New xeon E5-xxxx processors not on the list

[rakkenes]

Hi there, I have not seen any of the Xeon E5 processors on this list, for example E5-2630... I wish someone would add these.

[David (PassMark)]

They are rare CPUs are the moment. So we don't have very many samples for them all to get into the charts.

Results to date are,

Intel Xeon E5-1650 @ 3.20GHz 13,578
Intel Xeon E5-2630 @ 2.30GHz 10,401
Intel Xeon E5-2640 @ 2.50GHz 11,169
Intel Xeon E5-2670 @ 2.60GHz 15,567
Intel Xeon E5-2690 @ 2.90GHz 16,498

[Richard (PassMark)]

Some of the Xeon E5's names were getting filtered by our bad naming detection mechanism. The algorithm has been updated and they should start appearing in the charts.

We have also edited the names of the Xeon's E3 to add a "-" (hyphen) between E3 and Model Number. The CPU's name that we read off of the processor itself does not have it. This should make it easier for our users to locate the series on the CPU list.

[r4dius]

Hi, for some reason, bench for dual "e5-2630" (15,901) is lower than dual "e5-2620" (17,639),
plus dual "e5-2630" is only 3/4 of single "e5-2630" (10,401) x 2
Is there a problem with this ?

[Richard (PassMark)]

The benchmark results are an average of all baselines submitted by our users. When the number of samples are low, bad baselines can affect the CPU Mark more than a CPU with numerous results. Scores can also be affected if the results is from a 32 bit OS rather than a 64 bit OS. 64 bit OS system generally tend to score higher.

In this case, looking at the raw results themselves. There were three submissions for We have 3 samples for E5-2630, with one submission being really bad (32 bit OS, incorrect number of cores detected). This is probably from a poorly setup machine. We have removed the faulty submission and the chart should update overnight. The remaining submission is from a 64 bit machine scoring in around 19,800 and a 32 bit machine scoring around 16,000. For the E5-2620, we have 2 results from 64 bit machines each around 17,500 mark.

From the limited data, you can probably conclude that a properly set up E5-2630 on 64 bit platform will perform better than E5-2620. Once we get more submissions, the scores will flatten out and there will be less of an impact individual baselines.