News AMD deep-dives Zen 5 — Ryzen 9000 and AI 300 benchmarks, Zen 5, RDNA 3.5 GPU, and XDNA 2 microarchitectures

[bit_user]

Isn't that how people use their PC's?

Eh, no. I think someone knows if they need to run heavily-threaded apps and will want to see that figure by itself, when trying to choose a product which runs them either fast or efficiently (depending on which they care about).

For a person who almost never runs heavily-threaded apps, they'll want to just see performance and efficiency on lightly-theaded workloads.

In other words, by mixing together the two classes of apps, TechPowerUp presumed a certain weighting between the two. That effective weighting is going to be right for almost no one. So, it would be better for them to provide separate averages and then let users do their own weighting.

 

[TheHerald]

AMD is still way better than Intel in most power targets that matter for consumers and data center

I'm on amazonde right now. 300 to 350€ gets me either a 7700x, a 7800x 3d or a 13700k (they don't have a 7900x for whatever reason). I'm insanely confident the Intel chip will be much more efficient (ie. at the same power target) in both single and multithreaded workloads than the two amd chips. But let's all pretend otherwise for god knows what reason.

 

[bit_user]

But you don't seem to bother when the mismatched comparisons are done the other way, do you?

Why do you say that?

I agree ISO power are the only relevant comparisons

I disagree. Insisting on iso-power comparisons is:

1. Needlessly stringent. All we need is for CPUs to be roughly in the same ballpark, in both performance and power. These roughly align with their market segments.
2. Not how how virtually anyone actually runs these CPUs, in the real world. So, it provides data with less relevance than if you'd just run CPUs of a similar class in their typical customer configurations.

- the problem is when you do that Intel wins, in both MT and ST efficiency - in the vast majority of segments.

On the MT efficiency point, this disagrees. The 7950X and 7900X beat the i9 and i7, respectively.

​

At below 300€ you got the 7600 and the 7700 from amd and the 13600 and the 13700 for intel. (245$ for the 7700x, 265 for the 13700).

The PCPartPicker pricing shows the i7-13700K to be selling for about half way in between the 7900X and 7700X. It's more awkward to compare, when the prices don't match up neatly, but that kinda does align with the data.

 

[TheHerald]

I just went to amazon.com cause they just have more models available.

At below 300€ you got the 7600 and the 7700 from amd and the 13600 and the 13700 for intel. (245$ for the 7700x, 265 for the 13700). Both Intel chips are more efficient than the 2 amd chips at iso power in both ST and MT.

At below 400€ we got the 7800x 3d (385$), the 14700kf (349$) and the 7900x (342$). The 14700kf is a lot more efficient than the 7800x 3d, and probably a bit more than the 7900x (not sure, scratch that if you disagree). So how the heck are we saying amd chips are more efficient? I just dont get it.

 

[TheHerald]

On the MT efficiency point, this disagrees. The 7950X and 7900X beat the i9 and i7, respectively.

​

​

Only problem is the i7 competes with the R7, and according to your graph the i7 literally roflstomps it. The 7900x competes (in pricing) to the new 14700k which isn't in the graph.

 

[jeremyj_83]

Ah, but you had to find a workload that could make them both reach that number, and then limit them to it. So, what you really mean is that you can create a scenario in which the Intel CPU isn't generating more heat, but runs at cooler temps (due to things like fewer bottlenecks at the junction between the die and IHS).

That 250W is also beyond the PPT for AM5 with a 170W TDP (230W PPT). Therefore you are pushing the 7950X even further beyond the FV curve than a 230W PPT already does.

 

[TheHerald]

That 250W is also beyond the PPT for AM5 with a 170W TDP (230W PPT). Therefore you are pushing the 7950X even further beyond the FV curve than a 230W PPT already does.

LOL, ok let's lock both to 200 watts then. Doesn't change the point or the outcome

 

[bit_user]

I'm on amazonde right now. 300 to 350€ gets me either a 7700x, a 7800x 3d or a 13700k (they don't have a 7900x for whatever reason).

According to PCPartPicker, the 7900X is currently going for $330.

https://pcpartpicker.com/product/bwxRsY/amd-ryzen-9-7900x-47-ghz-12-core-processor-100-100000589wof
​

I'm insanely confident the Intel chip will be much more efficient

The data says otherwise.

 

[TheHerald]

According to PCPartPicker, the 7900X is currently going for $330.

You are comparing EU prices to US prices. According to part picker the 13700kf is at 266, only 20$ more than the 7700x, which it is competing against. Why are you not accepting this? What's gonna happen if we all admit the obvious, intel is more efficient at ISO power due to the immense core count advantage. What the heck guys...
https://pcpartpicker.com/product/w8...700kf-34-ghz-16-core-processor-bx8071513700kf

 

[bit_user]

Only problem is the i7 competes with the R7,

No, it doesn't. The prices don't match up. The i7-13700K costs about half way between the 7700X and 7900X.

