AMD Radeon Software Can Overclock Your Ryzen CPU Now, Too

An anonymous reader quotes a report from PCWorld: The latest version of Radeon Software adds an unusual (and welcome) new twist: The ability to automatically overclock your Ryzen processor if you're rocking an all-AMD gaming desktop. Yes, your GPU software can speed up your CPU now, too -- and it can do it all with a single click. [...] The addition of Ryzen auto-overclocking in Radeon Software 21.9.1 continues the theme, and might just allow you to ditch AMD's separate Ryzen Master tool if you're running a Team Red graphics card. They'll need to be newer hardware, though, as the feature currently only supports AMD's latest Ryzen 5000 CPUs and Radeon RX 6000 GPUs.

AMD's blog describes how to use the new tool: "To access this easy-to-use feature, open up Radeon Software using the hotkey 'ALT' + 'R', navigate to the 'Performance' tab found at the top of the window, and select 'Tuning' in the sub tab directly below it. If you have the latest generation of AMD Ryzen and Radeon product installed on your system, a 'Tuning Control' section should appear for your system, allowing you to select 'Auto Overclock' to increase performance on both your processor and graphics card. We also have a new tuning section for CPUs, allowing you to overclock just your CPU. When the feature is selected, the system will ask for a restart and once you are back in Windows, you will be good to go!" "Radeon Software 21.9.1 also adds official Windows 11 support and the ability for Radeon RX 5000-series GPUs to tap into Smart Access Memory," adds PCWorld. "AMD also took the time to tout FidelityFX Super Resolution's rapid uptake. The DLSS rival is now supported in 27 games, with Arkane's awesome-looking Deathloop set to launch this week with native FSR support in place."

You can download these new drivers here.

Great

[DrMrLordX]

. . . now can we buy 6000-series video cards as MSRP?

According to a brief search for video cards:

6600XT: $695
6700XT: $890
6800: $1500
6800XT: $1700
6900XT: $1710 (though these go up to $3000 lol)

Ridiculous!

Frak them!

[antdude]

I just don't bother anymore and use onboard GPUs. Yeah, they're slow and not great.

Doing this in software feels icky

[guruevi]

Overclocking should be physically limited imho, too many people simply don't know what they're doing and then get all sorts of glitches or even burn out components months later without understanding why.

Either your computer works at full speed or it doesn't, overclocking in software is ripe for abuse, especially if it's just coming in a software update, not what an IT organization originally may have expected the software to do.

Back in the day we modified our multipliers and voltages, 486 motherboards had multipliers of 1.5, 2 and 3 and could swap between 3.3V and 5V by setting them using pins on a motherboard. And if you went too fast on your FSB then the things on your ISA bus would start glitching out because they used a divider. So we overclocked our SCSI and 512k video cards at the same time we were overclocking our CPU. And we had to calculate on pieces of paper what combinations would likely work. I'm getting old.

Re:

[thegreatbob]

A little before my time, but I have done a whee bit with second-gen Pentiums. Brings back memories of fiddling with PLL settings via SpeedFan, MBM5 (maybe misremembering this one's abilities), etc. on my PII-P4 era OEM hardware. Similar issues with PCI/AGP cards as bus speeds are still set by dividers from the FSB clock. Honestly, the de facto 100MHz base clock of today is a blessing.

Re:

[DrMrLordX]

Overall, software overclocking is safer than using the UEFI. If you get crashy behavior without taking down the entire OS, you can dial back the OC without even having to reboot.

I've OCed AMD platforms since my old Sempron 2800+ (Palermo, not Thoroughbred) all the way up to my current 3900x. It's much easier now to find "safe" voltage limits and tinker around with clocks than it was in the old days. Nearly everything is unlocked from AMD now, so no need to increase fsb/htt/bclk. About the only way to r

Re:

[thegarbz]

Either your computer works at full speed or it doesn't

Well that underlying theme of your comment is about 20 years out of date. The reality is modern computers do not have a "full speed". They have a rated base speed with the actual full speed being dynamically calculated based on a large number of factors such as available thermal headroom, core temperature, current limits, and power limits. This is why swapping a CPU cooler on a system left entirely at stock default settings can cause a benchmarkable performance change.

Your computer has been software overclo

Re: Doing this in software feels icky

[Type44Q]

While you clearly have little experience overclocking, you have even less worth saying on the matter. Why post at all??

Re:

[guruevi]

I have plenty of experience overclocking, when I was young, yes I know you can do it, but I also know that it should take some experience to know what you are modifying.

Now that I have to support hundreds of the wretches, I hate that the boards give the option to do it right from the front page, every few months there is someone that just wanted to see if they could get some more out of an old machine, or their business machine didn't game well enough, and end up completely bricking it.

At best because chang

Why?

[bradley13]

OK, so the manufacturer tests their chips when they come off the line. They determine what clock rate the chip can support, and sell it accordingly.

Over clocking means that at least some part of the chip - probably one or more cores - will no longer work reliably.

Why do I want to do this?

Re:Why?

