News AMD announces Zen 5 Ryzen 9000 processors launches in July — four new Ryzen 9, 7, and 5 processors with a 16% IPC improvement

[Admin]

AMD ripped the covers off four new Zen 5-powered Ryzen 9000 processors in the Ryzen 9, 7 and 5 families today at Computex 2024 in Taipei, Taiwan, touting that the new Zen 5 architecture will deliver a 16% increase in instructions per cycle (IPC) throughput when the chips come to market in July.

AMD announces Zen 5 Ryzen 9000 processors launches in July — four new Ryzen 9, 7, and 5 processors with a 16% IPC improvement : Read more

 

[Deleted member 2731765]

I find it very odd that AMD still hasn’t divulged any pricing info yet !

But anyway, in all likelihood, I expect the Ryzen 9000 CPUs to cost around the same ballpark as the current Ryzen 7000 chips.


Edit:

Btw, thanks for the article. A nice read. Some slight typos need to be be corrected though.

 

[oofdragon]

Woohoo 9800X3D on the way!

 

[FiatDemise]

Seems not much reason to upgrade if you have a Ryzen 9 7900? New 12 core Ryzen 9 9900X has 120W TDP vs 65W.

 

[Deleted member 2731765]

Article:

""AMD says the new AM5 motherboards now support higher EXPO memory overclocks .......... We suspect that there has been at least a marginal increase in DDR5 transfer rates for stock processor settings as well.""

Yes. It appears that DDR5-5600 JEDEC speeds are being natively supported now, and Infinity Fabric clock has been raised to 2400 MHz on Zen 5 (vs 2000 MHz on Zen 4).

This was actually confirmed via some leaked Gigabyte slides before which I shared. Sorry for the blurry pics, but we can see that the Infinity Fabric clock of FCLK is now rated at up to 2400 MHz. Slide also confirms CCD on 4nm.

• Memory Native Speed by default - 5600 MT/s (vs 5200 on Zen 4).
• EXPO Speeds- up to 8000 MT/s+
• FCLK Speeds- 2400 MHz (Fabric)

 

[hotaru251]

Seems not much reason to upgrade if you have a Ryzen 9 7900?

i mean very rarely is it ever worth the cost to just go up 1 generation.


I just hope the USB4 is good & reliable as thats 1 feature I kind of am jealous about intel having :|

 

[saunupe1911]

Can it make a 4090 work? That's all I want to know. My next CPU purchase needs to put a 4090 to work and not come close to bottlenecking it.

 

[TheBeastInside]

AMD also has the Ryzen 7 5800XT with the same six cores and 12 threads

I imagine you mean 8 cores and 16 threads.
I'm not sure what real incentives exist to upgrade these days, at least on the CPU and Mobi side of things.
The value for money on a full upgrade seems low to me..

 

[bit_user]

The drop in base clocks and TDPs (except 9950X) are pretty surprising, as are the drops in TDPs. Not raising the boost clocks is an even bolder move. Hopefully, the CPUs will spend more time running closer to their max boost, or else the only improvement they're getting is from their IPC increase.

 

[TerryLaze]

The drop in base clocks and TDPs (except 9950X) are pretty surprising, as are the drops in TDPs. Not raising the boost clocks is an even bolder move. Hopefully, the CPUs will spend more time running closer to their max boost, or else the only improvement they're getting is from their IPC increase.

The drops in TDP are the most obvious thing since TSMC is stating that the 3nm needs 35% less power for the same performance.
Same for the IPC increase of about 15%.
The boost clocks we couldn't have predicted but also if that's all TSMC is good for what can AMD do about it?!

https://en.wikipedia.org/wiki/3_nm_process

TSMC has stated that its "3 nm" FinFET chips will reduce power consumption by 25–30% at the same speed, increase speed by 10–15% at the same amount of power and increase transistor density by about 33% compared to its previous "5 nm" FinFET chips.

https://pr.tsmc.com/english/news/2986

TSMC’s 3nm process is the most advanced semiconductor technology in both power, performance, and area (PPA) and in transistor technology, and a full-node advance from the 5nm generation. Compared with the 5nm (N5) process, TSMC’s 3nm process offers up to 1.6X logic density gain and 30-35% power reduction at the same speed, and supports the innovative TSMC FINFLEX™ architecture.

 

[thestryker]

I'm mostly curious about the IHS and what performance looks like across locked TDPs with this generation. I'd like to seriously consider going AMD again, but only if the ridiculous IHS situation is resolved as I have no interest in voiding warranty to deal with it.

While I've never cared about socket longevity in the past AMD effectively confirming at least Zen 6 support (2027+) on AM5 it's on my radar.

The drop in base clocks and TDPs (except 9950X) are pretty surprising, as are the drops in TDPs. Not raising the boost clocks is an even bolder move. Hopefully, the CPUs will spend more time running closer to their max boost, or else the only improvement they're getting is from their IPC increase.

I'm pretty sure you're spot on about how much time they spend boosting. Of course they've also mentioned IHS improvement and Zen 4 didn't exactly stick to TDP so there's plenty of room for extra performance.

The drops in TDP are the most obvious thing since TSMC is stating that the 3nm needs 35% less power for the same performance.
Same for the IPC increase of about 15%.
The boost clocks we couldn't have predicted but also if that's all TSMC is good for what can AMD do about it?!

https://en.wikipedia.org/wiki/3_nm_process

https://pr.tsmc.com/english/news/2986

They're not using N3 for Zen 5.

 

[Colif]

i mean very rarely is it ever worth the cost to just go up 1 generation.

i jumped from 3600XT to 5800x3d. that was worth it. But you right, upgrading every generation is a waste of money in most cases... and yet we have people who do around here... their money

Nice to see AMD make the XT mean something on the new ones... mine didn't do much.

 

[derekullo]

Can it make a 4090 work? That's all I want to know. My next CPU purchase needs to put a 4090 to work and not come close to bottlenecking it.

Many things can make a 4090 work.
Depending on what you plan on using the computer for, the cpu may not even matter.
You can never remove all bottlenecks.
A system without bottlenecks would have infinite performance 😛

For gaming, I tend to turn g-sync/v-sync on and as long as my game's fps is running at the max refresh rate of my monitor then any supposed bottlenecks don't matter.

 

[baboma]

As mentioned elsewhere, it makes sense for 9950X to get the full-fat 170W/230W power for benchmarking claims. Unsurprisingly, the few benchmark numbers shown are from 9950X. More surprising (but not really) is that it's pegged against 14900K and not KS.

Speaking for the mainstream, which I'm a part of, I have zero interest in flagship hype, and would be more inclined to see benchmark numbers for 9600X/9700X, both against their 7600X7700X predecessors, matching increased IPC against lowered base clocks--and more pertinently, against 14600K/14700K.

Given RPL & RPR's demonstrated superiority in productivity benchmarks against Zen 4, my expectation is that Zen 5 will achieve rough parity with RPR at the midrange, but with better power efficiency.

Of course, the showdown everyone is waiting for will be against ARL, but with its Q4 release, we'll have to wait some more months for that to happen.

The other aspect I'm a bit surprised about is that desktop parts get no NPU love. With all the noise AMD & Intel are making about AI, and CoPilot+ in particular, I'd expect some consideration for desktop. I'm wondering if that's because XDNA2 (AMD's NPU) is tied to the RDNA3.5 iGPU? And since desktop only gets wimpy iGPU, then XDNA2 is a nonstarter?

In any case, QualComm has made some noises about moving into desktop territory, so I expect desktop CPUs will get the NPU craze later, if not sooner. Perhaps Arrow Lake will provide a clearer outlook for desktop.

 

[Notton]

16% uplift with lower clocks was as expected from the rumors floating around.
There was no way it was going to be 20~30%.
I have no need for it, but 9600X3D sounds interesting.

 

[thestryker]

The other aspect I'm a bit surprised about is that desktop parts get no NPU love. With all the noise AMD & Intel are making about AI, and CoPilot+ in particular, I'd expect some consideration for desktop. I'm wondering if that's because XDNA2 (AMD's NPU) is tied to the RDNA3.5 iGPU? And since desktop only gets wimpy iGPU, then XDNA2 is a nonstarter?

The NPU would have to be located in the IO die and it appears AMD made no fundamental changes to that for desktop parts. I assume that we won't see desktop NPUs until Zen 6 (or a Zen 5 refresh if Intel has added theirs, but I doubt Intel will), or potentially Zen 5 APUs.

Personally speaking I hope they never waste space on it, but if the cost is worth it to them it'll happen.

 

[Amdlova]

The 65w catch My eye... it's the TDP power I currently use.
Well let see how the new chipset behavior.
And what intel will have for us.

 

[Silas Sanchez]

Well this sucks, 16% to me isnt worth the cost, work & potential risk of something going wrong in the update. I can't believe how sensitive my PC is in updating &/or changing stuff, ridiculous really.
Its not the right way to go AMD, get the clock speeds higher, 16%? it doesn't cut it. On the otherhand, air coolers no doubt leave a small amount of performance on the table, so this might be more air cooler friendly. Still 16% isn't enough, and it wont do anything for these crazy fast gpus that are struggling to do 30fps in high resolutions and extreme graphics.
It will be interesting to see how faster frequency plays out with air coolers since the kinematics of the CPU-cooler interface are the limiting factor in heat transfer.

