MSI Ryzen Gameboost Overclocks Revealed (Up to 4.4 GHz)

[AliG]

http://wccftech.com/amd-ryzen-proces...uency-bundles/

I think that gives a pretty clear indicator of what frequencies we can expect from each respective chip line. For reference, Stage 11 on 6900K is 4.5-4.7 GHz and 7700k is 5.2 GHz. My understanding is those are actually pretty aggressive overclocks on the Intel side, so I think 4.4 GHz is probably a realistic ceiling for stock voltage.

Lower than I'd want, but still way better than expected (recall the ES chips were originally running at 2.8 GHz).

[zalbard]

Surprised to see the 4 core variant to be a worse overclocker.

[AliG]

Surprised to see the 4 core variant to be a worse overclocker.

I've been saying for weeks it's all dependent on whether they have native or cutdown chips. These values seem to confirm that AMD has 1 chip, and is simply relegating that ones that fail their voltage leakage tests to the lower core variants.

Having said that, I'm not surprised at all (though disappointed a little).

[vario]

I've been saying for weeks it's all dependent on whether they have native or cutdown chips. These values seem to confirm that AMD has 1 chip, and is simply relegating that ones that fail their voltage leakage tests to the lower core variants.

Having said that, I'm not surprised at all (though disappointed a little).

Yea, i was pretty sure thats the way its gonna play out.We may see native quads later on though if ryzen pans out.
Either way, this makes me kinda optimistic, that maybe 4.2-4.4 clocks are achievable without chilled water.

[AliG]

Yea, i was pretty sure thats the way its gonna play out.We may see native quads later on though if ryzen pans out.
Either way, this makes me kinda optimistic, that maybe 4.2-4.4 clocks are achievable without chilled water.

I would assume those clocks were for air cooling/AIO water just based off the Intel calibrations. From I recall, 5.2 GHz is aggressive but fairly common for a 7700k.

[chew*]

I would be curious to see there voltage per frequency target. Should be same regardless of vid. Total voltage will obviously vary chip to chip but the voltage applied should be consistant regardless of vid.

[AliG]

I would be curious to see there voltage per frequency target. Should be same regardless of vid. Total voltage will obviously vary chip to chip but the voltage applied should be consistant regardless of vid.

It's my understanding that XFR actually dynamically increases voltage until you hit the pre-determined thermal and power limits.

It does make sense for a motherboard to have a pre-determined voltage curve though (probably polynomial in nature as given how fast voltage scales with frequency).

[chew*]

That just raises more questions. Is that a fixed speed boost a turbo boost or an xfr boost. Guess we will find out eventually.

[AliG]

That just raises more questions. Is that a fixed speed boost a turbo boost or an xfr boost. Guess we will find out eventually.

This is independent of XFR. I would consider buying this feature with a 1700, but not an 1800X (I'd just let XFR overclock or do it myself at that point).

The one advantage though is you can use the tool in game to crank up the juice temporarily. You may run into thermal issues, but you could always just dial it down at that point (though XFR does this as well).

[Ket]

Lower clocks for CPUs with less cores could also simply just be MSI artificially limiting the user to try and nudge potential customers in to buying a more expensive board that has higher auto-overclocking preset limits. Certainly not the first time I've seen this done. Asus even do it with memory on their midrange mainboards, nerf memory compatibility to force you to get a more expensive board. I well and truly proved this way back in the P45 days with the P5Q boards. Don't know if Asus stopped doing it after I uncovered it, but probably not.

On a more positive note if MSI aren't artificially limiting anything it goes to show that with a good cooler even Octa CPUs should hit 4.6GHz without needing to fight too hard for it.

[minirant] I can't stand these knobs / predefined "profiles" people can twist / press / just load and go. People with zero experience and / or knowledge use them then all of a sudden they think they are l33t for fiddling with a knob. Things like this degrade the standard of enthusiasts. Drastically. [/minirant]

[iddqd]

i love that the knob goes to 11

[AliG]

Lower clocks for CPUs with less cores could also simply just be MSI artificially limiting the user to try and nudge potential customers in to buying a more expensive board that has higher auto-overclocking preset limits. Certainly not the first time I've seen this done. Asus even do it with memory on their midrange mainboards, nerf memory compatibility to force you to get a more expensive board. I well and truly proved this way back in the P45 days with the P5Q boards. Don't know if Asus stopped doing it after I uncovered it, but probably not.

On a more positive note if MSI aren't artificially limiting anything it goes to show that with a good cooler even Octa CPUs should hit 4.6GHz without needing to fight too hard for it.

[minirant] I can't stand these knobs / predefined "profiles" people can twist / press / just load and go. People with zero experience and / or knowledge use them then all of a sudden they think they are l33t for fiddling with a knob. Things like this degrade the standard of enthusiasts. Drastically. [/minirant]

I don't think so, the Intel calibrations suggest stage 11 is actually a very aggressive overclock for traditional cooling.

[Ket]

Time will tell, MSI always tend to use a lot more voltage than needed for predefined overclocks regardless. I'm definitely getting that itch to see how well Zen clocks with the Wraith spire cooler but at the same time I can't help thinking why not just bundle the Wraith max with all CPUs it would surely be cheaper overall for AMD than having to tool for 2 different coolers.