*Official Benchmark Thread* - Post it here or it didn't happen :D

[tps3443]

12 minutes ago, Mr. Fox said:

My eyes are peeled and I am waiting with baited breath. 😄

 

Im curious of your results! Especially after delid. 

[Mr. Fox]

21 minutes ago, tps3443 said:

Im curious of your results! Especially after delid. 

I will probably be lazy again about the delid. If my temps without a delid are anything similar to what the 12900KS is after a delid I might not do it right away, or maybe not at all if the chiller gets the job done well enough. Of course, everything is contingent on it being a good sample and not an exchange candidate.

[tps3443]

54 minutes ago, Mr. Fox said:

I will probably be lazy again about the delid. If my temps without a delid are anything similar to what the 12900KS is after a delid I might not do it right away, or maybe not at all if the chiller gets the job done well enough. Of course, everything is contingent on it being a good sample and not an exchange candidate.

You planning to bin for a gold sample 13900K? 

[jaybee83]

22 hours ago, Papusan said:

Another GeForce RTX 4090 16-pin Adapter Bites The Dust

published about 1 hour ago

 

Great balls of fire!!!

 

https://www.tomshardware.com/news/another-geforce-rtx-4090-16-pin-adapter-bites-the-dust

 

I wonder how long it will take before we see a statement on status from Nvidia. 
 

Nvidia launched an investigation into the matter when the first case surfaced, and with good reason, since meltdowns are starting to become widespread. The user feedback shows that the problem may reside with the design of the 12VHPWR power connector. Bending the cables too close to the connector seemingly causes some terminals to loosen up, leading to uneven mating.

 

great, now i got that song from Top Gun stuck in my head @goodnes gracious, great balls of fire! 🤣

but its good that Nvidia takes these reports seriously. what im hoping for is an improved connector with sturdier cables / 90 degree connector / thicker plastic isolation which will be offered as a free replacement by the Nvidiots and the AIBs. fingers crossed!

 

21 hours ago, Reciever said:

Then you know, make it 4?

 

lulz. the funny think about this is that they still do 4 via those octopus adapter cables. but in that case Nvidia and the AIBs can push the blame to the consumer for not being up to snuff with sub ATX 3.0 PSUs, such peasants who dare buy a 4090! but with native 4x8 pin would be an embarrassment, of course...who cares about safety margins, its all about thin n light baby! even with bigass brick style gpus 😄

 

20 hours ago, Papusan said:

The PSU's quality can be ok but they cut down on the wire gauge thickness. More metal increase prices and thinner wires make it easier for (less savy) people to route the cables properly in their chassis. Smaller chassis the worse. Should have been at least 16 gauge thickness on all (PCIe) cables. 15 thickness would be more proper (higher safety margin/higher current). 

 

 

Nvidia investigating reports of RTX 4090 power cables burning or melting theverge.com
https://www.theverge.com/2022/10/25/23422349/nvidia-rtx-4090-power-cables-connectors-melting-burning

 

I wonder how much this will affect the supply of 4080 and more of the 4090 cards coming weeks. Maybe the new new sibling of 3060 will help nvidia and it's partners to keep up sales of Nvidia graphics in this difficult  time. 

 

The Nvidia fire crack continue... So appropriate with some fire music.

 

Update 10/25/2022 4:30 pm PT

A fourth case has allegedly surfaced on a YouTube video(opens in new tab) from the Bang4BuckPC Gamer channel. The user allegedly purchased a prebuilt gaming PC from CyberPowerPC with a GeForce RTX 4090. The power connector for the graphics card seemingly melted, and the user returned the PC to CyberPowerPC. We've reached out to the system builder for comment.

 

 

yep thats exactly what i was afraid of, cheapass prebuilt systems with "pseudo-highend" hardware being sloppily slapped together causing fire hazards in unsuspecting people's homes, yay....

 

14 hours ago, Mr. Fox said:

Here is some interesting information about A-die versus M-die timings and the impact of performance. A-die is only better when it can handle good timings, but apparently sometimes it does not.

 

 

 

huh so A-Die silicon lottery spead is wider than M-Die but its peaks can reach higher levels than M-Die?

 

10 hours ago, Mr. Fox said:

Block installed. Man, what a huge improvement.

