What's new
Carbonite

South Africa's Top Online Tech Classifieds!
Register a free account today to become a member! (No Under 18's)
Home of C.U.D.

Ryzen 3600 upgrade very underwhelming

Status
Not open for further replies
Just putting this out there.

Go have a look at my thread history and you will see that I had a thread on some issues I experienced with my Ryzen 5 2600.

In all benchmarks the cpu and gpu performed excellent, but in games my fps was very low and I had a lot of stuttering.

Turns out it was a faulty pci slot on the mobo. I figured this out by moving the gpu from pci slot 1 to pci slot 2 and it performed excellent.

Hope you can find the source of the issue.

Sent from my VOG-L09 using Tapatalk
 
Just putting this out there.

Go have a look at my thread history and you will see that I had a thread on some issues I experienced with my Ryzen 5 2600.

In all benchmarks the cpu and gpu performed excellent, but in games my fps was very low and I had a lot of stuttering.

Turns out it was a faulty pci slot on the mobo. I figured this out by moving the gpu from pci slot 1 to pci slot 2 and it performed excellent.

Hope you can find the source of the issue.

Sent from my VOG-L09 using Tapatalk
I will give this a whack thanks for the reply :)
 
Now correct me if I'm wrong, I'm not a technical person, but is a 520watt Bronze certified power supply not basically 520w minus 15%? Because it's up to 85% efficient? Meaning we are left with around 450 watts give or take?

Provided its not a el cheapo that claims 520w, when in fact they are meaning total wattage of PSU, it would be capable of supplying 520w on the 12v rail, whilst drawing around 612w from the wall (85% efficient). 92% efficiency would see 565w from the wall.
 
Provided its not a el cheapo that claims 520w, when in fact they are meaning total wattage of PSU, it would be capable of supplying 520w on the 12v rail, whilst drawing around 612w from the wall (85% efficient). 92% efficiency would see 565w from the wall.

So that’s how it works! Thanks for the answer :)
 
Bios settings

CPU Ratio - Auto
XMP applied - 3200mhz CL16-20-20-20-38 @ 1.35V
PBO enabled

Run 1:

Cinebench R15

Single Core - 189

Multithread - 1505

Cinebench R20

Single core - 471
Multithreaded - 3376 (this went down)


Bios settings

CPU Ratio - 42
Manual - 3200mhz CL16-19-19-19-38 @ 1.35V
PBO enabled
CPU Vcore - 1.13750

Run 2:

Cinebench R15

Single Core - 190

Multithread - 1512

Cinebench R20

Single core - 473
Multithreaded - 3449

That 2nd run results seems on par more or less with what you should be getting


I don't reckon it's your CPU.

Out of curiosity - what are your temps when running Cinebench? Liquid cool that sucker 😁
 
That 2nd run results seems on par more or less with what you should be getting


I don't reckon it's your CPU.

Out of curiosity - what are your temps when running Cinebench? Liquid cool that sucker 😁
temps are toasty, too much actually :(

AIO or tower cooler in my plans ASAP
 
temps are toasty, too much actually :(

AIO or tower cooler in my plans ASAP
Good idea.
And if you're in any way inclined to OC it's way better on liquid PLUS with a decent OC you can somewhat match Ryzen 3600x performance


Personally not a fan of OC, seems too much effort to gain a marginal increase, but the AIO certainly will help that stock 3600 at least not hit high thermals and throttle, if thats an issue now or could be in the future on air cooling.
 
temps are toasty, too much actually :(

AIO or tower cooler in my plans ASAP

Are you using the stock cooler? I had kak performance and temps using the stock cooler. When i upgraded it to the Noctua it was waaaay better.
For my overclock I'm running XMP on the memory and then PBO and auto overclock disabled. I set all cores to 42 and cpu voltage to 1.3. tried less voltage but my overclock wasn't stable.

Sent from my iPhone using Tapatalk
 
Just putting this out there.

Go have a look at my thread history and you will see that I had a thread on some issues I experienced with my Ryzen 5 2600.

In all benchmarks the cpu and gpu performed excellent, but in games my fps was very low and I had a lot of stuttering.

