Geekbench of Intel Core i7-12700 Matches the Ryzen 7 5800X

schmidtbag

2021-08-25 15:23

H83:

I still think that the little/big setup doesn´t make sense to desktops, specially after reading that it needs to be properly supported by the OS.

I've explained the benefits multiple times, including at least once to you. What doesn't make sense to me is why so many people (not necessarily you) insist it isn't a good idea for desktops. So long as you're running Windows 10 or Linux 4.0+ (maybe even something older), the scheduler should have sufficient competence to utilize the little cores, since both OSes already have implementations due to ARM. It will take some time before the benefits really kick in, but the same can be said of every CPU technology. This is a necessary way forward.

Webhiker:

If this is the case. Then the smartest move from AMD would be to make the Zen 3+ AM4 compatible and not release it with AM5.

I don't know what differentiates Zen3+ but no matter what way you look at it, I don't think it should be AM4: * If the only differences are things like DDR5 support or an IO chip with different capabilities, then it wouldn't be compatible with AM4 anyway. * If the only difference is just smaller transistors, I don't see that being worth preserving a dead socket. It's not like AMD is struggling to sell anything. * If there are changes more significant than a die shrink but less significant than DDR5, then it would be another headache of BIOS updates, which nobody wants to deal with for an EOL motherboards.

Kaarme:

Well, three would be even better, but unfortunately it's impossible for the current x86-64 environment. At least that might happen in the GPU market.

VIA: https://i.ytimg.com/vi/ZEmEtPkd4IQ/hqdefault.jpg (Yes, VIA, you are)

#5941020

Imglidinhere

2021-08-25 15:31

Caesar:

Any comments from Ryzenians?:p

I mean... this is why I didn't believe the initial 12900K leak where it was scoring higher than the 5950X. So you mean to tell me that the 12700k can't beat the 5800X, but the 12900K magically, with only an extra 4 efficient cores added on, can SOMEHOW beat a CPU with twice as many cores/threads as the 5800X? Nah. Doesn't add up. xD

#5941027

Venix

2021-08-25 16:09

is geekbench aware of the 4 little cores ?...from what it seems is that 8cores 16 threads + 4 small cores are 2% slower than the 5800x ? so the 4 extra little cores is there to cover up for the multithread ? , is geekbench not using em at all so they just sit around? , on the other hand the cpu might be strictly constraint on 65 watt tdp so everything remains on the 3 ish ghz range to stay there .. and with unleashed restraints that can jump to 95 or even 125 tdp and those score shoot right up ..... so what case of the above is it ? i guess we have to wait for the release and review !

#5941033

schmidtbag

2021-08-25 16:21

Venix:

is geekbench aware of the 4 little cores ?...from what it seems is that 8cores 16 threads + 4 small cores are 2% slower than the 5800x ? so the 4 extra little cores is there to cover up for the multithread ? , is geekbench not using em at all so they just sit around? , on the other hand the cpu might be strictly constraint on 65 watt tdp so everything remains on the 3 ish ghz range to stay there .. and with unleashed restraints that can jump to 95 or even 125 tdp and those score shoot right up ..... so what case of the above is it ? i guess we have to wait for the release and review !

Most likely the little cores were doing nothing. If the little cores were being used too, the performance would be abysmal compared to the 5800X. Intel has some catching up to do but not that much.

#5941037

Venix

2021-08-25 16:26

schmidtbag:

Most likely the little cores were doing nothing. If the little cores were being used too, the performance would be abysmal compared to the 5800X. Intel has some catching up to do but not that much.

yeah i suspect the little cores where not used too ... if they where this is not looking good . but we simply do not know yet

#5941057

waltc3

2021-08-25 17:15

Caesar:

Any comments from Ryzenians?:p

Certainly...It's easy to be suspicious when you see results for just one benchmark...;) Caveat Emptor.

#5941060

metagamer

2021-08-25 17:22

This runs cool (unlike the 5800x) and is priced well and we have a solid CPU.

#5941062

Webhiker

2021-08-25 17:26

schmidtbag:

I I don't know what differentiates Zen3+ but no matter what way you look at it, I don't think it should be AM4: * If the only differences are things like DDR5 support or an IO chip with different capabilities, then it wouldn't be compatible with AM4 anyway. * If the only difference is just smaller transistors, I don't see that being worth preserving a dead socket. It's not like AMD is struggling to sell anything. * If there are changes more significant than a die shrink but less significant than DDR5, then it would be another headache of BIOS updates, which nobody wants to deal with for an EOL motherboards.

Yes the AM4 platform is a dead platform, but it's a very well known platform. Adopting a BIOS for it shouldn't be that hard especially it the only change will be V-Cache. If they do release Zen 3+ and release it on AM5, then I don't see the relevance, since Zen4 will come out shortly after. Personally I would consider Zen 3+ on AM4 but if launched on AM5 I would wait until Zen 4 before deciding what to do. If Zen 3+ with v-cache gives the performance uplift everyone is expecting (>20%) then the Zen 4 launch might not be all that impressive.

#5941067

Mineria

2021-08-25 17:43

Venix:

is geekbench aware of the 4 little cores ?...from what it seems is that 8cores 16 threads + 4 small cores are 2% slower than the 5800x ? so the 4 extra little cores is there to cover up for the multithread ? , is geekbench not using em at all so they just sit around? , on the other hand the cpu might be strictly constraint on 65 watt tdp so everything remains on the 3 ish ghz range to stay there .. and with unleashed restraints that can jump to 95 or even 125 tdp and those score shoot right up ..... so what case of the above is it ? i guess we have to wait for the release and review !

i7 12700K comes with 12 cores, 8+4 and 20 threads, so 12/20. Max turbo at 5Ghz for a single and 4.7Ghz for all 12 cores. 125-228W.

metagamer:

This runs cool (unlike the 5800x) and is priced well and we have a solid CPU.