 

[bit_user]

You are comparing EU prices to US prices.

I'm using US prices. Both Intel and AMD are US companies and the US is the biggest market for them. Those prices should basically be considered normative.

Furthermore, I'm citing a price engine that aggregates over multiple stores and shows pricing history over time. You're quoting spot prices without even providing links, so I can't tell whether those are even new or from authorized resellers.

If you want to make a compelling argument, you must start with good data.

 

[bit_user]

BTW, we haven't been talking about Zen 5 for a long time, because certain individual(s) decided to take us off-topic on a soon-to-be-irrelevant tangent that we've hashed out several times already.

If you're an AMD fan (or neutral): Please don't take potshots at Intel, here. The Intel guys will definitely seize the opportunity to distract from the impressive announcements around Zen 5. And if they try to throw shade at Zen 4, just don't take the bait. This is a pitched battle they know how to fight, and they'd much rather we get mired in the same old debates than talk about Ryzen 9000.

 

[jeremyj_83]

Yeah, that's the usual braindead take I keep seeing being spread around. Only problem is, it's just not true, lol. Even if you compare the 7950x to the 14900k, the only segment that AMD actually has somewhat of a lead, it's very not true. At 125w the 14900k scores around 31k in CBR23. The 7950x at 65w scores 22k. But whatever, as I've noticed, facts don't matter when it comes to hating on Intel so keep at it.

Now if you go further down the stack, oh boy. I'd like to see a 65w 7700x against a 125w 13700k. Wouldn't you? 😍

You say things that are only 1/2 truths. Lets keep using CBR23 as an example. Compare the CBR23 numbers from Techpowerup and Anandtech's "A lighter touch" article. We see that the numbers in CBR23 are all very close with each other (13900k vs 14900k is basically irrelevant) across the power levels stated. Therefore we can also assume that with the 7950X values at stock being similar that the numbers for the different TDPs will also be the same. At every similar TDP AMD is ahead of Intel in performance to the point where Intel needs 125W TDP to be faster than AMD at 65W TDP. Oh and I don't care about the personal numbers that you come up with in your head for the 7950X getting 22k at 65W. There is no data to support your claims outside of your proprietary knowledge. We do know that Intel gets only 22K at 65W but AMD is at 31k or 125W Intel level.

 

[TheHerald]

You say things that are only 1/2 truths. Lets keep using CBR23 as an example. Compare the CBR23 numbers from Techpowerup and Anandtech's "A lighter touch" article. We see that the numbers in CBR23 are all very close with each other (13900k vs 14900k is basically irrelevant) across the power levels stated. Therefore we can also assume that with the 7950X values at stock being similar that the numbers for the different TDPs will also be the same. At every similar TDP AMD is ahead of Intel in performance to the point where Intel needs 125W TDP to be faster than AMD at 65W TDP. Oh and I don't care about the personal numbers that you come up with in your head for the 7950X getting 22k at 65W. There is no data to support your claims outside of your proprietary knowledge. We do know that Intel gets only 22K at 65W but AMD is at 31k or 125W Intel level.

I knew youd be citing anandtech 😍

Try to read the whole review. Need help? Let me help you, and im quoting

"The catch here, however, is that the AMD platform as a whole was far more lax in sticking to its programmed PPT values, as evidenced by yCruncher power consumption. Despite setting the 7950X to 65 W, we still measured 90.3 W under that workload. So on the assumption that translates to CineBench, the 7950X's power efficiency gains aren't as impressive; we're looking at 42% of the power consumption for 81.1% of the performance, or a power efficiency of 1.93x over stock."

TLDR, the 7950x wasn't actually drawing the power you see on the graph. It was drawing way more than that. It doesn't score 31k at 65w, I can assure you. Here is some data from club386 that actually measures PPT and not TDP like anandtech

No amount of data will convince you but at least I tried

 

[TheHerald]

I'm using US prices.

You are using US prices for the 7900x but then you are comparing it to the EU prices I used for the 13700k. The 7900x in the US costs as much as the 14700 (which doesn't exist in your graph). The 13700kf is a lot cheaper than the 7800x 3d, a lot cheaper than the 7900x and 20$ more expensive than the 7700x. Which it completely decimates in efficiency.

If I had no idea and just read the posts around here I'd buy the 7700x over the 13700kf because it's more efficient at ISO power. Then I'd release what an idiot I am for listening to you people...

 

[bit_user]

You are using US prices for the 7900x but then you are comparing it to the EU prices I used for the 13700k.

Nope. I looked at PCPartPicker for all my prices.

 

[TheHerald]

Nope. I looked at PCPartPicker for all my prices.

And according to Pcpartpicker 7900 = 7800x 3d = 14700.