Turns out it was a faulty pci slot on the mobo. I figured this out by moving the gpu from pci slot 1 to pci slot 2 and it performed excellent.

Hope you can find the source of the issue.

Sent from my VOG-L09 using Tapatalk
This is likely the problem as the 5500xt had worse performance when running PCIe gen 3 x8 Vs. PCIe gen4 x8 as a result of bandwidth. A faulty PCIe bus will probably cause bandwidth issues similar in nature to the 5500XT problems.
 
Bios settings

CPU Ratio - Auto
XMP applied - 3200mhz CL16-20-20-20-38 @ 1.35V

You do know that you're not tweaking the memory if you apply XMP? XMP just loads generic profiles and isn't going to change the speed or the latency of the RAM beyond that so you'll see zero gain.

Memory tweaking is disabling XMP, setting the voltage to a manual value and then playing with both the speed and timings.

You need to use this to help you manually configure the timings
 
You do know that you're not tweaking the memory if you apply XMP? XMP just loads generic profiles and isn't going to change the speed or the latency of the RAM beyond that so you'll see zero gain.

Memory tweaking is disabling XMP, setting the voltage to a manual value and then playing with both the speed and timings.

You need to use this to help you manually configure the timings

The calculator is awesome but I want to stress if it worked with Micron D-Die chips, it gives me only a safe option (not the fast one) which is the exact same timings as the XMP profile.

I have been manually tweaking and testing myself over the past few days:

3000Mhz at CL15-18-18-18-38 @ 1.350V FLCK - 1500Mhz
3200Mhz at CL16-18-18-18-38 @ 1.350V FLCK - 1600Mhz
3466Mhz CL17-19-19-19-38 @ 1.375V FLCK - 1733Mhz

Problem is I was really not seeing any difference in FPS, min max or average.

Only difference was in synthetics.

I did run into alot of boot loops and had to clear CMOS quite a lot, these do not overclock well from what I have seen.
 
Are you using the stock cooler? I had kak performance and temps using the stock cooler. When i upgraded it to the Noctua it was waaaay better.
For my overclock I'm running XMP on the memory and then PBO and auto overclock disabled. I set all cores to 42 and cpu voltage to 1.3. tried less voltage but my overclock wasn't stable.

Sent from my iPhone using Tapatalk
I am super keen to get this OC working but stock is not cutting it sadly, end of the month I am probably picking up a 240MM aio.
 
@ash515
A mate @Llama58 has the same setup except for the GPU (different brand)
I'm sure he wouldn't mind running some test to compare but i am inclinded to think its not the chip or the gpu but probably the board.

Also if i read correctly, 65 was your rough max which isn't terrible at all.

I think maybe try swapping the PCIE slot.
If i had a Ryzen setup i would gladly offer for you to come through and teat but running Intel so ram is the only thing i could help with.
 
sorry if someone mentioned it (painful to read through the whole article with the fanbabies derailing the thread) but have you retested that card in another machine?

I know some reviewers have said how bad the Gigabyte G1 cooling fitment is, maybe during the upgrade process the cooling to the VRM's has shifted and wont allow the card to keep sustained clock speeds?!

...and on the timing settings, dont underestimate its value. I got increased my Dota 2 FPS by 20% just with faster/manually tweaked memory.
 
whelmed-1430938192.gif
 
@ash515
A mate @Llama58 has the same setup except for the GPU (different brand)
I'm sure he wouldn't mind running some test to compare but i am inclinded to think its not the chip or the gpu but probably the board.

Also if i read correctly, 65 was your rough max which isn't terrible at all.

I think maybe try swapping the PCIE slot.
If i had a Ryzen setup i would gladly offer for you to come through and teat but running Intel so ram is the only thing i could help with.


That would be awesome, struggling to find direct comparisons.

I tried swapping PCIE slots saw a 10fps average reduction, moving back to the first PCIE brought back to what I usually getting.

sorry if someone mentioned it (painful to read through the whole article with the fanbabies derailing the thread) but have you retested that card in another machine?

I know some reviewers have said how bad the Gigabyte G1 cooling fitment is, maybe during the upgrade process the cooling to the VRM's has shifted and wont allow the card to keep sustained clock speeds?!

