Ryzen 5 1600 Overclock

[norrisadam96]

I tried to overclock my Ryzen 5 1600 to 3.85ghz and got it at about 1.37 volts. However, I’ve seen videos of people that got theirs to 3.8 and achieved 1260+ cinebench scores. Mine at 3.85 only got 1153. Why is mine 100 points lower if it’s .05ghz faster?

 

[Darkbreeze]

First of all, what you see on Youtube is never guaranteed to be accurate. Half of what you see on there is faked, that is without question. Even if it is accurate though, you cannot assume what one person is able to achieve with their CPU is what you will be able to achieve. Every processor is unique. They ALL have different characteristics and performance capabilities. Even two successive CPUs coming off the assembly line will not necessarily be capable of the same clocks or the same stability at any given clock. Some will be better or worse. That's simply the nature of the silicon lottery.

As far as what others are able to benchmark at the same clocks as you are at, that is also a game of variables. They may have different memory, with different memory clock speeds, or better, more stable memory at the same clock speed. They may be using an SSD compared to your mechanical hard drive, or an NVME M.2 SSD compared to your SATA SSD. Or simply the same TYPE of storage drive, but a faster model.

They may have a motherboard that performs better over the PCI bus or has a higher power phase with better VRMs that is simply more stable than yours. They may have a much higher quality power supply that provides far superior voltage stability with less ripple and noise, creating a more stable overclocking environment to start with.

There could be literally hundreds of potential reasons why you are not able to gain the same performance characteristics as somebody else EVEN if you have ALL the exact same hardware specifications. Perhaps they've tweaked their overclock with different sub-settings with which you are not familiar. Manual overclocks that are dialed in are almost always more stable and somewhat better performers than automatically derived overclocks.

I would not worry about what others have, I would worry only about what you have and making sure that YOUR overclock is stable and stays within the prescribed thermal envelope. That is what is most important. Take you time, do the research, find out what small tweaks have resulted in successful gains for others and try them, one at a time, not all at once, to see what works and what does not.

If you are not willing to take the time to do that and simply want quick and easy results through automatic settings, then you have to be willing to accept only what that offers and be satisfied with the results, otherwise, you'll need to put in some work to figure out where you can make changes that are effective for your specific configuration.

 

[norrisadam96]

Thanks for the reply. I understand what you’re saying about the many variables that come into play and people on YouTube faking things. However I’m referencing max yuryev, Paul’s hardware, and tech deals. These are reputable channels that do not lie. Also, I do not think the storage information you provided is relevant to this. Cinebench is purely a cpu test that will show the same results regardless of storage (however I have a samsung 960 evo nvme m.2 and an 850 evo 2.5”). I also somewhat understand the silicon lottery, but I thought that the same voltages and clock speeds, provided the cpus could handle it, would have closer performance benchmarks.

 

[norrisadam96]

Don’t get me wrong, I really appreciate your comment and this is also my first official build so I have a ton to learn, and could be totally wrong

 

[btmedic04]

with Ryzen, memory speeds effect cpu performance because its tied to the die (CCX) interconnect (infinity fabric) speed. (IE how fast will one die be able to get data off of the other die.) Also memory timings will affect your score. so say you compare 2 identical r5 1600 systems and they are overclocked to 3.8ghz, but one has its memory set to ddr4 3200 and the other has its memory set to ddr4 2133, there is a good chance the system with the faster memory will score higher on cinebench. now all of this is in addition to what darkbreeze stated above. every cpu is a little different and have different limits and performance characteristics. some have stronger memory controllers that will allow faster memory speeds, others will be able to clock their cores higher. your score is in the ballpark of where it needs to be. dont be disappointed if your particular cpu score in one synthetic benchmark isnt the highest one out there, those folks have tweaked every aspect of their system which is fun to do, however it can be time consuming. ultimately its up to you to decide if thats something you want to try

 

[Darkbreeze]

I guess the best way to put it is like this. I have, personally, taken two different systems, both with FX-8320's, both with the same revision of Sabertooth 990fx motherboard, both with the same 16GB of Kingston Hyper X memory modules, both overclocked to the exact same 4.5Ghz with 1.35v on each setup. Same coolers, open bench with no case, using clean installations of Windows 8.1 on identical 250GB Samsung 850 EVO SATA SSDs and both using the same R9 280 GPU card.

There was consistently about a 4-6% higher score on one system in virtually every benchmark I threw at them, which included about ten of the most popular benchmarks like 3Dmark, PCmark, Cinebench, Sisoft Sandra and Passmark. There was no reason for the difference in performance that I could ever find. It was simply, there. I have no answer or explanation for it.

So it's sort of irrelevant anyhow as both were within the ballbark of where they should have been. Had I increased one system to 1866mhz memory and left 1600mhz DIMMs in the other, that probably would have widened enough to see on synthetic benchmarks, but probably not enough to see in real world use. I sincerely doubt you'll see the difference between your score and theirs in real world use either.