• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Overclocking becoming a thing of the past?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
I found the base clock of my 2700X to be much more efficient than overclocking or using thr PBO features. I figured that out by F@H if not a slight underclock from that and even lower voltage. Runs a lot cooler I may add.

Aren't Ryzen overvolted by default to make sure the boost works properly at all times ? Seen many posts and some reviews saying it hits 1.4v+ on auto.
 
It's not overvolted. It's a standard voltage range. For 1000/2000 Ryzen it was 1.45V. For 3000 it's 1.4-1.45V, depends on the chip. 3600 has 1.425V or something if I'm right. This is the reason why it's better to set voltages manually when you care about lower temps ... especially in small PC cases. For example 4-4.1GHz can be acheved at about 1.25-1.30V. Max boost for lower chips will be 4.1-4.2GHz at 1.4V+. Single cores only but it will cause throttling with stock AMD cooler or any other small cooler.
 
They way I look at "high" voltages on AMD, (similarly clock headroom on Intel) is that they want their CPUs to be reliable even under what we would see as more abusive conditions. Higher temps, cheap power delivery. That headroom helps provide that. We as enthusiasts might say, we don't need that much headroom, and can optimise closer to the limit where it is still stable under our more controlled conditions.
 
Your perception isn't Intel's perception. They had stated it isn't overclocking if you don't flip the overclocking flag inside the CPU, which then lets you fiddle with more stuff. Basically, if you run a CPU as supplied, turbo and all, it isn't overclocking in their eyes. Also, without looking at the range, I think there are still many recent if not current CPUs that don't boost at all. They're fixed clock (apart from dropping for power save of course).

Im not in the Intel section but if it has a base frequency, anything over that is a boost or over base frequency or overclocked beyond base frequency.

Just because a cpu does it by itself doesnt mean its not an overclock simply because they worded it that way.

If the base clock was 4ghz and no boost, its a 4ghz chip.

But both manufacturers have base clocks.
Why then word it base clock. Why not high end power state instead.

Intel.... Meh give me AMD. Not interested nor ever was with Intel.

- - - Auto-Merged Double Post - - -

Oh and high voltages on AMD.... Is not 1.4v... Ive had mine closer to 1.6v lol.
But that just makes mad heat lol.
LN2 1.6v and up.
 
The factory boost clocks are considered stock for both AMD and Intel. If the box says 4 GHz base and 4.4 Ghz boost... then it is expected to reach the boost clocks from the factory/out of the box. In other words, stock. It actually does mean that since a CPU does it automatically (from the factory it boosts) it isn't an overclock. An overclock is beyond the listed specs... base and boost are listed specifications.

The words base clock are there to define the minimum clock speed it can run while boost/turbo is the maximum for however many cores.

RE: the voltage, we aren't talking your Zen+ CPU and voltage under LN2... the context is Ryzen 3 and ambient cooling.
 
Last edited:
Given this is not Intel forum I'll save the long reply. Suffice to say, base clock has a specific meaning. See it as a minimum, which doesn't exclude running at higher clocks without overclocking. I don't know if AMD uses the term in the same way, but given their more advanced boost function it is more difficult than ever to give a meaningful boost clock.
 
Funny that the XFR boost is a base clock and generally only one core. Advertised for my 2700x at 4.3ghz.

So I suppose a single core advertised at a boost clock could be considered stock frequency.
I can agree to that yes.
 
SAme with multiple cores in tiers.. if 4c fo 4.2 and 2c does 4.3, that is alllllllllllllllllllll considered stock.
 
What about ‘overclocked’ GPUs that come out of the box with higher boost speeds? The boost speeds are ‘out of the box’ but they are ‘overclocked’ compared to the reference card.

Go!
 
What about ‘overclocked’ GPUs that come out of the box with higher boost speeds? The boost speeds are ‘out of the box’ but they are ‘overclocked’ compared to the reference card.

That's overclocked. It is usually even in the model name. Think the way to look at it is that boost alone doesn't mean it is OC or not. You can have boost and not be OC, but you can certainly have boost and be OC too.
 
What about ‘overclocked’ GPUs that come out of the box with higher boost speeds? The boost speeds are ‘out of the box’ but they are ‘overclocked’ compared to the reference card.

Go!
If the box says xxxx MHz then that is THAT card's stock clock.
 
