I don't really understand the results of the test.

I'm running:
Asus E-Gaming board 299X, i9 7900x, 32GB, Strix 1080 Ti 11Gb, SD Samsung 960
Just about all other benchmarks puts the i9 7900k, with 1080 Ti, in the absolute highest FPS and 3d benchmark tests there is at least ATM. More youtube testing against various configurations including the new Thread ripper confirms as well.

Yet VOmarks are being posted by some of 25% higher score with less spec.

My i9 7900x VOmarks score: 14567 which appears to be the combination between the reading below:
7307(1080p) / 7260(4k)

What is the 4x4 images suppose to look like when the benchmark is running ?
Smooth or should it look like time laps photography ?

I assume it's suppose to look like time laps but if not then something is wrong.
Running it in demo mode, the application crashed and I had to end task for the VOmarks.

But the 4x4 benchmark test completed with the results of 14, 567

I'm not sure if the monitor or it's refresh rate would have anything at all to do with it, but the Gaming monitor is 165Hz Agon with G-sync

Captain, the 4x4 is the actual test, where the demo mode is the same as 4x4 except to show what a single frame is doing. No, it shouldnt be time-lapse unless your PC is having trouble rendering everything - and why ~should~ you expect it to be?

Regarding stability issues and the oddly substandard performance, dunno, but sounds like a major driver issue to me.

Captain, did you run these other benchmarks on your machine? Or are you going based on benchmarks of other people's systems you've seen online?

If it's the latter, I'd suggest trying some other benchmarks, perhaps?

The i9 7900 might not be faster than the best i7 in VendettaMark (honestly, I don't know), but it should be pretty fast. Your score sounds like 1) you're using the integrated intel GPU instead of the 1080? Or 2) You have a serious thermal issue and the CPU is clocking down to prevent destruction.

But, those are just guesses? Something is definitely wrong, and I'm not inclined to immediately suspect the benchmark itself.

@Luxen
I don't know

@incarnate.
I don't have any integrated graphics card on this board
I did base my purchase on online reviews and testing. Both real world gaming tests reviews and benchmark test.

Microcenter did benchmark and stress test it for me and I did run the tests they had on the computer and they seemed to line up with what I saw online.

I am curious about how SD drives might effect benchmark tests or performance.
Most of the reviews online post specs to a degree but many leave out what types of drives they use.
Also Windows 7 test vs Windows 10 seem to actually be different which I don't really understand.

Anyhow I do know that with this SD super fast drive real life experience just folder browsing is poor compared to SATA drive.
Network browsing is sluggish, and there are settings littered throughout Windows 10 that need to get turned off to increase performance of file browsing and other simple things.

There is very possibly something inside that I don't know about that needs tweaking or something.

Recent update has the simplest of tasks like (Shut Down) for the menu item at the start button changed functionality.
So now Windows 10 can't even shutdown manually you literally have to turn off / uncheck quick start some function added to update.

Otherwise the new update shutdown takes you back to login screen again. Sooooo stupid of a feature that they just don't bother to tell you about.

Anyhow, I'll mess with a few things to see if I can improve the performance of the test and stop Demo from crashing the application.

I only was able to run the 4x4 and not the demo.
Is there a crash error log for the VO bench ?

Here is what it looked like during the test.

This is a private link not public youtube just FYI

https://youtu.be/ZtPyfi3cUCY

2FPS is what the nvidia fps counter says and the xplit counter, but it's not showing up in the video.
The results however indicated much higher in spite of the 2FPS viewable bench.

How long does the Demo run before you get results

the 4x4 is a short run, but the demo seems to cycle on and on forever and never produces a result as far as I can tell.

Thats right - the demo isnt actually testing; it cycles continuously just to show you what the real tests do. Doesnt stop till you hit the escape key

I am curious about how SD drives might effect benchmark tests or performance.

While we do some preliminary IO testing, we do not count that in the benchmark score. So, the SD should have little or no impact on VendettaMark.

