Intel Core i5-12400 Early Review Dubs it a Game Changer

[btarunr]

The upcoming Intel Core i5-12400 processor could be a game changer in the mid-range, according to an early gaming performance review by Igor's Lab, which landed simulated the chip by disabling the E-cores, and setting the right clock speeds and power values. Based on the smaller H0 silicon of "Alder Lake-S," which physically only features six "Golden Cove" CPU cores, and no "Gracemont" E-core clusters, the i5-12400 ticks at 2.50 GHz, and 4.40 GHz boost frequency, with 65 W base power, and 117 W maximum turbo power (MTP).

Testing reveals that this MTP value lends the processor some stellar energy-efficiency numbers, and the chip strikes a performance/Watt sweetspot. Igor's Lab, however, recommends that for the best efficiency, the i5-12400 should be paired with DDR4 memory. In its testing, DDR4-3733 (with Gear 1) was used. Gaming benchmarks put out by Igor's Lab shows that the Core i5-12400 trades blows with the AMD Ryzen 5 5600X "Zen 3" in a number of games, beating it in several of them by virtue of higher IPC of the "Golden Cove" cores, and beating the i7-11700K "Rocket Lake" 8-core/16-thread processor at a fraction of its power-draw. A word of caution, though, is that the i5-12400 was simulated on a C0 silicon, possibly the i9-12900K, and the real i5-12400 die may not have the same refinements or electrical characteristics. Even with the E-core cluster disabled, the L3 cache size isn't the same (30 MB vs. 18 MB). Catch the review in the source link below.



View at TechPowerUp Main Site

 

[Guwapo77]

"A word of caution, though, is that the i5-12400 was simulated on a C0 silicon, possibly the i9-12900K, and the real i5-12400 die may not have the same refinements or electrical characteristics."

In other words, take these benchmarks with a grain (bucket) of salt... What is the purpose of dropping benchmarks like these when the consumer product doesn't reflect the stellar results posted above?

 

[Melvis]

I wouldnt call it a "Game changer" but more of a competitive product that trades blows with the 1yr old 5600X. Coool?

 

[Crackong]

Finally free of error cores
Will consider getting a 12400 for family PC

 

[piloponth]

Not exactly sure if 6core is future-proof enough. I would like to see a 8core Alder Lake with no Ecores released.

 

[Garrus]

I wouldnt call it a "Game changer" but more of a competitive product that trades blows with the 1yr old 5600X. Coool?

Exactly. Intel was so far behind here. So far all Alder Lake means is AMD has to lower prices, they don't even need new products honestly except they want to keep their average selling price high. Motherboards are typical Intel double pricing versus B550, a joke. I like my Alder Lake system but it was not cost effective and the entire point of Alder Lake is the e cores, which you get ZERO with the 12400.

 

[Nesters]

Correct term is "price changer" not "game changer". This is Intel catching up.

 

[Mussels]

After 11th gen, they had to go up

Igors simulation might be "best case" figures here, with reality falling short

 

[Jism]

Exactly. Intel was so far behind here. So far all Alder Lake means is AMD has to lower prices, they don't even need new products honestly except they want to keep their average selling price high. Motherboards are typical Intel double pricing versus B550, a joke. I like my Alder Lake system but it was not cost effective and the entire point of Alder Lake is the e cores, which you get ZERO with the 12400.

The chipset only Intel deliveres is obviously another milking cow for Intel. Refresh it with each generation and your looking at a steady profit.

AMD on the other hand pretty much houses everything inside the CPU itself. All you need is a compatible AM4 board and your good to go.

 

[Arcdar]

So if they bring out a 12400 with a GOOD silicon and 30mb cache it would be a great mid-range-cpu? Awesome.

Sadly, that will never happen as if the silicon would be that great intel wouldn't use it on that cpu and will not give mid-range cpu's "that much cache" :/

 

[Raendor]

12400, just like the rest of 12th gen, has higher ipc than 11th gen, thus some minor advantage over 11700 in some games. The latter though has 8 cores, which is in turn more beneficial for other games and productivity. Plus the benchmark is done at 720p and 1080p. All in all, even 12700 vs 11700 doesn't show any significant advantage in gaming. CPUs are not that exciting in terms of gen to gen uplifts unlike GPUs, and it's been the case for last 10 years at least.

 

[Selaya]

The chipset only Intel deliveres is obviously another milking cow for Intel. Refresh it with each generation and your looking at a steady profit.

AMD on the other hand pretty much houses everything inside the CPU itself. All you need is a compatible AM4 board and your good to go.

except ... if you were an early adopter w/ X370 I guess ... sucks to be you! Unless you're into BIOS modding, or something. Then again, a Z170 can be modded to work w/ a 9900K too so ...

 

[Vayra86]

What the hell kind of nonsensical news post is this?

Clickbait crap, that's what. Simulated on a different CPU, because we were dying to read this 'early review' that isn't actually a review?
Game changer, because it reaches margin of error differences with the rest of the midrange>top end of AMD/Intel CPU stacks? At a 'mere' 117W?! Because somehow that's a great number for CPUs now? Lmao

Its a ripple in the pond of lakes, at best. Its this the new hotness now, these early reviews? Can we kill it with fire tyvm? Launch day reviews are bad enough as it is, more often than not lacking critical info because everyone's rushing to be first. Previews are worse - they're nothing other than advertorials nobody pays for (I hope...).

 

[igorsLAB]

Is it really that hard to read the linked articles before writing such nonsense?

Apart from the fact that the emulation was done with a normal retail CPU (not an Intel sample) in the form of the i5-12600K, I also randomly tested the results with a qualified sample (QYHX) of the i5-12400 for plausibility and it was within the tolerance range of normal CPUs. However, I did not test the original for collegial reasons, even though I did not sign any NDAs. The BIOS settings up to the performance values are 1:1 the same as those of the non-K CPU.

First lines of this review:

Today we would like to anticipate one of the upcoming launches in January and test the really interesting Intel Core i5-12400 in gaming (for now). Of course, there will be a follow-up with real application benchmarks soon. Even though I’m not under NDA, I’ll be fair and deliberately use an emulated CPU and none of the available engineering or qualified samples. However, I took the liberty of testing some of my results for plausibility with a QYHX to make sure I got everything to fit.

A bit later:

Even if the Core i5-12400, which will be available in January, might not look quite so fancy in the CPU lottery – for today’s test I have set a current Core i5-12600K in terms of base clock (2.5 GHz) and turbo clock (4.4 GHz, 4.2 GHz as well as 4.0 GHz allcore) including the voltages and power limits (PL1 65 watts, MTP 117 watts) exactly as the specs, the BIOS and XTU say for the real non-K processor, but not set the PL1 to the PL2. That would NOT have had any impact in the end, I tested that too of course. Just like higher PL values in the professional applications. But let us surprise you today, and show you what is still possible.


The originally not unlocked E-Cores could be completely deactivated in the BIOS also with the Core i5-12600K, so that I could test a perfect (emulated) copy of a not yet launched i5-12400 here. The fact that the cache is minimally smaller in the QYHX has virtually no relevance in practice. The benchmark results of my DDR4 setting match those of the QYHX within tolerance almost shockingly perfectly, but they are still theoretically under NDA, at least for most. So today there are “only” the values of my blueprint. But very detailed in three resolutions with the usual metrics. Actually, any other reviewer could have done that, but they didn’t.

This isn't clickbait, but a shitload of work. Before Intel pays me anything, hell freezes over

 

[Vayra86]

Is it really that hard to read the linked articles before writing such nonsense?

Apart from the fact that the emulation was done with a normal retail CPU (not an Intel sample) in the form of the i5-12600K, I also randomly tested the results with a qualified sample (QYHX) of the i5-12400 for plausibility and it was within the tolerance range of normal CPUs. However, I did not test the original for collegial reasons, even though I did not sign any NDAs. The BIOS settings up to the performance values are 1:1 the same as those of the non-K CPU.

First lines of this review:

A bit later:


This isn't clickbait, but a shitload of work. Before Intel pays me anything, hell freezes over

I'm still struggling to see the value of pre-empting a real review. And struggling with the headline of being somehow a game changer, this CPU.

I didn't read the review because I like to read about the actual product, post release. I've learned a few things over the years regarding that constant need and want for attention to generate ad revenue. As far as the swearing-infested post in between... yeah. That social media-drive is exactly the core of that issue. Its the reason we think somehow an earlier review is somehow a good thing. I've killed all social media accounts over six years ago, its the cesspool of the internet.

Sorry for being all difficult and I appreciate your hard work, but really, this is just yet another CPU.

 

[Garrus]

Is it really that hard to read the linked articles before writing such nonsense?

Apart from the fact that the emulation was done with a normal retail CPU (not an Intel sample) in the form of the i5-12600K, I also randomly tested the results with a qualified sample (QYHX) of the i5-12400 for plausibility and it was within the tolerance range of normal CPUs. However, I did not test the original for collegial reasons, even though I did not sign any NDAs. The BIOS settings up to the performance values are 1:1 the same as those of the non-K CPU.

