7800X3D will be an utter failure of a CPU

Despite all the goofy shit going around AM5 mbs aren't imploding, 7000 series CPUs aren't burning whole cities down, the sky isn't falling and the world is still turning. Man...what a close call!

All of the x3d chips are great options.

Seriously. It's not to say that the GN was bullshit (it wasn't), but it does feel like a case of hyperbole for clicks.
When the videos first hit I started to second guess my purchases, but my new setup has been great. Basically exactly what I wanted.
With the crazy sales going on, I wouldn't hesitate to recommend the same thing to anyone.
 
Eh, the "stay tuned" bit from GN pretty much confirmed the cpus weren't ticking time bombs.

It'd be incredibly negligent of GN to pull that if there was a serious issue that was so bad it had to be addressed asap.
 
Extra datapoints :

1) Been running the 7900X setup in my sig since last Nov, still on the early Gigabyte F3c BIOS that I flashed then.
No issues with EXPO on, VSOC behaves itself (<= 1.3V, usually ~1.29V max under various loads)

2) Recently built another 7900X setup (the MC combo, so ASUS B650E-F mobo)
Immediately flashed to the 1413 non-beta BIOS, turned EXPO on, no issues with VSOC (<= 1.3V, typically ~1.28V)
Only annoying thing is an extra-long POST process (Gigabyte is instant, this ASUS takes +10 seconds or so)
 
(<= 1.3V, usually ~1.29V max under various loads)
it should still be lower than that. try manually setting it to 1.2/1.25v and see if its stable
boot delay is the ram training, gb isnt* doing as thorough of a job, i assume.

edit; isnt, not is
 
Last edited:
it should still be lower than that. try manually setting it to 1.2/1.25v and see if its stable
boot delay is the ram training, gb is doing as thorough of a job, i assume.
Haven't had the time to mess around with both setups, but figured I could get away with 1.25VSOC since I'm only running 6000-CL30

Am familiar with the initial RAM training on AM5, but this delay is on every cold boot for the ASUS.
Seems to be a common complaint on ASUS AM5 mobos (that and the dumb Armoury Crate BIOS option)
Non-existent on the Gigabyte (cold boot to Windows login prompt in < 10s)
 
Haven't had the time to mess around with both setups, but figured I could get away with 1.25VSOC since I'm only running 6000-CL30

Am familiar with the initial RAM training on AM5, but this delay is on every cold boot for the ASUS.
Seems to be a common complaint on ASUS AM5 mobos (that and the dumb Armoury Crate BIOS option)
Non-existent on the Gigabyte (cold boot to Windows login prompt in < 10s)
dig in the boot options and make sure there are no delays set for drives, make sure that your boot drive is the only boot option, might help lower it. check if you have ram training options, some have auto, minimal full etc.
 
dig in the boot options and make sure there are no delays set for drives, make sure that your boot drive is the only boot option, might help lower it. check if you have ram training options, some have auto, minimal full etc.
On the todo list. That setup was created in a hurry (long story - replaced the core 5900X components chasing a random reboot issue, turned out to be a defective Seasonic Focus GM PSU of all things)
Wasn't going to rip the 7900X out, and reinstall the 5900X, so what the heck, let's call it an impromptu upgrade! :)

Only flashed the 1413 BIOS, and validated the same PBO -30 curve + 135W PPT limit setup as with my first 7900X, and called it good (for now)
 
I've giving it a go right now. It runs fine, but man it still seems buggy. I've had the lighting go haywire (making it tough to see/do anything) twice and sometimes when leaving a 30fps cutscene the game will just stay at 30fps.

EDIT: I ended up full-on bailing. This kept happening to me at random times (usually after going to one of the menus). When I finally made it through the chapter, I discovered that they shut the damned movie servers off. You can watch the movies on YouTube, but that combined with the wonky lighting/colors was too much. It's honestly too bad. The part that I played seemed fine. A little bit like Max Payne, Uncharted, and Control.
Steam or Microsoft Store version? The episodes were still working on the Steam version, last I checked.
 
