New AMD Chipset Drivers Tested on Ryzen 9 3900X

[Jkk79]

So what, all CPU's are supposed to "idle" at full baseclock almost, with 1.45v?, that doesnt sound right
Of course, i checked with ryzen master, cpu-z, Hwinfo64, by themselves without any of the others running meanwhile

was under the impression that it should go sub 1 volt for idle?

Ryzen 3000 series go all the way down to 0.2V Vcore at idle. At least my 3900x does.
But you first need to close all the programs that abuse the windows system timer resolution. I have identified a few, like Steam and Telegram.
The default timer resolution is at 15.6ms and that's the fastest any background process should ever request.
But some, for whatever reason, request 1ms resolution and then you start seeing these crazy idle voltages, which indeed translate into wattages as I have measured with a power monitor device.
For me having a Steam or Telegram on means about 20W more system power usage at idle.

You can identify these offending background processes with running the following command in admin command prompt: powercfg /energy /duration 5
And then open the file it creates and look at the results.
I didn't come up with this information, an user 'el0j' in reddit pointed me to the right way.

 

[TheLostSwede]

B450 Tomahawk with 3700X. 7C02v1A BIOS and 1.07.29.0115 chipset driver.

The slider made no difference either.

I'd suggest waiting for the new AGESA, based on my own results, that made much more of a difference than the driver, which I though didn't do anything at first.

 

[BiggieShady]

Hey @W1zzard, on page 2 section about games at 4k, you end with "... especially at the higher resolutions, which is are relevant for gamers."
You have both "is" and "are", gotta pick one ... either high resolutions are relevant or performance increase is relevant
Maybe you are just missing a slash between is and are ...

I don't always do the grammar police, but when I do, I do it for really unimportant stuff

 

[W1zzard]

I don't always do the grammar police

Fixed. Please do keep posting them, or send in PM. Bta and I must have read this conclusion like a dozen of times. It's amazing how easy it is to miss these little things.

 

[purecain]

Thanks for the report W1zzard, very much appreciated as usual.

 

[notb]

The CPUs are fine. It's just RDRAND that's messed up, but since it's a new instruction, there isn't much code out there to make use of it.
It would have rendered my Linux install unbootable, since I like to keep up with software changes, but I haven't switched yet.

Funny how RDRAND is a new instruction, but Ivy Bridge is an ancient architecture.
If AMD can't do it properly, why bother?

It's really obvious that AMD is rushing these releases more and more with each gen. Inevitably, "finewine" is becoming a thing in Zen.

 

[bug]

Funny how RDRAND is a new instruction, but Ivy Bridge is an ancient architecture.
If AMD can't do it properly, why bother?

It's really obvious that AMD is rushing these releases more and more with each gen. Inevitably, "finewine" is becoming a thing in Zen.

It was just an oversight. As far as I have read, there's an additional operation you're supposed to do to check whether you got a valid random of not. Sometimes this check says you did get a valid value when in fact you didn't. It should have been caught in testing, but that doesn't mean Zen2 was rushed.
Crap happens. And we have CPU microcode precisely because manufacturers recognized the more complex the CPU becomes, the more likely it is for things like this to slip through and they needed a way to provide fixes without replacing the physical CPU.

Of course, it sucks when you rush to be an early adopter and show your support only to be bitten by a bug like this. But like I said, crap happens.

 

[B-Real]

Hehe, with this launch it seems AMD decided that since they weren't going to fall as far from the hype as usually (apart from the frequencies, of course), to just leave beta testing to the general population and somehow they are mostly getting away with it. I can't imagine if Intel did something similar - Internet would probably explode

Who told anything about frequencies from AMD? Oh yes, so you only heard rumours.

Getting away with it? The problem was shown by users, and they released a driver to fix it. Is it a problem that Intel doesn't get away with security issues? I know it's the same problem as this one. Oh wait. Well, I read forums, but most time I see Intel or NV trolls hanging around.

 

[Axaion]

What kind of sound card do you have?

