BANANA Benchmark Results Comparison

Update with the latest public physic optimization available.
i7-5820K @ 4.2 GHz & 2666MHz Ram. Good enough.

 

1.36 volt is crazy high for that small overclock!
im pretty sure you can easily go .2 volts lower.
its is generally not recommended to go over 1.3 volt or the lifetime of your cpu reduces drastically.

 

Here`s my BananaBench results using an Intel Core i7 4770 @3.4 GHz Turbo`ed to 3.9 GHz with 16 GB of DDR3 Ram

 

The processor is not overclocked yet - have not messed with it at all yet.

 

I did some slight changes to CPU speed and RAM timings on Ryzen 5 1600. Lifted the CPU speed from 3.5 --> 3.7 GHz and dropped RAM timings from 16-16-16 to 15-15-15.

Max JBeams rised from 175 MBeams/s --> 185 MBeams/s. Pretty significant from such small changes.


However, I began to hear coil whining from motherboard, not from GPU. This is slightly suprising and somewhat alarming, I have never ever withnessed anything like that. None of the other benchmark softwares I try give any coil whine. And the coil whine only occurs, when the CPU is calculating the MBeams, everytime it stops and starts again, there's slight pause/silence inbetween.

 

oh i see - ive read of people running the thing at 4.5 - 4.7 ghz at that voltage so that looked a little hight to me.
but if thats default there shouldnt be an issue.

 

Yeah, I plan to overclock it in the summer, so that I know the fans will be able to cope. (I have a large cooler, so it shouldn't be a problem, but better safe than sorry.)

 

If anyone is having problems getting the results because it closes itself, modify the bananbench.bat file and add pause to the end instead of EXIT.

 

I'm a happy litle camper: I got mah Ryzen Setup/ Upgrade today, with the following specs:

• Asrock B350 PRO4 MicroATX Mobo
• R5 1600
• 2x4 GB DDR4 Corsair Vengeance LPX 3000Mhz
• 128 GB SSD

For about 470 euro's (could shave off a few cents if I went with items out of stock, but hey... im impatient).

First thing I wanted to do is run the bananabench, here the results (fully stock clocks, no bios update and no ram tinkering).

Result:


And during the benchmark:



I'm pretty happy bout that , comming from a laptop with the i7 4710HQ which throttles like crazy.

Not sure if I want to apply a OC already, since this seems to perform enough that I aint bothered by it.

Stock cooler didn't spin up or became louder during the bench (which suprised me a little), meaning that there is room for an oc, right? I probably will tinker around with the ram a bit, since it's running at just 2133 Mhz.

 

We have to wait, until MB manufacturers provide more optimized bioses with better memory support, so the performance will go up from that. I have the same CPU, running at 3.7 GHz on stock voltages and RAM is running at 2666 MHz, getting 185 JBeams/s. And no heat issues or anything else, so whenever you need, it's easy to amp it up a bit to gain slightly more power.


Sorry for offtopic, let's move to specific topic.

 

Just an assumption, sometimes the vrm's can cause coil whine, maybe something weird is happening with your vcore?
https://www.reddit.com/r/techsupport/comments/3abfaa/coil_whine_nearin_vrms/

 

Yep, I get the exact same thing and have done on both my old FX 8350 system as well as my current X99 Intel Xeon system. Both at stock settings. I think its just a side effect of the benchmark, maybe just because of what part of the CPU it stresses in particular. I haven't had any issues caused by it and while it doesn't happen in other benchmarks I don't think it will do any harm at all.

 

Allright, so no need to worry. But I could just in case RMA the MB just before the warranty goes out, if the coil whine still continues.

 

Here's mine. Running on my main Desktop PC.

 

Yay, so I updated my bios for the Asrock PRO4 Micro-Atx to P2.40 (latest version as of this post). I was able to set my memory without messing with voltages or timings to 2666 Mhz (default timings are at: 16-16-17), got the following result:

This is just at stock speeds (all cores at 3.4Ghz).


Interesting enough, before the ram speed change/bios update I didn't have any coil whine, like mentioned by DD-Indeed:

Interestingly enough, the higher the load the lower the frequency of the coil whine(during this benchmark). And as mentioned, not present during other cpu intensive benchmarks.

Edit*: During Unigine Valley Benchmark I hear coilwhine at different frequencies as well, espescially in the credit screen after closing the bench :?.

 

Wow 183 @ 3.4GHz and 20 vehicles with room to spare?

My 3770K can only do 163 @ 4.7 GHz and 17 vehicles..

how's that for efficiency to performance ratio for ryzen.

I want to build a ryzen system so bad lol

Have you tried 3.9 GHz OC?

 

Not yet, I might. Now I am curious as well... lemme have a go with it, im on stock cooler tho so hopefully it won't get to hot.

Edit:

Well, my curious mind went ahead and tried to push it to it's limits. Tried to go to 4 Ghz even, but even with 1.45 on the vcore it crashed during cinebench . I got it to not crash and stay alive at 3.9Ghz @1.35v tho.

So now the results:

During the bench:


And the result:



Quit impressive , tho my package wattage went to 100 watts... where with 3.4 i had it around 65, not sure if its worth the 5% ish increase. I recon, with even higher memory clocks/tighter timings i could get really near to that 200 tho .

Edit 2:

Guess what... I beat it... the 200 mark:



, messed around with the timings, ryzen really likes them tight. Couldn't get it beyond > 2667 tho.

 

AMD A10-6700 Richland CPU, @ 4.2Ghz

20 vehicles, Insane.

 

That's how many it tests. You can run 5 maximum before going below 100% simulation time.

 

Oh, I feel dumb now.