• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

AMD Ryzen 7 5800X3D

You could probably ask if ASUS can lend you one of their Beta/Alpha BIOS to test the chip out further ~
www.techpowerup.com

MSI X570 Beta BIOS Enables Ryzen 7 5800X3D Tweaking

A user (via Komachi_Ensaka(opens in new tab)) on the MSI forums(opens in new tab) has shared a beta firmware for MSI's MEG X570 Unify motherboard that unlocks several previously unavailable options for the Ryzen 7 5800X3D. Before the Ryzen 7 5800X3D's launch, AMD had confirmed that overclocking...
www.techpowerup.com www.techpowerup.com
Pretty sure they're also working on easing some of the restrictions on this, not unlike MSI.

Also the mandatory "Try try this at your own risk" warning :D
 
Amaze said:
That is extremely low. The best I could do so far with the PBO curve tool was 15013.
github.com

How-to-undervolt-AMD-RYZEN-5800X3D-Guide-with-PBO2-Tuner/README.md at main · PrimeO7/How-to-undervolt-AMD-RYZEN-5800X3D-Guide-with-PBO2-Tuner

Get the Most out of your 5800X3D using PBO Curve Optimizer! - PrimeO7/How-to-undervolt-AMD-RYZEN-5800X3D-Guide-with-PBO2-Tuner
github.com github.com
Click to expand...

Yes it was.

Using PBO2 tuner, I get 4350 all core in Cinebench and about 14700 points.
Still I haven't managed to get this level of performance tweaking only the BIOS settings.
Enabling the PBO in BIOS and using the PBO2 tuner, I managed to get a reasonable level of performance out of it.
 
Last edited:
gffermari said:
Yes it was.

Using PBO2 tuner, I get 4350 all core in Cinebench and about 14700 points.
Still I haven't managed to get this level of performance tweaking only settings the BIOS settings.
Enabling the PBO in BIOS and using the PBO2 tuner, I managed to get a reasonable level of performance out of it.
Click to expand...
That's more like it :toast:
 
Shatun_Bear said:
Very impressive. Not sure why they are stealth launching this, it matches or beats the limited ediition behemoth power guzzling 12900KS for a much, much lower price.

Low stock maybe?
Click to expand...
It does NOT match or beat the KS when the latter is in fact 2% faster on average while certainly not using much more power while gaming (this myth needs to stop). Oh and it absolutely demolishes it in everything else at a comparable energy efficiency:
efficiency-multithread.png
 
loool mate.
We're talking about a 2 years old tech 8 core cpu with some added cache at 4.5Ghz at best and you for a 800e 5.5Ghz monstrosity with pcie5 and ddr5.

The 3D is not worth it because it's expensive for not being an all around cpu. A theoretical 5900X3D would be that.
The KS is just not worth it.
 
A whopping 2% eh.
Either way, those power consumption numbers are demolished by the optimized curve that users have discovered.
 
gffermari said:
loool mate.
We're talking about a 2 years old tech 8 core cpu with some added cache at 4.5Ghz at best and you for a 800e 5.5Ghz monstrosity with pcie5 and ddr5.

The 3D is not worth it because it's expensive for not being an all around cpu. A theoretical 5900X3D would be that.
The KS is just not worth it.
Click to expand...
Of course the KS isn't worth it for most people, even if they are assembling a high end gaming system, BUT it is the best in pretty much everything down the line, while the X3D is only the best gaming Ryzen, while getting smoked even by the lowly (in comparison, especially price wise) 12600k in everything else. It should be a $300 chip and I suspect the price will eventually get there...or they'll just stop making them after all the hardcore AMD enthusiasts will already have them and no one else will take a second glance at it anymore. Either way, we'll see, but honestly even at 300, it would be a somewhat difficult sell to someone buying an all new platform...
 
HenrySomeone said:
Of course the KS isn't worth it for most people, even if they are assembling a high end gaming system, BUT it is the best in pretty much everything down the line, while the X3D is only the best gaming Ryzen, while getting smoked even by the lowly (in comparison, especially price wise) 12600k in everything else. It should be a $300 chip and I suspect the price will eventually get there...or they'll just stop making them after all the hardcore AMD enthusiasts will already have them and no one else will take a second glance at it anymore. Either way, we'll see, but honestly even at 300, it would be a somewhat difficult sell to someone buying an all new platform...
Click to expand...

We 5950X owners say hi and send our heartfelt regards.

Amaze said:
A whopping 2% eh.
Either way, those power consumption numbers are demolished by the optimized curve that users have discovered.
Click to expand...

I still find it amazing that as power efficient as Zen 3 is, there's still room for improvement in them. It's a shame that the 5800X3D doesn't seem to have gone through the standard binning process, they just test if the cores work and ship the processor, which is why you're seeing so many of them doing -25 to -30 in the curve optimizer. Those values would indicate that the processors are quite horribly binned, but given the fixed v/f curve that makes sense. If eventually unlocked (even if extraofficially), it'd be interesting to see how different X3Ds behave, as the other Ryzen SKUs seem to live and die by their binning quality (e.g. 5950X's and very narrow CO range).

My personal 5950X only does -2 all-core, even if setting -2 and then -3 to the cores identified as the best ones will still result in a crash. The PBO scalar setting introduces crashes above 1x and it also hates to have the XFR frequency range increased, even if by just 50 MHz. It's an excellent bin CPU, even without PBO it tends to clock itself very, very generously, which is why I kind of get cold feet about upgrading it. Might just wait for the chance to buy something like a C8DH or C8E cheaper once they're no longer the flagships :p
 
Last edited:
Amaze said:
A whopping 2% eh.
Either way, those power consumption numbers are demolished by the optimized curve that users have discovered.
Click to expand...
Definitely not whopping, no, but it's not the other way around like the guy states and just like I said above - for purely gaming neither of these two is a smart choice. Also, not sure what kind of optimized curve you're talking about.
 
HenrySomeone said:
Definitely not whopping, no, but it's not the other way around like the guy states and just like I said above - for purely gaming neither of these two is a smart choice. Also, not sure what kind of optimized curve you're talking about.
Click to expand...

hothardware.com

Maximizing Ryzen 5000 Performance With AMD Curve Optimizer

AMD's new AGESA update adds simple controls for complex overclocking and under-volting that can bring nice performance gains.
hothardware.com hothardware.com

This. On Intel processors you still do traditional multiplier-based overclocks, but on AMD, you can simply tweak and reprogram the processor's reliability algorithm using the provided curve optimizer function to extract every last bit of performance your processor has, with 25 MHz granularity, since the multipliers on Ryzen are adjustable in 0.25x steps. The best part is that it does that by itself, all you need to be concerned with is keeping it cold. Due to being locked and programmed with a fixed curve, the X3D seems to have very, very loose binning. PBO2 Tuner works but it's not even *supposed* to, Ryzen Master and BIOSes won't allow CO adjustment even to negative voltage points.
 
AMD should really have kept negativce CO for X3D. It does not harm the cache and actually improves (atleast multithreaded) performance.
I mean they had a whole year to work on it and test it with various settings.
 
Yup, as soon as Intel's 10nm chips finally came out, AMD has already started to drift towards their once common practices, namely taking too long to get something new out and even then, it's not properly polished. It probably won't be long before they'll be OCing their chips to high heavens (in the manner of their black edition skus) in hopes of catching up or rather not falling behind even more. My bet would be after Meteor Lake lands next year, but could be even sooner as 13900k with its 24 cores is bound to be a multi thread monster!
 
HenrySomeone said:
Yup, as soon as Intel's 10nm chips finally came out, AMD has already started to drift towards their once common practices, namely taking too long to get something new out and even then, it's not properly polished. It probably won't be long before they'll be OCing their chips to high heavens (in the manner of their black edition skus) in hopes of catching up or rather not falling behind even more. My bet would be after Meteor Lake lands next year, but could be even sooner as 13900k with its 24 cores is bound to be a multi thread monster!
Click to expand...

Hey you can't say that. you will hurt about 60% of TPU members feelings /s
 
well you're flatlining 80C, you're thermal throttling


for <100Hz gaming, get an i5 or R5 and go play games.
You only need the faster chips (especially the x3D) for high refresh rate gaming, where the 99% lows are smoother.

1655771171796.png
 
fevgatos said:
Ive tested on my pc, 6000c30 1t that im running right now is up to 40% faster than 4800c40 in cyberpunk and sotr. Haven't tested more games, but its kinda obvious there is a big difference
Click to expand...
W1zzard posts actual proof,. That's kinda obvious, what your saying is 99.9% your opinion until you back it up with 100% proof.
Until then it's lips flapping.
 
TheoneandonlyMrK said:
W1zzard posts actual proof,. That's kinda obvious, what your saying is 99.9% your opinion until you back it up with 100% proof.
Until then it's lips flapping.
Click to expand...
Actual proof... As in, pretty bars made on paint. Dont have any problem with his reviews, but don't tell me they contain actual proof, cause they just dont. Bars made on paint is not proof.
 
fevgatos said:
Actual proof... As in, pretty bars made on paint. Dont have any problem with his reviews, but don't tell me they contain actual proof, cause they just dont. Bars made on paint is not proof.
Click to expand...
Right your going on ignore.

You need to realise, your proof is actually nothing, word's.

If I and many others didn't think the likes of w1zzard uses actual benchmarks, collates results , then presents those results in an informative way to PROVE they're opinion and any statements, this site wouldn't still be here for you to flap your mouth in while showing zero proofs.

I know which opinion I'm backing, and you going on ignore should tell you it's not yours.
 
You must log in or register to reply here.
Back
Top