On my end my X-Fi Titanium HD (in MSI B450 Tomahawk) with the latest driver (FRL_PCDRV_L11_3_00_2022) is working great without a fault on Windows 10 1903.

its an AE-5, seems pretty random which creative cards works or not lol

Ryzen 3000 series go all the way down to 0.2V Vcore at idle. At least my 3900x does.
But you first need to close all the programs that abuse the windows system timer resolution. I have identified a few, like Steam and Telegram.
The default timer resolution is at 15.6ms and that's the fastest any background process should ever request.
But some, for whatever reason, request 1ms resolution and then you start seeing these crazy idle voltages, which indeed translate into wattages as I have measured with a power monitor device.
For me having a Steam or Telegram on means about 20W more system power usage at idle.

You can identify these offending background processes with running the following command in admin command prompt: powercfg /energy /duration 5
And then open the file it creates and look at the results.
I didn't come up with this information, an user 'el0j' in reddit pointed me to the right way.

Yeah, gonna have to have a look at that.
Ryzen master says all cores but one is in sleep (new ryzen master version from yesterday i think?), and the one thats still active on about 205mhz, yet i was still on 1.45v with 95% minimum, and 0.93v on 5% minimum, plus idling 6c lower.

Shrug, guess ill wait till creative updates drivers so i can use 1903, so i can at least clean install yet again haha

 

[HenrySomeone]

Who told anything about frequencies from AMD? Oh yes, so you only heard rumours.

Getting away with it? The problem was shown by users, and they released a driver to fix it. Is it a problem that Intel doesn't get away with security issues? I know it's the same problem as this one. Oh wait. Well, I read forums, but most time I see Intel or NV trolls hanging around.

Bullshit, their own advertised boost clocks are but a pipe dream for most users and even then they are generally only achieved for like miliseconds and let's not even delve into max all-core OC, because that is just laughable

Funny how RDRAND is a new instruction, but Ivy Bridge is an ancient architecture.
If AMD can't do it properly, why bother?

It's really obvious that AMD is rushing these releases more and more with each gen. Inevitably, "finewine" is becoming a thing in Zen.

Yeah, I think they fully realize that when desktop Ice Lake strikes, the game will be pretty much over so they are trying to get get as much headstart as they possibly can, screw optimisation and compatibility

 

[bug]

Bullshit, their own advertised boost clocks are but a pipe dream for most users and even then they are generally only achieved for like miliseconds and let's not even delve into max all-core OC, because that is just laughable

Yeah, I think they fully realize that when desktop Ice Lake strikes, the game will be pretty much over so they are trying to get get as much headstart as they possibly can, screw optimisation and compatibility

I don't know, but this seems to indicate it holds up the boost clock pretty well: https://www.anandtech.com/show/14605/the-and-ryzen-3700x-3900x-review-raising-the-bar/5

 

[EarthDog]

I don't know, but this seems to indicate it holds up the boost clock pretty well: https://www.anandtech.com/show/14605/the-and-ryzen-3700x-3900x-review-raising-the-bar/5

There are a lot of people having trouble hitting peak boost for any amount of time. This includes me and thelostswede among other random forum users.

 

[bug]

There are a lot of people having trouble hitting peak boost for any amount of time. This includes me and thelostswede among other random forum users.

Well, that sucks then. Are you this isn't about the amount of cooling provided?

 

[HD64G]

Ryzen Master is better than the BIOS in tweaking and oc for now at least.

 

[EarthDog]

Well, that sucks then. Are you this isn't about the amount of cooling provided?

Same behavior on a 2x120mm AIO and 3x120mm AIO and using a 3700X, not even the beefy 3800x or 3900x. This is also across 5 different motherboards.

I haven't had a chance to check the new AGESA, but will today. It improved swede's, but still hasn't hit the marked boost.

 

[bug]

Same behavior on a 2x120mm AIO and 3x120mm AIO and using a 3700X, not even the beefy 3800x or 3900x. This is also across 5 different motherboards.

I haven't had a chance to check the new AGESA, but will today. It improved swede's, but still hasn't hit the marked boost.

Strange. People fret over NVMe SSDs not sustaining their write speeds, you'd think you'd hear more about CPUs doing the same...

