• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

ryzen 5 1600 or 1600x

Please, I've x2 1700s with the best board available and the fastest ram available and any higher than 3200 was too much for it to be able to boot 100% of the time despite being solidly stable.
I also have a 1300x on a msi tomahawk board. I have a 7970 running with that.
I've got a laptop with an A10 APU in it.

That's a lot of products from a company I hate:)

What I do hate is people singing AMDs praises for releasing a product that struggles to beat haswell in gaming and has a **** poor memory controller.
I hate that people fail to recognise that this platform still has issues.
Even more some of this same stuff comes from people that don't even own one themselves and have no idea of any of the quirks that present themselves when overclocking.
Take a look in the ryzen owners thread, I and others have experiences bios updates that breaks previously stable overclocks. Cold boot issues are still a thing, even this long after launch.
So no I don't hate AMD, I hate misinformed people spreading BS and pretend that ryzen is faultless.

to be honest gavin, like i said before i respect people opinions and I know that some cpu's mobo's, ram etc will work better with one set up than another and thats why if tried to gather as much info as i could on my build from people.

the only thing i didnt get an answer on was the coolmaster masterliquid 240 lite cooler working with the ryzen so i took a gamble
 
You did the right thing by asking but you'll get opinions from people that dont even have one yet go recommending it based on a few graphs from a youtuber.
That cooler will be fine, the 1700 under a corsair h100i v2 never went above 75c using the silent profile when overclocked to 3.8ghz.

You will be fine, just dont go expecting too high of an overclock and prepare for hit and miss performance in older games.

very true, im not expecting to be able to sore to everest in one day on overclocking even if i just squeeze a few more out of her i'll be happy and from what ive seen on the b350 it has an option to select your ram speed you have inserted so it maintains at that level
 
It does yeah, you might be lucky and 3200 will just work. But there are other tweaks and things to try if it doesnt.
Some people just end up at 2933 because that particular CPU in question cannot do it. The joys of silicon lottery :)

end off day i know ill have a stable system and one that will last
 
Part of getting high RAM speeds seems to be down to the silicon lottery given the amount of variance we see even with Samsung B-die, so ymmv in that regard.
I have one of the earlier samples, from week 9, and getting high ram speeds has been hit and miss for me so far.
when you say high ram speeds are you meaning at what your ram should be running at, but due to overall system it downgrades it ?

i always thought with ram that having to identical sticks at the same it trhen auto balances it out so you have a level playing field ? or is that not the case in todays ddr4 world ?
 
You want to aim for 3200mhz on the ram, if this is not achievable you will need to go to a slower speed like 2933,2666 etc.

right okay, but if thats the case why dont we all just but the 2933 or 2666 ram why go for the higher one at more money ?

is it cause we hope we can hit those peaks ?
 
can one of you guys answer a question on the cooling, when adding the fans on the rad do you bolt below or above the rad as ive heard different things like go above the rad so its sucks all heat out ?
 
Yeah, I'm having issues getting to rated speed (3200Mhz C16) so I'm using 2933Mhz C16 right now, which to be fair is rock stable after some extensive stress testing (2 passes of memtest86, 1100% HCI memtest coverage on 90% of the RAM in Windows, 3+ hours of Prime95 Blend).
The performance difference seems somewhat negligible for gaming (1~2% fewer FPS than 3200 C16), but this is with a GTX970.

this is where i get lost on testing, ive never done anything like that before, so thats a whole new world to me
 
Not unless you can get a lot of negative pressure but you need silly amounts of air flow. Fans in push pull can work well.

It's also worth trying to balance the intake of air as much as possible.

right okay well my case has two fans at front, one at back then off course the cooler at top, at front will be 2 corsair sp120 led fans
 
Might be worth having the fans at the front running a little faster than the rest. Is the PSU pulling any air from inside the case?

That would be 4 exhausts and 2 intake all 120mm?

all 120mm yes im not 100% how fans work in general like i say as its been a while since i rebuilt, so if you can elaborate in simple terms
 
That is a good card, normally i would recommend the RX 580 purely because of better new API (DX 12 / Vulkan) future-proofing but because of the Coin Mining craze they are getting too expensive, now the 1060 does make more sense but whatever you do 'DO NOT' :) get the 3GB one... it is slower than the 6GB one, it has less Shaders and 3GB is nothing like enough Memory.

ok thanks will do
 
It's not that difficult nowadays, I was just being overzealous.
Easy way to do it is memtest64 to validate your memory settings. You can run it in Windows and it will just use however much memory you have free for testings (be sure you close everything you have running). If you can get to >10 loops on >14000MB without any errors then you should be fine (it shouldn't take too long, <30 minutes).
To validate your CPU settings I'd suggest installing HWInfo64 for monitoring, it should offer the most accurate data and I'd suggest leaving Prime95 on "Blend" running overnight (usually if you don't get any errors after 3+ hours, you should be 99% fine). A faster way to do it would be to get Realbench and have it do a 30~60 minute stress test using all of your available RAM. You can do shorter stress tests if you want to risk things or longer if you want to make sure it's bedrock stable, but usually 3+ hours of Prime95 and/or 30 minutes of RealBench should be good enough to find most issues. Be sure to monitor temperatures and voltages while doing the stress testing.

The voltage limit which you should not pass is 1.425v for the CPU Vcore (with LLC Auto). Try 1.2v with 3.6Ghz or 1.3v with 3.8Ghz and see if you're stable. Most Ryzens should be able to reach 3.8Ghz with <1.35v based on overclocking lists.
For the memory you just need to select DOCP in the bios and set it to 3200, the 8pack kit should run it without too much hassle, but if you have any stability issues/errors with memtest64, be sure to ask in the Ryzen thread.

Memtest64 link: https://www.techpowerup.com/memtest64/
HWInfo64 link (get the 64 bit version): https://www.hwinfo.com/download.php
Prime95 link (get version 29.3, 64bit): https://www.mersenne.org/download/
Realbench link: https://rog.asus.com/rog-pro/realbench-v2-leaderboard/
right okay ill have too look into this more
 
If you are having the rad at the top as an exhaust then you want the fans below it. From there they will be blowing the hot air out of the radiator.

If you have the radiator front mounted intake then you want the fans in front so they will blow cool air onto the radiator.
would i be best top mounting then like the others have said ?
 
Back
Top Bottom