AMD's Robert Hallock Confirms Lack of Manual CPU Overclocking for Ryzen 7 5800X3D

[tabascosauz]

It clearly is - it's the first product to hit the market with a brand-new technology; it's relatively limited in scope (one SKU, the last product for a five-year-old platform, etc.) and they have announced no plans for further models for this platform with the feature. Definitely a test drive.

Almost definitely. It doesn't make sense on all SKUs, but a wider roll-out on a chip more thoroughly adapted to this (with a separate cache voltage rail, for example) would make a lot of sense. Something like every tier from x6xx or x8xx upwards having a 3D cache-enabled top-end SKU would make sense (i.e. 7600 65W, 7600X3D 105W, 7800 65W, 7800X3D 105W, etc.). This would make a lot of sense if they can make the cache die on 7nm even when the CCDs move to 5nm, as that would free up capacity to churn out more cache dice.

Just to be clear: no multiplier-based, fixed frequency and voltage OC is not "no OC". PBO and Curve Optimizer are still ways of overclocking, and they seem to be supported here. They also deliver better results in general on Zen2 and Zen3 than old-school OC techniques.

Haven't finished watching the video yet, but did he confirm PBO is actually around? Generally even Auto PBO quickly enables higher Vcore peaks (1.5V+) not possible at stock. Yes CO is fine, but seeing as the current AGESA completely removed Boost Override (only to be later reintroduced), sounds like they were prepping their firmware for the 5800X3D by removing the whole shebang. Without Override there's not much point to CO since all Vermeer chips hit their stock global limits easily

But knowing AGESA they have an abysmal record of keeping features consistent (especially vendors like GB that release and pull a billion beta BIOSes weekly). So if the V-cache really is that sensitive to voltage, then that sounds like a recipe for disaster..........but I wonder if it's that the cache really can't handle the voltage, or that heat density is too much under certain heavy loads?

Think Intel have had separate cache clock and voltage for like a decade now - obviously a bit more involved with V-cache but I'd fully expect this whole deal to be resolved on AM5. 5800X3D is just an experiment searching for guinea pigs

Currently on normal L3, cache freq is very closely correlated to core effective clock at all times, so curious to see if they have or will eventually decouple cache clock

 

[Taraquin]

I wish it could use negative CO atleast since it requires no more voltage, that could boost perf by 5%+

 

[freeagent]

That seems fair enough I suppose. I did not make the connection between cache and voltage. But even still.. this is a modified 5800X, that has been gimped so it doesn’t kill itself.. they shouldn’t be charging 5900X money for it.

 

[stimpy88]

I think the evidence shows that AMD have just found something out about this CPU. I think they are seeing them fail in a short amount of time. For proof, I say that a few days ago, AMD demanded that all the MB makers release a new emergency BIOS with OC support disabled. Meaning that it was previously enabled for the last couple of months, while they have been sampling and testing.

We have also been told that there were thermal issues earlier on, and that they had to downclock it to make it run at a more suitable temperature. AMD maybe got a little bit carried away by the new packaging tech, and thought they could glue on some cache, and all would be well, and it would be a cheap solution.

The benchmarks will be interesting. But I hope that cannibalizing the sales of the already existing 5800x, as well as the 5900x and possibly 5950x were worth AMD's experiment.

 

[SL2]

Yeah we'll be missing out on that sweet, juicy GHz OC everyone is talking about..

It's the age old I-want-something-for-nothing argument, just because you're only willing to pay extra for more cores and nothing else.

If it beats the 5900X in gaming, and I'm not saying that it does (I know the demo used 12 core models), then why not charge a premium for it? Hint: You're not paying for more cores, you pay for extra cache.
It's for gaming, you can't overclock manually.. if you don't like it then don't buy it.

I think they are seeing them fail in a short amount of time. For proof, I say that a few days ago, AMD demanded that all the MB makers release a new emergency BIOS with OC support disabled.

Just no. That's not a proof, that's a guess at best.
If the 3D cache was anywhere close to unreliable like you suggest then AMD wouldn't launch Milan-X this month.

This CPU won't cannibalize anything. Why? Well read this thread and you'll get a few hints. Nobody thinks this is a bargain so far at $450.
Besides, this CPU is supposed to have limited availability, although we'll see about that..

 

[Valantar]

I think the evidence shows that AMD have just found something out about this CPU. I think they are seeing them fail in a short amount of time. For proof, I say that a few days ago, AMD demanded that all the MB makers release a new emergency BIOS with OC support disabled. Meaning that it was previously enabled for the last couple of months, while they have been sampling and testing.

