AMD Athlon X2 7750 Black Edition: dual-core Phenom at the ready

[Ghostbuster]

Link HEXUS: AMD Athlon X2 7750 Black Edition: dual-core Phenom at the ready (http://www.hexus.net/content/item.php?item=16614) vs Intel Pentium Dual Core E5200 vs AMD X2 6000+ :cool:

More..

THG - Phenom Recycled: Athlon X2 7000-Series : AMD Introduces Dual-Core Phenom (http://www.tomshardware.com/reviews/athlon-x2-phenom,2104.html)
PCGH - Athlon X2 7750 BE: AMD-CPU mit K10-Technik (Kuma) im Test (http://www.pcgameshardware.de/aid,670335/Test/Benchmark/Athlon_X2_7750_BE-_AMD-CPU_mit_K10-Technik_Kuma_im_Test/?page=1)
Fudzilla - AMD Athlon X2 7750 does fine though it's no match for Intel, yet (http://www.fudzilla.com/index.php?option=com_content&task=view&id=10921&Itemid=40)
Legit Reviews - AMD Athlon X2 7750 and 5050e Dual-Core Processor Review - Phenom Cores Make It Into Dual-Core Processors (http://www.legitreviews.com/article/850/1/)
Matbe - Athlon 7750 & 7550: les Phenom dual core (http://www.matbe.com/articles/lire/1251/athlon-x2-7750-and-7550-les-phenom-dual-core/)
TweakTown - AMD Athlon X2 7750 BE 2.7GHz CPU – Phenom goes Dual-Core (http://www.tweaktown.com/reviews/1687/amd_athlon_x2_7750_be_2_7ghz_cpu_phenom_goes_dual_ core/index.html)
X-bit labs - AMD Launches “Phenom X2”: AMD Athlon X2 7750 Black Edition Review (http://www.xbitlabs.com/articles/cpu/display/athlon-x2-7750.html)
PC Watch - ようやく登場したK10デュアルコア 「Athlon X2 7750 Black Edition」 (http://pc.watch.impress.co.jp/docs/2008/1215/tawada159.htm)
HT4U - AMD Athlon X2 7750 und 7550 - Familienzuwachs mit "Kuma" (http://ht4u.net/reviews/2008/amd_athlon_x2_7000_kuma/)
DriverHeaven.net -AMD Athlon 7750 Black Edition and Asus M3A78-T - Introduction (http://www.driverheaven.net/reviews.php?reviewid=681)
bit-tech.net - AMD Athlon X2 7750 & 7550 CPUs: (http://www.bit-tech.net/hardware/2008/12/15/amd-athlon-x2-7750-7550-cpus-review/1)
Planet 3DNow! - AMD "Kuma" - Des Athlons Neue Kleider (http://www.planet3dnow.de/vbulletin/showthread.php?p=3809304#post3809304)
MadBox PC - AMD lanza Athlon X2 7000 series (Kuma) (http://www.madboxpc.com/amd-lanza-athlon-x2-7000-series-kuma/)
Guru3D - AMD Athlon X2 7750 BE review (http://www.guru3d.com/article/amd-athlon-x2-7750-be-review/)

[Bobsama]

I think they need to start designing some silicon dual-core Phenoms. But even so--these chips show promise. For the price, they really can compete.

[AliG]

perhaps, but it is definitely way too late to the market, this should have been out a year ago if amd wanted to make any money off it imo. At least phenom II seems to be something good to hope for

[Spectrobozo]

they need to improve the power consumption of the k10... it's just bad, even the dual core version...
+-30w more than a 6000+ at idle is ridiculous, +- 80w more than a e5200 load is not good to, but overall not bad... a 45nm version with 6mb l2 and overclocking more....

[ColonelCain]

I think they need to start designing some silicon dual-core Phenoms. But even so--these chips show promise. For the price, they really can compete.

I feel the same. Cutting one core for tri-core can be understood, but cutting two for a dual-core is a little bit much.

