AMD X2 6500+ coming

[oohms]

I can't believe they would be using phenom X4 cores to make dual core phenoms.... what a waste

[freeloader]

I can't believe they would be using phenom X4 cores to make dual core phenoms.... what a waste

It's hardly a waste if the core(s) were defective to begin with. That's just good business. Get something from nothing.

[RejZoR]

Actually all do this so why complaining? Graphic card vendors do that for ages. Chips with dead pipelines end up as lower end GPU's with fixed BIOS so it doesn't try to use dead pipes.

[savantu]

Your own pics already explained the issue well. And I'm not kidding.

Did it occur to you that the extra core might be responsible ?

[Macadamia]

Did it occur to you that the extra core might be responsible ?

I was implying Conroe/Allendale/Wolfdale in Crysis.

And IPC obviously depends on whether the app is cache-friendly. Games show that with the 2MB L3 left for itself, Kuma seems to perform better compared to Conroe on an overall scale. See how the 2.6Ghz Allendale struggles to get in pace with a 2.33Ghz Conroe.

The Cinebench R10 runs are controversial- The first slide shows a 20% deficiency for the 2.4Ghz Q6600 vs 9600. The new version pits the 2.3Ghz 9650 at 10% slower.

Decide yourself.

[Shintai]

I was implying Conroe/Allendale/Wolfdale in Crysis.

And IPC obviously depends on whether the app is cache-friendly. Games show that with the 2MB L3 left for itself, Kuma seems to perform better compared to Conroe on an overall scale. See how the 2.6Ghz Allendale struggles to get in pace with a 2.33Ghz Conroe.

The Cinebench R10 runs are controversial- The first slide shows a 20% deficiency for the 2.4Ghz Q6600 vs 9600. The new version pits the 2.3Ghz 9650 at 10% slower.

Decide yourself.

I think you mix up 800Mhz FSBs and 1333FSBs too.

[MrMojoZ]

Actually all do this so why complaining?

Because it is AMD doing it in this case and AMD punches babies.

[Rammsteiner]

I think you mix up 800Mhz FSBs and 1333FSBs too.

Do you really want to act like... 'that' or what

Then just take a look at the Phenoms ffs? From a clear point of view, which does not need a single piece of 'fanboyism', you can see the following:

X4 2.4Ghz = 41.8fps
X3 2.4Ghz = 42fps
X3 2.3Ghz = 42fps

Thus, how hard is it to see that a Phenom X2 2.3Ghz would actually perform faster than any other Phenom in non multi-threaded environments, or at least in Crysis?

And if the two salvaged cores are not affecting anything and the two active cores do work for 100%, 3Ghz+ shouldnt be any problem at all. Now how can people still say this wouldnt outperform a K8?

And oh yes, as gOJDO noted, the graphs are a bit weird so to say so make sure you read them correctly. It took me 3 times before I noticed it as well.

[Shintai]

Do you really want to act like... 'that' or what

Then just take a look at the Phenoms ffs? From a clear point of view, which does not need a single piece of 'fanboyism', you can see the following:

X4 2.4Ghz = 41.8fps
X3 2.4Ghz = 42fps
X3 2.3Ghz = 42fps

Thus, how hard is it to see that a Phenom X2 2.3Ghz would actually perform faster than any other Phenom in non multi-threaded environments, or at least in Crysis?

And if the two salvaged cores are not affecting anything and the two active cores do work for 100%, 3Ghz+ shouldnt be any problem at all. Now how can people still say this wouldnt outperform a K8?

And oh yes, as gOJDO noted, the graphs are a bit weird so to say so make sure you read them correctly. It took me 3 times before I noticed it as well.

2.3 vs 2.4..~4% (Insert margin of error)

[Donnie27]

because its a black edition. And you dont buy a black edition unless your going to play with the multi... otherwise, buy the cheaper locked dual cores. and With a 2.3ghz phenom dual, matching pretty much a 3.0ghz brizbane, its safe to say that the 6500+ will easily outstrip the 6400+ once the multi is upped a bit.

Its like buying a qx9650 and leaving it at stock multi..and upping the FSB... it makes no sense! a black edition is meant to be OC'd hence the huge tDp.

As others have said, then call it a Phenom X2=P

[Donnie27]

go to link in villians post

It wins one out of three and it is faster? Clearly not enough info there.

[sdsdv10]

I can't believe they would be using phenom X4 cores to make dual core phenoms.... what a waste

It's hardly a waste if the core(s) were defective to begin with. That's just good business. Get something from nothing.

Except if the money you get for selling the limited chip is less than what you can get as writing it off (AMD's business taxes) as a loss....

[villa1n]

As others have said, then call it a Phenom X2=P

obviously phenom is the name designated to triple and quad offerings, and amd would like to differeniate dual core offerings by keeping the athlon name alive and associated with x2 and value offerings, like they have on their mobile line.

