My 1700X is running higher than my OC.

[Dash Lambda]

I got a 1700X a few days ago, spent the past couple days dialing in a good overclock, and after passing my stability tests at 3.95Ghz, I set it to a Pstate so my chip wasn't running full-tilt all the time. I get into Windows, open up my hardware monitor, and run Cinebench to make sure the OC is in place -All cores run at 3.95Ghz under load.

Then, after I let it be for a while, I open up the monitor again, and I see that the maximum clock hit by every core is ~4.05.

 

I have no complaints, but I'm still curious what's causing this. I imagine it would be XFR, but I'm not sure entirely what XFR takes into consideration when boosting...

[SageOfSpice]

Sounds like it's just normal clock inaccuracy.

[Dash Lambda]

Just now, SageOfSpice said:

Sounds like it's just normal clock inaccuracy.

... No, clocks don't just randomly go 100Mhz higher.

[kreechrr]

50 minutes ago, Dash Lambda said:

... No, clocks don't just randomly go 100Mhz higher.

But Ryzen X CPU's do have XFR (the AMD Turbo Boost) that only kicks in if cooling allows. usually, it's as much as 100MHz.

[SageOfSpice]

3 hours ago, Dash Lambda said:

... No, clocks don't just randomly go 100Mhz higher.

Funny, guess I'm hallucinating.

[Jumper118]

3 hours ago, Create585 said:

But Ryzen X CPU's do have XFR (the AMD Turbo Boost) that only kicks in if cooling allows. usually, it's as much as 100MHz.

Not once you manually overclock it. 

[Dash Lambda]

6 hours ago, Jumper118 said:

Not once you manually overclock it. 

Well, when you directly set the clocks (and I think voltages) to manual and change them, it disables all the dynamic clock stuff -But is there any confirmation as to what's disabled and what's left on when you change a Pstate?

 

7 hours ago, SageOfSpice said:

Funny, guess I'm hallucinating.

Sorry, I should have explained myself more: The bus clock is usually not exactly what it's set to, in my case it's dropped a maximum of 300Khz. With a clock multiplier of 40, that means variations in the bus clock can throw off the chip's operating frequency by as much as 12Mhz. Now, given that none of the components governing frequency are perfect, I'd consider as much as a 20Mhz error to be within reason -But that's still quite large.

That is, unless my understanding is wrong at some point.

 

Though, now that I look at the logs in my hardware monitor, I see something else strange: My bus clock is set to 100Mhz, it has a minimum of 99.7, an average of 99.8, and a maximum of 102.1. That's quite a jump. It never dropped by such a large amount, so I've gotta imagine that's XFR adjusting bus clock when it sees the right conditions. I always just assumed XFR changed the multiplier, so that's kind'a surprising.

[Ryzen init bro]

On 07/08/2017 at 5:53 AM, Dash Lambda said:

I got a 1700X a few days ago, spent the past couple days dialing in a good overclock, and after passing my stability tests at 3.95Ghz, I set it to a Pstate so my chip wasn't running full-tilt all the time. I get into Windows, open up my hardware monitor, and run Cinebench to make sure the OC is in place -All cores run at 3.95Ghz under load.

Then, after I let it be for a while, I open up the monitor again, and I see that the maximum clock hit by every core is ~4.05.

 

I have no complaints, but I'm still curious what's causing this. I imagine it would be XFR, but I'm not sure entirely what XFR takes into consideration when boosting...

I can confirm this happening on my system too, it is xfr.

its because the base clock will go up to 102 or something. It also boosts your ram speed  

[TacoSenpai]

On 8/7/2017 at 5:39 AM, Jumper118 said:

Not once you manually overclock it. 

Agreed can confirm I've played around with OCing my 1800X which at stock boosts to 4.1 on single core i believe but once I tested it at 3.9 it stopped boosting at least from what AMD Ryzen Master was telling me.