Some years ago when there were just 5% improvements gen to gen, the word around was that "all low hanging fruits have been picked", and there won't ever be large jumps in performance anymore. Yet, somehow, these days, both companies manage to jump 15-20% gen to gen in both single and multithreaded. Looks like there are plenty more fruits to be picked...

wavetrex:

Some years ago when there were just 5% improvements gen to gen, the word around was that "all low hanging fruits have been picked", and there won't ever be large jumps in performance anymore. Yet, somehow, these days, both companies manage to jump 15-20% gen to gen in both single and multithreaded. Looks like there are plenty more fruits to be picked...

When you run out of low hanging fruits you just invent new low hanging fruits. Programs and rules change all the time, programs get more complex and suddenly more cache helps more now then it did 1-2 years ago, so that is now a low hanging fruit compared to last year. Filling a 13900k with E-cores was something that did not make sense 2 years ago because schedulers and programs got confused by different core speeds and instruction sets. AMD has now changed socket, so they are free to change the rules that previously was locked with the old socket, for example higher power usage is now allowed, so ramping up the CPU power is now a low hanging fruit, that previously was locked by max socket power and maybe also a better thermal transfer on the new heatspreader compared to the old design. Also AMD chiplet design suddenly opened up for other things that previously was not so easy to do. Intel does the same with the new 350W boost mode, easy performance by breaking the rules they themselves have set. Threadripper came with performance that was not possible before and the only reason it exists is because they had excess EPIC server production capabilities, so that was suddenly a low hanging fruit to sell something they already produce to new customers they did not deliver to before. New generations of memory and PCI also opens up for extra performance and change the rules a little. There is also the small tricks they do in the CPU like predicting what you do next and things like that to increase performance, it was a low hanging fruit at the time, but the security holes that was discovered later points to it not being a low hanging fruit afterall.

wavetrex:

Some years ago when there were just 5% improvements gen to gen, the word around was that "all low hanging fruits have been picked", and there won't ever be large jumps in performance anymore. Yet, somehow, these days, both companies manage to jump 15-20% gen to gen in both single and multithreaded. Looks like there are plenty more fruits to be picked...

Amazing what real competition can do to benefit consumers. It's because of this, that we need both AMD and Intel to be strong and competitive. If just one completely just dominates the market, performance stagnates.

tty8k:

Since multithread is what matters in 3D render, this 7700x gets literally destroyed by its counterpart the 13700k as well as 12700k. But hey it finally beats i5 12600k!

Lol.

Just because both products have a "7" in it's name, doesn't mean they are direct competitors. We have to wait for official prices, to see where it they fall in. Simple example. The 5700X costs around 299€. The 12700K costs around 450€. Clearly they don't compete in the same range.

Horus-Anhur:

Just because both products have a "7" in it's name, doesn't mean they are direct competitors. We have to wait for official prices, to see where it they fall in. Simple example. The 5700X costs around 299€. The 12700K costs around 450€. Clearly they don't compete in the same range.

5700x came late, 5800x was 450 at launch. 7700x early leaks put it at $480 same as 13700kf. how are they not direct competitors ?

cucaulay malkin:

5700x came late, 5800x was 450 at launch. 7700x early leaks put it at $480 same as 13700kf. how are they not direct competitors ?

Because the market adjusted to new PCU releases.

Horus-Anhur:

Because the market adjusted to new PCU releases.

in other words they'll have to drop the price if intel turns out faster than ryzen 7 for the same price, that's it. same as 10700kf was direct competition to 5800x but cost 100eur less because it was worse. https://www.notebookcheck.net/AMD-Ryzen-7-7700X-fails-to-hold-up-against-the-Core-i7-13700K-despite-performing-25-faster-than-the-Ryzen-7-5800X-in-Cinebench-R20.642695.0.html

cucaulay malkin:

in other words they'll have to drop the price if intel turns out faster than ryzen 7, that's it.

Exactly. From what we have seen so far from leaks, Zen4 and Raptor lake might be close in single thread. But in heavily threaded applications, Intel probably has a lead. So AMD will have to adjust their prices accordingly to performance. The number or name that a CPU has, bears no influence in price.

Horus-Anhur:

Exactly. From what we have seen so far from leaks, Zen4 and Raptor lake might be close in single thread. But in heavily threaded applications, Intel probably has a lead. So AMD will have to adjust their prices accordingly to performance. The number or name that a CPU has, bears no influence in price.

tbh if 13600kf gets 8 e-cores it might match 7700x too

cucaulay malkin:

tbh if 13600kf gets 8 e-cores it might match 7700x too

Could be. It's curious how 5 years ago, it was AMD slapping more and more cores on their CPUs. But now it's Intel. Though it's those e-waste cores.