The only actual benchmark is when you use the "Benchmark" button, giving the 4x4 grid of frames.

The "Demo" mode is NOT a benchmark, it simply is a visually prettier way of seeing what's happening in benchmark mode.

Your scores seem to indicate something is pretty wrong with your system. But, if you don't believe VendettaMark, I'd suggest downloading another major benchmark, maybe one from 3DMark, and use that as another point of comparison?

Fwiw, I'm also running a 960 m.2 drive and a g-synch monitor (34" 1440p@100hz) Going for a higher CPU overclock and a lower gpu clock/memory boost, I got just shy of 19,000 composite. Given the score posted in vendettamark for a lower spec system around 17,000... weird.

I suppose something could be wrong with my system, but I'd expect things like Prime95, ycruncher, burntest, Memtest86, and the various graphics benchmarks I've run to show it. It's not thermal limiting, both because I have watercooling and because neither the GPU nor the CPU reflects any real increase during a Vendettamark run.

@incarnate

Whatever man.

Whenever I come to VO to see how things might be progressing and considering trying it out again, it is always suggested there must be something wrong with my system or internet. This time even before actually installing the game. There is of course never anything wrong with anything else. Just my internet or my system all the time.

Thanks for clearing this up. LOL
Good Grief !

@Dr. Lecter I'm running same drive you have too

Actually also had the 35 inch version of the AGON 100HZ gaming screen too, but took it back for the AGON AG271QG 165HZ screen

That wide screen was awesome for gaming but not good for broadcasting due to the aspect ratio and black bars subject that is almost impossible to work around. Also it made my trading charts look all stretched out.

Whenever I come to VO to see how things might be progressing and considering trying it out again, it is always suggested there must be something wrong with my system or internet. This time even before actually installing the game. There is of course never anything wrong with anything else. Just my internet or my system all the time.

Wow. Okay. I was trying to help you.

I'm not claiming our benchmark is perfect, or bug-free, but it has been heavily tested by the internal benchmark teams at NVIDIA, AMD, Intel and various others.

All I did was suggest you try running a 3DMark test to corroborate. But, suit yourself.

I suppose something could be wrong with my system, but I'd expect things like Prime95, ycruncher, burntest, Memtest86, and the various graphics benchmarks I've run to show it.

Well, benchmarking is a technically challenging subject. The point of VendettaMark is to test the entire system at once, stressing it in much the same way as a game. Most synthetic benchmarks are much more specific, testing only one thing at a time. As a result, they can benefit from far better cache coherency and other factors, due to a simpler set of code loops at the time.

There's also a certain degree of run-to-run variability in VM. We improved it quite a bit (before release), by working around Windows 10's rather chaotic thread scheduler, which was moving threads between cores very quickly and making run-to-run testing less coherent (NVIDIA reported that one to us). But even with that, the number of different things going on at the same time, and Windows's tendency to choose to do random things at random times, can mess with the results.

Generally, it's best to run a benchmark when the system is cool (no recent activity), like rebooting and letting the system settle for ~5 minutes or so.

We also suspect we may be more memory intensive than some other benchmarks. But, we don't have a broad enough sampling of hardware/testing to say as of yet.

At any rate, 19000+ is pretty good, by my standards? It might not be as high as you might like, but I don't think it necessarily means anything is wrong. Captain's numbers were a lot lower and.. stranger, given his configuration.

I'm just surprised to see so little improvement over the 17750 score reflected for a significantly lower-end rig on the VM top scores page (Intel(R) Core(TM) i5-6400 CPU @ 2.70GHz, NVIDIA GTX 980). Maybe I should bump my ram up to 32 or 64gb. My CPU and GPU already smoke those specs and yet I'm not beating its score by more than 1000 points during a good run.