[Ghostbuster]

perhaps, but it is definitely way too late to the market, this should have been out a year ago if amd wanted to make any money off it imo. At least phenom II seems to be something good to hope forYou should be probably wondering what happened to all those dies with 2 out of 4 cores were defective (from months ago)... Possibly AMD must have stockpiled them for this launch. :D

Another review is up at X-Bit Labs: X-bit labs - AMD Launches “Phenom X2”: AMD Athlon X2 7750 Black Edition Review (http://www.xbitlabs.com/articles/cpu/display/athlon-x2-7750.html) this time against E5300. ;)

[Shintai]

Power consumption is just plain bad, and I mean seriously very very bad :(

And performance...not exactly good either. 80$ segment.

[Rammsteiner]

Power consumption is just plain bad, and I mean seriously very very bad :(

And performance...not exactly good either. 80$ segment.
What do you expect from a 77 Euro CPU:confused:

IMO it's pretty nice. Nice toy in general, cheap and K10 architecture. But to be honest it's a tad late to introduce them.

However power consumption is indeed something to work on, but you can still mess around a bit with P-states and CnQ if you want it to consume less, that can make a big difference really:up:

[Shintai]

What do you expect from a 77 Euro CPU:confused:

IMO it's pretty nice. Nice toy in general, cheap and K10 architecture. But to be honest it's a tad late to introduce them.

However power consumption is indeed something to work on, but you can still mess around a bit with P-states and CnQ if you want it to consume less, that can make a big difference really:up:

The issue was that AMD competes so badly as they do atm.

And are you sure they didnt use CnQ?


The systems were configured in exactly the same way as during our performance tests. Enhanced Intel SpeedStep and Cool’n’Quiet 2.0 power-saving technologies were activated. The CPUs were loaded using Prime95 25.8 utility.

The 7750 is an utter joke honestly. 36W more idle and 78W more loaded than the E5300. Or about 30W more both idle and load than the X2 6000. Not exactly cheap in the long run is it?

Not really sure if its an upgrade or downgrade vs the 6000....

[Rammsteiner]

The issue was that AMD competes so badly as they do atm.

And are you sure they didnt use CnQ?



The 7750 is an utter joke honestly. 36W more idle and 78W more loaded than the E5300. Or about 30W more both idle and load than the X2 6000. Not exactly cheap in the long run is it?

Not really sure if its an upgrade or downgrade vs the 6000....
Why do you think I said 'mess' around with CnQ and P-states. Enabling it isnt 'messing'.

And from my pov any K10 is worth it over K8 really. As said, it aint a killer CPU, but for the money and regarding the K10 architecture, it's definately a nice toy.

If you care about power, then get a Via or something:rolleyes:

[AliG]

I disagree, you're better off getting a fast 65nm k8 cpu instead because they consume much less power and offer close to the same performance and will cost less. This cpu is a joke, I mean come on selling something that's double rebandaged? People have on nvidia from doing that but its ok for amd? This should have never seen the market, especially this late in the game.

As I said, right now deneb seems to be the only glimpse of hope for amd since shanghai hasn't done much for them yet

[Nedjo]

Not exactly cheap in the long run is it?


pfff come one... this is desktop CPU for home machines, and price of power consumption for that kind of system is laughable!

[Rammsteiner]

I disagree, you're better off getting a fast 65nm k8 cpu instead because they consume much less power and offer close to the same performance and will cost less. This cpu is a joke, I mean come on selling something that's double rebandaged? People have on nvidia from doing that but its ok for amd? This should have never seen the market, especially this late in the game.

As I said, right now deneb seems to be the only glimpse of hope for amd since shanghai hasn't done much for them yet
You forget that nVidia named their 8-series as 9-series like it was an entirely new design. AMD basicly introduced a Phenom X2 line, that's completely different.

Also there's like a 3 Euro difference between an Athlon X2 6000+ and a Phenom X2 7750BE, to be honest that ain't a lot cheaper. As I said multiple times now, this ain't a superb CPU. It's a nice toy, and for me personally any K10 is worth it over K8.

I could ask the exact same thing, why is i7 better than C2Q while it uses more power and doesnt offer a lot more in gaming? Phenom X2 is faster clock for clock than K8, although we knew already K10 was faster than K8. But as I already said as well, it should have been launched earlier really. But well, this line will become 45nm soon enough, then it's even better;)

[gallag]

I could ask the exact same thing, why is i7 better than C2Q while it uses more power and doesnt offer a lot more in gaming?

Because it does not use more power on a system level and on most other things it is a lot faster?

[STaRGaZeR]

I could ask the exact same thing, why is i7 better than C2Q while it uses more power and doesnt offer a lot more in gaming?

