Intel Core i5-12400 Engineering Samples Surface on eBay, Riddled with Compatibility Issues

[btarunr]

A rather big chunk of Intel Core i5-12400 "Alder Lake" engineering samples (ES) hit eBay. The i5-12400 is an upcoming mid-range processor Intel is planning to release in Q1-2022. The 6-core/12-thread processor only features 6 "Golden Cove" performance cores, and lacks Efficiency cores (which is probably a good thing for gamers). Each of the six cores has 1.25 MB of L2 cache, while they share an 18 MB L3 cache.

VideoCardz warns that the ES chips out in the wild could be riddled with compatibility issues with Z690 motherboards that are in the market. Apparently, there are two revisions of i5-12400 unreleased doing rounds, C0 and B0, with the former being a QS or qualification sample, and the latter an ES or engineering sample. The two differ in maximum boost frequency—4.40 GHz vs. 4.00 GHz. They also differ with S-SPEC codes of QXDY and QYHX. Even with production versions of firmware and Intel ME, retail Z690 motherboards don't seem to guarantee compatibility with these samples. You are probably better off waiting for retail versions of these chips.



View at TechPowerUp Main Site

 

[Dristun]

Don't see how having less cores is good for gamers - nobody I know turns all other apps off when they start playing as if they're going to bench. I mean, sure, if AMD suddenly starts offering 10 big cores for the price of 12600K it's another story but we're not there yet. (I hope we do get there though!)
On that note - would be an interesting test actually: running popular games while having FF with 10 tabs, discord stream and iTunes on for example and see if the scheduler does its job then. Haven't seen anyone properly test whether it really understands what's in focus which imho is far more interesting than seeing it juggle cores for one task for absolute maximum perf.

 

[Fouquin]

That's a normal occurrence with ES chips. It's why any listings for ES/QS chips have a list of tested motherboards and their BIOS versions. This has been the case for the last decade at least, with a few examples going back to early Nehalem chips.

 

[ExcuseMeWtf]

Don't see how having less cores is good for gamers - nobody I know turns all other apps off when they start playing as if they're going to bench. I mean, sure, if AMD suddenly starts offering 10 big cores for the price of 12600K it's another story but we're not there yet. (I hope we do get there though!)
On that note - would be an interesting test actually: running popular games while having FF with 10 tabs, discord stream and iTunes on for example and see if the scheduler does its job then. Haven't seen anyone properly test whether it really understands what's in focus which imho is far more interesting than seeing it juggle cores for one task for absolute maximum perf.

Not having E-cores specifically might be a good thing for gamers, due to DRM compatibility issues.

 

[bug]

Don't see how having less cores is good for gamers - nobody I know turns all other apps off when they start playing as if they're going to bench. I mean, sure, if AMD suddenly starts offering 10 big cores for the price of 12600K it's another story but we're not there yet. (I hope we do get there though!)

Games aren't heavily multithreaded, eliminating the chance for a game thread to end up on an E core is good for gamers.

On that note - would be an interesting test actually: running popular games while having FF with 10 tabs, discord stream and iTunes on for example and see if the scheduler does its job then. Haven't seen anyone properly test whether it really understands what's in focus which imho is far more interesting than seeing it juggle cores for one task for absolute maximum perf.

As proven by benchmarks, even when dealing with a single type of task, the scheduler still does the wrong thing. That's a pretty good indictor of what will happen to more heterogeneous loads. It will be dealt with in time, schedulers will be patched. I wouldn't be surprised if some problems can be patched by apps themselves, by using the proper thread priorities (though it's not always that simple).

 

[Operandi]

That second one really dosn't sound too sure about it self does it?

 

[N3utro]

A full batch of ES getting leaked? Someone at Intel or one of their partners is getting fired.

 

[bug]

A full batch of ES getting leaked? Someone at Intel or one of their partners is getting fired.

An ES's life usually ends in its destruction. That's most likely where they "leak". At this point, probably nobody cares about it. The only possible damage is them being sold to unsuspecting buyers as retail chips.

 

[freeagent]

I had a few ES chips that were good enough to be retail, or better.. too bad these are duds.. maybe a more mature bios would help? Probably not though..

 

[newtekie1]

I had a few ES chips that were good enough to be retail, or better.. too bad these are duds.. maybe a more mature bios would help? Probably not though..

I think Linus just did a video one of these ES chips. It turns out the IME was updated between the first batch of ES chips and the QS/Retail chips. This made newer BIOSes incompatible with the older ES chips. So it's not likely we'll see a BIOS update fix this.

 

[bug]

I had a few ES chips that were good enough to be retail, or better.. too bad these are duds.. maybe a more mature bios would help? Probably not though..

Last ES batches will be like that, but earlier batches might not. It is only logical