7800X3D, 7900X3D, 7950X3D..

[Bazz_Mulder]

@some1 Do you have data of that benchmark but in 4K? I know in 4K cpu limit is low but that is nearest to vr in DCS I can think off.

[some1]

Using this benchmark in 4k is pointless, it will only show GPU limits.

In VR CPU frametimes went down by about 15%. 

[Fakum]

On 4/22/2023 at 1:54 PM, BitMaster said:

I don't want to hold you guys back tuning the X3D chips, it sounds like some fun but there is this serious risc of burning it out.

https://www.tomshardware.com/news/amd-ryzen-7000x3d-burnout-reports

@BitMasterI appreciate the heads up! All my parts are rolling in, expect to be able to build by end of next week. I did read that link you provided (thank you again) and I did read the other posts here that followed. Being basically inconclusive, what would be the BIOS version recommended with current info? My initial plan was to update the BIOS before I did the build, but now Im hesitant.

 

CPU: AMD 7800X3D

MOBO: ASUS ROG Strix B650E-F Gaming

RAM: G.Skill Trident Z5 NEO Series 64GB (2 x 32GB) DDR5 6000 CL30-40-40-96

SSD: M.2: SK hynix Platinum P41 2TB PCIe NVMe Gen4 M.2

GPU: MSI Gaming GeForce RTX 4090 SUPRIM Liquid X 24G

POWER SUPPLY: FSP Hydro PTM PRO 1200W ATX 3.0 PCI-E GEN-5 80 Plus Platinum

TOWER: Corsair 5000D AIRFLOW

CPU AIO COOLER: ARCTIC Liquid Freezer II 280

WINDOWS 10 Pro

 

Edited April 24, 2023 by Fakum

[EightyDuce]

8 minutes ago, Fakum said:

@BitMasterI appreciate the heads up! All my parts are rolling in, expect to be able to build by end of next week. I did read that link you provided (thank you again) and I did read the other posts here that followed. Being basically inconclusive, what would be the BIOS version recommended with current info? My initial plan was to update the BIOS before I did the build, but now Im hesitant.

 

CPU: AMD 7800X3D

MOBO: ASUS ROG Strix B650E-F Gaming

RAM: G.Skill Trident Z5 NEO Series 64GB (2 x 32GB) DDR5 6000 CL30-40-40-96

SSD: M.2: SK hynix Platinum P41 2TB PCIe NVMe Gen4 M.2

GPU: MSI Gaming GeForce RTX 4090 SUPRIM Liquid X 24G

POWER SUPPLY: FSP Hydro PTM PRO 1200W ATX 3.0 PCI-E GEN-5 80 Plus Platinum

TOWER: Corsair 5000D AIRFLOW

CPU AIO COOLER: ARCTIC Liquid Freezer II 280

WINDOWS 10 Pro

 

 

They rolled back 1410 back to 1409 dated 4/21.

That being said, the issue "seems" to stem form automatic voltage being set by the overclocking module of the high-end boards so boards like the B650E-F are "safe". 

I still rolled my bios back to 1409 to be on the safe side. 

[Fakum]

hmmm, so my board (B650E-F) is not considered a high end board because its not a "-E" (Extreme)? so I should be ok ( i realize this is all still inconclusive). 

[EightyDuce]

34 minutes ago, Fakum said:

hmmm, so my board (B650E-F) is not considered a high end board because its not a "-E" (Extreme)? so I should be ok ( i realize this is all still inconclusive). 

Hahaha, well, not exactly. High end I meant those with BCLK/ECLK/AI OC unlocked. Which are the tippy top boards. Only B650 board with those features is the Gigabyte Aorus Master. Basically boards with auto tuning of voltages and EzOC features. 

 

Anyway, Derbauer just posted this video. It's by no means conclusive as he mentions himself, but the issue may involve EXPO profiles and specifically SOC voltage they may try to set automatically. Go in your bios and double check SOC voltage. Mine has been manually set to 1.25v from when I was tuning the memory. 