:nono:
http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3453&p=3

[Mr.BSEL]

One reason for the huge power consumption vs the e5200 is the 790fx chipset its using vs the e5200 p45 chipset.

http://www.xbitlabs.com/articles/cpu/display/athlon-x2-7750_3.html#sect0

[Rammsteiner]

Because it does not use more power on a system level and on most other things it is a lot faster?


:nono:
http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3453&p=3
That ain't a C2Q by my knowledge, that's on a server based platform.

This (http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3448&p=9) however is, from your same link different page:rolleyes:

Dont get me wrong, i7 looks damn nice. But funny how we crap about power consumption on K10 X2 while it seems to be allright for i7 while actually the i7 doesnt offer a lot regarding gaming.

[Frodin]

One reason for the huge power consumption vs the e5200 is the 790fx chipset its using vs the e5200 p45 chipset.

http://www.xbitlabs.com/articles/cpu/display/athlon-x2-7750_3.html#sect0

The AMD chipsets are very power efficient, the 790FX uses around 10w, so it has to be something else.

[Shintai]

That ain't a C2Q by my knowledge, that's on a server based platform.

This (http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3448&p=9) however is, from your same link different page:rolleyes:

Dont get me wrong, i7 looks damn nice. But funny how we crap about power consumption on K10 X2 while it seems to be allright for i7 while actually the i7 doesnt offer a lot regarding gaming.

Gee...

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3453&p=3

:rolleyes:

Overall FASTER and LESS POWER CONSUMPTION.

[Macadamia]

lol @ using overvoltaged QX9770.
You guys NEVER read the fine print, eh?



Oh, and what about this?
http://www.fudzilla.com/images/stories/2008/December/Reviews/AMD7750/amd_7750_power.gif

[Shintai]

lol @ using overvoltaged QX9770.
You guys NEVER read the fine print, eh?



Oh, and what about this?
http://www.fudzilla.com/images/stories/2008/December/Reviews/AMD7750/amd_7750_power.gif

Where does it say it was overvolted? And now using fudzilla? :rofl:


I confirmed that I didn't have a particularly low power Core i7-965 by testing multiple chips, and Intel confirmed that my QX9770 fell within the middle of its distribution for power characteristics of all QX9770s. It looks extremely probably that at the same TDP level, Nehalem has the ability to be much more power efficient than even Penryn - all without so much as a die shrink, remember that both of these CPUs are built on the same 45nm process.

[STaRGaZeR]

That ain't a C2Q by my knowledge, that's on a server based platform.

This (http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3448&p=9) however is, from your same link different page:rolleyes:

Dont get me wrong, i7 looks damn nice. But funny how we crap about power consumption on K10 X2 while it seems to be allright for i7 while actually the i7 doesnt offer a lot regarding gaming.

Server platform? It's a C2Extreme QX9770 vs a Corei7Extreme 965, and C2Quad Q9450 vs a Corei7 920 :confused:

You can clearly see that at "low" frequencies Nehalem is not so power efficient when gaming, but stop right there. If you use anything other than games and/or you cranck up the frequency, well you see the result. The problem with Phenom (X2, X3 and X4) is that it underperforms in almost everything, and Corei7 underperforms in almost nothing, both compared to Penryn.

And yes, I crap about Nehalem's power efficiency in games at 2,66GHz, why not. I'm not a fanboy, remember? However I'm not going to use it only in games or at 2,66GHz so... ;)

[Ghostbuster]

lol @ using overvoltagedOh, and what about this?
http://www.fudzilla.com/images/stories/2008/December/Reviews/AMD7750/amd_7750_power.gifAs the others were sayin' Kuma is a little power hog... even undervolting it could not reduce its idle power consumption. On load doesn't look good at all... ;)

Maybe you should provide link to the article: Fudzilla - AMD Athlon X2 7750 does fine though it's no match for Intel, yet (http://www.fudzilla.com/index.php?option=com_content&task=view&id=10921&Itemid=40) :yepp:

[Macadamia]

Where does it say it was overvolted? And now using fudzilla? :rofl:

The QX9770 has a higher VID compared to other relatively low-flying, ~60W using Yorkfields.

IIRC a Q9650 used 10W less than an i7 920.




