Intel Core i5-12600

[Taraquin]

It's not that easy to compare. To overclock 5600X properly (even if it's overclocking itself), cheaper boards may not be enough (crappy VRM and everything). But yes, when the CPU price and performance are so close together, I'd make the choice based on where I can find the cheaper motherboard that meets my requirements.

Even a basic B450 can OC 5600X with ease, it uses very little power (even at 4.8GHz@1.32v it uses max 115W during load). 5800X and higher is tougher.

 

[AusWolf]

It looks like a nice little gaming CPU, although the 12500 and 12400 seem to offer better value for the money. I didn't expect to start thinking about upgrading my 11700 for the next couple of years, but now I'm actually considering buying something like this just for platform longevity. Only my wallet is shaking its head.

 

[bug]

Even a basic B450 can OC 5600X with ease, it uses very little power (even at 4.8GHz@1.32v it uses max 115W during load). 5800X and higher is tougher.

Oh they'll all overclock, I'm sure of that. But you won't always hit overclock levels you see people bragging about on the Internet.

 

[Valantar]

It looks like a nice little gaming CPU, although the 12500 and 12400 seem to offer better value for the money. I didn't expect to start thinking about upgrading my 11700 for the next couple of years, but now I'm actually considering buying something like this just for platform longevity. Only my wallet is shaking its head.

Question: isn't a pre-emptive upgrade for platform longevity a bit... backwards? If anything, that way you're only denying yourself access to whatever features are new when you actually need an upgrade, no?

 

[Taraquin]

Oh they'll all overclock, I'm sure of that. But you won't always hit overclock levels you see people bragging about on the Internet.

You can do a +200 pbo and co, this will net you 4.85GHz allcore and 4.5-4.65GHz allcore at 76W limit or 4.6-4.75GHz allcore on 88W limit. With manual OC there is very little to gain beyond that, often less. Even a basic A320 board can handle 88W in cinebench

 

[AusWolf]

Question: isn't a pre-emptive upgrade for platform longevity a bit... backwards? If anything, that way you're only denying yourself access to whatever features are new when you actually need an upgrade, no?

Point taken. 11th gen it is, until I truly need an upgrade.

 

[newtekie1]

I don't agree with that. According to current benchmarks, there are no tangible power savings whatsoever. Maybe the scheduler isn't smart enough to make proper use of the E core in light-load scenarios, or maybe it's bugged, but currently the E cores definitely do not lower power draw.

The difference between E-cores only and P-Cores only at idle is about 10w. They do make a difference.

Plus, I'm not aware of regulations targeting CPUs or PCs specifically.

There are currently 6 states in the US that regulate idle power consumption of PCs. And high end gaming PCs are hard to get under those limits with high end graphics cards pre-installed. Every watt counts there.

 

[bug]

The difference between E-cores only and P-Cores only at idle is about 10w. They do make a difference.

No, it isn't: https://www.techpowerup.com/review/intel-core-i5-12600/20.html
Only 2W between 12600 and 12600k. 12600 is clocked a bit lower, if they were clocked the same the difference would probably be 2.5W.

There are currently 6 states in the US that regulate idle power consumption of PCs. And high end gaming PCs are hard to get under those limits with high end graphics cards pre-installed. Every watt counts there.

I very much doubt the targets are that hard to meet (unless you want to put some numbers on that claim), everything is pretty damn good at idling these days.

 

[Valantar]

No, it isn't: https://www.techpowerup.com/review/intel-core-i5-12600/20.html
Only 2W between 12600 and 12600k. 12600 is clocked a bit lower, if they were clocked the same the difference would probably be 2.5W.

I very much doubt the targets are that hard to meet (unless you want to put some numbers on that claim), everything is pretty damn good at idling these days.

Also, remember that these are whole system power measurements. That's easily within margin of error for a measurement like that.

 

[newtekie1]

No, it isn't: https://www.techpowerup.com/review/intel-core-i5-12600/20.html
Only 2W between 12600 and 12600k. 12600 is clocked a bit lower, if they were clocked the same the difference would probably be 2.5W.

Yes, it is. https://tpucdn.com/review/intel-core-i9-12900k-e-cores-only-performance/images/power-idle.png

Even if you overclock the processor, with the E-cores enabled is almost 10w of power savings on the high end chips at idle thanks to the E-cores. And the difference would likely be 1w or so higher if the P-Cores had HT enabled. And yes, these laws are so strict that 1w actually matters.

I very much doubt the targets are that hard to meet (unless you want to put some numbers on that claim), everything is pretty damn good at idling these days.

There are literally manufacturers that had to pull gaming PC models with high end graphics cards from those markets because they couldn't meet the requirements. If you want hard numbers, manufacturers had to pull models from the market that idled at just 66w because it didn't meet the standards. In fact, the cap for idle power is, essentially, 60w for most desktop computers including gaming computers. They use a KwH/yr calculation, but it essentially amounts to if you plug 60w idle into the calculation, it gives you 60KwH/yr, which is the limit. This means the 12900K(and 12700k) test system without E-Cores and with HT turned off, still wouldn't meet the requirements. However, the 12900K system, overclocked, with E-cores still enabled actually does. The 12600 level is right on the edge, and anything lower doesn't really matter so Intel didn't care about stripping the E-cores from those.