 

Still too early to tell but some things to check on in the meanwhile. 

 

TLDR no one really knows. 

Edited April 24, 2023 by EightyDuce

[Hoirtel]

Ah new adopter blues. Maybe my case being out of stock has been a good thing. Now thinking it should all stay in the boxes! Not watched this yet but should also be a good point if view.

 

[dugite57]

https://www.tomshardware.com/news/amd-ryzen-7000-burning-out-root-cause-identified-expo-and-soc-voltages-to-blame

Latest update on the matter of burnting chips

[EightyDuce]

5 hours ago, dugite57 said:

https://www.tomshardware.com/news/amd-ryzen-7000-burning-out-root-cause-identified-expo-and-soc-voltages-to-blame

Latest update on the matter of burnting chips

They seem to be leaning hevy on the derbauer video/thoughts, which he stated are just an educated guess at that point. 

 

If it is an SOC issue, that's easy to control by just manually setting voltage. 

[Lurker]

How long do these 7XXX systems take to boot? I heard some pretty terrible things. 

[PhilSkyfire]

11 minutes ago, Lurker said:

How long do these 7XXX systems take to boot? I heard some pretty terrible things. 

Hardware Unboxed had some tests in a comparison video:

 

null

[Th3ChosenOn3]

8 minutes ago, Lurker said:

How long do these 7XXX systems take to boot? I heard some pretty terrible things. 

Mine takes about 20 seconds on my ASRock board with two sticks of 32GB DDR5 overclocked to 6200 MHz.

Boot times are dependent on the motherboard and the memory configuration and can take up to a minute, but motherboard venders are working to get boot times quicker.

Gigabyte boards have the fastest boot times at the moment I think.

[Lurker]

I haven't had a PC boot in 20 seconds since going from Harddisk to SSD. (And this is SATA SSDs, not even NVME) 

This might not be important for people who run their machines 24/7 but I like to shut down my PC when it's not in use. 

[AngleOff66]

Asus statement on damaged CPU.

https://videocardz.com/newz/asus-statement-on-damaged-ryzen-7000-cpus-mentions-amd-expo-and-soc-voltage

I think about new builders/system owners who may not understand/be aware of these motherboards that can and do push the voltages up without warning the user.  Isnt it the responsibility of the MB manufacturer to make sure that their voltage bosts do not fry the cpu/ram etc etc?

[EightyDuce]

40 minutes ago, Lurker said:

I haven't had a PC boot in 20 seconds since going from Harddisk to SSD. (And this is SATA SSDs, not even NVME) 

This might not be important for people who run their machines 24/7 but I like to shut down my PC when it's not in use. 

While I miss my sub 10 second boot times of my old 7700K, I don't see how an extra 10-15 seconds has any impact unless you're restarting your computer constantly. 

It does get to be a little annoying when Overclocking/tuning memory and need to reboot a lot. But day to day, at least to me, doesn't seem that bad. 

[Fakum]

Im still wrapping my head around this, and my parts are trickling in still. This makes me re-think something I noticed the other day. My MOBO is still in the box but I did remove its CPU protection cover to place my CPU into it, but when I removed the protection cover, I noticed 1 small spot that was reflecting the light differently then the rest, I didnt see anything real obvious so I shrugged it off and inserted the CPU. When this matter emerged, I decided to remove the CPU and check it out again. Although this "spot" is not in the same location as what I see in those burnt chip videos, I wonder if I should be concerned about this spot and send this board back for another? I dont know if this is a concern or not? Lower Left Spot, It doesnt show up as much in the photo as it does in person. In person, it really stands out. Thoughts?

null

 

Edited April 25, 2023 by Fakum

[EightyDuce]

1 hour ago, Fakum said:

Im still wrapping my head around this, and my parts are trickling in still. This makes me re-think something I noticed the other day. My MOBO is still in the box but I did remove its CPU protection cover to place my CPU into it, but when I removed the protection cover, I noticed 1 small spot that was reflecting the light differently then the rest, I didnt see anything real obvious so I shrugged it off and inserted the CPU. When this matter emerged, I decided to remove the CPU and check it out again. Although this "spot" is not in the same location as what I see in those burnt chip videos, I wonder if I should be concerned about this spot and send this board back for another? I dont know if this is a concern or not? Lower Left Spot, It doesnt show up as much in the photo as it does in person. In person, it really stands out. Thoughts?

null

 

 

From the picture kinda looks like the pin may be bent. 

