Monday, August 29th 2022

AMD Announces Ryzen 7000 Series "Zen 4" Desktop Processors

AMD today announced the Ryzen 7000 series "Zen 4" desktop processors. These debut the company's new "Zen 4" architecture to the market, increasing IPC, performance, with new-generation I/O such as DDR5 and PCI-Express Gen 5. AMD hasn't increased core-counts over the previous-generation, the Ryzen 5 series is still 6-core/12-thread, the Ryzen 7 8-core/16-thread, and Ryzen 9 either 12-core/24-thread, or 16-core/32-thread; but these are all P-cores. AMD is claiming a 13% IPC uplift generation over generation, which coupled with faster DDR5 memory, and CPU clock speeds of up to 5.70 GHz, give the Ryzen 7000-series processor an up to 29% single-core performance gain over the Ryzen 5000 "Zen 3."

At their press event, AMD showed us an up to 35% increase in gaming performance over the previous-generation, and an up to 45% increase in creator performance (which is where it gets the confidence to stick to its core-counts from). The "Zen 4" CPU core dies (CCDs) are built on the TSMC 5 nm EUV (N5) node. Even the I/O die sees a transition to 6 nm (N6), from 12 nm. The switch to 5 nm gives "Zen 4" 62 percent lower power for the same performance, or 49% more performance for the same power. versus the Ryzen 5000 series on 7 nm. The "Zen 4" core along with its dedicated L2 cache is 50% smaller, and 47% more energy efficient than the "Golden Cove" P-core of "Alder Lake."
The "Zen 4" CPU core gets a bulk of its 13% IPC gain from the core's front-end, followed by load-store, branch-prediction, and execution engine. The company also doubled the size of the per-core L2 cache to 1 MB. The core introduces support for AVX-512 instruction set. Eight cores share a 32 MB L3 cache on a CCD. The 6-core and 8-core SKUs in the Ryzen 5 and Ryzen 7 series, come with a single CCD, whereas the 12-core and 16-core Ryzen 9 parts come with two.
AMD introduces a brand new socket with Ryzen 7000, Socket AM5. This is a resilient 1718-pin LGA, with the ability to delivery up to 230 W of power, and comes with next-generation I/O that includes DDR5 and PCIe Gen 5. Physically, the coolers are compatible with Socket AM4 thermal solutions, so you can carry over your old coolers. AMD is promising to launch future generations of Ryzen processors that are AM5-compatible going up to 2025 at least.
There will be four chipset choices with Ryzen 7000, these include the X670E and X670 in the high-end; and the B650 and B650E in the mid-range. Motherboards with X670/E debut in September, and the B650/E in October. AM5 is the first platform with CPU-attached NVMe Gen 5, and the company predicts the first Gen 5 SSDs should arrive by November. We confirmed with AMD that they are not artificially limiting the performance of processors running on the B-Series chipsets vs the X-Series chipsets. The difference between B650 and B650E is that B650E offers support for PCIe Gen 5 for graphics cards and SSDs, while B650 non-E supports PCIe 5.0 SSDs, and PCIe 4 GPUs. AMD is introducing a new memory profile technology called EXPO that eases memory overclocking. It is a royalty-free technology, and includes memory settings specific to the AMD architecture. You are of course able to use Intel XMP-compatible DDR5 memory modules, these might just not have the most perfect settings out of the box. As many as 15 memory kits are being launched at speeds of up to DDR5-6400, from various manufacturers.
The AMD Ryzen 5 7600X is a 6-core/12-thread processor with 4.70 GHz nominal clocks. up to 5.30 GHz boost, 105 W TDP, and is being launched at $299. The Ryzen 7 7700X is 8-core/16-thread, clocked at 4.50 GHz, with up to 5.40 GHz boost, 105 W TDP, and is being launched at $399. The Ryzen 9 7900X is 12-core/24-thread, clocked at 4.70 GHz, with up to 5.60 GHz, 170 W TDP, and is being launched at $549. The top 7950X is 16-core/32-thread, clocked at 4.50 GHz, with up to 5.70 GHz boost, 170 W TDP, launching at $699. All SKUs available to purchase on September 27, 2022. This is an on-shelf date, not a preorder date (we have that confirmed personally).

