Intel Core i5-12600K

[Luminescent]

The press coverage of these new Intel cpu's sounds like Tomshardware "Just buy it" coverage of Nvidia RTX.
I fear that if AMD gets access to TSMC 5nm we could see Intel to consume triple the power for same performance.

 

[The red spirit]

That's for 8 Performance cores and 8 Atoms against 16 Performance cores though. The 12900K is dumb because it's marketed as a 24-thread part trying to beat a 32-thread part and the only way it can do this is by pushing the clocks and power envelope well beyond what is reasonable. 350W for a CPU is obscene outside of LN2 overclocking IMO.

What's a more useful comparison is the 12600K with E-cores disabled vs the 5600X. Both use similar amounts of power and operate in their sweet-spot clock/voltage ranges. Intel simply have the IPC advantage with Alder Lake compared to Zen3 right now. Whether that's architecture or DDR5 is unknown at this stage and what I (and a lot of other people) am waiting for is a like-for-like battle using the same speed DDR4.

I think that all this situation is really dumb and Alder Lake is just not great. They have a small edge over AMD, but at horrendous power use and heat output. You know, it's not like they are the only ones releasing stuff. If AMD released their own version of big.LITTLE design, I think that they would have an edge and, importantly, soon that power consumption and heat output will look prehistoric.

 

[The King]

That's for 8 Performance cores and 8 Atoms against 16 Performance cores though. The 12900K is dumb because it's marketed as a 24-thread part trying to beat a 32-thread part and the only way it can do this is by pushing the clocks and power envelope well beyond what is reasonable. 350W for a CPU is obscene outside of LN2 overclocking IMO.

What's a more useful comparison is the 12600K with E-cores disabled vs the 5600X. Both use similar amounts of power and operate in their sweet-spot clock/voltage ranges. Intel simply have the IPC advantage with Alder Lake compared to Zen3 right now. Whether that's architecture or DDR5 is unknown at this stage and what I (and a lot of other people) am waiting for is a like-for-like battle using the same speed DDR4.

This type of comparison is incorrect. AMD and Intel have different architectures for one.

When the 6C-12T 5600X Zen3 CPU was released it beat AMDs own previous Zen2 3800X 8C-16T in both CPU and Gaming tests..
Comparing two different Gens and saying one is better because of cores numbering X is not a valid comparison when one is new and one is old.

Lets not forget the 5800X with its 8C-16T laying a smack down on Intel 10C-20T 10900K

Saying Alder lake has 24threads and it should not be compared to a previous Gen 32 thread CPU is not a valid point.

 

[The red spirit]

The press coverage of these new Intel cpu's sounds like Tomshardware "Just buy it" coverage of Nvidia RTX.

To some extent all tech media lies. GoodOldGamer on YT, disclosed that, in his video what it means to be techtuber. That's just how they stay afloat. Some sell out more, some sell out less, but nearly all of them sell out and those that don't have very limited content coverage or don't last long.

 

[lightning70]

The most efficient processor in the 12600k series in terms of power consumption. Frankly, I prefer 12600k over 5600x. This model will be the processor that I intend to buy after this introduction. More efficient than 12900k. It's superior to the 5800x in terms of power, and the power consumption difference at full load isn't excessive. 175W to 190W Cinebech test.

 

[Chrispy_]

Alder Lake is just not great. They have a small edge over AMD, but at horrendous power use and heat output.

That's plainly incorrect! Energy usage for a task directly translates to power use and heat produced:

Stock 12600K = 10KJ
Stock 5600X = 9.6KJ

Stock 12700K = 9.7KJ
Stick 5800X = 10KJ

That seems pretty competitive to me. I've already agreed that the 12900K is pushed too far, but that's the exception to the rule and the other chips in the Alder Lake family and likely the locked SKUs coming later like the 12400F etc will also have competitive power efficiency. You can also make Zen3 use 250W if you force enough voltage through it with an aggressive overclock or sloppy PBO+ setting - that's not representative of the Zen3 lineup as a whole though.

Saying Alder lake has 24threads and it should not be compared to a previous Gen 32 thread CPU is not a valid point.

It's a valid point because it's undisputed fact that a 12900K has 24 threads and a 5950X has 32 threads.
The fact that in many tests a 12900K beats a 5950X despite a thread disadvantage only serves to reinforce my original point that Intel has the IPC advantage with Alder Lake.

Intel have officially priced the 12900K higher than the 12-month old original MSRP of the 5900X and close to the current street price of the 5950X so the comparison is not one I'm making, but one that Intel and the real-world performance results are making.

I'm not even trying to say that the two architectures or generations are the same, because they're obviously not. What kind of idiot would even try and claim that?

 

[The King]

It's a valid point because it's undisputed fact that a 12900K has 24 threads and a 5950X has 32 threads.
The fact that in many tests a 12900K beats a 5950X despite a thread disadvantage only serves to reinforce my original point that Intel has the IPC advantage with Alder Lake.

Intel have officially priced the 12900K higher than the 12-month old original MSRP of the 5900X and close to the current street price of the 5950X so the comparison is not one I'm making, but one that Intel and the real-world performance results are making.

I'm not even trying to say that the two architectures or generations are the same, because they're obviously not. What kind of idiot would even try and claim that?

If you want to get technical the 12900K has 8+8cores total 16cores TDP 125W and the 5950X 16 physical cores. TDP 105W. How each architecture is configured is down to design. HT or no HT.
The higher TDP 16 physical core CPU wins. That's amazing.

 

[Chrispy_]

If you want to get technical the 12900K has 8+8cores total 16cores TDP 125W and the 5950X 16 physical cores. TDP 105W. How each architecture is configured is down to design. HT or no HT.
The higher TDP 16 physical core CPU wins. That's amazing.

I've said in a pre-launch thread that I suspect two E-cores do a better job than one P-core using SMT. I've had plenty of experience with Tremont laptops and Atom servers and those cores are genuinely capable.

The software running has no concept of a physical core, only a logical core (thread), so talking about threads instead of cores does matter, but not as much as the overall performance/Watt and IPC for the CPU as a whole. A new CPU could be a hypothetical 64x E-core monster or a mad single-core CPU with otherworldly IPC and the configuration of that CPU simply wouldn't matter - all that actually matters is how well it performs against competing products on the market and how energy-efficient is it when doing so. The i5 and i7 are competitive with the Ryzens in both performance/Watt and have a performance advantage. The only place where Alder Lake comes into question is when comparing the 12900K to the 5950X instead of the 5900X, since the 5950X is a much bigger step up from the 5800X that the 12700K is compared to.

The 12900K absolutely nails the 5900X to the wall when it comes to performance and it does so at sensible power consumption that is competitive when looking at task energy. The fact that Intel have chosen to overclock and overvolt the snot out of it to chase down the 5950X is purely marketing; Just because they can doesn't mean that they should and anyone buying a 12900K who isn't concerned about those last couple of percent should just reduce the TDP to ~225W and enjoy the massive efficiency gains of not boosting at 350W.

 

[R0H1T]

That seems pretty competitive to me. I've already agreed that the 12900K is pushed too far, but that's the exception to the rule and the other chips in the Alder Lake family

I think you'll find a lot depends on the boards too, the unlimited turbo is good at winning benches but with hotter tropical climates, like for 2 billion+ pop in Asia, or anywhere without AC it'll run into a thermal wall without adequate cooling! I still think it will get better with win11 updates, but as ratirt opined that may push the performance down ~ maybe a lot?

Yes I know this applies to AMD as well.

 

[chrcoluk]

Scheduling issues not unexpected, my experience on Android phones is similar.

Your review stated foreground apps get P cores, background E cores, that explains your mysql observation as mysql server runs in the background non interactive. Does the thread director allow overrides to be configured?

wPrime however, what's going on there?

 

[Agent_D]

Scheduling issues not unexpected, my experience on Android phones is similar.

Your review stated foreground apps get P cores, background E cores, that explains your mysql observation as mysql server runs in the background non interactive. Does the thread director allow overrides to be configured?

wPrime however, what's going on there?

Yea, if you look at the 12900k bench; with e-cores disabled, the SQL stuff jumps hugely in performance. Same with wPrime.

 

[W1zzard]

Scheduling issues not unexpected, my experience on Android phones is similar.

Your review stated foreground apps get P cores, background E cores, that explains your mysql observation as mysql server runs in the background non interactive. Does the thread director allow overrides to be configured?

wPrime however, what's going on there?