 

loving me some hardware pr0n 😍 awesome temp improvements there! while we are on the topic of crossflashing vBios, i got a question for you and the rest:

ive been trying to crossflash the vBios on my Suprim X 4090 and it results in a black screen every time. system still works, it boots into windows, but no display output. other users on overclock.net have reported successful cross-flashing on the Suprim X, so im wondering whats up in my case? along the same lines, i have noticed that display output comes up only veeeery late on my system after power on. basically, whenever i want to enter bios, i need to switch to the iGPU, otherwise ill always miss the right moment to press Del / F2 to enter Bios. funkily, it doesnt help if i just spam it the whole time, i can see via the Mobo Q Codes that Bios is successfully entered but display output never comes up. its only initialized once i am on the windows logon screen. so....what gives?

could a manual switch to PCIe 4.0 on the gpu slot help out? ive read that "Auto" setting might cause some funky troubles sometimes.

 

in any case, any input appreciated guys!

 

3 hours ago, tps3443 said:

Hey, now Corsair and others needs to release a “Fire extinguisher for ATX cases” it’ll be sold as an accessory with included mounting, and it has 120/140 mounting holes. It can bolt on the back 120 slot of a case for quick access just in case your RTX4090 catches fire.

Miniature fire extinguisher.

 

yes id like one please 😛 

 

1 hour ago, Mr. Fox said:

My eyes are peeled and I am waiting with baited breath. 😄

 

 

are you though? 😄 similar to lots of tech youtubers, i could understand if after some extended period of time things tend to get....repetitive, i guess? basically, every new gen brings the same old questions, same perf gains, etc. plus, the added barricades from the OS side and crappier games tend to restrict actual use of new hardware. understandable if one gets jaded in that context.

 

1 hour ago, Mr. Fox said:

I will probably be lazy again about the delid. If my temps without a delid are anything similar to what the 12900KS is after a delid I might not do it right away, or maybe not at all if the chiller gets the job done well enough. Of course, everything is contingent on it being a good sample and not an exchange candidate.

 

haha just like above, one might get more and more frugal as time passes. just being "content" and not crazily excited anymore like a youngling 😄 

 

[Mr. Fox]

23 minutes ago, jaybee83 said:

ive been trying to crossflash the vBios on my Suprim X 4090 and it results in a black screen every time. system still works, it boots into windows, but no display output.

Have you tested with all of the display output ports. The behavior may change depending on which port your monitor is connected to. I have done a fair amount of cross-flashing of vBIOSes and the "address" (for lack of a more technically accurate term) can change from one brand to the next. Example here is the ASUS vBIOS I am using. ASUS uses 2xHDMI and 2xDP. EVGA uses 1xHMDI and 3xDP. So, to use all three of my monitors requires that one of them be connected to HDMI. I believe your Suprim X has three DP like EVGA so you might find success using a different port. If I leave all three monitors connected to DP one of them stops working because the ASUS firmware is designed to support two, not three, and by using one monitor on HDMI all three work.

23 minutes ago, jaybee83 said:

haha just like above, one might get more and more frugal as time passes. just being "content" and not crazily excited anymore like a youngling 😄 

I still desire crazy things, maybe even more than I used to, and want to play the role of mad scientist when it comes to PC tech. But, the rash of bad luck and disappointing purchases that I have had in the past few years has taken some of the starch out of my collar because I am not wealthy and can't afford to replace the things that I break. To be honest, I really can't afford some of what I buy to begin with, but it is an unhealthy addicition of sorts. So, it is more like caution to avoid disaster because I don't have the financial resources needed to throw caution to the wind. I certainly would if I could though.

---

Edit: Speaking of things I buy that I cannot afford...

Edited October 26, 2022 by Mr. Fox
13900K arrival

[tps3443]

10 minutes ago, Mr. Fox said:

Have you tested with all of the display output ports. The behavior may change depending on which port your monitor is connected to. I have done a fair amount of cross-flashing of vBIOSes and the "address" (for lack of a more technically accurate term) can change from one brand to the next. Example here is the ASUS vBIOS I am using. ASUS uses 2xHDMI and 2xDP. EVGA uses 1xHMDI and 3xDP. So, to use all three of my monitors requires that one of them be connected to HDMI. I believe your Suprim X has three DP like EVGA so you might find success using a different port. If I leave all three monitors connected to DP one of them stops working because the ASUS firmware is designed to support two, not three, and by using one monitor on HDMI all three work.

