Overclocking Results
Before we go too far into this section, there are a number of things that need to be put on the table. Voltage tuning posed a large roadblock since every one of the beta programs we had access to misreported the stock voltage on the R9 and R7 cards. This resulted in excess voltage being pushed to the core so while the clock speeds were spectacular, they never did pass our stability tests as temperatures eventually rose to astronomical levels.
AMD’s reference heatsinks aren’t all that great either since they’re quickly overwhelmed by the temperatures pumped out by an overclocked core. Sure, excellent results are possible when their fans are spinning for all they’re worth but no sane-minded gamer would put up with that racket. It should be interesting to see how non-reference designs hold up but for the time being, we had only have the two AMD designs.
With this in mind, until MSI, ASUS or someone else releases a stable tool, we used AMD’s Catalyst Control center for all overclocking while the reference cooling remained in place.
R9 270X Results
Considering the R9 270X is based off of the Pitcairn XT core, we had some relatively high expectations for it and it didn’t disappoint. Our sample hit a core speed of 1144MHz which represents a 94MHz increase without touching the voltage. The memory eventually peaked at 6084MHz which is a substantial increase for GDDR5 modules rated at 5Gbps. All in all, we were happy with how the 270X reacted to our ministrations and with some additional voltage, it should go even further provided cooling is sufficient.
R7 260X Results
The R7 260X just can’t catch a break. In our standard testing suite it was manhandled by the similarly priced GTX 650 Ti Boost and overclocking was nearly impossible. We eventually settled on a core speed of 1138MHz while the memory fared little better at 6888MHz before its error correction routines kicked in. Perhaps additional voltage would have helped the situation but certainly not with the stock cooler struggling to keep temperatures in check at our relatively pedestrian frequencies.
These poor results may point to AMD pushing the silicon as far as it will go since 1.1GHz is already quite high for a Bonaire XT core that was initially clocked 100MHz lower.
Overclocking Results
Before we go too far into this section, there are a number of things that need to be put on the table. Voltage tuning posed a large roadblock since every one of the beta programs we had access to misreported the stock voltage on the R9 and R7 cards. This resulted in excess voltage being pushed to the core so while the clock speeds were spectacular, they never did pass our stability tests as temperatures eventually rose to astronomical levels.
AMD’s reference heatsinks aren’t all that great either since they’re quickly overwhelmed by the temperatures pumped out by an overclocked core. Sure, excellent results are possible when their fans are spinning for all they’re worth but no sane-minded gamer would put up with that racket. It should be interesting to see how non-reference designs hold up but for the time being, we had only have the two AMD designs.
With this in mind, until MSI, ASUS or someone else releases a stable tool, we used AMD’s Catalyst Control center for all overclocking while the reference cooling remained in place.
R9 270X Results
Considering the R9 270X is based off of the Pitcairn XT core, we had some relatively high expectations for it and it didn’t disappoint. Our sample hit a core speed of 1144MHz which represents a 94MHz increase without touching the voltage. The memory eventually peaked at 6084MHz which is a substantial increase for GDDR5 modules rated at 5Gbps. All in all, we were happy with how the 270X reacted to our ministrations and with some additional voltage, it should go even further provided cooling is sufficient.
R7 260X Results
The R7 260X just can’t catch a break. In our standard testing suite it was manhandled by the similarly priced GTX 650 Ti Boost and overclocking was nearly impossible. We eventually settled on a core speed of 1138MHz while the memory fared little better at 6888MHz before its error correction routines kicked in. Perhaps additional voltage would have helped the situation but certainly not with the stock cooler struggling to keep temperatures in check at our relatively pedestrian frequencies.
These poor results may point to AMD pushing the silicon as far as it will go since 1.1GHz is already quite high for a Bonaire XT core that was initially clocked 100MHz lower.
Last edited: