Question How in the world has AMD got the Ryzen 7600X and 7700X priced same when they are inferior even in P cores only compared to 13600K and 13700K

[Wolverine2349]

I mean the Ryzen 7700X is an 8 core CPU and Ryzen 7600X is a 6 core CPU. And the 7700X is $399 and 7600X is $299.

Intel has the Core i7 13700K priced at $399 and Core i5 13600K priced at $399. And those CPUs have better P cores being 8 and 6 core counterparts with slightly better IPC than Zen 4 and can clock as high or higher with similar power usage. And for those who do not like e-cores (I am one of them, but I love Intel P cores) can disable them and you get better 6 and 8 core CPUs form Intel Raptor Lake than AMD Ryzen. And for those who want e-cores you get then as well for the same price and better P cores of equal core counts.

SO what is AMD thinking and they still have not budged on the prices of the 7600X and 7700X. They are pricing the like their 6 and 8 Zen 4 cores are better than Intel's Raptor Cove cores of equal count even though they are not any better and in fact not as good?? Or is that debatable??

The Ryzen 7900X and 7950X prices make more sense as then you get more than 8 strong cores and AMD has those by the balls who want more than 8 cores and do nit want to go hybrid route. SO yeah 7900X and 7950X prices make sense.

But 7600X and 7700X are almost a ripoff unless you just have not have AMD as they do nothing better than 13600K and 13700K for exact same price and have slightly weaker P cores and no additional e-cores for those that like the e-core options (And for those that do not it is easy peasy to disable and you get the better 6 and 8 core chips for the same price)

Its puzzling to me AMD is behaving as if they are still superior in all ways like they were with Ryzen 5000 from November 2020 to November 2021 when Intel was of no competition on core count nor per core IPC performance which was only for 1 year. I mean AMD is still much smaller and was underdog for years and hard to believe they think they can act they are premium brand in the 6 and 8 core CPU segment when the 7600X and 7700X are worse than Intel counterparts even with the e-cores off.

Your thoughts

 

[igor_kavinski]

so why did Intel fuse it off for those that disable e-cores and use P cores only

Here's your answer: https://www.igorslab.de/en/efficien...-the-returned-command-set-in-practice-test/2/

Only P-cores with AVX-512 aren't able to beat P+E-cores so they must have decided that allowing AVX-512 didn't have any benefit to the average user. Plus, it gives them the added benefit of forcing AVX-512 users onto their server/workstation platform. Of course, thanks to AMD, AVX-512 users aren't left out in the cold anymore.

 

[igor_kavinski]

I'm still noit convinced this hybrid thing as implemented by Intel has a justification on desktop.
It has a justification on silicon area, sure but otherwise...

AMD forced Intel's hand with the 5900X/5950X. If AMD had kept their consumer CPUs to 8C/16T only, Intel probably wouldn't have felt the need to use E-cores on desktop.

 

[Wolverine2349]

AMD forced Intel's hand with the 5900X/5950X. If AMD had kept their consumer CPUs to 8C/16T only, Intel probably wouldn't have felt the need to use E-cores on desktop.

I wish that were the case as 8 P cores only even better without having to deal woth the e-cores. Yeahg productiviuty and super parallel apps benefit form more than 8 cores and as many as you can throw at them, but 8 cores is more than en

Here's your answer: https://www.igorslab.de/en/efficien...-the-returned-command-set-in-practice-test/2/

Only P-cores with AVX-512 aren't able to beat P+E-cores so they must have decided that allowing AVX-512 didn't have any benefit to the average user. Plus, it gives them the added benefit of forcing AVX-512 users onto their server/workstation platform. Of course, thanks to AMD, AVX-512 users aren't left out in the cold anymore.

Well yeah of course in infinite threaded workloads that can parallel to as many CPU cores as you can throw at it. But why disable AVX when it was there and could be used with e-cores off. Was it an accident and not tested even if it worked and Intel does not want to get burned by people using what was not tested on the P cores.

 

[Hulk]

Because, on that very issue; it all seems to be a quick hack, that hybrid thing.
If they know all along the e-cores wouldn' be able to run AVX512, why take the hassle of developping the P-cores with it ?
Because the hybrid things is an afterthought. A pity, it could have been a lot better.

I'm still noit convinced this hybrid thing as implemented by Intel has a justification on desktop.
It has a justification on silicon area, sure but otherwise...

Good question as to why the P's have AVX512 and the E's do not. It does seem kind of wasteful unless the AVX512 structures require an insignificant amount of die real estate. If this is the case then the redesign of the P's, which my need AVX512 for other duties outside of the hybrid parts, might not have been worth the die area savings?

Just guessing I'm sure IntelUser2000 would have a good answer here though.

 

[Markfw]

Good question as to why the P's have AVX512 and the E's do not. It does seem kind of wasteful unless the AVX512 structures require an insignificant amount of die real estate. If this is the case then the redesign of the P's, which my need AVX512 for other duties outside of the hybrid parts, might not have been worth the die area savings?

Just guessing I'm sure IntelUser2000 would have a good answer here though.

My guess is that when the design was made for the P-cores, they did not have all the information, like how much power it took, and how much power is required to overtake Zen4. As things progressed, that when they upped the wattage, and then added the e-cores to save power, but still win the MT benchmarks. Then they figured out that when e-cores are working, avx-512 must be disabled. And lastly, since they did not want to have the user have to reboot, disable the e-cores, then boot again, they just disabled them.

I think that's what I am hearing, and I would agree with it. All about beating Zen 4 in benchmarks, whatever it takes. 350 watt power be dammed, full speed ahead !

 

[jpiniero]

Because, on that very issue; it all seems to be a quick hack, that hybrid thing.
If they know all along the e-cores wouldn' be able to run AVX512, why take the hassle of developping the P-cores with it ?

Mentioned this before... Intel added instructions to let software quickly understand what core type it is on. Intel realized well into development that wasn't going to be good enough and abandoned it... but it was well past the point where taking AVX-512 out would be too much effort. Especially when Alder Lake was probably put on ice for awhile with the 10 nm debacle.

Intel may have also thought they would release Xeon E server models with the small cores disabled but ended up not bothering I guess. If you look at Dell's website they are happily selling Servers with Comet Lake and Rocket Lake.

 

[Kaluan]

The jump in performance from Bulldozer to Zen 1 was between 50-60% IPC gain. That was how bad AMD stuff was before Ryzen. Big jumps were Zen 2 and Zen 3. I am not knocking Zen 4, but the jump from Zen 3 to Zen 4 is not as significant as people would have hoped it would be. When 3D v-cache arrives for Zen 4. People will be happy with Zen 4.

What a ridiculous statement.

5950X to 7950X is a massively bigger leap in both heavily and lightly/single threaded workloads than 3950X to 5950X. I think people are too absorbed in gaming performance and vs Intel metas and straw man their own perceptions of what big gen on gen gains are.

Did they blow Intel out of the water with vanilla Zen4 in a trifecta fashion a-la Zen3 vs 10th/11th gen? No.
But downplaying the huge 20/30/45% gaming/single/MT gains Raphael has over Vermeer, when Vermeer over Matisse only managed a fraction of that is pretty delusional.

 

[Kaluan]

What a ridiculous statement.

5950X to 7950X is a massively bigger leap in both heavily and lightly/single threaded workloads than 3950X to 5950X. I think people are too absorbed in gaming performance and vs Intel metas and straw man their own perceptions of what big gen on gen gains are.

Did they blow Intel out of the water with vanilla Zen4 in a trifecta fashion a-la Zen3 vs 10th/11th gen? No.
But downplaying the huge 20/30/45% gaming/single/MT gains Raphael has over Vermeer, when Vermeer over Matisse only managed a fraction of that is pretty delusional.

I'll steel man my own argument here.

13th gen, outside of multithreaded, has relatively modest gains over 12th. Does that mean it's not impressive overall? No.

Case closed.

 

[DAPUNISHER]

To further curb stomped the premise of this thread. Both Newegg and Amazon have sold out of the 7950X. Only available from 3rd party now.

 

[Wolverine2349]

To further curb stomped the premise of this thread. Both Newegg and Amazon have sold out of the 7950X. Only available from 3rd party now.

7950X has 16 strong cores and the 7900X has 12 strong cores. Intel does not have options yet with more than 8 strong cores without adding on weaker cores to help.

If you want a productivity monster and more than 8 strong cores and do not want to deal with hybrid route, AMD is the way to go by far.

The 7700X and 7600X made less sense.

 

[Hulk]

To further curb stomped the premise of this thread. Both Newegg and Amazon have sold out of the 7950X. Only available from 3rd party now.

Obviously the 7950X is selling well because on Microcenter it costs $0.01 less than the 13900K.
Just joking of course but I love how in short order they met at the same price point!

 

[DAPUNISHER]

Obviously the 7950X is selling well because on Microcenter it costs $0.01 less than the 13900K.
Just joking of course but I love how in short order they met at the same price point!

Competition FTW. AMD has not made these price cuts permanent however. They expire before new year's. Hopefully they cave to market and consumer pressure and the new pricing stays. Or gets even better when the 3D hits. Because you know there will be a premium for it.

 

[Thibsie]

7950X has 16 strong cores and the 7900X has 12 strong cores. Intel does not have options yet with more than 8 strong cores without adding on weaker cores to help.

If you want a productivity monster and more than 8 strong cores and do not want to deal with hybrid route, AMD is the way to go by far.

The 7700X and 7600X made less sense.

7700x and 7600x are pretty much competitive with 13rd gen what are you talking about ?

 

[Wolverine2349]

7700x and 7600x are pretty much competitive with 13rd gen what are you talking about ?

Well Intel has the 13700K and 13600K which have equal strong core counts and those P cores from Intel can clock higher and have slightly better IPC. Though yes 7600X and 7700X are not far behind and have AVX512 so they are competitive when priced right. But when they were priced equally it was strange. The 7900X and 7950X can go back to original price points as they have more strong cores and a big market for those especially who want WIN10 and no hybrid or want to ruin lots of VMs which I am not sure e-cores can do well, but the 7600X and 7700X should stay where they are as their cores are slightly weaker at same clock speed than Intel's except for lacking AVX512 plus they do not have the additional e-cores for those that want them ( I am not one of them lol but I do like Intel's P cores better as they are in fact better even if not by a lot in most cases power usage be darned as e-cores can just be disabled)

 

[poke01]

Companies exist to make profit. If they sell everything at cost or loss and they offer no services afterwards then they lose money.

 

[Hans Gruber]

What a ridiculous statement.

5950X to 7950X is a massively bigger leap in both heavily and lightly/single threaded workloads than 3950X to 5950X. I think people are too absorbed in gaming performance and vs Intel metas and straw man their own perceptions of what big gen on gen gains are.

Did they blow Intel out of the water with vanilla Zen4 in a trifecta fashion a-la Zen3 vs 10th/11th gen? No.
But downplaying the huge 20/30/45% gaming/single/MT gains Raphael has over Vermeer, when Vermeer over Matisse only managed a fraction of that is pretty delusional.

The 7950x is basically a server chip. There is a limited demand among PC users. Most of us do not encode video or try to save the world doing folding projects and trying to cure cancer. The need for 16core/32threads is limited.

If what you say is true about Zen 4 and the 7950x. Why has AMD been making huge price cuts on the 7950x? You are not factoring in the cost and performance of DDR5 vs. 3800mhz/4000mhz DDR4 memory with tight timings. Nobody is denying Zen 4 is a significant performance increase over Zen 3. Most people are simply sitting out Zen 4 until 3D v-cache is available or they are waiting for Zen5.

 

[Markfw]

The 7950x is basically a server chip. There is a limited demand among PC users. Most of us do not encode video or try to save the world doing folding projects and trying to cure cancer. The need for 16core/32threads is limited.

If what you say is true about Zen 4 and the 7950x. Why has AMD been making huge price cuts on the 7950x? You are not factoring in the cost and performance of DDR5 vs. 3800mhz/4000mhz DDR4 memory with tight timings. Nobody is denying Zen 4 is a significant performance increase over Zen 3. Most people are simply sitting out Zen 4 until 3D v-cache is available or they are waiting for Zen5.