Can it make a 4090 work? That's all I want to know. My next CPU purchase needs to put a 4090 to work and not come close to bottlenecking it.

On the one end of the spectrum, if gaming at 4K full ray tracing then the GPU will be fully loaded and struggling to get acceptable frame rates. In the middle of the spectrum, with 2K run of the mill games and add in DLSS then you want as much CPU power as possible, particularly higher sustained frequency, but better game design and software and architecture can no doubt do what higher frequency does. And at 1080p the CPU will severely limit the 4090 in most cases unless the game has god like path tracing or graphics.

 

[TerryLaze]

i jumped from 3600XT to 5800x3d. that was worth it. But you right, upgrading every generation is a waste of money in most cases... and yet we have people who do around here... their money

Nice to see AMD make the XT mean something on the new ones... mine didn't do much.

That's a generation PLUS the x3d cache and the cache is what really makes the big difference, at least for gaming, in the games that take advantage of it.

 

[Marnad]

Well, I believe I will do an overhaul and upgrade from my current Ryzen 5700X CPU.

I should see a significant performance bump with a new 800 series mobo, 32 GB of DDR5, and a 9700X (?)

 

[A Stoner]

I am running my first AMD processor in decades. The 7950X3D and am very satisfied with it. I wonder if the next generations of X3D will overcome some of the downsides of the current generation. It would be nice if they could either allow overclocking or get the stack on both dies and better yet, do both.

 

[salgado18]

It's curious how the TDP dropped so much since Zen 4, except for the 16-core CPU. Does it mean the current chip is actually constrained by its power limit, and the new model can work faster also because it is more efficient? I'm curious now.

Also, the 5900XT should be named 5950, without the X, to mean it is a cheaper 16 core and keep numbering conventions, but I lost all hope in that front by now.

 

[jeremyj_83]

The drops in TDP are the most obvious thing since TSMC is stating that the 3nm needs 35% less power for the same performance.
Same for the IPC increase of about 15%.
The boost clocks we couldn't have predicted but also if that's all TSMC is good for what can AMD do about it?!

The article says they expect AMD to be using N4P not N3.

Process node improvements	Performance	Power Efficiency	Density
TSMC N5 (Zen 4) vs N4P (Zen 5)	+11%	+22	+6%

 

[jeremyj_83]

It's curious how the TDP dropped so much since Zen 4, except for the 16-core CPU. Does it mean the current chip is actually constrained by its power limit, and the new model can work faster also because it is more efficient? I'm curious now.

Also, the 5900XT should be named 5950, without the X, to mean it is a cheaper 16 core and keep numbering conventions, but I lost all hope in that front by now.

Compared to the same chips from Zen 4 we see lower base clocks on Zen 5. That can make a difference in the TDP. We also saw that for Zen 4 there wasn't much of a benefit to have TDPs higher than 105W as the performance increase going from 105 > 125 > 170W was minimal on the 7950X. In Cinebench it was a 3% and 7% uplift vs 105W respectively for the higher TDPs. That was around best case increase compared for adding power.

 

[philged]

Couple of typos in the "AMD Zen 5 Ryzen 9000 'Granite Ridge' Gaming and Productivity Benchmarks" section.

We begin to see differences in the lesser Ryzen 9, 7, and 5 families. Aside from the massive 50W TDP decrease to 120W, most of the Ryzen 9 9900X’s specs match the previous-gen model. The 30% TDP reduction represents a tremendous improvement in power efficiency, especially given that AMD also claims substantial gen-on-gen performance improvements. However, AMD lowered the base clocks by 300 MHz. AMD says users can continue using the auto-overclocking Precision Boost Overdrive (PBO) feature to unlock higher performance by uncorking the power limits. For now, this chip lines up against Intel’s $389 Core i7-14700K.

The Zen 5 Ryzen 9 9700X gets a 100 MHz bump to a 5.5 GHz boost while most other specs remain the same. However, the lower TDP rating of 65W is a 40W reduction compared to the prior-gen Ryzen 9 7700X despite the similar core counts. Base clocks have been reduced by 700 MHz, which obviously helps to rein in the TDP. This chip will face off with Intel’s $380 Core i7-14700K.

The i7-9700K should probably be the same price. In the 2nd paragraph it should be R7 9700X instead of R9.