13700 ~= 7700.

How is AMD more efficient? 😆

It's a funny argument them cause it implies that dropping prices equates to increased efficiency, which is silly.

 

[stuff and nonesense]

Man it's not complicated. If my single tower air cooler u12a can dissipate 330 watts on a 14900k and only 200 watts on a 7950x, it's blatantly obvious that intel chips are easier to cool.

Your U12a has a contact plate designed for the Intel IHS, the Intel and AMS IHS contact faces are different shapes. Please review differences with the d15-G2, 3 contact face designs, they make a difference wrt Intel and AMD.
The 7950x ramps up to 95 °C and stays there, in its optimal design envelope. If it takes 200W to get there then so what, it takes 200W. Better cooling.. slightly higher clocks closer to the chip’s design limits (assuming stock).

 

[jeremyj_83]

I knew youd be citing anandtech 😍

Try to read the whole review. Need help? Let me help you, and im quoting

"The catch here, however, is that the AMD platform as a whole was far more lax in sticking to its programmed PPT values, as evidenced by yCruncher power consumption. Despite setting the 7950X to 65 W, we still measured 90.3 W under that workload. So on the assumption that translates to CineBench, the 7950X's power efficiency gains aren't as impressive; we're looking at 42% of the power consumption for 81.1% of the performance, or a power efficiency of 1.93x over stock."

TLDR, the 7950x wasn't actually drawing the power you see on the graph. It was drawing way more than that. It doesn't score 31k at 65w, I can assure you. Here is some data from club386 that actually measures PPT and not TDP like anandtech

No amount of data will convince you but at least I tried

Except that data is VERY flawed. The 7900 which has a stock TDP of 65W scores 24665 in CBR23. It shouldn't score higher in MT than the 7950X at the same TDP. But no amount of data will ever convince you that you are wrong but I tried. Overall you bringing up Zen 4 power, temps, and efficiency is still a straw man fallacy as that isn't relevant to Zen 5.

 

[TheHerald]

Except that data is VERY flawed. The 7900 which has a stock TDP of 65W scores 24665 in CBR23. It shouldn't score higher in MT than the 7950X at the same TDP. But no amount of data will ever convince you that you are wrong but I tried. Overall you bringing up Zen 4 power, temps, and efficiency is still a straw man fallacy as that isn't relevant to Zen 5.

The 7900 doesn't draw 65w either, TDP is NOT PPT. Just because amd is fooling you with those numbers doesn't mean it fools everyone else. According to the gnexus review, the 7900 draws 86 watts, not 65.

Computerbase has tested on every possible configuration btw, here is the data

The 13900k is the most efficient chip they've ever tested at 45w btw. You can also clearly see the i5 beating the r5 handily and the i7 beating the R7 at most configurations.

So if efficiency is your TOP priority, you buy an Intel. Period. Both ST and MT efficiency they have the lead.

 

[-Fran-]

Wait, why is price being used as part of an efficiency discussion?

Are we talking "price efficient" now? The jump from one can of worms to another is olympic in nature. Gold medalist class.

Regards.

 

[jeremyj_83]

The 7900 doesn't draw 65w either, TDP is NOT PPT. Just because amd is fooling you with those numbers doesn't mean it fools everyone else. According to the gnexus review, the 7900 draws 86 watts, not 65.

Computerbase has tested on every possible configuration btw, here is the data

The 13900k is the most efficient chip they've ever tested at 45w btw. You can also clearly see the i5 beating the r5 handily and the i7 beating the R7 at most configurations.

So if efficiency is your TOP priority, you buy an Intel. Period. Both ST and MT efficiency they have the lead.

Considering you don't listen and Bit User and I have said MULTIPLE times this is a straw man argument I am forever done with you. Please go away troll.

 

[TheHerald]

Wait, why is price being used as part of an efficiency discussion?

EXCELLENT question. I keep wondering the same, but that's how bit_user wants to compare. Not even MSRP, but current prices.

 

[TheHerald]

Considering you don't listen and Bit User and I have said MULTIPLE times this is a straw man argument I am forever done with you. Please go away troll.

How is it a strawman argument? Computerbase put all cpus at the same power and run their MT test suite. The 13900k knocked it out of the park. Just because you don't like the results doesn't make them false.

 

[bit_user]

Wait, why is price being used as part of an efficiency discussion?

Trying to group products by market segment. I maintain that it doesn't make sense to compare products in different market segments, which TheHerald has been doing in the past.

IMO, the comparison that's actually useful to people is to look at performance and efficiency of products in the same/adjacent market segments, as a user would typically configure them. Comparing anything else is a bit silly. Also silly: looking only at efficiency, in isolation of things like price and performance, because nobody who would seriously consider a 7950X3D would just change their mind and buy a i5-13400F, instead. Probably not even a i9-14900T, either.