There is a large DC community out there. We do a lot of things. Trying to help cure cancer is one. Trying to find cures and more effective vaccines for covid is another. Trying to find comets or errant asteroids that may crash into earth and end humanity is another. I could go on. If you ever need any of these things, you will appreciate what we are doing, but its obvious that your childish take in insulting us as a community is what you would rather do. The world does not revolve around gamers. Also, my best 5950x will not run over about 3733, for the most part its 3600.

Also, I really don't know the numbers, but you see that when people are trying to figure out what is important in a new computer, quite a few mention encoding.

 

[Exist50]

7950X has 16 strong cores and the 7900X has 12 strong cores. Intel does not have options yet with more than 8 strong cores without adding on weaker cores to help.

If you want a productivity monster and more than 8 strong cores and do not want to deal with hybrid route, AMD is the way to go by far.

The 7700X and 7600X made less sense.

No one should be buying a CPU based on some arbitrary judgement of its core config, but rather its performance in real world workloads. And despite the insistence (or perhaps hope...) from some on this forum, hybrid has proven very robust and useful in the vast majority of real world productivity workloads.

I can't wait for AMD to fully embrace hybrid as well so this argument can die.

 

[Carfax83]

I think a lot of the consternation surrounding Zen 4 on enthusiast forums (especially this one which has a strong pro AMD presence) stems from the absurd hype that Zen 4 had before the AMD presentation. I remember the claims of 25% IPC increase over Zen 3 (that's IPC, not overall performance for those that don't know the difference) which when combined with a process node advantage and a purportedly massive clock speed increase was predicted to smash Raptor Lake into a pulp in every aspect. That was the expectation, but of course the reality was much different.

Zen 4 is still a great CPU however, just not as good as it was hyped up to be. Raptor Lake was the opposite in that it had initially low expectations, which it easily outperformed. Which goes to show that it's often better to be the underdog than the expected champion.

 

[Wolverine2349]

No one should be buying a CPU based on some arbitrary judgement of its core config, but rather its performance in real world workloads. And despite the insistence (or perhaps hope...) from some on this forum, hybrid has proven very robust and useful in the vast majority of real world productivity workloads.

I can't wait for AMD to fully embrace hybrid as well so this argument can die.

Well if it is embraced and there are options for more strong cores fine. But some do not like it and want hybrid arch. It can cause compatibility issues with lots of things and you need WIN11. With WIN11 it works well as mostly set and forget. But WIN10 is a different story.

Also those who run lots of VMs and need equal resource allocation between them the big.little not good at all and having same core type with no hybrid is way way better. I do not do either and run games and like 8 super fast cores so Raptor Lake with e-cores off and fast clocked P cores it is for me.

 

[Hans Gruber]

I think a lot of the consternation surrounding Zen 4 on enthusiast forums (especially this one which has a strong pro AMD presence) stems from the absurd hype that Zen 4 had before the AMD presentation. I remember the claims of 25% IPC increase over Zen 3 (that's IPC, not overall performance for those that don't know the difference) which when combined with a process node advantage and a purportedly massive clock speed increase was predicted to smash Raptor Lake into a pulp in every aspect. That was the expectation, but of course the reality was much different.

Zen 4 is still a great CPU however, just not as good as it was hyped up to be. Raptor Lake was the opposite in that it had initially low expectations, which it easily outperformed. Which goes to show that it's often better to be the underdog than the expected champion.

AMD was expecting a much greater uplift in performance going from 7nm to 5nm silicon. That didn't happen and they decided to go the Intel way of massive increases in TDP. This reduced the massive efficiency advantage that AMD had over Intel in previous generations. I do not understand why TSMC is increasing the price of their silicon. The gains they are achieving are not significant enough to justify the prices they are going to charge for 3nm. Hot-spotting is going to be even worse when 3nm CPU's are released. Not to mention that silicon is not efficient at conducting or dissipating heat. Serious cooling issues are around the corner for Zen5.

 

[Mopetar]

AMD was expecting a much greater uplift in performance going from 7nm to 5nm silicon. That didn't happen and they decided to go the Intel way of massive increases in TDP.

I don't think that's the reason in full. AMD could have kept TDP the same with AM5 and while the numbers wouldn't be quite as large, they still wouldn't be bad, especially considering how many watts Intel chips can guzzle.

