PCMark05 likewise confirms the slight gain the 1333MHz FSB brings a boost of a mere percentage point to most tests except Memory, where those extra cycles give you a four per cent increase on PCMark scores. The FSB factor is also highlighted by our SiSoft Sandra readings.
Incidentally, we put the difference in PCMark05 CPU scores between the E6700 and E6750 down to the inherent errors in PCMark05's readings - had we run the benchmarks more time, these errors would be averaged out and we'd get CPU scores for each processor that are near enough identical.
SiSoft Sandra memory results
Bandwidths in MBps
Longer bars are better
Latencies in ns
Shorter bars are better
When we were testing the E6750 we got the impression that it was throwing off noticeably less heat through the CPU coolerthan the E6700 was. We were using a Zalman CNPS8000 cooler with the fan turned down to a reasonably low speed to keep the noise down.
We decided to investigate further by running Intel's Thermal Analysis Tool. The software monitors the temperature of both cores on a dual-core processor and has the facility to drive the cores at full load. Unfortunately it gets confused by a quad-core CPU so you can't compare the figures for the QX6800 and QX6850 with the E6700 and the E6750 but even so they make for interesting reading.
Intel Thermal Analysis Tool results
Temperatures in °C
When idle, the E6750 runs 26°C cooler than the E6700 and is 19°C cooler under full load. Those figures are clearly being read from the thermal diode in the processor core so you won't see a 20°C drop in the heatsink temperature, but there is a very noticeable difference in air temperature which is, of course, the reason we ran this check in the first place.
Verdict
The combination of reduced price, increased performance and lower heat output makes the Core 2 Duo E6750 a compelling buy. For the time being we're not sold on the P35 chipset or DDR 3 system memory, but that's fine as we're biding our time and waiting patiently for X38.
Intel Core 2 Duo E6750 2.67GHz desktop processor
COMMENTS
Thermal Analysis Tool
I'm not taking the numbers as holy writ however I only ran the tool as I could feel the expelled air was cooler on E6750 than with E6700.
Had i said 'I think it runs cooler' I would fully expect readers to ask how much cooler. This may not be a full answer but I hope it's better than nothing.
your all wet
the e6700 and e6750 are overclocked chips it so happens
that it's OK to do that they can take it but I hate to see people
say this is a bug etc or it's the errata all chips have these things
get over it you would never know of it if they hadn't published
the information I am using one of those chips it's a delight
but when you increase the heat and the voltage you are going
to decrease stability and lifespan of all components live with it.
One other thing these chips will be much cheaper much sooner
than anything AMD ever made.
Intel Thermal Analysis Tool
My understanding is that this utility is only intended to be used with mobile CPUs. It cannot be relied upon to be accurate for other Intel products ( in fact, the consensus seems to be that it is way off).
Doesn't it seem like the faster bus *should* make more of a difference?
Seems to me that the faster bus should have a larger impact than it does (1 to 2%?)
So my question is, can we ID what the real bottleneck is, and maybe the next gen chip can address that?
Still Buggy?
So, have Intel sorted out any of the bugs in the Core Duo, some of which are "interesting" (for "interesting", read "errr ....I'll try something else, thanks")
http://www.geek.com/images/geeknews/2006Jan/core_duo_errata__2006_01_21__full.gif
