenom1 like we did the p4d.

Very well said

It doesn't matter how many good things you say, one "bad" word and its like in left4dead when a boomer pukes on you, the hord is going to rape you.

Absol-Frakkin"-lutely! QFT!

[Helmore]

I see a lot of "Intel fanboys" also saying that if deneb pans out like it looks like it will it might be there next system. Strange that.

Tell me, when are you an Intel or an AMD 'fanboy'? Only a very few can be considered to be a 'fanboy', most are not. I think pretty much everyone is a 'fanboy' of their wallet or a performance 'fanboy', for the latter you will most likely have to choose Intel and for the former it depends a bit on what you expect of you system.

[Solus Corvus]

Actually I'm surprised that it took 5 pages to happen. People get too emotionally invested in their hardware and it becomes impossible to discuss anything rationally.

[Particle]

Tell me, when are you an Intel or an AMD 'fanboy'? Only a very few can be considered to be a 'fanboy', most are not. I think pretty much everyone is a 'fanboy' of their wallet or a performance 'fanboy', for the latter you will most likely have to choose Intel and for the former it depends a bit on what you expect of you system.

I would say you're a fanboy when you are so obsessed with a company that you find yourself purposely posting in threads about the team-other-than-the-one-you-like and flamebait. We've been seeing a lot of this on the AMD side. When I cruise the Intel stuff I don't see much of it.

But no matter which side they like, people like that are fanboys.

[Rammsteiner]

That is because you guys aren't mentally unbalanced.

Hahaha, clueless

If deneb really rocks it will be amazing to see how fast the review sites become unbiased again and stop being paid Intel pumpers in some peoples eyes lol. Reviews and benches will start to matter again and performance of the cpu will actually become important again, AHHH good times.

If you would have read several arguments correctly according this issue, you would see benchmark results were often not even being crapped with but something else. But since you couldnt care to read those arguments back then, Ive no single reason to post them right now

Anyway, cut it. Either do care or do not care, but dont talk a bit here and be ignorant a bit there.

You always say the same to justify the lack of perfomance.

Oh wait, so SuperPI actually does care?

Phenom underperforms in almost everything compared to a similar clocked C2Q, not only in SuperPi. And C2Q are usually clocked much higher.

Ah, now I get it. Some numbers are apparently 'the' justification not to get a K10? Besides that, there's still a huge difference between underperforming and performing less really. Ive ran my 9850BE stock for several months, and Ive yet to discover any performance issues. Of course higher clocks are very welcome, but dont act like it's crap because the numbers how it performs less

Oh I missed this one. So being a "Legend" automatically implies he's right? Please answer this question of you don't mind.

No, Tony has to be independed, so he actually cant just go spread out lies really. Also the chance he's right is a lot higher than the average poster in this thread

I was shocked. I thought Tony was joking or being sarcastic. Funny how when AMD ran it faster, it was important.

SuperPI is only... relevant when you compare eventual steppings or what so ever. Then still it's questionable. When AMD ran faster it wasn't of any performance for me really. Also Ive yet to see any proof of the group of people you're aiming at

[Donnie27]

Either way, I'm hoping AMD has it right and they are NOT just blowing smoke. IMHO, AMD doesn't have to *win, they just need to play a great game. Sorry guys, I don't see folks posting that Phenom II sucks. Slower doesn't mean slow at all. Phenom is a very good processor that was born at the wrong time. Only Fanboys think AthlonXP was faster than the Northwoods or that Northwoods where faster than Athlon64 when most of the time they weren't. Neither of the slower processors sucked=P

Phenom and the Nehalem fights are just starting and folks are already at each other's throat. I personally like the Phenom II but I have a problem calling it P2 or P-II, sorry! Maybe call it Ph-2 or PhII. Folks have the right to think it will overclock just as others think it will not. Please folks, accept the others opinion as nothing but that. This, "I must attack folks for saying blank about blank" is a JOKE, a very bad JOKE!

Some of us said Nehalem would overclock, others said it wouldn't. Now whatever happens with Ph-II, I hope the Nay Sayers don't slip off and hide as they did with Nehalem:wink:

[Donnie27]

SuperPI is only... relevant when you compare eventual steppings or what so ever. Then still it's questionable. When AMD ran faster it wasn't of any performance for me really. Also Ive yet to see any proof of the group of people you're aiming at

That can be said about all benchmarks. But check this out? I've NEVER been a proponent or fan of SuperPi=P You see whatever you want to see or not. No need to roll you eyes LOL! You are one of the people you seem to not see and that my friend is absolutely you're biggest problem.

[[XC] Synthetickiller]

I know discussed this earlier, but does anyone have any conformation if phenom II OC's will be hindered by 64-bit OS's?

I really hope not. I'm not going back to 32-bit anything.