And in this case FUD is actually right. There's no possible way that the 7750 would use THAT much extra power without having a 125W TDP. But fine, since you obviously trust websites with Core i7 ads at the background... your choice.

[Hornet331]

The QX9770 has a higher VID compared to other relatively low-flying, ~60W using Yorkfields.

IIRC a Q9650 used 10W less than an i7 920.


and performes worse on most apps. ;)





And in this case FUD is actually right. There's no possible way that the 7750 would use THAT much extra power without having a 125W TDP. But fine, since you obviously trust websites with Core i7 ads at the background... your choice.

All reviews i have seen (op) about the 7750, show that it consumes ~30W more @ stock then a X2-6000+.... so all those sites are "paid intel bumpers"?

[Rammsteiner]

Gee...

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3453&p=3

:rolleyes:

Overall FASTER and LESS POWER CONSUMPTION.
Errr, that link was already posted:p: Where as I replied with the link I posted since QX977x ain't really a guideline for C2Q.


Server platform? It's a C2Extreme QX9770 vs a Corei7Extreme 965, and C2Quad Q9450 vs a Corei7 920 :confused:
Can I remind you what socket QX977x uses? Right, skt 771, that's a server platform. Core i7 965 uses just like 920, 1366. QX977x is nowhere inline with any skt 775 C2 CPU, at all. It's mainly aimed at ST for epic e-peen rage;)


You can clearly see that at "low" frequencies Nehalem is not so power efficient when gaming, but stop right there. If you use anything other than games and/or you cranck up the frequency, well you see the result. The problem with Phenom (X2, X3 and X4) is that it underperforms in almost everything, and Corei7 underperforms in almost nothing, both compared to Penryn.

And yes, I crap about Nehalem's power efficiency in games at 2,66GHz, why not. I'm not a fanboy, remember? However I'm not going to use it only in games or at 2,66GHz so... ;)
Phenom underpeforms? Excuse me? Been running 9850BE at stock for 4 months and a 9950BE stock for 1.5 months and Ive yet to suffer from any 'under performing' results really:rolleyes: Compared to Penryn, it ain't doing that bad. At least, at stock.

Core i7 running at higher frequencies, I cant comment on that since QX977x is just a stupid comparison. Get some skt 775 QX CPU instead to compare:rolleyes: When OC'ing comes into this, of course Phenom loses since it's simply not scaling well. But we knew this for how long already? Not like Phenom cant be OC'd though, 3Ghz at default Vcore isnt a rare thing at all but nothing to write home about either.

[gryle]

Errr, that link was already posted:p: Where as I replied with the link I posted since QX977x ain't really a guideline for C2Q.


Can I remind you what socket QX977x uses? Right, skt 771, that's a server platform. Core i7 965 uses just like 920, 1366. QX977x is nowhere inline with any skt 775 C2 CPU, at all. It's mainly aimed at ST for epic e-peen rage;)


QX9770 is socket 775: http://processorfinder.intel.com/details.aspx?sSpec=SLAWM

[Hornet331]

http://ht4u.net/reviews/2008/amd_athlon_x2_7000_kuma/

another review, thought its german.

[Rammsteiner]

QX9770 is socket 775: http://processorfinder.intel.com/details.aspx?sSpec=SLAWM
What the:confused:

Think I got messed up with QX9770 and QX9775. Nevermind then:p:

[STaRGaZeR]

Can I remind you what socket QX977x uses? Right, skt 771, that's a server platform.

Not really LOL.


Phenom underpeforms? Excuse me? Been running 9850BE at stock for 4 months and a 9950BE stock for 1.5 months and Ive yet to suffer from any 'under performing' results really:rolleyes: Compared to Penryn, it ain't doing that bad. At least, at stock.

This is the kind of response I don't really want to read. I know you're not stupid, so please don't act as such. You know what underperform means, but in case you don't want to remember it I'll tell you: it involves comparison. One thing underperforms compared to other when its perfomance is lower than the thing you're comparing to. There is no such thing as underperforming with only one thing, you need two and you need to compare them to decide which one is faster. Phenom is not slow by itself, Phenom is slow compared to Penryn (and Kentsfield). In consequence, Phenom underperforms compared to Penryn. I have not suffered any "under performing" as you call it with my Banias laptop, and yet it underperforms like hell compared to any Core2 or K10. So please, enough with that self-justificative responses.


