Ryzen 9000 officially announced for July 31st Release

[TonyCarter]

Oh dear...more 'meh' than the 9600/9700X...but this time there's no benefit to tweaking PBO/power...and they use the same core parking process as the 7900/7950X3D CPUs (which require a Windows reinstall if you change your mind). <5% performance boost, for 15% more cost!

So, I guess the recommendation will remain as 7800X3D for gaming and 7950X for multi-core productivity!

 

[SpyderTracks]

Oh dear...more 'meh' than the 9600/9700X...but this time there's no benefit to tweaking PBO/power...and they use the same core parking process as the 7900/7950X3D CPUs (which require a Windows reinstall if you change your mind). <5% performance boost, for 15% more cost!

So, I guess the recommendation will remain as 7800X3D for gaming and 7950X for multi-core productivity!

I still think this is all part of them rushing the release before things were ready.

I think there’s more work to be done on windows scheduler and possibly chipset drivers that’s going to improve performance.

I still wouldn’t ever buy based on any expectations or promises, always go by proper third party reviews.

I just don’t believe this is all these chips have got.

 

[polycrac]

I'm still on a 2700X, was waiting to upgrade to this range. I guess I can hang on and see what the 9800X3D is like.

 

[SpyderTracks]

I'm still on a 2700X, was waiting to upgrade to this range. I guess I can hang on and see what the 9800X3D is like.

I would, there are benefits to the 9000X3D cpus outside of simply new architecture in that apparently they’re not restricted in frequency like previously. Obviously all this is conjecture at this point but I would hold out if you can.

 

[Scott]

It's all a bit of a damp squib in honesty.

Noone on Zen4 should be upgrading at all. With Zen3, there's definitely an argument that it's worth upgrading..... but it's an expensive upgrade. Would you recommend Zen5 over Zen3 for an upgrade though? It's time for people to grab a bit of a bargain and jump to Zen4 for productivity, Zen5 doesn't have the value yet.

The 9800X3D needs to be the saving grace. The gaming market is what makes all the noise. So far this release hasn't really dented the 5800X3D, never mind the 7800X3D.

They've not done anything wrong, they've pulled an Intel tbh.... but it's the state Intel is in that allowed this to happen. I hoped for more from AMD though. The lies spoken about this release from them are just as shady and mis-representative as Intel are (hardware issues aside).

 

[TonyCarter]

I don't know why they bother with misleading claims, as they must know they'll be rumbled as soon as the review samples go out.

Even if it was just to increase their market capitalisation based on expected growth figures, investors will be annoyed that the growth failed to materialise anyway.

It's almost the AAA game developer model...promise the world, under-deliver, and promise to fix it with patches over the next 3 years!

I'll stick to Apple...you know you're getting 'bent over' on the price, but at least you go into the deal with eyes wide open

 

[Ekans2011]

Another Ryzen's 9000 interesting analysis by Moore's Law.

 

[TonyCarter]

Seems there's a 'elevated admin' account that releases more performance - which you really don't want to be using that as your daily driver - but the same uplift seems to occur on Zen4 as well as Zen5, so the generational uplift stays about the same.

 

[SpyderTracks]

There’s definitely more to come from this ASSUMING Microsoft want to work with AMD like they did on the Ryzen 3000 series optimisation.

Look at Wendell’s findings on Linux, these are very real repeatable results.

edit: sorry, see this was already posted

The snarky side of me says that Intel paid Microsoft a wedge to make the scheduler incompatible but I realise that may be unfounded outside of the matrix

 

[TonyCarter]

 

[Ekans2011]

 

[TonyCarter]

What AMD don't seem to understand is that by telling everyone they're doing their testing wrong, and then following that new test process, the testers are still no seeing the generational uplift that AMD is claiming.

The only way the testers can get the same uplift that AMD is claiming is if they compare the previous generation CPU using their own test method, to the current generation CPU using AMD's convoluted test method...as if they test them both using either of the same test methods they end up with the same single-digit uplift.

Even Level1 Techs said not to compare Linux FPS to Windows FPS, because the game will not be delivering exactly the same graphic quality on both systems due to underlying graphics subsystems - so the 300FPS gained at Ultra on Linux is not getting the same graphical fidelity as 250FPS gained on Ultra on W11. All you can do is see the % uplift of the 7700X vs 9700X on Linux, then compare that to the % uplift of the same CPUs on W11 using the same elevated admin, PBO, memory tuning, standing on one leg, speaking tongues technique you have to use for Zen 5. In L1Techs video he says there's only about a 6% difference like-for-like between W11 and Linux when graphics fidelity is at the same level on both systems...and Linux just has a slightly smaller system overhead.