AMD Ryzen 9 7950X3D

[JustBenching]

You: "Why bang on about ST supremacy when unlocked 13xxx are highly inefficient?"


E-Cores are useful for AVX/AVX2 with 128-bit use cases and benchmarks.

Current E-Cores have three ports 128-bit AVX2 hardware with 256-bit AVX2 software compatibility. Unofficial "AVX-128" refers to AVX's 128-bit mode subset.


For market segments that prioritize power efficiency, refer to https://www.notebookcheck.net/AMD-s...te-much-lower-power-consumption.698349.0.html

Ryzen 9 7945HX beats Intel Core i9-13980HX despite much lower power consumption

Cinebench R23 Multi Score

Unlimited TDP
Core i9-13980HX = 33052
Ryzen 9 7945HX = 34521

100 Watts TDP
Core i9-13980HX = 26507
Ryzen 9 7945HX = 33487

55 Watts TDP
Core i9-13980HX = 19478
Ryzen 9 7945HX = 26191

​

That's a laptop cpu. In desktop Intel is king of efficiency

 

[Panther_Seraphin]

In desktop Intel is king of efficiency

Yet nearly every graph I see shows Intel being one of the worst at efficency....Fast sure......Efficent though?

 

[ratirt]

Yet nearly every graph I see shows Intel being one of the worst at efficency....Fast sure......Efficent though?

Not to mention. The 7950x3d (especially 3d with low power consumption) is one of the most efficient CPUs to date if you consider tier of the processor and performance/watt or any other metric showing power consumption and yet some people still insist that Intel's power hog 13900K is the most efficient. It is mind bugling even if it is shown in graphs and charts these people still say Intel is more efficient probably by reading these charts backwards? I dont really know but it is a phenomena that has been observed and is getting stronger with time.

 

[JustBenching]

Yet nearly every graph I see shows Intel being one of the worst at efficency....Fast sure......Efficent though?

Cause you are watching unlocked with unlimited power 13900k reviews. There are a ton of Intel skus that are more efficient than AMD's best, like 13900t,13700t,12900t, 13900,12900,13700 etc. And the list goes on

 

[ratirt]

Cause you are watching unlocked with unlimited power 13900k reviews. There are a ton of Intel skus that are more efficient than AMD's best, like 13900t,13700t,12900t, 13900,12900,13700 etc. And the list goes on

No OC but rather stock measurements that crashes your statements about Intel's efficiency.
There are more efficient CPUs from Intel but these are slower than 13900K so irrelevant to the argument.
You can also lower power consumption if you get a lower tier AMD cpu but that is irrelevant. 7900X is using just as much power as 12700K. 7900X3D will use less power than 7900X for sure but it's performance will not be much lower than a 7950X3D.

 

[Ayhamb99]

Intel needs to come out with a similar sku to the X3D chips because the performance results demonstrate how most games benefit more from extra cache instead of higher clocks, maybe with Meteor Lake since they're moving to a chiplet design they might release a sku like that............. The 7950X3D will be a great option if one does a lot of heavy multi-threaded workloads and also games a lot. The power consumption is also way better compared to the stock 7950X and the 13900K

 

[JustBenching]

View attachment 286206

No OC but rather stock measurements that crashes your statements about Intel's efficiency.
There are more efficient CPUs from Intel but these are slower than 13900K so irrelevant to the argument.
You can also lower power consumption if you get a lower tier AMD cpu but that is irrelevant. 7900X is using just as much power as 12700K. 7900X3D will use less power than 7900X for sure but it's performance will not be much lower than a 7950X3D.

When you keep posting the unlocked and power unlimited k skus against locked power limited zen 4 skus, I know you know you are wrong.

As I've said repeatedly, lots of intel skus wipe the floor with zen 4 in efficiency out of the box. That's just a fact, whether you want to admit facts or not up to you. I'm totally against out of the box efficiency comparisons, I think those should be done at iso power, in which case amd wins, but since you seem to prefer out of the box comparisons, intel is the efficiency king with the skus I mentioned already.

