Ryzen Owners Zen Garden

[Count von Schwalbe]

compared to the <60ns i managed on the 5800x, this is a tad slower
buuuut its getting there for zen2
View attachment 285251

Isn't the IMC the same on the chiplet-based Zen 2 and 3 CPUs?

So it would be pretty much down to process maturity.

 

[Mussels]

Isn't the IMC the same on the chiplet-based Zen 2 and 3 CPUs?

So it would be pretty much down to process maturity.

its weaker but not by much - this is an early 3700x

hitting a wall where tiny adjustments result in no POST, this might be her sweet spot

Interestingly this board shows TRFC differently in the bios, manual entries stop post

I was wrong. Bios shows 312 - different again

final tweaks done on gpu

gaming values, 1440 144fps

 

[The King]

its weaker but not by much - this is an early 3700x

View attachment 285253


hitting a wall where tiny adjustments result in no POST, this might be her sweet spot

Interestingly this board shows TRFC differently in the bios, manual entries stop post
View attachment 285254

I was wrong. Bios shows 312 - different again
View attachment 285255

final tweaks done on gpu

View attachment 285258

gaming values, 1440 144fps
View attachment 285259View attachment 285260

Try lowering both your SCLs to 4 or even 3.

Also you can try tFAW 28 which is usually calculated but multiplying tRRDS X 4.

Try all changes separately not at once.

 

[AusWolf]

I had a habit of hooking up a spare vandal power button to the clear CMOS header, saved a ton of time.

That's genius! Why have I never thought about that before?

 

[tabascosauz]

That's genius! Why have I never thought about that before?

If you do mem OC for an extended period of time it's inevitable before you will grab a button to hook up to it. Much more convenient and also safe - Aorus AX has the jumper jammed in close to a bunch of other components, I've had a few near misses using a screwdriver before.

Or buy a board with a rear I/O button that way you also don't have to worry about where to route the button, if you like to keep your case closed

 

[Space Lynx]

im excited for 7900x3d reviews. I'm not expecting to be blown away, but I will be surprised if intel beats it in more than 3 or 4 games.

I think Ryzen will truly shine in late 2024 when we get 3nm chips.

 

[Chomiq]

I think Ryzen will truly shine in late 2024 when we get 3nm chips.

Never ending "truly shine" moments of Ryzen since Gen1

 

[AusWolf]

Never ending "truly shine" moments of Ryzen since Gen1

It's always the next generation that "truly shines". The bigger question is, why are we always unhappy with what we have, and chase numbers in our jobs (as in money) as well as in our hobbies until we die?

I say, everything shines that is fit for purpose. I'm just as happy with my 7700X than I am with the passively cooled Core i7 4765T in my bedroom HTPC.

 

[Space Lynx]

Never ending "truly shine" moments of Ryzen since Gen1

I mean I was just making a casual comment. Personally I think the Ryzen 5600 non-x I have now is a shining moment for Ryzen, I mean for $120 you have a chip that runs at 5.6 all core at 1.200 volts, barely uses 70 watts in most games, in demanding games I see it hit like 90 watts sometimes, and at 1440p, its honestly not that far behind the big boys in terms of fps. 10-20 fps in vast majority of cases. and in most games my frame cap of 160 at 1440p is already being hit anyway, so its eh

It's crazy to me you can build such a nice PC these days so cheaply. even if we pretend covid and inflation didn't happen, the fact you can get my system specs for so cheap is insane. its a golden age of PC building if you ignore like 5-7 top selling games that are horribly unoptimized and demanding like Cyberpunk 2077 and don't game at 4k*

 

[trparky]

Just updated to a new BIOS with AMD AGESA 1.0.0.5 C in it. Oh yeah, a lot faster boot times. It spends far less time at that scary black screen where you think it's boot looping or failing POST when EXPO is enabled.

It still won't sleep when EXPO is enabled though. ARG!

 

[P4-630]

Just leave this here

 

[HD64G]

+7% on average

 

[Nordic]

One game I play has a 30% performance increase with the 5800x3d. I am eager to see how these perform.

 

[Space Lynx]

One game I play has a 30% performance increase with the 5800x3d. I am eager to see how these perform.

yeah a lot of games I play FFXIV + Starcraft 2 for example have a big performance jump with the x3d cache.

 

[AusWolf]

Just leave this here

View attachment 285305

I know comparing to the competition is important for marketing, but I'm much more curious to see a comparison to non-3D cache Zen 4 CPUs.

Just updated to a new BIOS with AMD AGESA 1.0.0.5 C in it. Oh yeah, a lot faster boot times. It spends far less time at that scary black screen where you think it's boot looping or failing POST when EXPO is enabled.

It still won't sleep when EXPO is enabled though. ARG!

I'm not sure if updating is worth it in my case.

It was a struggle to get everything right on this version (the SoC voltage is too high when I first enable EXPO, it needs some playing around to come back down). I don't want to go through all that again.

 

[Kodehawa]

Just leave this here

View attachment 285305

I feel like they should use World of Warcraft for this chart. It has like 50% more on the 5800X3D vs 5800X, afaik.

 

[tabascosauz]

One game I play has a 30% performance increase with the 5800x3d. I am eager to see how these perform.

I understand that even if Vcache massively benefits old games, they can't exactly market their new CPUs with old games; however, given the 5800X3D's unchallenged dominance in MSFS2020 last time, I'm surprised it's not even on the slide for the 7950X3D.

Also, really do not care for the 7950X3D, show me how the 7800X3D performs. No scheduler/AGESA excuses for bad performance with that one. The leaked reviewer guide is all over the board and doesn't even have data for half the games vs. 7950X

Isn't the IMC the same on the chiplet-based Zen 2 and 3 CPUs?