 

[Valantar]

Yes, it is. https://tpucdn.com/review/intel-core-i9-12900k-e-cores-only-performance/images/power-idle.png

Even if you overclock the processor, with the E-cores enabled is almost 10w of power savings on the high end chips at idle thanks to the E-cores. And the difference would likely be 1w or so higher if the P-Cores had HT enabled. And yes, these laws are so strict that 1w actually matters.

Considering that the 12600 which has no E-cores idles at exactly the same power, that doesn't hold up to scrutiny. Most likely something happens with the power management of the 12900K when the E cores are disabled, causing it to idle at a higher power level. That kind of stuff is pretty common when changing the configuration of large-scale features like this, after all.

There are literally manufacturers that had to pull gaming PC models with high end graphics cards from those markets because they couldn't meet the requirements. If you want hard numbers, manufacturers had to pull models from the market that idled at just 66w because it didn't meet the standards. In fact, the cap for idle power is, essentially, 60w for most desktop computers including gaming computers. They use a KwH/yr calculation, but it essentially amounts to if you plug 60w idle into the calculation, it gives you 60KwH/yr, which is the limit. This means the 12900K(and 12700k) test system without E-Cores and with HT turned off, still wouldn't meet the requirements. However, the 12900K system, overclocked, with E-cores still enabled actually does. The 12600 level is right on the edge, and anything lower doesn't really matter so Intel didn't care about stripping the E-cores from those.

There was literally one example, which was an Alienware desktop that used an old, inefficient PSU and thus failed to meet these requirements, and would have passed easily if they had used a reasonably modern PSU. Please stop spreading FUD, and I'd recommend watching the GN video on this that I linked previously, as it explains the entirety of the "issue" at length. There is absolutely no way E-cores exist in order to pass idle power requirements, which is proven by the simple fact that CPUs with P-cores only idle at the same power levels.

 

[bug]

Yes, it is. https://tpucdn.com/review/intel-core-i9-12900k-e-cores-only-performance/images/power-idle.png

Even if you overclock the processor, with the E-cores enabled is almost 10w of power savings on the high end chips at idle thanks to the E-cores. And the difference would likely be 1w or so higher if the P-Cores had HT enabled. And yes, these laws are so strict that 1w actually matters.

I'm not sure what i should be looking at in that picture.
I have already sent you a link showing 12600 idling at 57W vs 12600k idling at 55W. I really don't know where you came up with 10W from.

As for manufacturers not being able to meet the 60W idle power, that's more about systems going crazy. I mean W1zzard's system seems to have no problem meeting that and it's pretty extreme with 32GB RAM, one one the most featureful motherboard, high-end CPU and GPU, water cooling at whatnot.
60W idle power... When I started building PCs, our only choice was between a 200W or a 240/250W PSU.

 

[Valantar]

I'm not sure what i should be looking at in that picture.
I have already sent you a link showing 12600 idling at 57W vs 12600k idling at 55W. I really don't know where you came up with 10W from.

As for manufacturers not being able to meet the 60W idle power, that's more about systems going crazy. I mean W1zzard's system seems to have no problem meeting that and it's pretty extreme with 32GB RAM, one one the most featureful motherboard, high-end CPU and GPU, water cooling at whatnot.
60W idle power... When I started building PCs, our only choice was between a 200W or a 240/250W PSU.

Not to mention that those tests are run with a 1200W PSU that ... well, doesn't perform admirably at idle (that's the 1000W version - the 1200W version is likely even less efficient at low loads).

That graph tells us that its efficiency in the ~50W output range (either 12V, minor rails, or a mix) is in the 70-75% efficiency range. As this is for DC output wattage, for a 57W AC idle reading, assuming a best-case 75% efficiency, that's 57/100*75=42.75W DC power, at that point - but more importantly, a quarter of the power wasted as PSU losses. For comparison, the ATX 3.0 PSU standard requires 60% efficiency from 10W or 2% output power, and recommends 70% at that level, where this PSU scores less than 60% in the linked review above. And as no PSU has a flat efficiency curve at its extreme low end, efficiency would still rise rapidly from this 10W point. In other words, these test setups are compliant, and are using older PSU designs than the most recent standard, which will further lower idle power draw through lowered PSU losses. If that same 42.75W DC draw saw 85% efficiency rather than 75, the AC load would be just 50W, or 53.4W at 80%. These clearly aren't massive differences, but are more than enough to ensure that a PC like this is perfectly compliant for years to come.

I don't think it's possible to go much further than this without starting to power gate entire AICs or onboard controllers on idle, which would be .... well, troublesome in practice. But these levels are perfectly attainable with even high end hardware today.

 

[InVasMani]

1500w PSU it's not much, but it's all mine...proceeds to use like 10% to 20% of rated wattage on average.

 

[Valent117]

just found a cheap used 9900k setup for my cousin
i didn't realize how much the 12th gen slam on previous gen, this is insane