AMD Readying Phenom II X3 740
- Thread starter PeterStoba
- Start date
Rastalovich
New member
We resigned to the idea that AMD are using cpus that failed an x4 test in someway, although they pass perfectly well with x3 ?
Even with it being the case that x4 can/could be achieved, the test is one that is high enough that although it fails in a lab condition, it would work very well in 99% of enthusiast situations ?
Otherwize why have a spare core hidden in there ? Or indeed manufacture the cpu as inherently 3 cores.
If it is the case, I reckon it's a good idea and u know the greens will like it.
Would make u wonder what Intel do with quads that may fail a lab test.
Even with it being the case that x4 can/could be achieved, the test is one that is high enough that although it fails in a lab condition, it would work very well in 99% of enthusiast situations ?
Otherwize why have a spare core hidden in there ? Or indeed manufacture the cpu as inherently 3 cores.
If it is the case, I reckon it's a good idea and u know the greens will like it.
Would make u wonder what Intel do with quads that may fail a lab test.
PeterStoba
New member
name='Rastalovich' said:
Yes it is odd because I've seen many people unlock the 4th core and it work perfectly, and also overclock the 4th core too, which does make you wonder how they are sold as X3 in the first place? Maybe the yields are high on quads and the demand is higher for the X3s, so they just disable a perfect core and sell it.
I've no idea what Intel do, can't sell them as Duals cos they are different architecture... Interesting point that.
Rastalovich
New member
Lab criteria can be enormously higher than anything even an enthusiast would define, but nevertheless be a level they need to reach to attain sellable status.
It's a good idea if they're using it.
It's a good idea if they're using it.
It's my understanding that Phenom II X2's and X3's aren't strictly failed X4's. One would expect the amount of (partially) defective CPU's from such a production line to be rather low and that if the production of PhII X2/X3's were solely reliant on a constant flow of such processors, there simply wouldn't be near as much of them on the market. I have no doubt that AMD need a certain quota of Dual and Triple Cores and the only way to go about it is to bin functional X4's. It doesn't sound like a great business case but not only does it work for AMD, it works for the enthusiasts (granted a small % of sales perhaps) that buy them. 
Note that AMD don't do this throughout their line up as that truly would be suicidal (Deneb based £28 Sempron 140 anyone?
) and have a separate native Dual Core CPU. On the same note, we are also expecting a native Quad without the Level 3 cache as well (Propus).
Note that AMD don't do this throughout their line up as that truly would be suicidal (Deneb based £28 Sempron 140 anyone?
) and have a separate native Dual Core CPU. On the same note, we are also expecting a native Quad without the Level 3 cache as well (Propus).
Rastalovich
New member
Thing that doesn't back the argument is that they don't have a production line for x3 cpus.
i.e. they don't make a cpu with x3 cores.
Looking at the business side of things, they certainly wouldn't make a x2 cpu and sell it as a single core, by design. Although theoretically they could.
i.e. they don't make a cpu with x3 cores.
Looking at the business side of things, they certainly wouldn't make a x2 cpu and sell it as a single core, by design. Although theoretically they could.
Indeed they don't and in order to fulfill PhII X3 stock, disabled Denebs need to be used, whether they were previously defective or not.
On another note, the 45nm Sempron series are indeed Regor cores (Athlon II X2's) with one core disabled. There have also been reports of them unlocking to duals.
On another note, the 45nm Sempron series are indeed Regor cores (Athlon II X2's) with one core disabled. There have also been reports of them unlocking to duals.
Similar threads