The complete slide-deck follows.
Add your own comment

195 Comments on AMD Announces Ryzen 7000 Series "Zen 4" Desktop Processors

#151
ratirt
fevgatosSure, let's do that.

Can you explain this?

To be fair, this one does not make a lot of sense without proper mapkey. 1st of all the bottom axis is what? There is literally no explanation of what it is. You can assume it is watts per core? The graph says watts(Less is better) so you can take from it that the Ryzen CPUs and 11th gen are better than 12th gen CPUs since these are less efficient? 12th gen CPUs have the highest score and in that chart you should go for, less is better.
Posted on Reply
#152
JustBenching
ratirtTo be fair, this one does not make a lot of sense without proper mapkey. 1st of all the bottom axis is what? There is literally no explanation of what it is. You can assume it is watts per core? The graph says watts(Less is better) so you can take from it that the Ryzen CPUs and 11th gen are better than 12th gen CPUs since these are less efficient? 12th gen CPUs have the highest score and in that chart you should go for, less is better.
Τhe translation is wrong cause its' a german site using google translate. The graph is score per watt. For example in cinebench you score 10.000 points at 100 watts, you divine 10000 / 100 and you get the score / watt. There are many workloads that in order for the 5950x to compete with a stock 12900k, it needs to be oced to hell and consume equally as much. Lightroom / photoshop / h264 etc.

Actually, phoronix run a test suite of 143 benchmarks. Not only was the 12900k the fastest CPU on the planet, it actually consumed the least amount of power in total to finish the whole test. So yeah...
Posted on Reply
#153
AM4isGOD
With new CPUs coming soon, it is pointless to still go on about how in efficient AL is now as you can change the record to how in efficient RL is and the stupid comments will start all over again. I just can't wait for Meteor Lake when Intel will shut the lot of you up.
Posted on Reply
#154
ratirt
fevgatosΤhe translation is wrong cause its' a german site using google translate. The graph is score per watt. For example in cinebench you score 10.000 points at 100 watts, you divine 10000 / 100 and you get the score / watt. There are many workloads that in order for the 5950x to compete with a stock 12900k, it needs to be oced to hell and consume equally as much. Lightroom / photoshop / h264 etc.

Actually, phoronix run a test suite of 143 benchmarks. Not only was the 12900k the fastest CPU on the planet, it actually consumed the least amount of power in total to finish the whole test. So yeah...
This contradicts with HWUB review and GN review. They have concluded that the 12900k in general is the fastest but in order to be the fastest you need to use much more power. Stock vs stock it does not look good. Tweaking helps but that's another story.




I don't really know where those contradictions are coming from with the efficiency you are talking about. Both reviews state that Intel 12900k is not an efficient architecture especially for MT workloads. So saying 12900k is efficient, kinda puts to question HWUB and GN reviews and conclusions which I understand are wrong in your eyes?
I get the 12900k (and others) can be tweaked but any given CPU can be tweaked so yeah...
Posted on Reply
#155
JustBenching
ratirtThis contradicts with HWUB review and GN review. They have concluded that the 12900k in general is the fastest but in order to be the fastest you need to use much more power. Stock vs stock it does not look good. Tweaking helps but that's another story.

I don't really know where those contradictions are coming from with the efficiency you are talking about. Both reviews state that Intel 12900k is not an efficient architecture especially for MT workloads. So saying 12900k is efficient, kinda puts to question HWUB and GN reviews and conclusions which I understand are wrong in your eyes?
I get the 12900k (and others) can be tweaked but any given CPU can be tweaked so yeah...
What do you mean they contradict? I was showing you autocad numbers and you are showing me blender. There is no contradiction, just different workloads.

Of course the 12900k out of the box is inefficient at heavy MT workloads. Actually, EVERY cpu will be inefficient when it's running heavy MT workloads at 5 ghz.. So the question is, if you want to run heavy MT workloads and you care about efficiency, why are you running them at 5ghz and 240w?