[Fakum]

I looked as best I could with magnifier, couldnt see well enough. Definitely something odd! Sending it back, Thank you Prime!

[BitMaster]

I hope "they" accept it back as RMA.

From what I know, bent pins are seen as a customer mishap and it will be almost impossible to proof it came this way.

Good Luck !

[EightyDuce]

2 hours ago, BitMaster said:

I hope "they" accept it back as RMA.

From what I know, bent pins are seen as a customer mishap and it will be almost impossible to proof it came this way.

Good Luck !

When returning to Amazon, less is more when defining a reason for a return. I've returned a motherboard and two CPUs and never had a remote issue with simply ticking the "defective" box and hitting submit. 

Now if it it's not defective and you know you borked it, well.. That's  between you and your ethics. 

[dugite57]

A bit of a response from AMD; "We are aware of a limited number of reports online claiming that excess voltage while overclocking may have damaged the motherboard socket and pin pads. We are actively investigating the situation and are working with our ODM partners to ensure voltages applied to Ryzen 7000X3D CPUs via motherboard BIOS settings are within product specifications. Anyone whose CPU may have been impacted by this issue should contact AMD customer support." -- AMD Spokesperson to Tom's Hardware.

https://www.tomshardware.com/news/amd-ryzen-7000-burning-out-root-cause-identified-expo-and-soc-voltages-to-blame

I conject that there will be a fix via bios coming soon, In the meantime I'll be letting mine just run at stock until this is sorted.  These CPU's are IMO pretty much overclocked by default. There's not much headroom like we saw in our old unlocked intel cpu's like my 4770k etc.  So for me the gains by overclocking offer a poor risk vs reward result.  The cache chips siting on top of the cpus compromise the cooling and it doesn't take muich of a voltage tweek to fry it.

[Hoirtel]

7 hours ago, dugite57 said:

A bit of a response from AMD; "We are aware of a limited number of reports online claiming that excess voltage while overclocking may have damaged the motherboard socket and pin pads. We are actively investigating the situation and are working with our ODM partners to ensure voltages applied to Ryzen 7000X3D CPUs via motherboard BIOS settings are within product specifications. Anyone whose CPU may have been impacted by this issue should contact AMD customer support." -- AMD Spokesperson to Tom's Hardware.

https://www.tomshardware.com/news/amd-ryzen-7000-burning-out-root-cause-identified-expo-and-soc-voltages-to-blame

I conject that there will be a fix via bios coming soon, In the meantime I'll be letting mine just run at stock until this is sorted.  These CPU's are IMO pretty much overclocked by default. There's not much headroom like we saw in our old unlocked intel cpu's like my 4770k etc.  So for me the gains by overclocking offer a poor risk vs reward result.  The cache chips siting on top of the cpus compromise the cooling and it doesn't take muich of a voltage tweek to fry it.

It seem the root cause of the issue lies in RAM overclocking and not directly CPU overclocking. Some ram expos require SOC voltage over 1.3 (which is on CPU) which seems to be the danger zone over extended periods. This is still early information so I am sure there will be more detail to come, the safe position at the moment is not not run any expo settings. This could be problematic as if the fix involves a hard lock on SOC voltage then it may make a whole load of EXPO settings invalid for life and the only way will be manual tuning or standard timings.

Considering vcache is supposed to be less sensitive to memory I would be interested how much this reduces performance.  I will be building my machine soon and will not run EXPO at this time, coming from my old PC I still expect better performance and I will see what happens. Bit annoyed at this whole EXPO/XMP thing, they all advertise it but when it goes wrong they say that it shouldn't be done.