[Donnie27]

Edit: It's the model number and the Name.

obviously phenom is the name designated to triple and quad offerings, and amd would like to differeniate dual core offerings by keeping the athlon name alive and associated with x2 and value offerings, like they have on their mobile line.

Yet, still seems a way to confuse folks. Yes, they've lied about the old models as well.

The 3000+ Skinny

AMD is shipping three new Barton-based CPUs: The Athlon XP 3000+ (running at 2.167 GHz), a new XP 2800+ (2.083 GHz), and the first XP 2500+ (1.833 GHz). All three include 512KB L2 cache and a 333-MHz front side bus. The 2800+ Barton-based CPU replaces an existing Athlon XP 2800+ chip that AMD shipped in limited quantities to five PC vendors in the fall of 2002.

The original 2800+ chip runs at 2.25 GHz, which is faster than both its replacement and the 3000+ chip. What gives? "The bottom line is the 2800+ model number is derived from application performance," says AMD spokesperson Damon Muzny. The new 2800+ uses more cache instead of more megahertz to reach a specific application speed level, he adds.

All 2800+ chips shipping in PCs should be Barton-based models by the time you read this, Muzny says. (And chip pricing will be the same, so there's no bargain-hunting opportunity.)

BS because it still ran slower on most tests even with more cache. Also the same price for a slower model with more cache but slower more times than not.

[MrMojoZ]

Yet, still seems a way to confuse folks. Yes, they've lied about the old models as well.

Like when Intel made commercials that said MMX made your internet faster?

[Bobsama]

Maybe 5900+ BE or something would make more sense.

[Rammsteiner]

2.3 vs 2.4..~4% (Insert margin of error)

It's indeed very close. However, you could say as well that the X3 2.4Ghz is about ~4% faster instead.

It wins one out of three and it is faster? Clearly not enough info there.

It's indeed not a very solid review, but considering it still runs only at 2.3Ghz and not ending at the last spot at all but actually tailing the X2 6000+ it's impressive.

But, to name it actually 6500 is indeed a bit confusing there since it clearly does not outperform the 6000+ at any application. Therefor it's the question how we should be reading the missing '+', whether it's worth it to call it a 6500 because it runs behind the 6000+, I dont know. I think we need to see more graphs before making a conclusion about that.

However, it's very clear though that the 6500BE shows more potential in it than even the 6400+, no doubt in that. But for general customers who're not into OC'ing we need a lot more graphs to actually be able to say it's indeed 'better'.

BS because it still ran slower on most tests even with more cache. Also the same price for a slower model with more cache but slower more times than not.

Well, it's been quite 'honest' to say that an extra 512KB L2 cache is about 200Mhz bonus since K8. Ive not been very active in hardware in the XP days though so I cant really comment a lot about that.

But, as far as I understand it your quote is about only Barton but with more cache and slower speeds but eventually with different PR ratings unlike AMD did with the K8. Also dont forget that at this point we're talking about a K8 vs K10 where as cache and Mhz will have a completely different effect.

Yet I cant say AMD was in their right to do that with the Athlon XP as far as I understand it from that quote. But then again, it's all about marketing in the end and it's been 6 years as well. It's not like AMD was the first and last one to pull that

[qurious63ss]

Like when Intel made commercials that said MMX made your internet faster?

Well this is an instruction set that made internet applications run a bit faster just like SSE does with certain applications, so technically they were telling the truth.

[eXa]

And if the two salvaged cores are not affecting anything and the two active cores do work for 100%, 3Ghz+ shouldnt be any problem at all. Now how can people still say this wouldnt outperform a K8?

If its an unlocked BE then yes.. otherwise, no.
u need high multi to really clock the phenoms.

[Rammsteiner]

If its an unlocked BE then yes.. otherwise, no.
u need high multi to really clock the phenoms.

Thus far the only 6500 is a BE. AMD only had one BE with locked multi's and that was the 6400+ so it shouldnt be much of a problem.

Also HTT is becoming less of an issue with SB750, but it seems to be a bit weird for now with RS780D since it looks like setting HTT higher OC's the IGP as well

[freeloader]

Except if the money you get for selling the limited chip is less than what you can get as writing it off (AMD's business taxes) as a loss....

Ahh, you got me there.

[Kuroimaho]

The other danger with the chip is if it becomes too successful and they run out of broken chips will they disable working cores to fill the demand ?

[Macadamia]

The other danger with the chip is if it becomes too successful and they run out of broken chips will they disable working cores to fill the demand ?

No?

People will likely grab the 8750BE then.

[Kuroimaho]

Could AMD say no to Dell ?

[Macadamia]

This is channel only. Every Black Edition CPU is.