Also you keep mixing architecture and product. How many times do I have to repeat myself, in order to compare archs you need to do it at same wattage. It blows my mind how you can't comprehend this. Your method of testing archs is shown to lead to contradictions. For example, Raptor lake will have a model named 13900t, it will be power limited at 35w at stock. That will without a doubt make it the most efficient CPU in existence at stock. According to your logic then, raptorlake will be the most efficient architecture, right?
Posted on Reply
#156
AM4isGOD
This supposed AMD thread is a car crash. Needs closing or a good clean with a flame thrower.
Posted on Reply
#157
ratirt
fevgatosAlso you keep mixing architecture and product. How many times do I have to repeat myself, in order to compare archs you need to do it at same wattage. It blows my mind how you can't comprehend this. Your method of testing archs is shown to lead to contradictions. For example, Raptor lake will have a model named 13900t, it will be power limited at 35w at stock. That will without a doubt make it the most efficient CPU in existence at stock. According to your logic then, raptorlake will be the most efficient architecture, right?
I keep mixing architectures and you compare different nodes and arch on a wattage basis and that is ok? You don't compare arch you compare nodes in my opinion. Arch comparison is only valid when you have 2 CPUs made on a same node. For example 10th and 11th gen Intel. Compare Zen 2 vs Zen 3 since both are made on a 7nm FF. AL will be good to compare with RL since these are done on a same 7nm node. Otherwise you compare one product to the other.
Anyway. I think that's enough. It is still AMD thread so maybe lets just stop here with your Intel efficiency problems.
Posted on Reply
#158
JustBenching
ratirtI keep mixing architectures and you compare different nodes and arch on a wattage basis and that is ok? You don't compare arch you compare nodes in my opinion. Arch comparison is only valid when you have 2 CPUs made on a same node. For example 10th and 11th gen Intel. Compare Zen 2 vs Zen 3 since both are made on a 7nm FF. AL will be good to compare with RL since these are done on a same 7nm node. Otherwise you compare one product to the other.
Anyway. I think that's enough. It is still AMD thread so maybe lets just stop here with your Intel efficiency problems.
The node is a fundamental aspect of the cpu's architecture that CANNOT be changed. TDP can be changed. Amd or Intel might as well release a 7940x at 50w or a 13900t at 35w (which they will btw). They can't change the node on a whim. So I asked you but you didn't answer, Intel will actually release a T version of the 13900 at 35w stock. Using your logic, doesn't that mean that raptor lake will be the most efficient architecture by far?
Posted on Reply
#159
Dyatlov A
AMD did forgot to compare their CPU with overclocked Intel 12400. I would be curious how is standing against, 150$ Intel 12400 @5.2GHz vs. $300 AMD 7600X.
Posted on Reply
#160
Valantar
Dyatlov AAMD did forgot to compare their CPU with overclocked Intel 12400. I would be curious how is standing against, 150$ Intel 12400 @5.2GHz vs. $300 AMD 7600X.
... why would a manufacturer ever compare their product to a non-stock competing product? That makes absolutely zero sense.
Posted on Reply
#161
Dyatlov A
Valantar... why would a manufacturer ever compare their product to a non-stock competing product? That makes absolutely zero sense.
Becasue from Intel can customers get extra performance, AMD is factory fully overclcocked.
Posted on Reply
#162
Valantar
Dyatlov ABecasue from Intel can customers get extra performance, AMD is factory fully overclcocked.
That is not only not true - there are tangible performance benefits to be had with PPT/EDC/TDC adjustments, PBO and Curve Optimizer - but you're actually asking why a manufacturer isn't comparing their product - at stock - against a competing product at non-stock settings. You see how utterly ludicrous that is, right? Reviewers might do OC showdowns, but manufacturers talk about the stock configurations of their products - otherwise they'd just be shitting on the settings they've themselves decided on.
Posted on Reply
#163
Dyatlov A
ValantarThat is not only not true - there are tangible performance benefits to be had with PPT/EDC/TDC adjustments, PBO and Curve Optimizer - but you're actually asking why a manufacturer isn't comparing their product - at stock - against a competing product at non-stock settings. You see how utterly ludicrous that is, right? Reviewers might do OC showdowns, but manufacturers talk about the stock configurations of their products - otherwise they'd just be shitting on the settings they've themselves decided on.
PBO, Curve optimizer brings 5% extra performance maybe? What i meant is i find ridiculous AMD compared their 7600X with Intel’s highest end in gaming and it is nice the $300 AMD 7600X can do the similar gaming performance as the super expensive Intel. However all these gimmick nice until people not knows the $150 Intel 12400 can gaming also as well.
Posted on Reply
#164
Valantar
Dyatlov APBO, Curve optimizer brings 5% extra performance maybe? What i meant is i find ridiculous AMD compared their 7600X with Intel’s highest end in gaming and it is nice the $300 AMD 7600X can do the similar gaming performance as the super expensive Intel. However all these gimmick nice until people not knows the $150 Intel 12400 can gaming also as well.
It absolutely can, but at stock, it's slower. So, if the 7600X matches or beats the fastest, then it's also faster than anything slower. That's what they're saying, neither more nor less. Overclocking is irrelevant outside of enthusiast circles anyhow - which is a small niche among pc gamers overall - so comparing against a cheaper overclocked SKU makes no sense either way - outside of enthusiast press and the like. Which isn't the target audience for a product release - the release targets the whole market, not just enthusiasts. I mean, do you complain when Intel presents a product that they don't show OC performance in their own graphs too? That would make slightly more sense than this demand at least. Overclocking potential is never sure, never a given, and this never used in marketing.
Posted on Reply
#165
RandallFlagg
Dyatlov APBO, Curve optimizer brings 5% extra performance maybe? What i meant is i find ridiculous AMD compared their 7600X with Intel’s highest end in gaming and it is nice the $300 AMD 7600X can do the similar gaming performance as the super expensive Intel. However all these gimmick nice until people not knows the $150 Intel 12400 can gaming also as well.
You really have to throw up a graphic for these people to understand, not that it will change their opinion.