Or they could just be bad at consistently pushing BIOS updates as needed - for which there is plenty of evidence.

We have also been told that there were thermal issues earlier on, and that they had to downclock it to make it run at a more suitable temperature.

There's been speculation as to that, but I can't recall that "we have been told" that. The reductions in base and boost clock are quite easily explained by the power limits being the same (105W/144W) but there being an additional die with 64MB of cache on it, requiring some power that the cores would otherwise have had access to.

AMD maybe got a little bit carried away by the new packaging tech, and thought they could glue on some cache, and all would be well, and it would be a cheap solution.

Cheap? This is clearly a lot more expensive than just selling a 5800X, and they'll be launching Zen4 for AM5 whether or not it includes 3D cache, so arguing that they did this because it somehow adds a cost benefit does not compute.

The benchmarks will be interesting. But I hope that cannibalizing the sales of the already existing 5800x, as well as the 5900x and possibly 5950x were worth AMD's experiment.

How does it cannibalize those sales? And if it does, and the buyers are happy with their chips, does it matter?

 

[TheLostSwede]

I think the evidence shows that AMD have just found something out about this CPU. I think they are seeing them fail in a short amount of time. For proof, I say that a few days ago, AMD demanded that all the MB makers release a new emergency BIOS with OC support disabled. Meaning that it was previously enabled for the last couple of months, while they have been sampling and testing.

We have also been told that there were thermal issues earlier on, and that they had to downclock it to make it run at a more suitable temperature. AMD maybe got a little bit carried away by the new packaging tech, and thought they could glue on some cache, and all would be well, and it would be a cheap solution.

The benchmarks will be interesting. But I hope that cannibalizing the sales of the already existing 5800x, as well as the 5900x and possibly 5950x were worth AMD's experiment.

That was around a months and a half ago. It took about a month for that information to leak.
None of it was an emergency, it was just another AGESA update from AMD and a beta version at that.

 

[chrcoluk]

OC is yesteryear. If you tweak, you undervolt, or just do custom p-states.

This seems quite innovative and will be cool to see how it plays out as consumers get hold of the chips.

We are actually lucky the cpu companies are open about their products like this, they could just say "15 % faster but its our secret sauce", as now of course Intel will probably make their own stacked cache on a future gen.

 

[TheLostSwede]

Or they could just be bad at consistently pushing BIOS updates as needed - for which there is plenty of evidence.

Actually, AMD is pushing out a lot of test/beta builds to the motherboard makers that we as users never get a whiff of. This was most likely supposed to be one of those, but as it goes, it couldn't be kept a secret for long.

We are actually lucky the cpu companies are open about their products like this, they could just say "15 % faster but its our secret sauce", as now of course Intel will probably make their own stacked cache on a future gen.

Just look at how it is in the Arm world. We just have to trust the SoC makers there, as most of them are more than unwilling to share details that gives any kind of insight into what their own secret sauce is that they've added to make a chip. AMD and Intel are very open with how their products work in comparison.

 

[Deleted member 24505]

So 5800x3D users will be beta testing the 3d cache for AMD in effect, how the tables turn ie ADL

 

[dgianstefani]

RAM/IF OC is still available and that's where all the performance is anyway so what's the problem?

 

[SL2]

So 5800x3D users will be beta testing the 3d cache for AMD in effect, how the tables turn ie ADL

No. Milan-X is launched before 5800X3D.

 

[QuietBob]

Bummer, why would you release an enthusiast SKU and lock out enthusiast features? While I perfectly understand the reasons behind AMD's decision, I can't help feeling a little disappointed. I guess limiting the voltage is essentially a safeguard. Just think of these ill-advised "experts" who would happily apply 1.5+ v to static overclocks, only because they saw some random dude doing it.

On the flip side, lower maximum Vcore and boost clocks should mean lower power consumption and temperature. And perhaps we'll see stable 2000+ MHz on the IF as well.

I'll probably still end up getting one. My current workloads would see some improvement from Zen 3 IPC lift, as well as having access to 8 physical cores.

 

[GerKNG]

and what has that anything to do with an unlocked multiplier?

 

[SL2]

RAM/IF OC is still available and that's where all the performance is anyway so what's the problem?

I dunno.

The amount of FUD in this thread is astonishing.

"Why should I pay for higher performance when it runs at lower clocks? Them clocks = my epeen. Lower clocks should mean lower price. I measure performance in Hz."

"This thing will break, and that's why AMD have decided to also destroy their entire server market this month."

"The extra cache is in fact a couple of 5G chips. I said it, which means it is the proof of what I said."

