Corsair Scimitar Elite RGB gaming mouse review
Cougar 700K Evo keyboard review
Aorus Liquid Cooler 360 review
AMD Ryzen Threadripper 3990X review
Corsair A500 CPU air cooler review
Palit GeForce RTX 2080 SUPER GamingPRO review
Lamptron SM436 PCI fan/RGB controller review
Alphacool Eisbaer Aurora 360 review
PowerColor Radeon RX 5600 XT Red Dragon review
GALAX HOF Pro PCIe 4.0 NVMe review
Ryzen 3000: AMD deliberately limited Boost behavior in favor of longevity, says Asus staff
An Asus employee gas mentioned that AMD has reduced the boost behavior of the Ryzen 3000 processors to a more moderate level as it was too aggressive and now is a bit more limited in favor of longevity.
It has been a bit of a discussion as to why motherboard differs in Boost frequencies with Ryzen 3000 on different motherboards. Now it is said by shamino (Asus employee) that AMD has limited the boost behavior in newer AGESA versions.
The motivation for this step would be that AMD apparently wants to ensure a longer life. As it says at least in the contribution of "shamino1978" in forums (via reddit.com).
"every new bios i get asked the boost question all over again, i have not tested a newer version of AGESA that changes the current state of 1003 boost, not even 1004. if i do know of changes, i will specifically state this. They were being too aggressive with the boost previously, the current boost behavior is more in line with their confidence in long term reliability and i have not heard of any changes to this stance, tho i have heard of a 'more customizable' version in the future"
As for the controversy surrounding the boost frequencies of the current Ryzen 3000 processors, have a peek here.
New Firmware update Reportedly solves Ryzen 3000 boot issues Linux - 08/14/2019 07:52 AM
We don't talk about Linus a lot, as the install base is small and not really the PC Gamers domain, however as it turns out Linux users have had Boot issues with Ryzen 3000. A problem that is now conf...
G.SKILL Releases Optimized DDR4-3800 CL14 Memory Kit for AMD Ryzen 3000 & X570 Platform - 07/31/2019 12:54 PM
G.SKILL is glad to announce Trident Z Neo DDR4-3800 CL14-16-16-36 4x8GB / 2x8GB memory kits for AMD Ryzen 3000 & X570 motherboard....
ASUS 300 and 400 AM4 Motherboards Fully Support AMD Ryzen 3000 Processors - 07/31/2019 08:18 AM
UEFI BIOS update delivers compatibility with 3rd Gen AMD Picasso APUs and Matisse CPUs with no loss of features, highlighting the ASUS commitment to customer care...
Destiny 2 AMD Ryzen 3000 Issues solved with chipset driver update - 07/29/2019 08:25 AM
Bungie’s Destiny 2 doesn’t currently work with on AMD’s latest CPU. It’s was not clear why this is currently the case. The problem cropped up a few days ago wh...
AMD Offers Free ‘Boot Kit’ for Struggling Ryzen 3000 Owners - 07/15/2019 07:12 AM
The launch of Ryzen Generation3 has been positive overall. The processors are hard to get as they are in high demand. Some of you purchased one, installed is on a Series 300 or 400 motherboard but nev...
mtrai
Senior Member
Posts: 1175
Joined: 2013-09-21
Senior Member
Posts: 1175
Joined: 2013-09-21
#5705095 Posted on: 08/28/2019 03:57 PM
To the guy that is writing about hundreds of hidden options, you know that setting them bad is worse than having them hidden?
I prefer a simple bios and a modding comunity if i want to go over it.
We all have to rely on guides and explanations for the 90% of those hundreds hidden settings anyway
Really you would need a guide on how to use the hidden fan controls? Incidentally the album I showed everything that could be unhidden....however for the modded bios I release to the wild do not have them all as some people just do not need to mess with or simply do nothing.
To the guy that is writing about hundreds of hidden options, you know that setting them bad is worse than having them hidden?
I prefer a simple bios and a modding comunity if i want to go over it.