Exactly, heat is the big downside of the 5800x, it can be controlled with tweaking but not as easy as with Intel CPU's where you just enable speed-stepping, AMD definitively could learn something there instead of keeping on insisting that it can take the heat. We all know that heat equals shorter lifetime and increased power usage, so bit of an sad excuse from AMD and their fanbase to look away from it while been part of the bashing on Intel CPU's with high temperatures in the past. I got mine running fine with low temperatures after hours of reading, investigating and tweaking, never ever had to do that with an Intel CPU, just using a proper cooler was enough to gain control and even have the CPU to lower frequency while using the performance power plan.

#5941068

schmidtbag

2021-08-25 17:45

Webhiker:

Yes the AM4 platform is a dead platform, but it's a very well known platform. Adopting a BIOS for it shouldn't be that hard especially it the only change will be V-Cache.

In theory, you're right. But if AMD never planned to have these Zen3+ models on AM4 then there could potentially be issues getting existing motherboards to boot up properly.

If they do release Zen 3+ and release it on AM5, then I don't see the relevance, since Zen4 will come out shortly after. Personally I would consider Zen 3+ on AM4 but if launched on AM5 I would wait until Zen 4 before deciding what to do.

I agree, though they also released Steamroller on AM4. From what I heard, Zen3+ isn't even official, regardless of socket.

If Zen 3+ with v-cache gives the performance uplift everyone is expecting (>20%) then the Zen 4 launch might not be all that impressive.

This is very true, though, with v-cache, it doesn't matter which socket it's on, the performance impact would make Zen4/AM5 models not as attractive to upgrade to. Motherboard manufacturers in particular are not going to be happy about reducing their sales because of old chipsets retaining popularity.

#5941070

tunejunky

2021-08-25 17:48

Kaarme:

When AMD couldn't compete during the decade of degeneration, Intel didn't really bother to develop anything but the production technology. Intel has found it hard to compete with Ryzen due to that negligence, but nobody sane would certainly wish for a future where Intel can't compete for many years, making AMD, in turn, turn lazy. We need a situation with two strong competitors, who all the time turn the tables. Well, three would be even better, but unfortunately it's impossible for the current x86-64 environment. At least that might happen in the GPU market.

and not a few at Nvidia planning for an ARM future for mainstream cpu and perhaps a few "top secret" projects that synergize existing assets...

#5941096

Toreslee

2021-08-25 19:07

D1stRU3T0R:

yes? Intel future CPU is as good as an imediately "last gen" Ryzen. Upgrades people, upgrades.

tell that to AMD GPU fan girls.

#5941102

cucaulay malkin

2021-08-25 19:32

Webhiker:

Yes the AM4 platform is a dead platform

if only all dead plafroms were like am4..... and btw intel's 65w means absolutely nothing 10900F is 65w and it'll chug over 200w

Toreslee:

tell that to AMD GPU fan girls.

tough but not unfair

#5941136

nizzen

2021-08-25 21:36

cucaulay malkin:

if only all dead plafroms were like am4..... and btw intel's 65w means absolutely nothing 10900F is 65w and it'll chug over 200w tough but not unfair

I got the 5900x to draw 350w, and it says 105w. Full stock 5900x draw 142w in cinebench. Just saying..... 😉 The funny part is that 5900x stock is hotter than 11900k with the same cooling 😀 I tested with Ek watercooling with 360 radiator.

#5941139

Horus-Anhur

2021-08-25 21:43

That big IO die, made in 12nm, does no favor to Ryzen.

#5941141

cucaulay malkin

2021-08-25 21:47

nizzen:

I got the 5900x to draw 350w, and it says 105w. Full stock 5900x draw 142w in cinebench. Just saying..... 😉 The funny part is that 5900x stock is hotter than 11900k with the same cooling 😀 I tested with Ek watercooling with 360 radiator.

what does 11900k draw in cinebench 220w ? 250w ? for 70-75% of 5900x's performance ? at the same price ? that is just sad. 105 to 140 is not 65 to 220

#5941147

nizzen

2021-08-25 22:03

cucaulay malkin:

what does 11900k draw in cinebench 220w ? 250w ? for 70-75% of 5900x's performance ? at the same price ? that is just sad. 105 to 140 is not 65 to 220

Non of the cpu's are sad. They do work great in games and programs. Maybe you are sad with no money to buy a cpu? Feels bad man....

#5941151

itpro

2021-08-25 22:14

People who purchase/build multiple rigs, or buy the latest and greatest by selling used hardware for their withering e-peen are even more sad as far as I know. Their lifetime isn't quite full. Feels so generic, with no purpose or goals.....

#5941155

alanm

2021-08-25 22:34

itpro:

People who purchase/build multiple rigs, or buy the latest and greatest by selling used hardware for their withering e-peen are even more sad as far as I know. Their lifetime isn't quite full. Feels so generic, with no purpose or goals.....

Its a hobby. If they can afford it, why not? And it sounds like fun and something I would do with an unlimited budget too. 🙄

#5941158

itpro

2021-08-25 22:59

alanm:

Its a hobby. If they can afford it, why not? And it sounds like fun and something I would do with an unlimited budget too. 🙄

I enjoy giving 5k$ per year for public university, subscriptions, books and online learning platforms, for entertainment and education. At least I endorse self-improvement. By making billion dollar corporations even more rich from a recycle hobby, I can't really agree. It reminds me kids asking for toys.