So then overclocking is in fact dead.

No need for overclock forums anymore I suppose.

Welp, back to the team cup then.
Bye
 
How is Ryzen 2 scaling with temperature? I was thinking of modding my geo system to maintain a 20c constant temperature. Would that coupled with maybe a little under volting wring out some extra Mhz? Any of the chiller guys in the new Ryzen yet?
 
Your perception isn't Intel's perception. They had stated it isn't overclocking if you don't flip the overclocking flag inside the CPU, which then lets you fiddle with more stuff. Basically, if you run a CPU as supplied, turbo and all, it isn't overclocking in their eyes. Also, without looking at the range, I think there are still many recent if not current CPUs that don't boost at all. They're fixed clock (apart from dropping for power save of course).

Intel perception is just that. There perception. They dont get to define anything except 14nm lol


But seriously i just looked at ice and comet lake and its pretty clear Intel must have dementia now.
 
If the box says xxxx MHz then that is THAT card's stock clock.

If on the box there is 'OC version' then it's 'factory overclocked', not stock. It's sometimes described like that in gfx card features. I guess that depends on the point of view. We can look at this as the guaranteed frequency which for a given graphics card model is its standard clock but not for the used GPU which in general specification has a bit different (lower) frequency.

The same for CPU and GPU, there is a base clock and boost clock which works in a clearly specified range. The boost clock is still stock as for this frequency GPU/GPU was designed.

Intel perception is just that. There perception. They dont get to define anything except 14nm lol


But seriously i just looked at ice and comet lake and its pretty clear Intel must have dementia now.

No matter what is Intel doing, they are not losing market shares as much as we may think. Problem with new AMD is that expectations were much higher and a lot of users are clearly disappointed. For most, replacing 1000/2000 Ryzen with 3000 is simply waste of money as most of these processors work about the same. There is a bit higher performance but for most things, users don't need it and can't see the difference.
Motherboards are another problem with new Ryzen. X570 cost as much as already overpriced Intel Z370/90. AMD used to be a cheaper alternative but right now it isn't. Even though we get more cores then most users won't see more than 4.2GHz so for many of them, Intel will be still a better option.
I'm not defending Intel as their moves recently are pathetic and I'm not a fan of Intel in any way. I'm glad that AMD is improving their products but I simply wish they released something new and significantly better as Ryzen 3000 or new graphics cards are not changing anything.
 
Last edited:
How is Ryzen 2 scaling with temperature? I was thinking of modding my geo system to maintain a 20c constant temperature. Would that coupled with maybe a little under volting wring out some extra Mhz? Any of the chiller guys in the new Ryzen yet?

Yep it's been tested: https://www.gamersnexus.net/news-pc/3492-ryzen-cpu-thermals-matter-coolers-and-cases
For their 3900X, they saw a 60MHz improvement in all core boost clocks going from 55C to 20C CPU temp and a 225MHz improvement going from 84C to 55C.
 
I remember the days from last decade when multiplier locked chips but FSB OC's possible on Intel chips were fun & challenging & what with diff memory dividers.. it was all good.
Heck, even with the Intel quad core/HT chips that followed, even they were capable OCers (1156,1155 etc) could get my old i7-860 tp 50% OC on air, fully benchable.
Even the AM3+ platform was fun to fiddle with using FX line or processors.. however..

I think the laws of physics are the biggest obstacle to OC potential on 14nm & less fabs today. Especially with the Ryzen range.

I seem to be doing OK but only a little bit with my 2600X, at least can do all core/threads OC to 4.3GHz for daily gaming with 1.40v & temps around 50C.
Would like to do 4.4Ghz consistently but the Vcore is a problem imo. With 1080P gaming, cpu performance counts.
 
Hotrod2go I'd be pretty proud of that OC. Best I can do is 4150MHz.

Thanks, its amazing how little OC headroom there is in ALL the Ryzen range, don't matter what series.

I can inch it a tiny bit more to 4.325Ghz for not only CinebenchR20 multithread test but also for daily gaming now on ALL cores/threads.

Funny thing is if it goes to 4.35Ghz & I raise the Vcore a tiny bit like as in the next increment in Vcore, CinebenchR20 multithreaded test crashes about half way through.

I wasn't sure if it was temps or just not enough vcore. Still experimenting cause' new bios updates can change stuff too.
 
Back