 

[ratirt]

When you keep posting the unlocked and power unlimited k skus against locked power limited zen 4 skus, I know you know you are wrong.

As I've said repeatedly, lots of intel skus wipe the floor with zen 4 in efficiency out of the box. That's just a fact, whether you want to admit facts or not up to you. I'm totally against out of the box efficiency comparisons, I think those should be done at iso power, in which case amd wins, but since you seem to prefer out of the box comparisons, intel is the efficiency king with the skus I mentioned already.

You said it repeatedly and you are repeatedly wrong and cherry picking things that align with your way of perceiving things. If you take a non K sku it will be slower no matter how you slice it than a 13900K. You do not provide any evidence but instead empty words yet again. non K skus are more efficient but are slower and because of this they can't match with 7900x3d because only 13900K can much its speed. Not to mention, even if you do look at the power consumption of the non K skus they still use more power than a 7950X3d and are slower.
In games uses less power than a 12700K, 13600K, 12600k even.

Beats 13400F which was claimed to be the most efficient

Then you have application and it uses so low power that is a tad more than a 12600K but the performance difference is enormous between the two.

Seriously.
Just stop or give some proof because you still did not give anything supporting your claims. That includes this thread

AMD RDNA4 Architecture to Build on Features Relevant to Gaming Performance, Doesn't Want to be Baited into an AI Feature Competition with NVIDIA. ​

I'm still waiting.​

I have videos on my channel showing 13900k with ecore usage on a 13900k

I'll show you when I'm home

 

[JustBenching]

You said it repeatedly and you are repeatedly wrong and cherry picking things that align with your way of perceiving things. If you take a non K sku it will be slower no matter how you slice it than a 13900K. You do not provide any evidence but instead empty words yet again. non K skus are more efficient but are slower and because of this they can't match with 7900x3d because only 13900K can much its speed. Not to mention, even if you do look at the power consumption of the non K skus they still use more power than a 7950X3d and are slower.
In games uses less power than a 12700K, 13600K, 12600k even.
View attachment 286216
Beats 13400F which was claimed to be the most efficient
View attachment 286217

Then you have application and it uses so low power that is a tad more than a 12600K but the performance difference is enormous between the two.
View attachment 286218
Seriously.
Just stop or give some proof because you still did not give anything supporting your claims. That includes this thread

AMD RDNA4 Architecture to Build on Features Relevant to Gaming Performance, Doesn't Want to be Baited into an AI Feature Competition with NVIDIA.​

I'm still waiting.​

But you are talking about efficiency, then you are talking about speed, whichever benefits your preconceived conclusion.

The 7950x 3d is slower than the 13900k, but more efficient. The 13900 or the 13900t is slower than the 3d, but also a lot more efficient. Intel is the desktop efficiency king OUT OF THE BOX. That's just a fact that you have to admit and the only thing that's stopping you is your bias.

 

[ratirt]

But you are talking about efficiency, then you are talking about speed, whichever benefits your preconceived conclusion.

The 7950x 3d is slower than the 13900k, but more efficient. The 13900 or the 13900t is slower than the 3d, but also a lot more efficient. Intel is the desktop efficiency king OUT OF THE BOX. That's just a fact that you have to admit and the only thing that's stopping you is your bias.

You know what I'm talking about so stop playing dumb. It all matters. You cant say efficient when you constrain it to 20watts and then say it is the fastest when the power consumption jumps to the roof. It all comes together. Otherwise it is cherry picking.
Slower? You call this slower? Fraction of a percent? Still claim Intel is more efficient? You've got my proof where is yours?

Show me the 13900 being more efficient than 7950X3d I dare you. Show anything even total bullshit but show it. Especially if you claim it is a fact so there is plenty of proof for that. Should not be a problem. Either show a proof or zip your lip and stop trolling.

 

[Arco]

Don't even bother. He is too deep in his echo chamber. Remember, arguing with someone intelligent is hard. But arguing with an idiot? That's straight-up impossible.

He won't accept any facts. Why even bother?