Eh, the "stay tuned" bit from GN pretty much confirmed the cpus weren't ticking time bombs.

It'd be incredibly negligent of GN to pull that if there was a serious issue that was so bad it had to be addressed asap.

Yeah I'm actually slightly disappointed with GN on this one. They made it seem like such a huge deal saying they got a motherboard to "self immolate" and all. If there really was a severe risk of fire hazard then the issue wouldn't have been put off and would be more like in the case of the NZXT H1. Instead we got a "stay tuned for our video guys!" which was a clear indicator that the issue wasn't so serious after all.
 
Yeah I'm actually slightly disappointed with GN on this one. They made it seem like such a huge deal saying they got a motherboard to "self immolate" and all. If there really was a severe risk of fire hazard then the issue wouldn't have been put off and would be more like in the case of the NZXT H1. Instead we got a "stay tuned for our video guys!" which was a clear indicator that the issue wasn't so serious after all.

typically what he'll do is give companies a chance to respond. so if they're actually being responsive to him and are willing the setup a meeting like he asked then he's probably holding off on publishing the video unless they keep pushing it back. he did the same thing with the H1 issue where they waited 30 days before posting it since NZXT was actually communicating with them. from what he's hinted at i'm going to guess he's going to get the meeting he requested with ASUS while they're in Taiwan but we'll see.
 
Steam or Microsoft Store version? The episodes were still working on the Steam version, last I checked.

Microsoft Store/Gamepass. I looked up the error it kept and Reddit was full and people with the same experience. Ditto with the wonky colors thanks to the forced DX12 version of the game.
 
typically what he'll do is give companies a chance to respond. so if they're actually being responsive to him and are willing the setup a meeting like he asked then he's probably holding off on publishing the video unless they keep pushing it back. he did the same thing with the H1 issue where they waited 30 days before posting it since NZXT was actually communicating with them. from what he's hinted at i'm going to guess he's going to get the meeting he requested with ASUS while they're in Taiwan but we'll see.

I forgot exactly how the entire H1 thing went down but did he ever give us something like "In the mean time, do ________ to prevent your PC from going up in flames" while we wait for NZXT to respond? The H1 did have a real fire hazard risk so I believe he did tell users what they should be doing while they wait for a response from NZXT. In the case of the 7800X3D there was no such thing given, probably because the actual risk was super low.
 
I forgot exactly how the entire H1 thing went down but did he ever give us something like "In the mean time, do ________ to prevent your PC from going up in flames" while we wait for NZXT to respond? The H1 did have a real fire hazard risk so I believe he did tell users what they should be doing while they wait for a response from NZXT. In the case of the 7800X3D there was no such thing given, probably because the actual risk was super low.
basically just said not to use it until the diy fix kit was made available and also gave the screw details to go buy your own non conductive screws yourself if you had a local hobby shop.

as far as the 7800X3D/ASUS issue, he did.. not use XPO and/or manually lower the SOC voltage below 1.3v until an actual fixed bios was released.. the risk of melting the cpu/socket wasn't high but it was a gradual damage until it formed the short thus melting the pcb/socket so even the short hadn't happened yet more then likely the cpu was/is already damaged if you had XPO enabled with a 7800X3D cpu on one of the ASUS x670e boards. whether or not any potential damage will effect performance or longevity of the CPU over time we won't know for a while.
 
Can someone give me the TLDR?

Looking to drop in a 7800X3D on an Asus x670E board running DDR5-6000 - am I good to go with the latest BIOS update? Ok to enable AMD EXPO memory settings or is that what caused all the aforementioned issues?
 
dig in the boot options and make sure there are no delays set for drives, make sure that your boot drive is the only boot option, might help lower it. check if you have ram training options, some have auto, minimal full etc.
Figured it out. Had to set "Memory Context Restore" to Enabled rather than Auto.
Didn't think it was memory training related since the extra delay was only ~8-10s (vs the 30s+ that everyone seems to talk about), but I guess ASUS' really screwed the pooch with their AM5 offerings
(lots of reports of this switch not working on some recent BIOS versions across multiple AM5 models, etc.)
 
