I have a 2.26B@3.2GHz, 188MHz fsb on a SiS661 running dual channel. Video card is a 5700 Ultra.
I just put together a fourth spare gaming rig using a mobile P4 1.6GHz (512k) @ 2400MHz on a P4S-800MX which only supports single channel DDR. Got the cpu for $29 shipped on ebay and the board refurbed for $29. You can't beat that price, it was so dang cheap I'm gonna build a fifth system.
That's when I started wondering whether I should invest a few more bucks on the fifth system and get a dual channel board. I benchmarked the 3.2GHz system with both single and dual channel and was very suprised on what I found out.
We keep hearing about how bandwidth hungry a P4 system is, so I thought there would be a huge difference.
Sandra memory bandwidth tests show a 50% boost, that's absolutely huge.
3DMark and gaming benchmarks show a 2.8% boost, almost too small to notice. I double checked the results and that's correct. Can anyone give a technical reason why the real world benchmarks show such a small performance increase from single to dual channel DDR for P4?
Of course the fifth system is gonna be the same as the fourth, not worth the extra $40 for a dual channel motherboard with decent overclocking options. These mobile P4's default to a 12x multiplier for the desktop and make a sweet bang for the buck, easily matching the incredible value of a overclocked lower range Sempron.
I just put together a fourth spare gaming rig using a mobile P4 1.6GHz (512k) @ 2400MHz on a P4S-800MX which only supports single channel DDR. Got the cpu for $29 shipped on ebay and the board refurbed for $29. You can't beat that price, it was so dang cheap I'm gonna build a fifth system.
That's when I started wondering whether I should invest a few more bucks on the fifth system and get a dual channel board. I benchmarked the 3.2GHz system with both single and dual channel and was very suprised on what I found out.
We keep hearing about how bandwidth hungry a P4 system is, so I thought there would be a huge difference.
Sandra memory bandwidth tests show a 50% boost, that's absolutely huge.
3DMark and gaming benchmarks show a 2.8% boost, almost too small to notice. I double checked the results and that's correct. Can anyone give a technical reason why the real world benchmarks show such a small performance increase from single to dual channel DDR for P4?
Of course the fifth system is gonna be the same as the fourth, not worth the extra $40 for a dual channel motherboard with decent overclocking options. These mobile P4's default to a 12x multiplier for the desktop and make a sweet bang for the buck, easily matching the incredible value of a overclocked lower range Sempron.
Facebook
Twitter