Ryzen 9 7950X allegedly 40% faster than 5950X in CPU-Z Multi-Threaded
Click here to post a comment for Ryzen 9 7950X allegedly 40% faster than 5950X in CPU-Z Multi-Threaded on our message forum
Krizby
DonMigs85
Looks like the IPC gain isn't really much over Zen 3 unless you use some of the new instructions. Most of the gainz seem to come from the higher clocks and bigger L2 cache. It's probably still a 4-wide design I'm guessing
cucaulay malkin
cucaulay malkin
https://www.purepc.pl/amd-ryzen-5-5600-vs-intel-core-i5-12400f-test-procesorow-do-1000-zlotych-ktory-jest-szybszy-i-bedzie-lepszym-wyborem?page=0,53
https://www.purepc.pl/amd-ryzen-5-5600-vs-intel-core-i5-12400f-test-procesorow-do-1000-zlotych-ktory-jest-szybszy-i-bedzie-lepszym-wyborem?page=0,55
https://www.purepc.pl/amd-ryzen-5-5600-vs-intel-core-i5-12400f-test-procesorow-do-1000-zlotych-ktory-jest-szybszy-i-bedzie-lepszym-wyborem?page=0,57
the only reason why I haven't bought one yet is 13400/13500 coming soon (with larger cache and e-cores).13500 is going to have 4.5GHz stock clock and 8 e-cores.hope it's ~220eur like 12500. or under 250. I can get 250eur I paid for 10700f back no problem. If I can get 13500 for selling 10700f and reuse my b-die ddr4, I'm buying it for sure. only a fool wouldn't.
12400f is absolutely ridiculous for gaming,it's faster than 9900k more often than not
TLD LARS
Witcher29
EDK-Rise
13900KF ES
Horus-Anhur
cucaulay malkin
tunejunky
and the clickbait of testing on unfinalized drivers begins (for AMD)...
a few notes:
1) no-one is going to buy the 7950X for a single application or (just) gaming especially since it's apparent there will be a Raphael-X (3D cache).
this is an enthusiast's SKU that will now be the poor man's TR
2) the leaks are very promising for both flavors and both camps have done impressive work. AMD is still kicking butt on efficiency and there will be no topping the 7950X for multi-threaded work loads (or tests like blender & transcode).
3) nobody should be citing these alleged benches, including those from Intel as it is just feeding the "content" monster of pre-mature speculation.
the real monsters for the new generation are the mid-range price battles. we know this gen that there is near parity for gaming (until Raphael-X obliterates all the other new cpus) so i think we should also look at the ecosystem and environment. Intel has a bit of an advantage in that they own 80+% of the OEM market but they require more power and emit more heat.
AMD is being a bit different in that they are differentiating the current market by motherboard and chipset - we have yet to see a single X670 (non E) or B series mobo but they all will need ddr5 - unlike the low end Intel. a lot of these bench tests are directly affected by which board is used under what circumstances because of the pcie lanes at the very least
i'm very curious about pcie 5.0 storage, but we all know that even the most basic NVME M.2 will work a treat for gaming.
nizzen
nizzen
Krizby
Silva
Trying to OC a CPU that OC by default is a bit meh. 10 years ago you could squeeze meaningful performance out of any CPU, depending on how lucky you were.
Today is meaningless trying to do that when both Intel and AMD are pushing for their hardware to do that for yourself and beat scores.
I personally like to undervolt and see less temperature, power consumption and noise for a negligible loss in performance.
Witcher29
mackintosh
Iβve been gaming for over 30 years and Iβm yet to run into a game that would BSOD because of a stable overclock. An unstable overclock - sure, many times. But a stable one? Nope. Not saying there arenβt any, just finding it hard to believe.
nizzen
TLD LARS
nizzen
cucaulay malkin
I've had 1600 cl8 @ 2133 cl9 then 4133 c19 at 4133 c16 and everything has been fine, maybe you're not setting vcsaa or vccio. correctly. if it's stable it's stable. crashes / freezes are part of the process of getting it 100% stable.