There is a near 100% success rate in overclocking non K chips to 5.1Ghz all core, a 15%+ OC on the 12400. All of HW Unboxed non K chips did this with no fuss.

The caveat here is that for the most part, you'll need a $250+ motherboard to do this in the US else you have to order from overseas where you can get models that will allow this for under $150 USD. Outside the US, this isn't an issue as far as I can tell.

HW Unboxed was able to get all 3 of their Non-K chips to do a stable 5.1Ghz OC. I've seen these running over 5.4Ghz. From what I can tell, a 12400 at 5.4Ghz is in the same performance ballpark as a 7600X.



Der8auer used DDR5-5200 C38 - cheap memory - on the 12400 @ 5.24Ghz and and in some cases it beat a 12900K equipped with DDR4-3600 C14 :



In this one averages for 5.24Ghz 12400 beat the 12600K and lost to 12700K/12900K, but the 1% lows were the highest and frame time consistency were the best on the 5.24Ghz 12400 :

Posted on Reply
#166
Dyatlov A
Thanks for understanding my point and i just need to add to your explanation, that i believe Asus Rog B660-f and g are available anywhere in the world and they cost about 220EUR. If we consider the AMD motherboards will not be cheap either, than the 12400 is a big win against 7600X.
Posted on Reply
#167
RandallFlagg
Dyatlov AThanks for understanding my point and i just need to add to your explanation, that i believe Asus Rog B660-f and g are available anywhere in the world and they cost about 220EUR. If we consider the AMD motherboards will not be cheap either, than the 12400 is a big win against 7600X.
I think as soon as this became known back in January, they began to sell out fast in the US, or maybe Intel told them to stop selling them here who knows.

B&H Photo and Newegg don't have them in stock, can't find one on Amazon either. Ditto for Microcenter. E-Bay has them for like $260-$300 stateside. Beyond that it's international, and if doing a budget build there's the ASRock PG Riptide at ~$150.
Posted on Reply
#168
Valantar
RandallFlaggYou really have to throw up a graphic for these people to understand, not that it will change their opinion.