 

[A Computer Guy]

But you are talking about efficiency, then you are talking about speed, whichever benefits your preconceived conclusion.

The 7950x 3d is slower than the 13900k, but more efficient. The 13900 or the 13900t is slower than the 3d, but also a lot more efficient. Intel is the desktop efficiency king OUT OF THE BOX. That's just a fact that you have to admit and the only thing that's stopping you is your bias.

Finding it difficult to find reviews for 13900 non-k with power consumption figures. Do you have any links?

For a bit of a laugh I found this...

 

[JustBenching]

Finding it difficult to find reviews for 13900 non-k with power consumption figures. Do you have any links?

For a bit of a laugh I found this...

View attachment 286250

Yeah, nobody reviews them, they only review the unlocked k parts so they are able to claim that amd is more efficient out of the box, which it isn't

You know what I'm talking about so stop playing dumb. It all matters. You cant say efficient when you constrain it to 20watts and then say it is the fastest when the power consumption jumps to the roof. It all comes together. Otherwise it is cherry picking.
Slower? You call this slower? Fraction of a percent? Still claim Intel is more efficient? You've got my proof where is yours?
View attachment 286224
View attachment 286226
Show me the 13900 being more efficient than 7950X3d I dare you. Show anything even total bullshit but show it. Especially if you claim it is a fact so there is plenty of proof for that. Should not be a problem. Either show a proof or zip your lip and stop trolling.

The 13900k is the fastest cpu, but the 3d is more efficient. The 13900 and the 13900t are the most efficient cpus out of the box, period.

 

[Panther_Seraphin]

Yeah, nobody reviews them, they only review the unlocked k parts so they are able to claim that amd is more efficient out of the box, which it isn't


The 13900k is the fastest cpu, but the 3d is more efficient. The 13900 and the 13900t are the most efficient cpus out of the box, period.

13900t I can see it being pretty good at efficency but you are giving up 700Mhz of top turbo and 1.9Ghz of base frequency.

I cant see the baseline 13900 being "efficent" as its turbo limits is just shy of 220 watts vs the 253 of the K.

 

[Vya Domus]

But you are talking about efficiency, then you are talking about speed, whichever benefits your preconceived conclusion.

You can't talk about efficiency without talking about speed, you genius.

Efficiency = speed / power.

 

[JustBenching]

13900t I can see it being pretty good at efficency but you are giving up 700Mhz of top turbo and 1.9Ghz of base frequency.

I cant see the baseline 13900 being "efficent" as its turbo limits is just shy of 220 watts vs the 253 of the K.

Ιts turbo (PL2) lasts for 54 seconds. For workloads more than 54 seconds (which basically all mt tasks are) it drops to 65w. I

You can't talk about efficiency without talking about speed, you genius.

Efficiency = speed / power.

Obviously, but efficiency has nothing to do with max speed "genius"

 

[A Computer Guy]

You can't talk about efficiency without talking about speed, you genius.

Efficiency = speed / power.

I could be wrong but I thought the formula was more like this for a given frequency and wattage...
efficiency = cpu_time / (run_time x number_of_cpus)

However there are different kinds of efficiency calculations that may be interesting and useful.

 

[Vya Domus]

Obviously, but efficiency has nothing to do with max speed "genius"

That statement literally doesn't mean anything. The point is speed is implicitly relevant when talking efficiency, whether it's "max speed" or min speed or whatever is irrelevant.

efficiency = cpu_time / (run_time x number_of_cpus)

I don't know what that formula is for but it looks like something that measures multicore scaling.

 

[JustBenching]

That statement literally doesn't mean anything. The point is speed is implicitly relevant when talking efficiency, whether it's "max speed" or min speed or whatever is irrelevant.


I don't know what that formula is for but it looks like something that measures multicore scaling.

I was replying to a comment saying that "the 13900t isn't the fastest cpu", which is obviously completely irrelevant when talking about efficiency. Please, stop being wrong every single time, youll end up being ignored.

 

[Vya Domus]

I was replying to a comment saying that "the 13900t isn't the fastest cpu", which is obviously completely irrelevant when talking about efficiency.

