DrMrLordX
Lifer
- Apr 27, 2000
- 21,637
- 10,855
- 136
This review looks bad, but let's be honest here: AMD has shipped a quad-core part and it works at stock speeds. It has bugs which seem to prevent some, if not all the B2 chips from overclocking well. Anyone looking for a drop-in quad replacement for a slow-ish X2 will probably like this thing. Anyone looking for a chip to run in a platform hosting 2-4 Radeon 3870 cards will like this thing.
Anyone looking to overclock will fear and/or hate it.
Seriously, I was thinking about getting one, lapping it, putting on some good ol X23, and seeing how far I could take the chip. Now I'm not so sure about that. Viditor said he was waiting on B3 chips and I can see why . . . these chips look risky for the enthusiast. Even if there are some X4-9500s and 9600s out there that will not display problems at speeds of 2.4 ghz or higher, they may still experience difficulties at higher clock speeds (as did Anand's review sample) and they may be virtually indistinguishable from dud OC chips. Power consumption on these chips isn't great either, and unless it improves on future steppings I see this as being a sticking point for some folks even after the B3 stepping chips become available.
AMD has not explicitly sold these chips as anything other than 2.2 and 2.3 ghz parts. Sadly they touted their overclocking tool during a release of chips that are not, in and of themselves, reliable overclockers.
Show me the B3 chips that can launch at stock speeds of 3 ghz. This launch is mostly a dud. Anyone who does get a good B2 chip that can run stable at 3 ghz will be lucky, I suspect.
His journalism requires no excuses. He told the truth about the ordeal he went through just to get a review sample that could be put through its usual paces. Compare his article to the mess on Tom's Hardware. Tom's didn't (and wasn't allowed to) run anything other than 3dMark06 after overclocking which prevented them from encountering the difficulties Anand had with the chip past 2.6 ghz. They even reviewed the X4-9700 which has since been recalled! They were also forced to label their power consumption test with the warning "Power Dissipation - No Reliable Data". The fact that the chip in their test chewed up more power at load than the QX9650 was still ugly.
Anyone looking to overclock will fear and/or hate it.
Seriously, I was thinking about getting one, lapping it, putting on some good ol X23, and seeing how far I could take the chip. Now I'm not so sure about that. Viditor said he was waiting on B3 chips and I can see why . . . these chips look risky for the enthusiast. Even if there are some X4-9500s and 9600s out there that will not display problems at speeds of 2.4 ghz or higher, they may still experience difficulties at higher clock speeds (as did Anand's review sample) and they may be virtually indistinguishable from dud OC chips. Power consumption on these chips isn't great either, and unless it improves on future steppings I see this as being a sticking point for some folks even after the B3 stepping chips become available.
AMD has not explicitly sold these chips as anything other than 2.2 and 2.3 ghz parts. Sadly they touted their overclocking tool during a release of chips that are not, in and of themselves, reliable overclockers.
Show me the B3 chips that can launch at stock speeds of 3 ghz. This launch is mostly a dud. Anyone who does get a good B2 chip that can run stable at 3 ghz will be lucky, I suspect.
Originally posted by: AlabamaCajun
Still doesn't excuse the journalism does it?
His journalism requires no excuses. He told the truth about the ordeal he went through just to get a review sample that could be put through its usual paces. Compare his article to the mess on Tom's Hardware. Tom's didn't (and wasn't allowed to) run anything other than 3dMark06 after overclocking which prevented them from encountering the difficulties Anand had with the chip past 2.6 ghz. They even reviewed the X4-9700 which has since been recalled! They were also forced to label their power consumption test with the warning "Power Dissipation - No Reliable Data". The fact that the chip in their test chewed up more power at load than the QX9650 was still ugly.