When I build my new machine and it asks for a name, it will be "shouldhaveboughtintel"

Source: 

Update: Just noticed that yesterday (25/4) another new ASUS bios has been released and all the last ones have been removed. This is a beta for now, info included is: 

SoC voltage for Ryzen 7000X3D series limited to a maximum of 1.30V to protect the CPU and motherboard.

Before running the USB BIOS Flashback tool, please rename the BIOS file (SX670EF.CAP) using BIOSRenamer.

Please note that this is a beta BIOS version of the motherboard which is still undergoing final testing before its official release. The UEFI, its firmware and all content found on it are provided on an “as is” and “as available” basis. ASUS does not give any warranties, whether express or limited, as to the suitability, compatibility, or usability of the UEFI, its firmware or any of its content. Except as provided in the Product warranty and to the maximum extent permitted by law, ASUS is not responsible for direct, special, incidental or consequential damages resulting from using this beta BIOS.

This includes a nice little note about using a beta BIOS which is currently the only option... Seems I won't be building at all now.

Edited April 26, 2023 by Hoirtel

[EightyDuce]

Yeah just saw the bios drop. So my take on it from my experience overclocking DDR5 on AM5, SOC voltage of 1.3v is actually quite a high amount. I don't have any EXPO kits on hand and have only done manual overclocks, but the most voltage on the SOC that I needed to use to hit 6200/2167mhz on the RAM/IF  between 3 sets of RAM kits and 3 different CPUs, was 1.25v. And that was on  1 of the 3,

the other two did it at 1.225v. The 7700X could do 6400 at 1.3v SOC. To run at 6000 you shouldn't need more than 1.2v SOC unless the memory controller on your CPU is terrible (silicon lottery). 

It seems the manufacturers are being very loose with the voltage as they are with timings when it comes to higher clock speeds. If in fact they push 1.3v as part on an expo profile is a bit baffling to me unless they did very little stability testing/tuning and just cranked the voltage up to get it stable. 

Its unfortunate that they took a shotgun approach and straight up locked max SOC at 1.3 instead of limiting 'auto" setting to no more than 1.3v an leaving the ability to manually set voltage alone, maybe slap a warning on it. 

Edited April 26, 2023 by EightyDuce

[SilentSierra]

I don't know to much about PC, but I considering purchase a Ryzen 7 7800X3D and ASUS ROG STRIX B650E-F or ASUS TUF Gaming X670-PLUS WIFI with 64GB DDR5 5200MHz.

Will I have this problem of burning the CPU and motherboard running in default memory profile? Or this stuff is on EXPO overclocking memories?

I don't intend to overclock the system.

Best regards.

[BitMaster]

vor 8 Minuten schrieb SilentSierra:

I don't know to much about PC, but I considering purchase a Ryzen 7 7800X3D and ASUS ROG STRIX B650E-F or ASUS TUF Gaming X670-PLUS WIFI with 64GB DDR5 5200MHz.

Will I have this problem of burning the CPU and motherboard running in default memory profile? Or this stuff is on EXPO overclocking memories?

I don't intend to overclock the system.

Best regards.

THAT is the dilemma we all face !

AMD and Intel advertise their CPU's with XMP and EXPO but you better not touch it, it's off limits and voids your warranty.

Just, all reviewers test those CPU's WITH those memory profiles enabled, even pushed further, way way beyond Warranty so to say.

If you run them at stock speeds, JDEC/SPD, you loose a lot of performance. In order to get that extra performance, the bit of cake that makes ALL the difference, you need to void

warranty and call yourself asn overclocker but all you did was enable the memory profile they all advertise with.

 

All reviewers should reach out and make a statement. Test all CPU's ONLY with standard RAM speeds until AMD and Intel come up with something better.

Why show results that are potentially dangerous for your Board, CPU and Cash !