 

[tabascosauz]

Just no. That's not a proof, that's a guess at best.
If the 3D cache was anywhere close to unreliable like you suggest then AMD wouldn't launch Milan-X this month.

EPYC is clocked way lower, if V-cache even remotely has a problem with longevity/power/heat in that envelope then we'd have a massive problem on our hands. Zen has never had a problem in that range. But for the past 4 generations AMD has been pushing Ryzen to the very limit of what its respective process can sustain and often struggling to reconcile the clocks and V-F it wants with the hardware it has, that again looks like the deal with the 5800X3D.

I guess we'll see. I certainly hope it's just a matter of thermals and AMD playing it safe for tbe first time in forever. Every generation we get increasing granularity and precision in terms of voltage domains and clock dividers - cache might be the next one.

 

[TheinsanegamerN]

They could have just locked down the voltage if that was the case, and not all OCing.

Nvidia locked down mobile GPU OCing over voltage concerns too, and the community ripped them a new one.

But when AMD does it its OK.

 

[SL2]

EPYC is clocked way lower, if V-cache even remotely has a problem with longevity/power/heat in that envelope then we'd have a massive problem on our hands. Zen has never had a problem in that range. But for the past 4 generations AMD has been pushing Ryzen to the very limit of what its respective process can sustain and often struggling to reconcile the clocks and V-F it wants with the hardware it has, that again looks like the deal with the 5800X3D.

I guess we'll see. I certainly hope it's just a matter of thermals and AMD playing it safe for tbe first time in forever. Every generation we get increasing granularity and precision in terms of voltage domains and clock dividers - cache might be the next one.

My point was that there's an idea that will break even when AMD tries to play safe, AKA more FUD.

 

[dont whant to set it"']

1.35V should be more than sufficient for an allcore clock of 4.6GHz ,thermals be kept in check that is.

 

[SL2]

They could have just locked down the voltage if that was the case, and not all OCing.

Nvidia locked down mobile GPU OCing over voltage concerns too, and the community ripped them a new one.

But when AMD does it its OK.

You're comparing a whole lineup to one single half-generation old, possibly limited in availability, almost niche, overprized SKU.

It's not like you don't have options. Drop the bitterness.

 

[Taraquin]

Yeah we'll be missing out on that sweet, juicy GHz OC everyone is talking about..

View attachment 240110

It's the age old I-want-something-for-nothing argument, just because you're only willing to pay extra for more cores and nothing else.

If it beats the 5900X in gaming, and I'm not saying that it does (I know the demo used 12 core models), then why not charge a premium for it? Hint: You're not paying for more cores, you pay for extra cache.
It's for gaming, you can't overclock manually.. if you don't like it then don't buy it.


Just no. That's not a proof, that's a guess at best.
If the 3D cache was anywhere close to unreliable like you suggest then AMD wouldn't launch Milan-X this month.

This CPU won't cannibalize anything. Why? Well read this thread and you'll get a few hints. Nobody thinks this is a bargain so far at $450.
Besides, this CPU is supposed to have limited availability, although we'll see about that..

5800X has lower latency than 5900X since it only has one ccd. Single core can reach 4.85GHz stock and 5.05GHz with pbo, 5900X only goes 50MHz highet. 5900X has about same powerbudget so allcore speed is generally lower than 5800X, combine thst with a bit worse latency and games run faster on 5800X, but games that utilize lots of cores prefer 5900X.

 

[Dr. Dro]

Non-issue, imo. Memory and Infinity Fabric settings remain unlocked, and if curve optimizer also works, this should be as versatile as the other Ryzen processors, IMO. The automatic tweaking algorithm on these processors is optimized to near perfection, and admittedly, above my own manual overclocking skills.

They could have just locked down the voltage if that was the case, and not all OCing.

Nvidia locked down mobile GPU OCing over voltage concerns too, and the community ripped them a new one.

But when AMD does it its OK.

When did that happen? Not to brag but... my mobile RTX 3050 can OC and it OCs hard... it will run >2.1 GHz... insane little chip

 

[DeathtoGnomes]

It looks like a lot of responses clearly show those who havent watched the video. So I partially agree with @Mats on this thread having a lot of FUD (and some whiney comments).

 

[dgianstefani]

Literally who OC's Ryzen clock, tune the memory and IF and that's it.

 

[ThrashZone]

It looks like a lot of responses clearly show those who havent watched the video. So I partially agree with @Mats on this thread having a lot of FUD (and some whiney comments).

Hi,
Did the video contain any test against 12900k ?
If not who cares