There is a near 100% success rate in overclocking non K chips to 5.1Ghz all core, a 15%+ OC on the 12400. All of HW Unboxed non K chips did this with no fuss.

The caveat here is that for the most part, you'll need a $250+ motherboard to do this in the US else you have to order from overseas where you can get models that will allow this for under $150 USD. Outside the US, this isn't an issue as far as I can tell.

HW Unboxed was able to get all 3 of their Non-K chips to do a stable 5.1Ghz OC. I've seen these running over 5.4Ghz. From what I can tell, a 12400 at 5.4Ghz is in the same performance ballpark as a 7600X.



Der8auer used DDR5-5200 C38 - cheap memory - on the 12400 @ 5.24Ghz and and in some cases it beat a 12900K equipped with DDR4-3600 C14 :



In this one averages for 5.24Ghz 12400 beat the 12600K and lost to 12700K/12900K, but the 1% lows were the highest and frame time consistency were the best on the 5.24Ghz 12400 :

None of this is wrong. The problem is that it's not the whole picture. For one, that HWUB screenshot is rather selective - in the games tested, roughly half saw results similar to that, and about half saw results where the lower end chips lagged much further behind (SOTTR, CP2077, Watch Dogs: Legion). There were still marked increases over stock, of course, but you're quite clearly picking one of the few results where the three chips are nearly tied. (I also find it rather hilarious how people just don't understand the concept of linking their sources. How hard can it be?)

On top of this: overclocking is not trivial. I entirely agree that if you're a budget constrained enthusiast, or have the time and willingness to teach yourself how to overclock, low end ADL will likely present a better value proposition than Zen4, at least until lower end Zen4 launches. However, if you don't then it doesn't, as overclocking potential that isn't used is utterly worthless. And, crucially, the vast majority of PC builders do not overclock. One thing is talking about the small niche we as enthusiasts live in; another is talking about the market in general, the advice we give to others, etc. For someone who isn't an enthusiast, who just wants something that works out of the box, Zen4 will likely be a very good option - and we'll have to wait for reviews to see just how it stacks up against RPL.

@Dyatlov A's "point" as they're saying now - which wasn't a point made, but a question - makes no sense in the context of a mass market product launch, but instead seems to believe that AMD when launching a new product is only communicating with enthusiasts. Which ... well, just isn't the case. And that is the problem of asking "why didn't they compare it to OC'd ADL?" - it's a nonsensical thing to ask outside of a small niche of enthusiasts. It makes sense to us, but applying that logic to AMD's statements just fundamentally fails to take into account how the world works.
Posted on Reply
#169
Mussels
Freshwater Moderator
fevgatosSure, let's do that.

Can you explain this?




Ah, when you are doing 24/7 rendering at 5ghz. What they fail to realize, is that any CPU asked to do heavy MT task at 5 ghz will be smoking hot and consume a truckload. That's why, if you are interested in these workloads, you POWER LIMIT the freaking CPU

I've asked but never got an answer, how many tons of LN2 would you need to run a 5950x at 5ghz all core in CBR23.
Another graph that i feel people cant understand because "less is better" isn't in big bold letters

the 11th gen do pretty well there, and then 12th just... flops horribly :/


People are mixing and matching their data, but the simple fact is intel fell behind in power efficiency something fierce - they wanted to win in performance no matter the cost
Posted on Reply
#170
ratirt
ValantarNone of this is wrong. The problem is that it's not the whole picture. For one, that HWUB screenshot is rather selective - in the games tested, roughly half saw results similar to that, and about half saw results where the lower end chips lagged much further behind (SOTTR, CP2077, Watch Dogs: Legion). There were still marked increases over stock, of course, but you're quite clearly picking one of the few results where the three chips are nearly tied. (I also find it rather hilarious how people just don't understand the concept of linking their sources. How hard can it be?)