So it would be pretty much down to process maturity.

IO die is the same, but the changes do generally net a ~5-10ns reduction in DRAM latency with the same memory kit. The UMC itself is not so responsible for latency penalties on the chiplet design

 

[Count von Schwalbe]

IO die is the same, but the changes do generally net a ~5-10ns reduction in DRAM latency with the same memory kit. The UMC itself is not so responsible for latency penalties on the chiplet design

Is that due to the CCX=CCD design? It just general IF improvements?

 

[tabascosauz]

Is that due to the CCX=CCD design? It just general IF improvements?

There weren't any real IF improvements. It's probably just the 4 to 8 core CCX. You can see a similar ~5ns (?) improvement between Renoir and Cezanne too, though smaller

 

[johnspack]

Anyone know how I could tighten my ram timings down to 15 or even 14? I've got sammy b-die single rank 3200mhz ram at 16 18 18 18.... yes it's old...
Ram is already at 1.35v... do I need to increase soc voltage or anything? Haven't tried yet as this thing doesn't have crash free bios, and
resetting bios is a bit of a pain....

 

[Mussels]

I understand that even if Vcache massively benefits old games, they can't exactly market their new CPUs with old games; however, given the 5800X3D's unchallenged dominance in MSFS2020 last time, I'm surprised it's not even on the slide for the 7950X3D.

Also, really do not care for the 7950X3D, show me how the 7800X3D performs. No scheduler/AGESA excuses for bad performance with that one. The leaked reviewer guide is all over the board and doesn't even have data for half the games vs. 7950X




IO die is the same, but the changes do generally net a ~5-10ns reduction in DRAM latency with the same memory kit. The UMC itself is not so responsible for latency penalties on the chiplet design

The used the same IO die but made changes to the IF

The latency changes tend to be because of more cores per die - having 3-4 cores at low latency vs 6-8 means you're far more likely to have less latency between the cores, and then they have higher clock speeds and more cache to boost it even more

Playing around with my 3700x has reinforced my views that once you max out a single CCX, you get a performance penalty in gaming - its got amazing performance in engines that use <4 primary threads, but once you pass that you either get the inter-ccx latency penalty or the SMT threads and their lower performance

It's still better than a 4 core alone because the OS and background tasks can use those cores and its' better to have lowered performance than 100% CPU usage, but i've definitely learned to focus on the most cores in a single CCX (or P cores on the intel side) for longevity with gaming performance

yeah a lot of games I play FFXIV + Starcraft 2 for example have a big performance jump with the x3d cache.

I've seen that as well, i've gone back and replayed everything again (all the campaigns and the coop missions) and noticed that i'm not seeing *any* late game slowdown, unless the other player lags out
It used to run atrociously with certain effects on screen, and it's just... not.

Try lowering both your SCLs to 4 or even 3.

Also you can try tFAW 28 which is usually calculated but multiplying tRRDS X 4.

Try all changes separately not at once.

Testing them now for you
I'm finding it interesting that settings either work, or no POST - theres very few settings that POST and are not stable (RAM at 3333 or higher boots, but has WHEA errors - so that's just IF/SoC, and a newer CPU would fix that)


SCL's working at 3 perfectly fine
Will try tFAW 28 after R23 finishes a 30 min run

tfaw 28 is fine too

one of these changes makes the BIOS lag for keyboard input, but no issues in windows. Odd.
Disabling CSM causes this. Known bug from multiple gigabyte boards, OH WHAT FUN.

TRFC lets me change it now, i think part of the issue was the keyboard lag screwing up some adjustments i was making
Updated benchies OTW
not much room left to tune

 

[Mussels]

Oh one silly update
I had lingering instability that felt really random, then I realized PBO settings still apply when you use per-CCX overclocking (which i had since it idles better)
PBO disabled had limited the wattages, so lowering the SoC voltage as i was tuning it let the CPU run harder with the same voltage, and had crashies


results are pretty much what is seen above, except the EDC value went from 90 to like 130

 

[trparky]

Changed out my case today, got myself a Lian Li LANCOOL II MESH C. Now that case has some air flow!!!

 

[tpa-pr]

It's always the next generation that "truly shines". The bigger question is, why are we always unhappy with what we have, and chase numbers in our jobs (as in money) as well as in our hobbies until we die?

I say, everything shines that is fit for purpose. I'm just as happy with my 7700X than I am with the passively cooled Core i7 4765T in my bedroom HTPC.

I managed to avoid the "unhappy with what we have" thinking by:
1) Setting a "realistic" ceiling of performance (for me it's 1080p @ 60fps).
2) Way WAY over-provisioning for that need.

And so far, it's worked I don't care about benchmark numbers, I care about whether or not I can get 60fps in my games (with maxed graphics). And I can, so I'm happy with my hardware and don't have any desire to upgrade. It's really neat that my PC can play anything and everything I throw at it. I'm actually giving serious consideration to selling off my consoles because the emulators can give me a far better experience than the original hardware itself!

That's not to say I don't understand the chase for higher numbers or that it's a bad mindset. It's just one I've avoided intentionally.

 

[Nordic]

It's always the next generation that "truly shines". The bigger question is, why are we always unhappy with what we have, and chase numbers in our jobs (as in money) as well as in our hobbies until we die?

I say, everything shines that is fit for purpose. I'm just as happy with my 7700X than I am with the passively cooled Core i7 4765T in my bedroom HTPC.

I'm all for not needing the absolute best hardware. I don't need to keep up with the Jones's. I do have a habit of playing unoptimized games that need maximum single threaded performance to get better than a slide show in fps.

Going from the 5950x to a 5800x3d would give me 30% more fps and I have been strongly considering it. A 7800x3d could provide 40-45% improvement which would finally provide nearly acceptable performance.