Thread director is not configurable, it also has no API. However, you can always manually set affinity for processes in task manager (or through your own code), which overrides whatever Thread Director and Windows Scheduler decide.

If was just "foreground" vs "background", then we wouldn't need Thread Director. Windows can make such a decision on its own. Just the fact that Thread Director exists is clear confirmation that Intel's vision goes beyond that.

 

[ncrs]

Thread director is not configurable, it also has no API. However, you can always manually set affinity for processes in task manager (or through your own code), which overrides whatever Thread Director and Windows Scheduler decide.

If was just "foreground" vs "background", then we wouldn't need Thread Director. Windows can make such a decision on its own. Just the fact that Thread Director exists is clear confirmation that Intel's vision goes beyond that.

Well... it seems that we have a modern Turbo button, according to AnandTech:

[...] There’s an option in the BIOS that, when enabled, means the Scroll Lock can be used to disable/park the E-cores, meaning nothing will be scheduled on them when the Scroll Lock is active. [...]

So at least there seems to be a workaround, but I don't know how common this BIOS option is.

 

[Why_Me]

Bring on the B660 boards.

Best buy incoming: sub-US$200 Intel Core i5-12400F beats the Ryzen 5 5600X

For now, the i5-12600K is the value king, but, if you can wait for the i5-12400F releasing in two months, you could get roughly the same performance for sub-US$200.

www.notebookcheck.net

 

[fevgatos]

You do realize you're quoting efficiency numbers for the CPU in what's generally a GPU heavy task i.e. gaming

I assume you also have the GPU locked at a certain frequency & normalized the results with other variables taken care of?

It's been known, at least to me, for some time now that Intel are extremely efficient in gaming (the 11th gen notwithstanding, those are trash). But for some reason, the internet is full of comments like "only 5% more performance in 1440p while consuming 100% watts", completely neglecting the fact that the cpus dont actually consume that much in gaming

 

[Mussels]

It's been known, at least to me, for some time now that Intel are extremely efficient in gaming (the 11th gen notwithstanding, those are trash). But for some reason, the internet is full of comments like "only 5% more performance in 1440p while consuming 100% watts", completely neglecting the fact that the cpus dont actually consume that much in gaming

It's because some games *do* use all that power.

Rimworld wont max out my 5800x, but DX12 titles sure make a good go at it

 

[Scrizz]

This is an interesting chip for sure.

 

[HD64G]

So, for budget gamers, no difference at all between 12600K and 5600X, or even 12700K and 5800X in performance. Just on power draw and platform's cost. As for the power users or pros, those CPUs are irrelevant.

 

[The King]

So, for budget gamers, no difference at all between 12600K and 5600X, or even 12700K and 5800X in performance. Just on power draw and platform's cost. As for the power users or pros, those CPUs are irrelevant.

Just shows what an awesome gaming CPU AMD made with Zen 3 over a year ago, within +/-95% gaming performance. of ADL at 1080p and the gap will be less @1440p.
I was never impressed with ADL gaming performance to be honest.

@HD64G Please post source article

 

[Ruru]

This looks to be the best one of these when thinking about price/performance and the power consumption.

Though we'll see what AMD has on its sleeve with Zen4, as it's not a surprise that this wins an year old AMD lineup.

 

[lexluthermiester]

So, for budget gamers, no difference at all between 12600K and 5600X, or even 12700K and 5800X in performance. Just on power draw and platform's cost. As for the power users or pros, those CPUs are irrelevant.
View attachment 223964

Please give this a watch.

And no.

 

[The King]

Please give this a watch.

And no.

 

[fevgatos]

So, for budget gamers, no difference at all between 12600K and 5600X, or even 12700K and 5800X in performance. Just on power draw and platform's cost. As for the power users or pros, those CPUs are irrelevant.
View attachment 223964

There is no difference in power draw in game. Stop making shit up

 

[The King]

There is no difference in power draw in game. Stop making shit up

Single threaded testing shows a different story on TPU. Maybe ask @W1zzard to stop making things up.

 

[HenrySomeone]

There is no difference in power draw in game. Stop making shit up

Admins should really start stepping in because the number of posts perpetuating this bullshit (using gaming performance delta with stress test power consumption) is getting out of hand. I can sort of see now how a not-so-well-informed, casual reader can then quickly get the impression that Intels are indeed power hogs in all circumstances.