When OC'ing comes into this, of course Phenom loses since it's simply not scaling well. But we knew this for how long already? Not like Phenom cant be OC'd though, 3Ghz at default Vcore isnt a rare thing at all but nothing to write home about either.

Phenom scales well, I don't see any faults there. The problem is that when you increase frequency in both K10 and Core2 the base differencies grow bigger. And the problem with Deneb is the same, but it has way more frequency headroom so AMD can mitigate it with that: you'd need something like 4,5GHz+ to beat a 4GHz Penryn, the same way it needs 3GHz to be somewhat competitive with Q9450 and Q9550.

[informal]

Phenom scales well, I don't see any faults there. The problem is that when you increase frequency in both K10 and Core2 the base differencies grow bigger. And the problem with Deneb is the same, but it has way more frequency headroom so AMD can mitigate it with that: you'd need something like 4,5GHz+ to beat a 4GHz Penryn, the same way it needs 3GHz to be somewhat competitive with Q9450 and Q9550.

That may very well be the case,but don't you think we need some real reviews with variety of apps to conclude what you said?Agena was pretty close to even Penryn in a lot of apps,but it trailed it in others too.It won't be black& white so that you can say 4.5Ghz will be needed to match 4Ghz Penryn.In certain apps it will do better per clock,in some worse,but the average is still unknown.But i say again,those numbers may indeed be true,we just need to wait for some real tests.

[informal]

He can't "attak me"^^,but he sure can attak Athlon X2 7750!1!1 :D

Now seriously,the chip is pretty good at stock clocks,IPC wise sits between 2.5 and 2.7 E5xxx series(as shown by bit tech and guru3d).What is not so good is power draw,but the question is how much is the chip to blame since FUdzilla did get some strange low(er) power draw numbers(lower than other sites at least).Maybe the 790FX chipsets in the reviews are using more power and it shows now with dual core K10s.OCing is pretty decent too,at least for 65nm Agena core.Hits between 3.2 and 3.4Ghz with 1.45V.which is surprisingly good.Price is not bad too,but it will go down for sure.

[Rammsteiner]

Not really LOL.
Already pointed that out really. Im embarrased enough already, dont lol over it:p:


This is the kind of response I don't really want to read. I know you're not stupid, so please don't act as such. You know what underperform means, but in case you don't want to remember it I'll tell you: it involves comparison. One thing underperforms compared to other when its perfomance is lower than the thing you're comparing to. There is no such thing as underperforming with only one thing, you need two and you need to compare them to decide which one is faster. Phenom is not slow by itself, Phenom is slow compared to Penryn (and Kentsfield). In consequence, Phenom underperforms compared to Penryn. I have not suffered any "under performing" as you call it with my Banias laptop, and yet it underperforms like hell compared to any Core2 or K10. So please, enough with that self-justificative responses.
Then I think we both have different opinions about underperforming. Comparing it to competition then to define it as underperforming is nuts IMO. K10 is better than K8. When a succesor performs worse, then I call it underperforming. It's not like an architecure is designed today, tomorrow a competitor releases a better one and thus they design a better one the day after and launch it. Although AMD and Intel do have to compete with each other price and performance wise, their advances are pretty much independend most of the time.


Phenom scales well, I don't see any faults there. The problem is that when you increase frequency in both K10 and Core2 the base differencies grow bigger. And the problem with Deneb is the same, but it has way more frequency headroom so AMD can mitigate it with that: you'd need something like 4,5GHz+ to beat a 4GHz Penryn, the same way it needs 3GHz to be somewhat competitive with Q9450 and Q9550.
Well, Phenom actually does better at higher clocks than Penryn scaling wise, at least in gaming though. But Yorkfield has a lot more head room and its advantage above Agena makes this like neglectible, for numbers people though. Although I used numbers a long time back to notice Agena does better at higher clocks than Yorkfield but well, everyone has got to be a bit hypocrite now and then:p:


loser CPU
At this point I feel so bad for Chad really now I called him the Italian Chad, Chad at least shown some... you know, a slight form of arguments.

Seriously mate, get a clue there:rolleyes:

[accord99]

Then I think we both have different opinions about underperforming. Comparing it to competition then to define it as underperforming is nuts IMO.
It's the only meaningful definition of performance, how well something performs against its current rivals. A Formula 1 car is incredibly fast versus virtually any other car in the world but if it's 1 second a lap slower than the fastest Formula 1 car racing at the same time, then it's slow.

