Tweet
Hi. I have used PassMark for years & recommended it to many others. However - i notice that my benchmarks are gone. I even paid for the service. Instead it seems to be a monopoly of 'high-scores' by the same 8 people who have repeating entries over and over and over and over. Why remove mine? If your system reads a piece of hardware in error - it DOES NOT disqualify me from my score. I worked hard to make an excellent machine, & i want my money back if this is just some type of 'sales' site that excludes anyone not running the newest / most expensive parts.
-
-
We nearly never delete benchmarks. Duplicates from the same hardware are excluded from the average results (but not deleted). Some obvious fake results are also deleted. But this would be well under 0.1% of submitted results.
There is no paid service for this. It isn't something we offer. We do sell the software, but not as a subscription and perpetual online storage isn't part of the paid software.
Or maybe you mean you were once on the leader board for the fastest PCs worldwide, but aren't anymore? Maybe due to a other people submitting faster benchmarks than you, maybe also they are submitting results from different runs from similar hardware? (i.e. duplicate hardware or nearly duplicate hardware).
Can you provide any additional details of what got deleted?
Baselines numbers would be the best. Or API key if you are using that.
-
I was able to find some of the benchmarks. But they all say "Excluded" - either too many repeat submissions, or Anomaly in results. My results were in the top 20, and also top 10 for this month, but now removed.
I have spent 100s of hours learning how to overclock, undervolting, timings, building, matching chip mfgrs., and now all my results are 'invalid'??
But the same people who have been the same repeat leaders seem to have no issues maintaining it. (one guy is up there 7 times in top 20)
My score was valid. IDT anyone else has created an open loop liquid cooled hard lined PC with the same parts as mine - so how could anyone know what it would run like with countless hours of tuning / tweaking?
Comment
-
i have an API Key :: <key removed from public post, see internal ticket #FNJ-94375-561 for details>Comment
-
Excluded doesn't necessarily mean invalid. Nor does it mean "gone" or deleted.
Excluded results are either
1) Valid results that come from the same machine in a short period. These aren't hidden nor removed from the site. They are just excluded from our graphs. For example we didn't want global averages to just be an mostly an average from one person / machine.
2) Dubious benchmark results that are impossible.
I looked up one of what we think are your results.
https://www.passmark.com/baselines/V...d=294715539362
The hard drive detected was "PNY CS2311 500GB SSD", which is a 3.5" SATA drive. It's performance can never be above 500MB/sec, due to the SATA interface.
But somehow the benchmark result for this PNY drive in your machine was 37,231MB/sec. So impossible.
So at least for this baseline file, the result is impossible. So the exclusion seems to be valid and the benchmark isn't valid. We can't mislead the general public to think that this PNY drive can do 37GB/sec.
It would be a valid result if your drive was flagged as a RAID setup or RAM drive.
But maybe this is bad detection of the drive under test by the software? I think this is the real issue. What was the process you were using to end up with the PNY drive benching at 37GB/s ?
Comment
-
Hi. Thanks for checking into this.
It should be reading the drive as a Samsung 9100 Pro (Gen 5 PCIe w/ Gen5 PCIe Slot).
I do have a RAID0, but its results are unpredictable, so i usually end up re-doing the test with the Samsung Gen 5 NVMe (But its in the Intel version of Disk reading/detection).
I have had a PNY drive, but it is not fast. It is just for emergency space if needed when doing asset creation / mods.Comment
-
Samsung 9100 Pro is a very fast drive, but it also can't do 37GB/sec. So there is something else going on.
37GB/sec is RAM drive speeds or disk cache (and even then very fast RAM and CPU are required to get a RAM drive to 37GB/sec).
Theoretically max of DDR5 at 6400MT/s is only 51GB/sec and this is before the (very significant) file system, disk device drivers and operating system overheads are considered.
Comment
-
Ram is at 9000MHz real-time, and CPU is almost 6.0Ghz. And the Samsung is set the same as those on the leader-boards of desktops are :that is FAT/ALT file system.
Just like corsair MP700 can not run 32R/31W unless the motherboard has an internal cache system (which ROG DOES) &/or they have an area used only to benchmark with that is an alt file system. Some use both obviously from their scores.
My point is that my system's results are valid. But it (benchmark) read the Samsung as a PNY drive for unknown reason. But if mine are not valid - then neither are the leaders on the top 20.
And hitting 37Gb might have been a one time thing, because i often hit 1300 for a read speed (extreme slow), and have to run the tests multiple times to get anywhere close to even the manufacturers advertised speeds.
I even turn off my PCs lights and disable almost every windows function possible to allow for best results, and i'm guesing that others do the same. So why is mine not possible, but their Gen5 parts just fine?Comment
Comment