What PSU are you using Captain? If it is the one that came in your case I would seriously consider replacing it. If you have a decent one then are you actually using the latest drivers supplied by your manufacturer or just whatever Windows installs? (Not just for GPU, but also install the *proper* drivers for your motherboard - the drivers that Windows install are getting better, but they are still general drivers that are considered to "just work" and may miss features that your hardware actually has)

As for the shut down thing you are complaining about, huh what? If I select power off, my PC powers off. If you do not have the power off option then you have either disabled it, or something you have installed has.

Anyway, hope you get it sorted, Id hate to have a rig like that under perform on any benchmark too!

[edit]

Seeing as the vendettamark windows binary is 64 bit, can we get a windows 64 bit build of VO yet? I have asked in the past and it really is not just "because I want one", some plugins I run can quickly exhaust the 2gb of ram they can allocate, especially when I load (for example) my roid database up while using preload. You could include both in the regular Windows download and check for 64 bit compatibility in the updater, and run the 32 bit binary on systems that fail the check, if you are worried about people downloading the wrong one.

Heck, you could just include a vendetta64.rlb in the download, and let us that want it rename the file if the above is too hard. You could even leave the initial .exe and updater as 32 bit. The binary + drivers is only like 5.9mb (give or take) and you already have the drivers compiled!

Lecter: I was referring to ram performance (frequency, cas latency, etc), not the total size of ram (and I'm not sure about the RAM performance impact, either). 16gb should be fine. I also recommend running the test when it's the only open app on the system, and full-screen.

Yoda: VM is a standalone app. We know we can deliver a 64bit Windows version of the game, but it would complicate our weekly build process, and roughly double the size of every update, along with the related bandwidth increase on our update servers.

We may do it eventually, but it's a much more complicated, delicate process, with a lot of testing and secondary ramifications involved. VendettaMark has none of those challenges.

Inc.

Fair enough - I'm running fairly fast but not top of the line stuff at its binned speeds via XMP, because I'm lazy. (4x4GB HyperX predator ddr4 3000 at CL15-17-17 T2). I've run this kit a bit faster (either 13-15-15 T2 or 15-16-17 T1, both around 1.4v) while staying stable, and will play around a bit more. I usually don't find much point to the marginal gains eeked out by manual OC of everything, plus the added headache of truly verifying stability. But even dick measuring has its place.

I've been running it as the only open program, in full screen, and killed any non-essential processes I can kill (fuck you, Cortana). The run to run variations can be pretty funny though. This morning I bounced from a 15,500 score to an 18,600 score for no apparent reason, definitely didn't change anything.

Any idea when VMark will get a real Top Scores section?

@yodaofborg Yeah it's a Corsair 850 lighted made for the Corsair Crystal 570X RGB ATX Mid-Tower Computer Case

I'm not sure which model without checking my receipt offhand but it was a custom build, with thermaltake water coolers, lighted corsair ram, Aura sync for everything including the GPU.

I guess there could be something wrong but I'm just not experiencing a problem. I can literally throw any game in ultra settings at this thing while broadcasting and recording at the same time and still run high FPS.

I don't doubt the VObench score necessarily but the fact that I'm watching 2FPS on 4x4 and get a moderately high score at all is most puzzling don't you think ?

@yodaofborg

As far as the "Shut Down" menu item for Windows 10, this is a new subject that just happened and is known via Windows 10 forums etc.

There was an update and so now the "Shut down" menu items on the computer do not actually shutdown. The default is something in the power management settings which says "turn on fast startup (recommended)" This of course normally means nothing to do with Shut down features at all.

It's in the "Control Panel\Hardware and Sound\Power Options\System Settings"

If this is checked off the computer does not shut down anymore for the latest windows 10 update

Mostly seems like it signs you out and turns off your display to a sleep state.

So you tap a key and have a login screen so you never actually shut off the computer.

This seems to be the new windows 10 feature which is really stupid and not even in the right location for a power management feature.

Now if this was a feature already included in Windows 10 but was uncheck by default then it would seem the new windows update enables this by default perhaps depending on the power settings you have selected.

I know it was lame not to let anyone know that your computer will never be able to shutdown from the "shut down" menu item if it's checked.

LAME Windows 10