Can someone give me the TLDR?

Looking to drop in a 7800X3D on an Asus x670E board running DDR5-6000 - am I good to go with the latest BIOS update? Ok to enable AMD EXPO memory settings or is that what caused all the aforementioned issues?
I would avoid ASUS boards but that's just me. MSI Tomahawk is a good alternative.
 
Figured it out. Had to set "Memory Context Restore" to Enabled rather than Auto.
Didn't think it was memory training related since the extra delay was only ~8-10s (vs the 30s+ that everyone seems to talk about), but I guess ASUS' really screwed the pooch with their AM5 offerings
(lots of reports of this switch not working on some recent BIOS versions across multiple AM5 models, etc.)
but it was "memory training related" , youre now skipping it which is basically the "minimal" option i mentioned. glad you found it though. boots faster now?
1685635378398.png
 
Can someone give me the TLDR?

Looking to drop in a 7800X3D on an Asus x670E board running DDR5-6000 - am I good to go with the latest BIOS update? Ok to enable AMD EXPO memory settings or is that what caused all the aforementioned issues?

You don't even need the latest BIOS if your SoV voltage is below 1.30v. Just manually set it to that or lower like 1.25v or even 1.20v and you are good to go with using EXPO. I've been running my 7800X3D since launch day and SoC voltage was sitting at 1.25v the entire time so I haven't even bothered with any BIOS updates at all.
 
You don't even need the latest BIOS if your SoV voltage is below 1.30v. Just manually set it to that or lower like 1.25v or even 1.20v and you are good to go with using EXPO. I've been running my 7800X3D since launch day and SoC voltage was sitting at 1.25v the entire time so I haven't even bothered with any BIOS updates at all.

Does lowering the SoC voltage prevent you from being able to run the tightest timings? I have some Corsair Vengeance DDR6000 rated for 30-36-36-76 when set to EXPO II. Will I still be able to run the same settings?
 
Does lowering the SoC voltage prevent you from being able to run the tightest timings? I have some Corsair Vengeance DDR6000 rated for 30-36-36-76 when set to EXPO II. Will I still be able to run the same settings?

1.25v should handle that just fine. 1.20v is a maybe but really the 1.35-1.40v that was being pushed before all this was completely overkill and not good for the CPU.
 
but it was "memory training related" , youre now skipping it which is basically the "minimal" option i mentioned. glad you found it though. boots faster now?
View attachment 573948
I know, just didn't think that it was since the extra delay wasn't that long (~8-10s)
Was set to Auto before, but it clearly was never being enabled (or it might have been when EXPO was not enabled)
 
I am not going to read this whole thread but did the chips end up being a total fail?

I am tempted to get a 79503d and give my son my 5800x or is that a waste of money right now?
 
I am not going to read this whole thread but did the chips end up being a total fail?

I am tempted to get a 79503d and give my son my 5800x or is that a waste of money right now?
Nope. The chips work just fine and the boards do too when they're not trying to overvolt the cpu.
 
Finally a couple of games that don't perform like ass. SF6 and D4 run like a dream on my PC. 7800X3D purchase still parked. Let's hope the Intel 6.2 Gigawatts CPU is the monster we deserve.
 
Finally a couple of games that don't perform like ass. SF6 and D4 run like a dream on my PC. 7800X3D purchase still parked. Let's hope the Intel 6.2 Gigawatts CPU is the monster we deserve.
Can you elaborate on the Intel cpu you mention forgive me , I have been living in a cave.
 
I am more concerned about heat rather than power. 180 watts doesn't seem like a ton of heat for a 360 AIO especially during gaming.
Anyways, we will see what's what in August. No point nabbing a 7800X3D about now. No game challenging me to do so especially at 4K.
 