 

[shmuck]

Ryzen Master is better than the BIOS in tweaking and oc for now at least.

Not with virtualization turned on

 

[EarthDog]

Strange. People fret over NVMe SSDs not sustaining their write speeds, you'd think you'd hear more about CPUs doing the same...

I can only tell you what I, and many others have seen. Most of the time people aren't even checking that anyway.

 

[HD64G]

Not with virtualization turned on

Valuable info to share

 

[Octavean]

Interesting,.....

And in other news today,.....

AMD RyZen 9 3900X CPU's don't seem to be in stock anywhere. If you can find one it will likely be priced higher then the MSRP of the upcoming RyZen 9 3950X 16 core / 32 thread at ~$750 USD.

So, slow clap for the AMD chipset drivers,......

Clap,......clap,.......clap,.....................................clap.

 

[advanced3]

Didn't change anything here.

Still high idle voltage and idle temps.

 

[notb]

Crap happens. And we have CPU microcode precisely because manufacturers recognized the more complex the CPU becomes, the more likely it is for things like this to slip through and they needed a way to provide fixes without replacing the physical CPU.

It's been less than a month since Zen2 and X570 launched. Do you think the amount of "crap" that happened is reasonable?

Sure, I know there are people (a lot here) that have spent this month updating, tweaking and benchmarking their shiny 3700X. And they loved it.
But wasn't AMD hoping for a broader audience? Weren't Zen CPUs supposed to be robust and mainstream one day?
Does AMD expect the 90% PC users they can't reach to take tweaking PCs as a hobby? Seriously?

And as usual: no love from OEMs at launch, but maybe in few months... ;-)

 

[Nordic]

Yes you also need to use Ryzen Balanced Power Plan, which automatically gets installed and activated on chipset driver install

Are there any gains from the Ryzen Performance Power Plan?
Are the gains from SMT on or off different with this update?

Sure, I know there are people (a lot here) that have spent this month updating, tweaking and benchmarking their shiny 3700X. And they loved it.
But wasn't AMD hoping for a broader audience? Weren't Zen CPUs supposed to be robust and mainstream one day?
Does AMD expect the 90% PC users they can't reach to take tweaking PCs as a hobby? Seriously?

According to a number of sellers, it looks like AMD is outselling intel by a lot right now. It is right after a release, so those of us who have been waiting for ryzen are all buying it right now.

 

[TheLostSwede]

It's been less than a month since Zen2 and X570 launched. Do you think the amount of "crap" that happened is reasonable?

Sure, I know there are people (a lot here) that have spent this month updating, tweaking and benchmarking their shiny 3700X. And they loved it.
But wasn't AMD hoping for a broader audience? Weren't Zen CPUs supposed to be robust and mainstream one day?
Does AMD expect the 90% PC users they can't reach to take tweaking PCs as a hobby? Seriously?

And as usual: no love from OEMs at launch, but maybe in few months... ;-)

Compared to my Ryzen 7 1700 and X370 board, this launch has been vastly improved. Has it been without faults? No.
However, things like RAM is working much better, there are seemingly fewer motherboard related issues and general compatibility is much better.

What AMD should've done, is made sure they'd worked closer with the board makers to work out a few more bugs with the AGESA before launch and apparently also work on their drivers a bit more. I'm willing to cut AMD some slack, as their budget is a lot smaller than Intel, but at the same time, they deserve shit for both the 3800X and 3600X, as neither chip seem to offer any real world benefit over their nearest product in the CPU stack. As such, both options feels like a bit of a cash grab, as I think a lot of people bought them based on the experience of the past two generations of Ryzen CPUs where the X parts really made a difference, not so much this time around.

Am I disappointed with my purchase? Not as such, but I do feel conned by the fact that I spent an extra $100 and ended up with a chip that seemingly performs worse than every 3700X.

 

[JaymondoGB]

Still waiting for an oppertunity to buy a 3900X, I think AMD gave way too much stock to every US based Youtuber or whatever going, and missed out on an oppertunity to serve the paying public.