And when you have a two company battle like in CPUs, it's pretty much a given that a CPU from one company will also be compared with a CPU from the other.



Well, Phenom actually does better at higher clocks than Penryn scaling wise, at least in gaming though.
That has to do with Yorkfield hitting the GPU wall much earlier.

[Rammsteiner]

It's the only meaningful definition of performance, how well something performs against its current rivals. A Formula 1 car is incredibly fast versus virtually any other car in the world but if it's 1 second a lap slower than the fastest Formula 1 car racing at the same time, then it's slow.
No, then it's slower than the competition, not slow, especially not if another team got a new car the day before with whatever technology that makes it go 20Km/h faster than any other car. That means this team is very good, not that all of a sudden the rest is slow or underperforming.

If this said team would introduce a new car which actually is 20Km/h slower than the previous car, uses even more fuel etc, then it's slow and underperforming, independend whether this car is still better or worse than the cars of the competition.


And when you have a two company battle like in CPUs, it's pretty much a given that a CPU from one company will also be compared with a CPU from the other.
Yeah, I can imagine that happening:ROTF: Still doesnt justify to call something like that though, at least not as a fact.



That has to do with Yorkfield hitting the GPU wall much earlier.
Could be, but not sure since there was a rising line still for both CPU's. If there was this GPU wall, wouldnt the FPS remain the same?

[STaRGaZeR]

That may very well be the case,but don't you think we need some real reviews with variety of apps to conclude what you said?Agena was pretty close to even Penryn in a lot of apps,but it trailed it in others too.It won't be black& white so that you can say 4.5Ghz will be needed to match 4Ghz Penryn.In certain apps it will do better per clock,in some worse,but the average is still unknown.But i say again,those numbers may indeed be true,we just need to wait for some real tests.

Sure, we have to see the actual numbers. Mine are pure speculation, but you can perfectly guess the margin, it'll be around 200-400MHz average depending on the application. At it will outperform even Core i7 in some of them. I think nobody will tell you otherwise.


Then I think we both have different opinions about underperforming. Comparing it to competition then to define it as underperforming is nuts IMO. K10 is better than K8. When a succesor performs worse, then I call it underperforming. It's not like an architecure is designed today, tomorrow a competitor releases a better one and thus they design a better one the day after and launch it. Although AMD and Intel do have to compete with each other price and performance wise, their advances are pretty much independend most of the time.

We have exactly the same definition. Look at your example, you're comparing the current processor with the predecessor. I'm not talking about comparing it with the competition, I'm talking about comparing it to whatever you want. You can't measure anything without a reference point, so don't insult Einstein ;). You can't say i7 is fast, in absolute. Therefore your previous post was BS, hence my response, as you were measuring speed without references.

[Rammsteiner]

We have exactly the same definition. Look at your example, you're comparing the current processor with the predecessor. I'm not talking about comparing it with the competition, I'm talking about comparing it to whatever you want. You can't measure anything without a reference point, so don't insult Einstein ;). You can't say i7 is fast, in absolute. Therefore your previous post was BS, hence my response, as you were measuring speed without references.
K, you lost me here now:p:

You said Phenom was underperforming compared to Yorkfield, where as I said it isnt underperforming because it's faster than K8. But well, I guess this whole thing is turning out in nothing once again:p: Lets just get back on-topic (unless you could make a quick line what Im missing here though:confused:).

[knopflerbruce]

loser CPU

Loser user, I think that's a more bullet proof statement;)

On-topic: THis CPU looks very nice, but does this mean AMD will stop making 90nm parts (Windsor) very soon?

[roofsniper]

loser CPU

ummmmmm do you have anything to say or should i just go trolling in the intel section posting "i7 sucks" on every page without saying anything else? if you don't think its a good cpu then why? personally i don't think its a bad cpu but its not a good one. it came way too late and if you compare it vs the x2 6000 at stock it is pretty even besides power consumption. but if we look at how both of them overclock this could very well be better than the 6000 and be amd's best dual core cpu.