...and on the timing settings, dont underestimate its value. I got increased my Dota 2 FPS by 20% just with faster/manually tweaked memory.

I did read that but to be honest my G1 has been solid since I got it, 66 degrees under full load stock fan curve.
My card boosts to around 1949Mhz max (trust me I have tried so many times to get this above 2000)

I out right crash in every game if I so much as increase the core by +50mhz this is also with power set to max and temp limit of 92.
 
That would be awesome, struggling to find direct comparisons.

I tried swapping PCIE slots saw a 10fps average reduction, moving back to the first PCIE brought back to what I usually getting

Ok that is expected as its an 8 or 4x slot so you will see a drop.
I'll ask him to go some runs tonight in heaven, Cinebench (although I think that your score is pretty much on the money) and also a 3dmark although he might need to download it so might not be tonight.

@Llama58 you up for helping this gent?
 
Or is something wrong with your hardware, or you ram is crap.

My suggestion is to get another set of ram. Get B-die or E-die.
Rev E-die can do far above 4000Mhz more easily than most b-die can. They are easier to drive than B-die. Sorry but I have both b-die and e-die and I'll get e-die every time as long as the price is lower for any mainstream build. Pushing my b-die to 3800 CL14 took A LOT more voltage than I was comfortable with for daily usage with severe diminishing returns. If you're a dedicated overclocker and do it for the numbers then by all means get 2x16gb b-die or use your kit. If you're a mainstream user trying to get good speed out of your system for daily usage then e-die is more than sufficient.

My humble opinion. I'm sure some people swear with b-die because they are overclocking enthusiasts but the reality is not much extra raw performance for quite the premium price.
That said. If you can get cheap well binned b-die by all means take it. It's good memory. If you want to run 4x8gb then take E-die. It's easier to drive for the memory controller. Higher chance of getting it above 3733Mhz and possibly to 3800Mhz. B-die is loading heavier on the memory controller.

Ok, here is with a Chiller for overclocking, not 24/7:
 
Hi All

Thanks to every one who replied and took the time to share advice or assist in tests / benchmark results.

I am going to close this thread as "solved", I am happy (now) I am on Ryzen there is so much to learn. :)

Just to clarify I did not fix the problem nor gain the "missing" performance, more focused on games that perform well and base results off repeatable tests such as synthetic tests.

I have been extensively at this problem but have come to this conclusion:

All synthetic benchmarks I have done are correct to other scores
AC Odyssey (game I mainly focused on, friends of mine with other setups have noticed a reduction in FPS)
The rest of my games are within margin of error to others

Losing sleep over this is pointless which is why I am going to leave it as it is and focus on using the PC.

If the opportunity arises to test faster ram say, B-die 3600mhz kit or 3800mhz kit I will and compare results.

I will be upgrading the CPU cooler ASAP and a new GPU on the cards this year.
 
Starting to question my decision to go with AMD after hearing your issue lol i am supposed to buy either a 2700x or 3600 this february lol already have other parts. I wonder waht the issue with your pc is
 
Starting to question my decision to go with AMD after hearing your issue lol i am supposed to buy either a 2700x or 3600 this february lol already have other parts. I wonder waht the issue with your pc is
I would suggest go with the Ryzen 3600 even though I did not see the results I wanted, it still is a very good CPU from what I have seen and read.
Also AMD made big improvements with the 3000 series.

Just buy a proper cooler the stock one is trash
 
So all in all what is your final verdict?
Was the "upgrade" worth it for 1) productivity and 2) gaming?

Final verdict:

I am happy to have latest gen parts (4 years + old parts start to get worrisome even though they were stable and well looked after)
Frametimes are night and day, my 4790K was a stutter fest the 3600 is incredibly stable
Game performance is a mixed bag but then again I am using a 1070... will see once I upgrade to a 5700XT or a 2070 super

Productivity:

Sadly I have not touched this yet. I used to make videos and do streaming the odd game, looking to get back into it.

I am excited to get a proper cooler on this 3600 and overclock it to 4.2 and start messing more with that as well as a new gpu.
 
Status
Not open for further replies

Users who are viewing this thread

Latest posts

Back
Top Bottom