We all have to rely on guides and explanations for the 90% of those hundreds hidden settings anyway
Really you would need a guide on how to use the hidden fan controls? Incidentally the album I showed everything that could be unhidden....however for the modded bios I release to the wild do not have them all as some people just do not need to mess with or simply do nothing.
K.S.
Senior Member
Posts: 2080
Joined: 2009-06-07
Senior Member
Posts: 2080
Joined: 2009-06-07
#5705206 Posted on: 08/28/2019 08:42 PM
Ryzen Master seems to conflict more often than not with most other well known software.
Ryzen Master seems to conflict more often than not with most other well known software.
Astyanax
Senior Member
Posts: 4697
Joined: 2018-03-21
Senior Member
Posts: 4697
Joined: 2018-03-21
#5705215 Posted on: 08/28/2019 08:55 PM
ryzen master should not be operated with any other software.
ryzen master should not be operated with any other software.
Fox2232
Senior Member
Posts: 9938
Joined: 2012-07-20
Senior Member
Posts: 9938
Joined: 2012-07-20
#5705254 Posted on: 08/28/2019 10:30 PM
Geekbench is famous for not reading clock correctly. Both Geekbench 3 and 4 mostly show base clock, therefore 3.8GHz.
If ti shows higher, then it is what someone set as manual all core clock.
Then there are some weird reading from time to time where it actually reads boost clock.
"processor_frequency": {
"minimum": 4252,
"maximum": 4370,
"median": 4354,
"mean": 4350,
"stddev": 15.43159011900798
or
{
"id": 68,
"name": "Processor Minimum Frequency",
"value": "4.27 GHz",
"ivalue": 4270,
"fvalue": 4270.0
},
{
"id": 69,
"name": "Processor Maximum Frequency",
"value": "4.39 GHz",
"ivalue": 4391,
"fvalue": 4391.0
}
Is that supposed to mean that there was around 125MHz difference between fully loaded 12C/24T tests and just 1C/1T loaded?
It should be noted, that programs like CPUz read clock from Core 0, and therefore are not showing currently fastest core (all cores available in tools > clock). Only thing that polls all cores properly is HWiNFO64 since it keeps min/max/avg values next to current.
It's collected it directly from geekbench - he provides the data along with the .xls he used to analyze the data - it's not like he polled people asking what their frequency is. Now there can be people with misconfigured systems but it's over 1000 entries for the 3900x - I feel like that's a decent sample size. That's not to mention that the behavior shown with later versions matches my experience and dozens of others on here and reddit and it also matches what ASUS is claiming.
Geekbench is famous for not reading clock correctly. Both Geekbench 3 and 4 mostly show base clock, therefore 3.8GHz.
If ti shows higher, then it is what someone set as manual all core clock.
Then there are some weird reading from time to time where it actually reads boost clock.
"processor_frequency": {
"minimum": 4252,
"maximum": 4370,
"median": 4354,
"mean": 4350,
"stddev": 15.43159011900798
or
{
"id": 68,
"name": "Processor Minimum Frequency",
"value": "4.27 GHz",
"ivalue": 4270,
"fvalue": 4270.0
},
{
"id": 69,
"name": "Processor Maximum Frequency",
"value": "4.39 GHz",
"ivalue": 4391,
"fvalue": 4391.0
}
Is that supposed to mean that there was around 125MHz difference between fully loaded 12C/24T tests and just 1C/1T loaded?
It should be noted, that programs like CPUz read clock from Core 0, and therefore are not showing currently fastest core (all cores available in tools > clock). Only thing that polls all cores properly is HWiNFO64 since it keeps min/max/avg values next to current.
CPC_RedDawn
Senior Member
Posts: 7764
Joined: 2008-01-06
Senior Member
Posts: 7764
Joined: 2008-01-06
#5705266 Posted on: 08/28/2019 11:49 PM
The entire point is that newer BIOS's from AMD have lowered it.. did you even read the article?
If it bothers you that much just flash back to the previous BIOS.
The entire point is that newer BIOS's from AMD have lowered it.. did you even read the article?