[G0ldBr1ck]

loser CPU

its really getting old friend. every thread I have read in the last week has been filled with the same crap from you.

think im gonna get together all your troll posts (will take quite a wile) and send it to a mod. Or you can just chill out and stop trying to make things unpleasent for others.

[Hell Hound]

ummmmmm do you have anything to say or should i just go trolling in the intel section posting "i7 sucks" on every page without saying anything else? if you don't think its a good cpu then why? personally i don't think its a bad cpu but its not a good one. it came way too late and if you compare it vs the x2 6000 at stock it is pretty even besides power consumption. but if we look at how both of them overclock this could very well be better than the 6000 and be amd's best dual core cpu.

Will the x2's get the hkmg treatment or not.Maybe 45nm x2 8400+ BE or something like that.:shrug:

[roofsniper]

Will the x2's get the hkmg treatment or not.Maybe 45nm x2 8400+ BE or something like that.:shrug:

idk about hkmg but there are making a 45nm dual core. regor. it is the athlon x2 235 or 240 at 2.7ghz and 2.8ghz respectively.

[Hell Hound]

Will that bring 4mb l2 with ddr2 1200 mem controller ?

[informal]

Will that bring 4mb l2 with ddr2 1200 mem controller ?

AFAIK,it will have 2x1MB L2(dedicated ,not shared),no L3 and will be AM3(backwards compatible with AM2+).It should be based on Deneb/Shanghai core and support up to ddr3-1333 officially and upwards unofficially .
It should be faster than current Athlon X2s based on Agena cores,somewhere in E8xxx league ipc wise.

[roofsniper]

AFAIK,it will have 2x1MB L2(dedicated ,not shared),no L3 and will be AM3(backwards compatible with AM2+).It should be based on Deneb/Shanghai core and support up to ddr3-1333 officially and upwards unofficially .
It should be faster than current Athlon X2s based on Agena cores,somewhere in E8xxx league ipc wise.

yea thats about it. heres a link that has some of the info. http://en.wikipedia.org/wiki/List_of_future_AMD_Phenom_microprocessors#.22Regor .22_.2845_nm.2C_dual_core.29

[Macadamia]

I don't think Regor is that much of a leap, though. It's just a very small K10.5 die.
Might be slower compared to current Kumas clock for clock because of lack of L3, but they definitely use much less power for one.

[roofsniper]

I don't think Regor is that much of a leap, though. It's just a very small K10.5 die.
Might be slower compared to current Kumas clock for clock because of lack of L3, but they definitely use much less power for one.

yea i bet regor will not be that popular. for amd users the low end seems to be becoming triple cores not dual cores.

[Ghostbuster]

Regor may end up as "EE" (Energy Efficient) processors. Time will tell.. ;)

[roofsniper]

Regor may end up as "EE" (Energy Efficient) processors. Time will tell.. ;)

i think kuma will as well. that l3 has to take up a decent amount of power which may be one of the reasons why they don't have it in kuma besides cutting costs.

[G.Foyle]

i think kuma will as well. that l3 has to take up a decent amount of power which may be one of the reasons why they don't have it in kuma besides cutting costs.

Kuma = cut Agena core... no way it has low power consumption under any conditions. Kuma gets beaten hard by Deneb (yes, two cores more) in terms of power consumption and efficiency.

[Macadamia]

From server demos, 2.5Ghz Shanghai has same idle power with either 1.8Ghz or 1.9Ghz Opterons.
It wouldn't be far to say that the Phenom Is are the same as 65nm Opterons.


From here http://www.madboxpc.com/review-amd-phenom-x4-9350e-x4-9150e/6/
We see that the EE 65W Phenoms take the same power in idle as they're in the same voltage bin. Same applies for tri-cores, we see a trend here.

And then there's this: http://images.anandtech.com/graphs/amdphenom9350e_070108145125/17176.png (I hope Shintai's looking, as it's his fave site)
There are 2 P-II X3s, 2.8Ghz and 2.6Ghz. Both probably are in the same voltage bin in terms of idle.

So if an 2.6/2.8Ghz X3 uses same power as E8400 at Idle, it's quite nice. At load it will use more though, can't ever avoid that as it's 3 vs 2.

[roofsniper]

Kuma = cut Agena core... no way it has low power consumption under any conditions. Kuma gets beaten hard by Deneb (yes, two cores more) in terms of power consumption and efficiency.

ah yes. :slap: ment to say propus.