AMD probably wanted more headroom with AM5 since they couldn't make earlier Zen CPUs draw more than the boards were built for. Of course they would have pissed off their board partners if they told them to make boards supporting a higher TDP, but didn't release chips that could actually hit those limits.

Personally I would have liked it more if AMD set the base TDP lower, but had an easily selectable mode that would draw up to the limits. I can understand why they didn't, and anyone who cares about efficiency can just tweak the settings.

I do not understand why TSMC is increasing the price of their silicon. The gains they are achieving are not significant enough to justify the prices they are going to charge for 3nm.

What's not to understand? There anyone else you can buy 3nm wafers from? TSMC is paying a lot for the equipment to even be able to produce those wafers so they're going to pass it on to their customers.

 

[Wolverine2349]

I don't think that's the reason in full. AMD could have kept TDP the same with AM5 and while the numbers wouldn't be quite as large, they still wouldn't be bad, especially considering how many watts Intel chips can guzzle.

AMD probably wanted more headroom with AM5 since they couldn't make earlier Zen CPUs draw more than the boards were built for. Of course they would have pissed off their board partners if they told them to make boards supporting a higher TDP, but didn't release chips that could actually hit those limits.

Personally I would have liked it more if AMD set the base TDP lower, but had an easily selectable mode that would draw up to the limits. I can understand why they didn't, and anyone who cares about efficiency can just tweak the settings.



What's not to understand? There anyone else you can buy 3nm wafers from? TSMC is paying a lot for the equipment to even be able to produce those wafers so they're going to pass it on to their customers.

I do not think it would have been bad but if clock speeds on Zen 4 were not increased, the performance uplift would have been underwhelming as IPC increased by only like 10% or maybe 13% and in some cases worse latency.

 

[Hans Gruber]

I don't think that's the reason in full. AMD could have kept TDP the same with AM5 and while the numbers wouldn't be quite as large, they still wouldn't be bad, especially considering how many watts Intel chips can guzzle.

AMD probably wanted more headroom with AM5 since they couldn't make earlier Zen CPUs draw more than the boards were built for. Of course they would have pissed off their board partners if they told them to make boards supporting a higher TDP, but didn't release chips that could actually hit those limits.

Personally I would have liked it more if AMD set the base TDP lower, but had an easily selectable mode that would draw up to the limits. I can understand why they didn't, and anyone who cares about efficiency can just tweak the settings.



What's not to understand? There anyone else you can buy 3nm wafers from? TSMC is paying a lot for the equipment to even be able to produce those wafers so they're going to pass it on to their customers.

Everybody would be happier if AMD stuck with their original/standard TDP from previous generations. Even if the top silicon only hit 5/2-5.3ghz the massive efficiency advantage to AMD would offset the Intel unlimited power to the chip philosophy.

Intel's argument against TSMC is the density of their silicon vs. Intel. I can't explain it in detail other than 3nm TSMC would be equivalent to 7nm of Intel silicon. Intel puts a lot more transistors on their silicon vs. what TSMC does. I like TSMC silicon. I am noticing the energy efficiency gains are really all that TSMC provides in their node shrinks.

Then there are those who rightly argue CPU's should have IPC gains as their driving force between generations of CPU design. Relying on an uplift in performance via a node shrink and increased energy efficiency no longer exists.

 

[Hulk]

The hybrid debate is .... interesting to say the least.

Currently Intel is 8+16 and AMD is 16+0 for their top of the stack parts.
Despite being at a transistor density and power disadvantage Intel is more than competitive across the board with 16+0. Yes, they pay a power cost to win the MT benches, but even backing down the power a bit the parts are very competitive.

If, currently, 8 big cores are enough for most applications that are not MT optimized then assuming applications are becoming MORE and not less MT capable, 8 big cores will be enough moving forward.

If Intel 7 can manage 8+32 in the same die as the current 8+16 then MT workload competition is going to be a real challenge for AMD if they don't move to a hybrid architecture. In fact Intel could possible consider 6+40.

Love it or hate it Intel has found a great alternative to squeeze a lot of compute into a small area. I'm quite sure if AMD had done it first there would be a lot more "brilliant" and "genius" adjectives thrown around. That being said if AMD still has the process advantage so if they moved to hybrid they would press Intel against a wall.