Using curve optimizer and turning off smt my 7800X3D has never exceeded 35 watts during gaming. Including VR MSFS 2020. I enjoy gaming a lot more when I'm not sweating profusely and VR in particular as the headset face pads just get disgusting.
 
After coming from a 165W 10940X (when stock at least, pretty sure I got it somewhere around 250W when overclocking), to a 105W 5800X3D (that I can't increase the power on), I'm not going to play Intel's game of "lets pump as many watts as we can in order to get more performance" - not anymore. It was cool (pardon the pun) at the time, but now I'm favoring high performance without needing to crank up the wattage to get there.

It's so nice to not need my Arctic Freezer II 420 to ramp up fans to a distracting level while I game.
 
I am more concerned about heat rather than power. 180 watts doesn't seem like a ton of heat for a 360 AIO especially during gaming.
Anyways, we will see what's what in August. No point nabbing a 7800X3D about now. No game challenging me to do so especially at 4K.

Well that 180 watts has to go somewhere, and it'll just go towards heating up your room. Doesn't seem like much I know but I was actually able to feel some difference in my room temperature when I upgraded from a 2080 Ti to 3090 which was dumping out an extra 100+ watts.
 
Well I suppose 35w on a 7900x3d is not bad but damn gpu pumping 300watts is sweltering heat.

My son has basically taken over my 5800x 6900xt box and I use my 3070ti Ryzen Lenovo Legion Laptop for gaming. With DLSS I'm easily way past 120 fps in Diablo 4 at 2k res on my Odyssey monitor.
 
Well that 180 watts has to go somewhere, and it'll just go towards heating up your room. Doesn't seem like much I know but I was actually able to feel some difference in my room temperature when I upgraded from a 2080 Ti to 3090 which was dumping out an extra 100+ watts.
Not to mention the noise generated by fans/pumps. I know when my AIO coolant gets heat saturated the fans have to work harder to get the temps stable.
 
Great video by HUB on 7800x3d vs 5800x3d. I couldn’t care less about Hogwarts now that I finished it. For rest of the games largely it’s a wash at 1440P and 4K.

Further confirms that even after a few months the perf diff in platforms isn’t crazy.
 
Great video by HUB on 7800x3d vs 5800x3d. I couldn’t care less about Hogwarts now that I finished it. For rest of the games largely it’s a wash at 1440P and 4K.

Further confirms that even after a few months the perf diff in platforms isn’t crazy.

On the flip side, if you play a lot of emulated games on Yuzu like I do then the performance improvement can be massive like in Tears Of The Kingdom.

https://yuzu-emu.org/entry/yuzu-progress-report-may-2023/

"The latest CPUs (Zen 4/13th gen Core, always speaking of desktop products) provide massive improvements in IPC, RAM bandwidth, and cache sizes. Where a Ryzen 7 5800X3D barely manages 55 FPS, a Ryzen 5 7600 reaches 90 FPS."
 
Well, I'm not an EE, but for most electronics, I'm pretty sure they're the same thing. Cunk reference for effect.


I'm an EE and heat dissipated is related (depends on) to the resistance of the circuitry and the current delivered. It's not the same thing.

Power delivered is simple P = IV. Power DISSIPATED (as heat) is } = I^2 * R.

1687121119723.png
 
I'm an EE and heat dissipated is related (depends on) to the resistance of the circuitry and the current delivered. It's not the same thing.

Power delivered is simple P = IV. Power DISSIPATED (as heat) is } = I^2 * R.

View attachment 577702
Don't your words and picture say two different things? If both are correct then Power Delivered = Power Dissipated. I can divide both sides by power and arrive at Delivered = Dissipated. Assuming we're not taking large portions of the power and converting it into some other energy state, like radio waves or other types of emitted radiation. I guess my comment should have pointed to CPUs, as probably north of 95% of the power draw will be waste heat torching your rooms.
 
Back
Top