I still desire crazy things, maybe even more than I used to, and want to play the role of mad scientist when it comes to PC tech. But, the rash of bad luck and disappointing purchases that I have had in the past few years has taken some of the starch out of my collar because I am not wealthy and can't afford to replace the things that I break. To be honest, I really can't afford some of what I buy to begin with, but it is an unhealthy addicition of sorts. So, it is more like caution to avoid disaster because I don't have the financial resources needed to throw caution to the wind. I certainly would if I could though.

 

It would be cool if every time a new launch came, we could all get a pallet of (100) samples to test from and keep the best few!

 

We send the rest of the duds back with a note that says “TO INTEL: Try harder next time” 
 

One thing I learned is, if all CPU’s were golden we wouldn’t know what a dud was. And then it wouldn’t be exciting at all if we got a great sample. 

[jaybee83]

8 minutes ago, Mr. Fox said:

Have you tested with all of the display output ports. The behavior may change depending on which port your monitor is connected to. I have done a fair amount of cross-flashing of vBIOSes and the "address" (for lack of a more technically accurate term) can change from one brand to the next. Example here is the ASUS vBIOS I am using. ASUS uses 2xHDMI and 2xDP. EVGA uses 1xHMDI and 3xDP. So, to use all three of my monitors requires that one of them be connected to HDMI. I believe your Suprim X has three DP like EVGA so you might find success using a different port. If I leave all three monitors connected to DP one of them stops working because the ASUS firmware is designed to support two, not three, and by using one monitor on HDMI all three work.

I still desire crazy things, maybe even more than I used to, and want to play the role of mad scientist when it comes to PC tech. But, the rash of bad luck and disappointing purchases that I have had in the past few years has taken some of the starch out of my collar because I am not wealthy and can't afford to replace the things that I break. To be honest, I really can't afford some of what I buy to begin with, but it is an unhealthy addicition of sorts. So, it is more like caution to avoid disaster because I don't have the financial resources needed to throw caution to the wind. I certainly would if I could though.

 

yes checking all display outputs was something i tried. however, would u recommend doing a fresh boot every time i switch to another output? i just did it while it was sitting on the windows login screen. but with every output the display would come out of standby for a few seconds, searching for a signal before finding none and going back to standby.

 

i get that @cant afford this stuff. especially with those neverending price increases. at some point, who do they expect to buy their stuff? maybe thats part of the long game plan to push everyone to cloud services, subsciption for everything! including hardware, ugh...

[Mr. Fox]

12 minutes ago, jaybee83 said:

 

yes checking all display outputs was something i tried. however, would u recommend doing a fresh boot every time i switch to another output? i just did it while it was sitting on the windows login screen. but with every output the display would come out of standby for a few seconds, searching for a signal before finding none and going back to standby.

 

i get that @cant afford this stuff. especially with those neverending price increases. at some point, who do they expect to buy their stuff? maybe thats part of the long game plan to push everyone to cloud services, subsciption for everything! including hardware, ugh...

Yes, as a matter of fact I have to reinstall the driver and reboot between vBIOS flashes. If I do not the higher power limit is not detected even though the driver installs itself. If I only flash and reboot something causes the driver to not update the power limit, or something else ends up being off. This is something I started experiencing with Turing and it is that way for me with Ampere. Even when everything might seem to work right on the surface, the performance is often adversely affected if I do not reinstall the driver.

On the Eluktronics Mech-15 I had to reinstall the drivers after cross-flashing or no external displays would function.

[jaybee83]

1 hour ago, Mr. Fox said:

Yes, as a matter of fact I have to reinstall the driver and reboot between vBIOS flashes. If I do not the higher power limit is not detected even though the driver installs itself. If I only flash and reboot something causes the driver to not update the power limit. This is something I started experiencing with Turing and it is that way for me with Ampere. Everything might seem to work right on the surface, but performance is adversely affected.

On the Eluktronics Mech-15 I had to reinstall the drivers after cross-flashing or no external displays would function.

 

aha! that might be the culprit. i have automatic driver install disabled and the driver im running is a custom one, not the standard crap direct from Nvidia. so might make sense for me to remove the driver first, flash vbios, then try again for a display signal.

Jay had a very similar issue with the STRIX OC 4090 where it wouldnt be properly initialized due to the driver. thanks for the info, definitely worth checking out!

 

main reason why id like to flash is the increased 600W TDP some other AIB cards are offering. im noticing that in a lot of instances my card bounces off the 520W stock max limit, so the 600 could help out in this regard!