First lines of this review:

A bit later:


This isn't clickbait, but a shitload of work. Before Intel pays me anything, hell freezes over

First image in this article: 1.6 percent faster versus the 5600x. At 720p.

That's not a game changer, that's all. Even the fact that it uses 27 percent less power isn't really much of a game changer either. This is no Apple device. I appreciate the hard work, but I don't think the tone is warranted. It sounds like "this is the best thing ever" to beat a 15 month old CPU (at release date) by 1.6 percent.

 

[Selaya]

Upcoming game changer in the mid-range? Intel Core i5-12400 Review and Gaming Benchmarks – more efficient without E-Cores and with DDR4 instead

igor's original headline, for reference

turns out omitting a single ? can sensationalise a headline a lot

 

[Vayra86]

Upcoming game changer in the mid-range? Intel Core i5-12400 Review and Gaming Benchmarks – more efficient without E-Cores and with DDR4 instead

igor's original headline, for reference

turns out omitting a single ? can sensationalise a headline a lot

Take note of the fact I said I didn't like the TPU article as it is a repost of an early review from another site, and the article on TPU is clickbaity. Exposure exposure exposure. Content or accuracy is secondary.

 

[Selaya]

classic case of lost in translation i guess
but yeah, these clickbaits w/o even linking to the original article are ... uhm ...

 

[Garrus]

Upcoming game changer in the mid-range? Intel Core i5-12400 Review and Gaming Benchmarks – more efficient without E-Cores and with DDR4 instead

igor's original headline, for reference

turns out omitting a single ? can sensationalise a headline a lot

True the question mark is pretty important in meaning. Still I believe Igor was just trying to make it entertaining, but I don't know, was he serious? From the article:

"The values are, despite the still so high PL1, so low compared to the Ryzen CPUs that I forgot to close the jaw dropped in amazement again. And yet it is reality (...) Whoever believed that they would experience their blue miracle with Alder Lake S here"

Blue miracle? My jaw dropped in amazement? Really?

Just to be clear I think the i7-12700k is the best CPU around right now, as AMD has not lowered their prices. But that's all. A Ryzen 5900X is good competition, and a 5600X is good competition to the 12400. Not a game changer imo. Cheers everyone, it's no big deal

 

[Selaya]

another case of lost in translation, if you were to read the german version of it (if you can understand it), it all makes a lot more sense

 

[Vayra86]

See I just looked at those charts there and mirrored them off the headline, and that was all that was required. Pictures > 1000 words... I really don't need to read yet another guy's overinflated text about the umpteenth iteration of a Core CPU. Heck its the main reason I love TPU's review style. No nonsense. Numbers. Data. A conclusion.

Ripples in the pond at best. The real game changer is the competition between AMD and Intel is on, which means cheaper CPUs hopefully for us. And if not, the market is even more broken than we thought. I know, it sounds boring. But sometimes its nice to just be realistic, instead of an extension of hypercapitalist marketing machines fighting for attention.

 

[lexluthermiester]

Not exactly sure if 6core is future-proof enough. I would like to see a 8core Alder Lake with no Ecores released.

For some perspective, I still run a 10 year old Xeon W3680(Socket1366 6c12t) based system as a general use & guest gaming system and it has no problems running whatever I throw at it at 1080p and will continue to be a common use system for several more years I'm sure. This new i5-12400 easily triples the performance of my W3680.

For anyone to say that a i5-12400 will not provide years worth of solid performance is just not in touch with computing in the real world.

 

[b1k3rdude]

As this non-existent chip was simulated on 12900, this is a non-review. And as f***ing usual, all the reviews out there including this one are all based of the same BAD data. The title of this new article should be "Simulated review...".

 

[Leshy]

Exactly. Intel was so far behind here. So far all Alder Lake means is AMD has to lower prices, they don't even need new products honestly except they want to keep their average selling price high. Motherboards are typical Intel double pricing versus B550, a joke. I like my Alder Lake system but it was not cost effective and the entire point of Alder Lake is the e cores, which you get ZERO with the 12400.

How it is possible to compare highend chipset and mid? OFC AL wont be cost effective (its K), if buy enthusiast grade components. If 12400 ll cost ~ 200€ and it ll be 10% faster as AMD counterpart for 320, its a great value. And hopefully there ll be good enought motherboard for this CPU around 70-90€.

As gamer i dont care if E cores are there or not. If ST performance is better (and it is 15-20% compared with 11400). I wan it.