[Anonymous Coward]

Because that's simply not true.
Yes, all chips are tested and binned according to their performance. But not all cores on the same die have the same performance limit. Current overclocking (at least with Ryzen) is per-core--so if you only have one core that doesn't clock as high you can still benefit by clocking the others higher. Everything is limited by current and thermal limits to prevent damage.
If you happen to exceed the limits of a cpu core, it might start throwing errors but is unlikely to be caused any permanent damage--not unless you forcefully run it it over current or over heated for an extended amount of time. The time it takes for the software to test the limits of each core and adjust is not long enough to risk any permanent damage--unless you've removed safeties or egregiously allowed too much current.
Pretty sure the AMD PBO software won't let you hurt it--they use both CPU, and motherboard current and thermal limit settings even when overclocking.

For gaming, sure, a 1% overclock is not useful. 5%, maybe, if the game is running at the limits maybe that gets you to some magical minimal framerate that makes it more playable.

However, for some workloads, a "mere" 5% gain can be very meaningful--something that takes 20 hours to compile, well that 5% could shave an hour off it.

Re:

[bobjr94]

Many chips can run with no issues overclocked for years, many times with no or little voltage increase. They are often under rated from the factory, capable of running faster but the sold as slower to keep the prices higher on the ones the do pick for the high end cpu's. Overclocking does not double the performance like it did 20 years ago, like when everyone got celeron 300a's and ran them at 450-500. People may get 5-10% at most on normal overclocking, not that much of an advantage now.

Re:

[tlhIngan]

OK, so the manufacturer tests their chips when they come off the line. They determine what clock rate the chip can support, and sell it accordingly.

False, chips are binned to what is needed. If you have a CPU that can do 4GHz, 3.5GHz and 3GHz, and you have a massive order for 3GHz CPUs, you only test that the CPUs can do 3GHz. Plenty of those CPUs might be able to do 3.5 or 4GHz, but since no one is ordering them, you only test them to 3GHz and mark them as 3GHz.

Now, if you have an order for say, 100 4GHz

Re:

[Calinous]

You could do small overclocks "for free" (basically on default components - PSU, motherboard, cooling, RAM, ...).
However, in some cases you paid less money for a high-end mainboard, PSU, RAM, cooling and a normal CPU than for a normal mainboard, RAM, PSU, ... and the highest end processor (and got the same performance).
There was an entire boutique industry of "pre-overclocked" gaming rigs - some of them were bought by big "beige PC" companies and jumpstarted their "gaming" divisions.

Re:

[Calinous]

Simplifying it, f Intel has to sell a million 4GHz processors and a million 3GHz processors, it will put silicon chips into a million 4GHz processors and a million 3GHz processors.
However, the lottery of the silicon production might produce one and a half million 4GHz processors out of the two millions produced. In this case, half the 3GHz processors are actually 4GHz processors.
One of the "mainstays" of microprocessor overclocking of the old days was simply finding a processor built in week X at plant Y (a

Re:

[thegarbz]

False. A manufacturer tests their system to a conservative minimum specification when it comes off the line. They make poor assumptions about the hardware you plug it into and thus *guarantee* that your state of the art high end system will run, as will Joe Averger's one with his Chiangwaxi motherboard.

The reality is from the base spec there are many opportunities to overclock and squeeze even double digit percent changes out of the system without any negative consequence what so ever. Heck you can squeeze

Gaming laptop

[King_TJ]

I just purchased a new 17" gaming laptop (a Maingear Vector Pro, which is apparently one of Micro Center's house brands). Nice machine with an RTX 3080 GPU, but this model has the Intel 11th. gen. Core i7 CPU in it. Another version of it is identical except for an AMD Ryzen processor vs. the Intel.

Among other things, I'm trying to configure this so it can dual boot into either Win 10 or into Linux on a second PCIe NVMe SSD installed in it (has 2 sockets for them, which is pretty nice). I guess my initial

Re:

[sconeu]

I've run Linux on AMDs (Original Athlon, Athlon FX, and Ryzen 5 3600) with no issues whatsoever.

Re:

[gweihir]

I've run Linux on AMDs (Original Athlon, Athlon FX, and Ryzen 5 3600) with no issues whatsoever.

Same here and a lot longer. It seems Intel is spreading FUD again. Intel CPUs do _not_ have any compatibility advantage. Also note that these Intel CPUs are using the AMD64 architecture, because Intel was incapable of coming up with a working 64 bit processor architecture for x86.

Re:

[Calinous]

Intel was totally capable to come up with a working 64 bit processor architecture for x86.
However, that would have (and had, in the end) killed their Itanium business (which would have been their monopoly).

Re:

[gweihir]

Intel was totally capable to come up with a working 64 bit processor architecture for x86.
However, that would have (and had, in the end) killed their Itanium business (which would have been their monopoly).

Complete uninformed bullshit. Intel failed several times to produce anything viable.

Re:

[drinkypoo]

I've run Linux on many AMD processors and on none of them did power management work quite correctly. Literally none. AMD has consistently failed to even provide enough information for kernel developers to get it to work right, while intel has consistently contributed code to the kernel to make it work right on their processors.

I haven't run Linux on a latest and greatest Ryzen system, my potato has an FX-8350 and my laptop (which was cheap) has a Ryzen 3. And it sometimes generates a free reboot on return f

Go AMD!

[RealNeoMorpheus]

All Hail Our Holy Lady Dr. Lisa Su!