[Mr. Fox]

Just now, jaybee83 said:

 

aha! that might be the culprit. i have automatic driver install disabled and the driver im running is a custom one, not the standard crap direct from Nvidia. so might make sense for me to remove the driver first, flash vbios, then try again for a display signal.

Jay had a very similar issue with the STRIX OC 4090 where it wouldnt be properly initialized due to the driver. thanks for the info, definitely worth checking out!

If I had to guess it would be something gets changed in the Windows registry when the driver package is reinstalled. Just speculation on my part, but seems logical.

[jaybee83]

Just now, Mr. Fox said:

If I had to guess it would be something gets changed in the Windows registry when the driver package is reinstalled. Just speculation on my part, but seems logical.

see, for this kinda stuff its good to have the community around 🙂 ive just been out of the desktop game way too long, but its fun learning the new stuff!

[Mr. Fox]

3 minutes ago, jaybee83 said:

see, for this kinda stuff its good to have the community around 🙂 ive just been out of the desktop game way too long, but its fun learning the new stuff!

Indeed. I love it for the same reason.

 

To add  yet additional insight to previous comments, when I say reinstall the driver I mean run DDU and clean install. I thought that something was wrong with the Mech-15 not supporting an external monitor after cross-flashing with a Clevo vBIOS but apparently DDU removed registry components that caused conflicts and doing that fixed the conflicts. If I installed the driver on top of itself it didn't work correctly versus the clean installation.

[tps3443]

@Mr. Fox

 

What is Sp?!!!!!!!!

 

 

Waiting anxiously. 

[jaybee83]

18 minutes ago, Mr. Fox said:

Indeed. I love it for the same reason.

 

To add  yet additional insight to previous comments, when I say reinstall the driver I mean run DDU and clean install. I thought that something was wrong with the Mech-15 not supporting an external monitor after cross-flashing with a Clevo vBIOS but apparently DDU removed registry components that caused conflicts and doing that fixed the conflicts. If I installed the driver on top of itself it didn't work correctly versus the clean installation.

 

yep, DDU is a must, thats the same on laptop and desktop haha 🙂 

 

and indeed! WHATS YOUR SPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP? go ahead and show @Talon and @tps3443 what a REAL golden sample looks like 😄 

[Mr. Fox]

4 hours ago, tps3443 said:

@Mr. Fox

 

What is Sp?!!!!!!!!

 

 

Waiting anxiously. 

 

4 hours ago, jaybee83 said:

 

yep, DDU is a must, thats the same on laptop and desktop haha 🙂 

 

and indeed! WHATS YOUR SPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP? go ahead and show @Talon and @tps3443 what a REAL golden sample looks like 😄 

I am working but will know something soon. I can't stop working to disassemble my work computer to find out, LOL. Soon means today.

5 hours ago, tps3443 said:

You planning to bin for a gold sample 13900K? 

No, only reserving the right to return or exchange a crappy sample rather than live with it. I couldn't go through a full-fledged binning process even if I wanted to because I do not live within driving distance of a Micro Center, and I wouldn't ask a friend to be my gofer.

[tps3443]

Windows 11 is killing my R23 numbers. I have gotten 5.9Ghz all P-Cores stable with reasonable power levels about 350 watts or so. But, my score numbers are low.. Only 43K+ My updated windows 10 OS was so much faster. This Windows 11 OS is just terrible with benching for sure. 

Any help would be greatly appreciated. I’ve got the chip, I’ve got the cooling, but my OS is a garbage bloated turd…

 

5.9 all P-Cores is FAST btw. Bios voltage is only 1.330V!!!! I’m starting to believe I do have a great chip. The simple fact that I can run R23 at 5.9. I believe the new cold plate is what has done all of this and made this possible. That last one was extremely clogged up. 
 

 

[jaybee83]

29 minutes ago, Mr. Fox said:

 

I am working but will know something soon. I can't stop working to disassemble my work computer to find out, LOL. Soon means today.

No, only reserving the right to return or exchange a crappy sample rather than live with it. I couldn't go through a full-fledged binning process even if I wanted to because I do not living within driving distance of a Micro Center, and I wouldn't ask a friend to be my gofer.

 

of course u can stop working for that, gotta set priorities!