If it bothers you that much just flash back to the previous BIOS.
Andrew LB
Senior Member
Posts: 1078
Joined: 2012-05-22
Senior Member
Posts: 1078
Joined: 2012-05-22
#5705419 Posted on: 08/29/2019 11:01 AM
Its just like watching politics on TV. Obama separates children from their parents because thats what happens when you're taken into custody for breaking the law... the media agrees and never mentions it. Trump does the same and.... CAGES! Concentration camps!! RARRR!!
Intel owners tend to not behave like that.
if any other company had done this people would be losing their crap but AMD does it and half the community just handwaves it away, some people trying to make it out as a good thing..
Its just like watching politics on TV. Obama separates children from their parents because thats what happens when you're taken into custody for breaking the law... the media agrees and never mentions it. Trump does the same and.... CAGES! Concentration camps!! RARRR!!
Intel owners tend to not behave like that.
Evildead666
Senior Member
Posts: 1281
Joined: 2003-09-14
Senior Member
Posts: 1281
Joined: 2003-09-14
#5705447 Posted on: 08/29/2019 12:04 PM
Its just like watching politics on TV. Obama separates children from their parents because thats what happens when you're taken into custody for breaking the law... the media agrees and never mentions it. Trump does the same and.... CAGES! Concentration camps!! RARRR!!
Intel owners tend to not behave like that.
No, they just hide the 1500W chiller behind the scenes and "Forget" to mention it.....
Its just like watching politics on TV. Obama separates children from their parents because thats what happens when you're taken into custody for breaking the law... the media agrees and never mentions it. Trump does the same and.... CAGES! Concentration camps!! RARRR!!
Intel owners tend to not behave like that.
No, they just hide the 1500W chiller behind the scenes and "Forget" to mention it.....
vbetts
Moderator
Posts: 14800
Joined: 2006-07-04
Moderator
Posts: 14800
Joined: 2006-07-04
#5705491 Posted on: 08/29/2019 01:17 PM
Its just like watching politics on TV. Obama separates children from their parents because thats what happens when you're taken into custody for breaking the law... the media agrees and never mentions it. Trump does the same and.... CAGES! Concentration camps!! RARRR!!
Intel owners tend to not behave like that.
Can we not bring politics into this, thank you.
Its just like watching politics on TV. Obama separates children from their parents because thats what happens when you're taken into custody for breaking the law... the media agrees and never mentions it. Trump does the same and.... CAGES! Concentration camps!! RARRR!!
Intel owners tend to not behave like that.
Can we not bring politics into this, thank you.
Denial
Senior Member
Posts: 12548
Joined: 2004-05-16
Senior Member
Posts: 12548
Joined: 2004-05-16
#5705495 Posted on: 08/29/2019 01:27 PM
Geekbench is famous for not reading clock correctly. Both Geekbench 3 and 4 mostly show base clock, therefore 3.8GHz.
If ti shows higher, then it is what someone set as manual all core clock.
Then there are some weird reading from time to time where it actually reads boost clock.
"processor_frequency": {
"minimum": 4252,
"maximum": 4370,
"median": 4354,
"mean": 4350,
"stddev": 15.43159011900798
or
{
"id": 68,
"name": "Processor Minimum Frequency",
"value": "4.27 GHz",
"ivalue": 4270,
"fvalue": 4270.0
},
{
"id": 69,
"name": "Processor Maximum Frequency",
"value": "4.39 GHz",
"ivalue": 4391,
"fvalue": 4391.0
}
Is that supposed to mean that there was around 125MHz difference between fully loaded 12C/24T tests and just 1C/1T loaded?
It should be noted, that programs like CPUz read clock from Core 0, and therefore are not showing currently fastest core (all cores available in tools > clock). Only thing that polls all cores properly is HWiNFO64 since it keeps min/max/avg values next to current.
I don't know about geekbench - I use HWINFO to monitor my frequencies though, I've never seen it record 4.6Ghz on my 3900x with neither of my boards or any bios. My Initial ASUS x570-E bios was the closest at 4.575 - my gigabyte hits 4.5 which is better than my ASUS did after it was updated.
