X3 445s: Denebs in disguise?

[DrMrLordX]

Take it for what it's worth, but the Egg is apparently rife with x3 445s that are Denebs with failed L3 and one failed core:

http://www.newegg.com/Product/Produc...82E16819103872

That means the x3 445 would be a relatively easy unlock to a 3.1 ghz Deneb at a price below that of the x2 555 BE. I don't know if 445s elsewhere are AC chips but it's worth investigating . . .

Also, there SHOULD be an x3 435 variant (http://www.cpu-world.com/CPUs/K10/AMD-Athlon II X3 435 - ADX435WFK32GM (ADX435WFGMBOX).html) floating around out there that may also be a Deneb in disguise, but I haven't found many etailers carrying them yet. Amazon has it listed as "currently unavailable": http://www.amazon.com/gp/product/B0...mp=1789&creative=9325&creativeASIN=B003PJ6PWA

Provantage has a C3 x3 440 available, but at a lousy price compared to the 445 on the Egg: http://www.provantage.com/amd-adx440...x~7AAMD2F9.htm

On a partly-related note, here's some info about which older x3 and x4 variants will (and won't) make it to market with a C3 stepping: http://www.cpu-world.com/news_2010/...Athlon_II_X2_processors_with_C3_revision.html

edit: woops! Looks like I posted this to PFI by mistake. Talk about a brain fart there. Thanks to the mods for moving it here with no muss or fuss.

 

[formulav8]

I'm kinda of forgetting about being in the unlocker club. None of the attempts were for my rig but for my side job of hardware testing purposes so it wasn't a huge deal.

I've tried 2 different 5000+ (2.2ghz) X2 chips and could get neither to unlock anything.

I've had one unlock success and that was a Sempron 140 that I unlocked the 2nd core and I was actually able to get a pretty stable 3.8ghz oc from.

The Sempron has actually been sitting around in a unmounted mobo doing nothing for along time now. I still don't know what to do with it....

I also remember buying a x3 440 for a customer and on the side I tried unlocking the 4th core and was a no go either.

So I only have a 25% success rate which I consider pretty bad odds.


Jason

 

[Sonikku13]

X2 5000+ CPUs are based on the K8 microarchitecture, and they don't unlock.

 

[WaitingForNehalem]

X2 5000+ CPUs are based on the K8 microarchitecture, and they don't unlock.

There is also a K10 version: http://www.newegg.com/Product/Produc...82E16819103716

 

[DrMrLordX]

So I only have a 25% success rate which I consider pretty bad odds.

My success rate is currently at 0%. What I find most interesting is that with these C3 versions of budget chips, they're still dumping Denebs into the Athlon II product lineup. People still pay what I would consider to be a premium for 555 BE and 550 BE processors, making the x3 445 an interesting specimen at a price about $10 below what you can expect to pay for a 555 BE online.

What is also interesting is that, at least in the case of the C3 Athlon II x2s and x3s, there is very little retail availability of the lower-end models (example: you can buy an x3 445 or x2 260 on the Egg, but not an x3 435 C3 or x2 220 C3). Some Athlon II parts did not make the transition to C3 at all (x4 620, for example), but for the low-end ones that did, they're not making it easy on buyers to find them.

 

[cbn]

What I find most interesting is that with these C3 versions of budget chips, they're still dumping Denebs into the Athlon II product lineup.

DrMrLordX,

In some other threads I have seen the terms "hot" and "cold" mentioned with respect to different ways AMD can run their fabs. Apparently running a fab "hot" can enhance certain yields while lowering other yields.

Can you or someone else briefly explain what this means?

 

[DrMrLordX]

Your guess is as good as mine. I'm not enough of an insider to know what that might mean.

 

[cbn]

Your guess is as good as mine. I'm not enough of an insider to know what that might mean.

I found this on a search. JFAMD has mentioned this "hot" and "cold" strategy before at least a couple of times.

http://forums.anandtech.com/showthread.php?p=29920178

Not necessarily sure on how much downclocking happens, it is very dependent on how you are running your fabs, what you are targeting, etc. combined with the overall market demand.

Let's say that you need to satisfy demand for a lot of high bin speed parts. You run your fab hot, your yields go down, and most likely you end up with parts that are netting out at their true speed.

If you are driving for volume, you will probably run your fab cold, get better yield but lower clocks.

In the first case, because of the fallout (and the fact that you are looking for higher clocks), you will probably have poor overclockers because each of the die is pushed closer to the limits (less headroom). This is regadless of the marked speed.

However, in the second, you could get more headroom because you might have more than your distribution of mid bins and some of that yield can be pushed lower.

But the to all of this is how many parts did you need to yield at what speed from the wafer?

The net of all of that is that you can never assume that you will get better overclocking from lower speeds. It has just as much to do with the recipe (input parameters) as it does with the distribution curves. It is probably true that you have better odds with lower speed, but you might not see as large of a proportional gain. Your mileage may (will) vary.

Processors are snowflakes, proceed accordingly.

It sounds like if AMD is needing to run fabs "hot" for whatever reason our chances of getting a chip to "core unlock" or "overclock" will be reduced.

Since Thuban has less competition from Intel (and the die is larger) maybe AMD is running fabs "cooler" for that product line? For this reason I hope AMD releases those Thuban quad cores. Maybe they will unlock better than the parts you have mentioned in this thread DrLordMrX?

 

[DrMrLordX]

The only reports I've heard of 960Ts being unlocked were not terribly good. Here, read for yourself (take it for what it's worth, it IS Tom's, but at least they had a 960T to play with): http://www.tomshardware.com/reviews/phenom-ii-x4-960t-zosma-core-unlocking,2627-2.html

Our sources at ASRock tell us that, out of 16 samples the company has tested, six are able to unlock to six cores. That's a 37% chance in a fairly small sample size.

Furthermore:

We tested this out a bit using our Phenom II X4 960T sample and found that hitting 3.9 GHz was not a problem for this 3 GHz chip using a 1.425V BIOS setting. Turning on the two locked cores forced us down to 3.6 GHz to avoid crashing as Windows loaded up, and the extra heat forced a voltage reduction to 1.4V.

Hot or cold, I don't know. The sample size is limited, and we may never see 940Ts or 960Ts tested in any significant number ever again.