How could it possibly be irrelevant, talking about efficiency while ignoring performance is nonsensical, that's what efficiency means, performance per watt. How can you not comprehend that ?

Please, stop being wrong every single time

Stop it with the autobiography.

 

[A Computer Guy]

I don't know what that formula is for but it looks like something that measures multicore scaling.

I thought for a moment that before arguing about efficiency it's best to first know if you agree on formula for calculating such so you know your are talking about the same thing. Otherwise one person may have a different idea and the discussion becomes pointless because the parties are talking and arguing about different things.

 

[JustBenching]

How could it possibly be irrelevant, talking about efficiency while ignoring performance is nonsensical, that's what efficiency means, performance per watt. How can you not comprehend that ?

Being the fastest is irrelevant to being the most efficient. If you can't understand that, well, can't help you

I thought for a moment that before arguing about efficiency it's best to first know if you agree on formula for calculating such so you know your are talking about the same thing. Otherwise one person may have a different idea and the discussion becomes pointless because the parties are talking and arguing about different things.

Efficiency is power used to finish a task. At least that's what im talking about. Intel has several skus that are kings of efficiency out of the box.

 

[Vya Domus]

I thought for a moment that before arguing about efficiency it's best to first know if you agree on formula for calculating such so you know your are talking about the same thing. Otherwise one person may have a different idea and the discussion becomes pointless because the parties are talking and arguing about different things.

Power efficiency is implicitly related to energy consumption which does not appear in that formula, I think it's pretty clear that it's not the same thing.

Being the fastest is irrelevant to being the most efficient.

Absolutely nobody thinks that expect you, like I keep saying efficiency is just another way of saying performance/watt. Hence comparing just efficiency figures is worthless, that only makes sense we you also look at speed independently because there is no objective criteria which can determine if for example a product which is 5% more efficient but twice as slow is a better or worse option. Nobody buys a CPU because it is the most efficient.

AMD and Intel could release CPUs running at 1 Ghz max and they'd have amazing efficiency yet nobody would be dying the get their hands on one of those.

 

[JustBenching]

Absolutely nobody thinks that expect you, like I keep saying efficiency is just another way of saying performance/watt. Hence comparing just efficiency figures is worthless, that only makes sense we you also look at speed independently because there is no objective criteria which can determine if for example a product which is 5% more efficient but twice as slow is a better or worse option. Nobody buys a CPU because it is the most efficient.

AMD and Intel could release CPUs running at 1 Ghz max and they'd have amazing efficiency yet nobody would be dying the get their hands on one of those.

I completely agree with you (surprisingly). That's why I find out of the box efficiency comparisons completely dumb. Efficiency comparisons should be done at equal wattage,, in which case sure, the 7950x is more efficient (only in mt workloads) than the 13900k, but not by much. Best case scenario is 19% in vray, but usually it's around 10-15% in most workloads. On the other hand the 13900k is much more efficient in mixed workloads. Gaming is the only task where the 13900k stinks, but frankly so does the normal 7950x.

 

[A Computer Guy]

Being the fastest is irrelevant to being the most efficient. If you can't understand that, well, can't help you


Efficiency is power used to finish a task. At least that's what im talking about. Intel has several skus that are kings of efficiency out of the box.

See I was thinking for example if CPU-A vs CPU-B were scaled to run at 35 watts @ 2GHz for a given task T and CPU-A completed in a shorter time frame it would be considered more efficient at that task.

Using your description and going back to look at the charts (https://www.techpowerup.com/review/ryzen-7800x3d-performance-preview/20.html) it seems many of the Intel parts are doing well for single threaded efficiency and more poorly for multi-threaded efficiency with a mix in gaming efficiency. Would you agree or disagree that these graphs fit your standard? And secondly do you have an example of which sku's that you believe are out of the box efficient as you are describing?

AMD and Intel could release CPUs running at 1 Ghz max and they'd have amazing efficiency yet nobody would be dying the get their hands on one of those.

Well that depends on the market and purpose I think.