Performance difference between 16 GB 1066 MHz DDR3 & 32

[drchristopher]

I have two W7-U 64bit machines.

Dell T5400 desktop,
Processor - one-X5450 Intel Xeon (L2 Cache-2x6MB, 3.0GHz, 1333MHz FSB)
RAM - 32 GB PC2-5300 (667 MHz ECC Fully Buffered)
Hard drives - two 1TB 7400RPM RAID 1 (mirrored drives)
Video card - nVidia Quadro FX580 512MB GDDR3

Dell M6400 laptop,
Processor - one quad-core Intel QX9300 (2.53GHz, 1067MHZ, 12M L2 Cache)
RAM - 16 GB (DDR3-1066 MHz SDRAM,)
Hard drive – one Samsung 256 GB SSD PM800 series
Video card – nVidia FX 3700 M (1-GB DDR3)

The M6400 is noticeably faster than the T5400.

1) The QX9300 does not have Hyper-Threading capacity so I do not think there is any technological advantage over the X5450. Any thoughts here?

2) Could 16 GB of DDR3-1066 MHz SDRAM RAM be a better work horse than 32 GB of DDR2 667 MHz ECC Fully Buffered RAM, but if it is not the RAM what else could it be?

3) Any thoughts on if the HD speed could (also) be behind the observed speed difference between the two machines?

4) If I added a second Xeon X5450 processor would that noticeably increase my observed speed?

Thanks everybody in advance! Appreciate your time.

drchristopher

[Tekxpert]

1. more than CPU technical attributes the effects by your points 2,3 & 4 wil have bigger effect.

2. ECC fully buffered ram has more latency than regular DDR3 due to the extra error checking, but when you combine the bonus of extra bandwidth from the DDR3 (8 bits per clock cycle)1066MHz compared to DDR2 (4 bits per clock cycle) 667MHz, the DDR3 1066 will be quicker no doubt.

3. The SSD on machine no.2 will definately show quicker response with execution of applications & OS. Mechanical interface of traditional HD is hardly a match for SSD with no physically moving parts.

4. The bottle neck would still be the chipset on the motherboard of machine no.1 in combination with the slow EEC DDR2 @ 667MHz and traditional mechanical HD, unless the applications you run sit in the cpus cache most of the time.