Horus-Anhur:

Could be. It's curious how 5 years ago, it was AMD slapping more and more cores on their CPUs. But now it's Intel. Though it's those e-waste cores.

still,they'll have those additional e-cores beat amd, so there's that, I bet you'd rather have them at the same price that amd sells a corresponding product than not, so why always complain.

cucaulay malkin:

still,they'll have those e-cores beat amd, so there's that, I bet you'd rather have them at the same price that amd sells a corresponding product than not, so why always complain.

But we are paying for them. They are not free. They occupy die space and cost to develop and to produce. My complaint is that we could have something more useful. Like more P-Cores or cache. Also, like we both have shown in several threads, these e-cores have limited use. A lot of applications don't use them. Especially games.

Horus-Anhur:

But we are paying for them. They are not free. They occupy die space and cost to develop and to produce. My complaint is that we could have something more useful. Like more P-Cores or cache. Also, like we both have shown in several threads, these e-cores have limited use. A lot of applications don't use them. Especially games.

Just get the 13400F? no E-cores, perform within 5% of 13900K in games and cost less than half? Use the left over money for better DDR5...

Krizby:

Just get the 13400F? no E-cores, perform within 5% of 13900K in games and cost less than half?

it'll get 4 c-cores now.

Horus-Anhur:

But we are paying for them. They are not free. They occupy die space and cost to develop and to produce. My complaint is that we could have something more useful. Like more P-Cores or cache. Also, like we both have shown in several threads, these e-cores have limited use. A lot of applications don't use them. Especially games.

but rpl-s is getting a massive cache size increse too lol https://www.pcgamer.com/intels-next-gen-raptor-lake-doubles-cache-sizes-and-matches-amds-zen-4-thread-count/

Krizby:

Just get the 13400F? no E-cores, perform within 5% of 13900K in games and cost less than half? Use the left over money for better DDR5...

That will probably be a great deal. But it will probably have a locked multiplier and a limit to memory speed.

Horus-Anhur:

Could be. It's curious how 5 years ago, it was AMD slapping more and more cores on their CPUs. But now it's Intel. Though it's those e-waste cores.

Intel certainly did a real 180. For so long they kept saying that no ordinary consumer will ever need more than 4 cores, and if they really must have something better, then they can pay considerable extra for HT. If a consumer is really crazy, they can multiply the money spent and go for the extreme editions and the incomprehensible eight cores. No wonder Intel kicked out its visionless old CEO shortly after Ryzen appeared, otherwise they couldn't have made such a change. In the current situation AMD most certainly can't sit on its hands. I wasn't sure what to think of the performance+efficiency cores tech, but it has proven out to be quite powerful. The max power consumption of Intel CPUs remains too high, though, especially in today's world when in Europe saving power is essential. I'm not altogether happy with the unstressed power consumption of my AMD system, though, so there are two sides to everything.

Horus-Anhur:

That will probably be a great deal. But it will probably have a locked multiplier and a limit to memory speed.

MSI will offer board with external clock gen on their B660 board that allow BCLK overclocking 😀

Horus-Anhur:

That will probably be a great deal. But it will probably have a locked multiplier and a limit to memory speed.

yeah but all z690/b660 boards with extrenal clock generator will be able to oc 13400 too

Krizby:

MSI will offer board with external clock.gen on their B660 board that allow BCLK overclocking 😀

yeah,value-wise intel is serving some crazy options these days. 13400f+ b660 ext clock gen board + your current ddr4 and you're matching 7700x spending maybe 400eur altogether,probably less.

cucaulay malkin:

but rpl-s is getting a massive cache size increse too lol https://www.pcgamer.com/intels-next-gen-raptor-lake-doubles-cache-sizes-and-matches-amds-zen-4-thread-count/

Not as big as you think. Because people are counting all the caches. It's just a 1MB increase in L2 cache for the P-Cores and another 1MB for each 4 cluster of E-cores. But because these CPUs also have more E-cores, the overall cache number has to increase with it. The thing is, we could replace several of those E-cores with more cache, and have a greater IPC improvement.

Krizby:

MSI will offer board with external clock gen on their B660 board that allow BCLK overclocking 😀

If Intel doesn't find a way to block it...