 

btw, reporting from a fully functioning "Strix OC 4090" 😄 ur pointers helped! reinstalling the driver helped bring back the display output after bios flash. 600W TDP FTW!

now to test and be on the lookout, should the Neptune 630W vBios ever drop...

[Papusan]

A new one. Nvidia is in deaper and deaper water. Can Jensen swim? My best guess...  Class action or recall of cards/in best case sending out new adapters. Nvidia's investigation team can't keep up with this pace. One per day is +60 before Christmas Eve. And there isn't many 4090 cards out there. 

 

 

I think my Asus 4090 TUF OC adapter is melting too

 

https://ibb.co/94Q97dX https://ibb.co/0q0RWPq https://ibb.co/YXK1d2h https://ibb.co/XYBHG5L

Had to disconnect my GPU to make space for the 13900k installation and I noticed that one of the pins looks melted. What should I do?

[electrosoft]

Curse of Ryzen dips strikes again.

 

New 10.0 patch which is the basis and engine for Dragonflight. I had hoped they had managed to address the Ryzen issues with those savage dips but they are still there.

 

Even with a 4090, same problems in WoW as with 5800x....sudden brutal 1% and min dips...

 

On my 12900k and 3090ti fps isn't as high (duh), but none of those brutal dips especially in PvP or in heavily populated areas (IE Oribos) and tons of CPU handled player data and physics.

 

5800X3D managed to smooth it out for the most part. Hopefully 7800X3D (or whatever they end up with) will sort that out again and then some.

 

 

[Mr. Fox]

On 10/24/2022 at 12:04 PM, johnksss said:

Damn. got a SP99 this time.

 

On 10/24/2022 at 3:41 PM, johnksss said:

 

Same here, SP99... but better P-cores and worse E-cores. 

 

Is that CPU Force 2 136 the same ASUS SP99 CPU?

[Talon]

26 minutes ago, Mr. Fox said:

 

Same here, SP99... but better P-cores and worse E-cores. 

 

Is that CPU Force 2 136 the same ASUS SP99 CPU?

 

Those are actually really strong P cores. 

 

Edit:

 

Test it and let us know! I expect since you're using high end water cooling it should go pretty far on those P cores. 

[Mr. Fox]

13 minutes ago, Talon said:

 

Those are actually really strong P cores. 

Yes they look really good so far. Probably not going to worry too much about the overall SP rating as much as the P cores. If 13th Gen anything like 12th the E-cores  are not going to overclock worth a damn anyway. Later tonight I will take it out of the Asus mobo see how it runs on the Dark. Before I take it out I'm going to see if the DDR4 memory overclock is any different on this generation or if it's still sucks the same. That would confirm that the issue is the motherboard is weak were memory overclocking is concerned.

[Talon]

10 minutes ago, Mr. Fox said:

Yes they look really good so far. Probably not going to worry too much about the overall SP rating as much as the P cores. If 13th Gen anything like 12th the E-cores  are not going to overclock worth a damn anyway. Later tonight I will take it out of the Asus mobo see how it runs on the Dark. Before I take it out I'm going to see if the DDR4 memory overclock is any different on this generation or if it's still sucks the same. That would confirm that the issue is the motherboard is weak were memory overclocking is concerned.

 

From what I'm hearing IMC is still a crap shoot at least with DDR4, some getting an extra bin or +133mhz over previous DDR4 stable. 4133 Gear 1 would be pretty monster though, 4266 likely a golden chip? 

[tps3443]

ddr5 7200 just showed up. 

 

@Talon

 

What is your CPU VDDQ voltage for DDR5 7200 XMP? 

[Mr. Fox]

34 minutes ago, Mr. Fox said:

Yes they look really good so far. Probably not going to worry too much about the overall SP rating as much as the P cores. If 13th Gen anything like 12th the E-cores  are not going to overclock worth a damn anyway. Later tonight I will take it out of the Asus mobo see how it runs on the Dark. Before I take it out I'm going to see if the DDR4 memory overclock is any different on this generation or if it's still sucks the same. That would confirm that the issue is the motherboard is weak were memory overclocking is concerned.

21 minutes ago, Talon said:

From what I'm hearing IMC is still a crap shoot at least with DDR4, some getting an extra bin or +133mhz over previous DDR4 stable. 4133 Gear 1 would be pretty monster though, 4266 likely a golden chip? 

DDR4 is absolutely a lot better than it was with 12900K and KS. It would not even POST at 4400 before.