On top of this: overclocking is not trivial. I entirely agree that if you're a budget constrained enthusiast, or have the time and willingness to teach yourself how to overclock, low end ADL will likely present a better value proposition than Zen4, at least until lower end Zen4 launches. However, if you don't then it doesn't, as overclocking potential that isn't used is utterly worthless. And, crucially, the vast majority of PC builders do not overclock. One thing is talking about the small niche we as enthusiasts live in; another is talking about the market in general, the advice we give to others, etc. For someone who isn't an enthusiast, who just wants something that works out of the box, Zen4 will likely be a very good option - and we'll have to wait for reviews to see just how it stacks up against RPL.

@Dyatlov A's "point" as they're saying now - which wasn't a point made, but a question - makes no sense in the context of a mass market product launch, but instead seems to believe that AMD when launching a new product is only communicating with enthusiasts. Which ... well, just isn't the case. And that is the problem of asking "why didn't they compare it to OC'd ADL?" - it's a nonsensical thing to ask outside of a small niche of enthusiasts. It makes sense to us, but applying that logic to AMD's statements just fundamentally fails to take into account how the world works.
No to mention, OC potential is an add-in thing to a CPU. Like a bonus which may be there or may not. Or if it is there, you can't know exactly how much potential your CPU has which cannot be treated as this is how the CPU will overclock because it may not happen. For saving money purposes this chip is great but it depends. Also, if you want to OC, you need a pricey mobo than the one you could have bought for the CPU to work.
It is like buying a GPU because of it's OC potential. It may not be there because it depends on the GPU. One can go crazy OC the other not really. Same with CPUs.
Posted on Reply
#171
JustBenching
MusselsAnother graph that i feel people cant understand because "less is better" isn't in big bold letters

the 11th gen do pretty well there, and then 12th just... flops horribly :/


People are mixing and matching their data, but the simple fact is intel fell behind in power efficiency something fierce - they wanted to win in performance no matter the cost
Nope, the translation is wrong. It's scores per watt, therefore higher is better. Zen 3 flips horribly. You would know if you bothered to read the actual review where igors comments on how ahead intel is in efficiency.

Zen 3 fell behind in power efficiency cause they wanted to win in performance apparently.
Posted on Reply
#172
ratirt
fevgatosNope, the translation is wrong. It's scores per watt, therefore higher is better. Zen 3 flips horribly. You would know if you bothered to read the actual review where igors comments on how ahead intel is in efficiency.

Zen 3 fell behind in power efficiency cause they wanted to win in performance apparently.
Following your notion which is wrong, 5000 series was here already when 12th gen showed up. Ryzen3 did not compete with 12th gen Intel.
Even if what you are saying is true and has a merit, wrong translation or whatever, Zen3 was released 2 years ago when 12th gen was not in the market so stating, "Zen 3 is behind in power because they wanted to win in performance" is at best misleading. Zen3 competed with 10th and 11th gen and comparing to those it was a very efficient and performed way better chip. It took Intel a bit to even compete with Zen3.
Posted on Reply
#173
JustBenching
ratirtFollowing your notion which is wrong, 5000 series was here already when 12th gen showed up. Ryzen3 did not compete with 12th gen Intel.
Even if what you are saying is true and has a merit, wrong translation or whatever, Zen3 was released 2 years ago when 12th gen was not in the market so stating, "Zen 3 is behind in power because they wanted to win in performance" is not at best misleading. Zen3 competed with 10th and 11th gen and comparing to those it was a very efficient and performed way better chip. It took Intel a bit to even compete with Zen3.
What do you mean "even if" it was true. It is true. You can actually check the review.

But you are right, zen 3 was competing with 10th and 11th. Now with zen 4 amd increased power consumption by a huge amount to compete with raptorlake, even though half their skus fail to even compete with alderlake.
Posted on Reply
#174
big_glasses

is this one also translated wrong....?
Pic from TPU 5800X3D review, power efficency.
Posted on Reply
#175
JustBenching
big_glasses
is this one also translated wrong....?
Pic from TPU 5800X3D review, power efficency.
Thats not autocad is it? Maybe if you read the discussion before replying you would know whats going on.
Posted on Reply
Add your own comment
Dec 22nd, 2024 01:06 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts