Why would my CPU mark not match your Benchmark for That CPU

[DPressm]

I would guess that it is a function of the other process that might be running in the background?

The chart shows my processor at: Intel Core i7-2860Q@2.5ghz 7960 with 24gb of RAM

But when running the test I get only 7675 ??? Did I get a bad chip?

Could it be that my power plan is throttling? I am running "High Performance" with min and max processor set at 100%


Dell Alienware 17x R3 i7-2860Q@2.5ghz with 24 Gb of 1600 mhz ram (all 24GB is 1600mhz) and AMD Radeon 6870M 750 Gb 7200rpm main drive and 320Gb 5400 Gb Secondary drive

FYI I found this machine on Dell outlet - 16Gb ram initially I upped it to 24Gb and added the second drive - $1,750 including tax + $279 for memory (to be reduced when I sell the obsoleted 4Gb) - I had the second drive from my old laptop.

And before you ask I develop OLAP databases (Essbase) - very large ones.

I will be submitting my full benchmark in an hour or so when the format for the second drive completes - It seems to be the only 2860Q so far so it should be easy to find.



Thank you.

[David (PassMark)]

You are 3% below the average. This is a pretty small margin. You might find that just running the tests again gains you a few %.

Certainly not an indication of a bad chip. By definition someone needs to score below the average. Can't all be above average

There are 53 other samples of machine with this CPU in our database.

If you are doing serious DB work get a SSD. In our DB work we get about 5x increase in the transaction per second from a SSD compared to a traditional HDD. Of course with 24GB of RAM you could setup a pure RAM drive and get maybe 20x the performance.

[DPressm]

Thanks and a follow up question

Thanks - Wow I had heard that the ssd's were only adding a 5-10% so that is suprising.

Would a Jump drive in a usb 3.0 port be as fast as an SSD?

I presume that an external SSD connected to the SATA port I was suprised to see on the side would be just as good as an internal sata.


Thanks for your quick response and help.

Actually while I have you any difference in performance hooking up two external monitors via the hdmi ports vs the mini-Display port to dvi to dvi splitter????

I wonder why when I downloaded the comparisons I did not find any 2860QM's ????

[David (PassMark)]

It of course depends on the drives, the mix of read / writes, the record sizes and how random the access is to the drive.

USB flash drives tend to do really badly writing small blocks. So the flash drive could be even worse than a internal HDD. Internal SSD are way faster than USB flash drives.

Best way to find out how your particular DB goes is to run some benchmarks.

Are you trying to show the same thing on 2 monitors, or have a extended desktop? If the spliter is what I think it is, you can't do an extended desktop.

[DPressm]

Thanks and another follow up (please mr president)

I feel like a reporter at a press conference but I really appreciate your answers.

1. Understand the usb flash drive answer but do I presume correctly that an external drive plugged to the sata port would be the equal of an internal sata drive (whether ssd or rotating)?

2. As for the mini display port to display dvi to dvi splitter - I have used the dvi splitter in the past to go from a nVidia quadro 440 card's DVI port to a splitter to two monitors (and actually it has two DVIs so I was able to drive four monitors total) all with extended desktop. Or is that something special?

3. I see (but do not understand) in the display port standard that it can have 1,2 or 4 data lanes - I presume that that means 1,2 or 4 monitors extended? This is with my AMD Radeon 6870M card.

4. Back to benchmarks I was suprised to see that my 320gb 5600 rpm disk was often faster than the 750 gb 7200 pm disk - I presume that this is dependent on the distribution of the random reads and writes across a larger region?

[David (PassMark)]

SATA works the same regardless of if the drive is in the box or out of the box.

The Quadro card was special I think. We had a similar one. It went from a custom connector (non standard) with lots of pins, to two DVI with a Y cable.

The display port 1.2 standard allows daisy-chain connections with multiple monitors. But you would need compatible monitors and the right video card. The number of lanes refers to the total bandwidth available.

For the drives, I wouldn't want to generalize. There might be many reasons. e.g. fragmentation, position of the files on the disk, differing data density, cache size, interface in use, etc..