If it bothers you that much just flash back to the previous BIOS.
I wrote four times throughout this thread, one of which you quoted, and even referenced a fifth time that the frequency thing DOES NOT bother me. It's the larger issue that AMD gets away with false advertising or missing/dropping features after release and this is just a continuation of that. So I'm not sure why you're telling me to flash my bios back unless you think that me flashing my bios is somehow going to stop AMD from doing this in the future.
Its just like watching politics on TV. Obama separates children from their parents because thats what happens when you're taken into custody for breaking the law... the media agrees and never mentions it. Trump does the same and.... CAGES! Concentration camps!! RARRR!!
Intel owners tend to not behave like that.
I don't agree with your point on politics - I'm also not sure I agree entirely with your point on Intel owners. I agree that some people that own Intel processors are fanboys and are frustrating to deal with but I don't find that nearly with all Intel owners. I think most people that own Intel products own them because they simply aren't getting that much more performance from AMD that they deem it worthy to swap or they just haven't upgraded their system yet and Intel was the default choice prior to Zen.
Geekbench is famous for not reading clock correctly. Both Geekbench 3 and 4 mostly show base clock, therefore 3.8GHz.
If ti shows higher, then it is what someone set as manual all core clock.
Then there are some weird reading from time to time where it actually reads boost clock.
"processor_frequency": {
"minimum": 4252,
"maximum": 4370,
"median": 4354,
"mean": 4350,
"stddev": 15.43159011900798
or
{
"id": 68,
"name": "Processor Minimum Frequency",
"value": "4.27 GHz",
"ivalue": 4270,
"fvalue": 4270.0
},
{
"id": 69,
"name": "Processor Maximum Frequency",
"value": "4.39 GHz",
"ivalue": 4391,
"fvalue": 4391.0
}
Is that supposed to mean that there was around 125MHz difference between fully loaded 12C/24T tests and just 1C/1T loaded?
It should be noted, that programs like CPUz read clock from Core 0, and therefore are not showing currently fastest core (all cores available in tools > clock). Only thing that polls all cores properly is HWiNFO64 since it keeps min/max/avg values next to current.
I don't know about geekbench - I use HWINFO to monitor my frequencies though, I've never seen it record 4.6Ghz on my 3900x with neither of my boards or any bios. My Initial ASUS x570-E bios was the closest at 4.575 - my gigabyte hits 4.5 which is better than my ASUS did after it was updated.
If it bothers you that much just flash back to the previous BIOS.
I wrote four times throughout this thread, one of which you quoted, and even referenced a fifth time that the frequency thing DOES NOT bother me. It's the larger issue that AMD gets away with false advertising or missing/dropping features after release and this is just a continuation of that. So I'm not sure why you're telling me to flash my bios back unless you think that me flashing my bios is somehow going to stop AMD from doing this in the future.
Its just like watching politics on TV. Obama separates children from their parents because thats what happens when you're taken into custody for breaking the law... the media agrees and never mentions it. Trump does the same and.... CAGES! Concentration camps!! RARRR!!
Intel owners tend to not behave like that.
I don't agree with your point on politics - I'm also not sure I agree entirely with your point on Intel owners. I agree that some people that own Intel processors are fanboys and are frustrating to deal with but I don't find that nearly with all Intel owners. I think most people that own Intel products own them because they simply aren't getting that much more performance from AMD that they deem it worthy to swap or they just haven't upgraded their system yet and Intel was the default choice prior to Zen.
Click here to post a comment for this news story on the message forum.
Senior Member
Posts: 1281
Joined: 2003-09-14
If Geekbench is collecting the correct data, then i'm fine with it.
As i stated in the other reply, some say Ryzen Master is the only tool to give correct values, some others say HWInfo.
if Geekbench does truly have the correct values collected, then fine.
If not, its pointless to analyse bad data, because the result is pointless, due to the bad data....