Intel 12th Gen Core Alder Lake for Desktops: Top SKUs Only, Coming November 4th +Z690 Chipset

[Dabombinable]

6 minutes ago, igormp said:

Linux already has had support for hybrid designs (see ARM CPUs), and intel has already sent many patches related to alder lake that are already in the upstream

So not that long then. Still...the architecture is vastly different to any ARM CPU with the same design principle.

[igormp]

Apparently DDR5 speeds are more disappointing than I thought:

 

 

Quote

If the motherboard has two memory slots total, then the maximum support is DDR5-4800 in any configuration.
If the motherboard has four memory slots total, then the maximum support is DDR5-4400 when two slots are filled with any memory.
If all four memory slots are filled, single rank memory will support up to DDR5-4000.
If all four memory slots are filled, dual-rank memory will support up to DDR5-3600.

 

[ARC1T3CT]

Why Intel????

[RejZoR]

3 minutes ago, ARC1T3CT said:

Why Intel????

So, basically ALL of the bragging of "UP TO 30% BETTER THAN AMD" is pretty much entirely and totally worthless and it'll in the end be like 5%. If at all.

 

Also this whole unboxing thing and other useless stuff almost a week before actual launch is the dumbest s**t ever. No one cares about theoretical would be, might be stuff. We just want to see hard numbers, not this worthless PR fluff.

[Zando_]

4 hours ago, cj09beira said:

so they finally stopped lying about the tdps, but....its still 240W,

that's way too much, seems close to a bulldozer moment if people were consistent, but who are we kidding they wont.

Been running rigs with CPUs that pull that or more, house hasn't melted yet. My 6950X at 4.2GHz seems to chug around 280-300W or more under full AVX allcore, but when gaming (the beefiest workload I actually do) my full system (specs in sig) pulls around 300-350W from the wall according to my kill-a-watt. 

[Forbidden Wafer]

4 minutes ago, RejZoR said:

So, basically ALL of the bragging of "UP TO 30% BETTER THAN AMD" is pretty much entirely and totally worthless and it'll in the end be like 5%. If at all.

Big oof from them, but they've been acting as a scammer for a while.

[cj09beira]

 

14 minutes ago, RejZoR said:

So, basically ALL of the bragging of "UP TO 30% BETTER THAN AMD" is pretty much entirely and totally worthless and it'll in the end be like 5%. If at all.

 

Also this whole unboxing thing and other useless stuff almost a week before actual launch is the dumbest s**t ever. No one cares about theoretical would be, might be stuff. We just want to see hard numbers, not this worthless PR fluff.

With ryan behind it its to be expected 

really the reviewers should refuse to show any manufacturer provided benchmarks.

[porina]

20 minutes ago, RejZoR said:

So, basically ALL of the bragging of "UP TO 30% BETTER THAN AMD" is pretty much entirely and totally worthless and it'll in the end be like 5%. If at all.

The earliest public reference I could find about the Win11 AMD performance impact was 7 Oct., nearly a week after Intel apparently did their testing. AMD themselves stated the Win11 performance at the time resulted in "Estimated performance impact of <3-5% in affected applications, with possible outliers in some games." Intel probably did not know of AMD-MS problem and tested what was available at the time. The MS patch did not exist at the time. The AMD updated chipset drivers did not exist at the time.

[Forbidden Wafer]

10 minutes ago, porina said:

The earliest public reference I could find about the Win11 AMD performance impact was 7 Oct., nearly a week after Intel apparently did their testing. 

And they didn't re-run the tests after the issue became public and got fixed, which doesn't look good for them. 

[leadeater]

1 hour ago, igormp said:

Apparently DDR5 speeds are more disappointing than I thought:

DDR5-4800 is 38.4 GB/s per channel (both 32 bit channels aka kinda per DIMM/64bit). DDR4-3600 is 28.8 GB/s per channel.

[Elisis]

28 minutes ago, Forbidden Wafer said:

And they didn't re-run the tests after the issue became public and got fixed, which doesn't look good for them. 

These things cost both money and time, for decidedly little gain here.

[Forbidden Wafer]

Just now, Elisis said:

These things cost both money and time, for decidedly little gain here.

Well, I'd say their reputation is worth a lot more than running some benchmarks, but apparently their PR is completely stupid.

[Elisis]

1 minute ago, Forbidden Wafer said:

Well, I'd say their reputation is worth a lot more than running some benchmarks, but apparently their PR is completely stupid.

You severely overestimate how much bad rep this will generate.

[porina]

28 minutes ago, Forbidden Wafer said:

And they didn't re-run the tests after the issue became public and got fixed, which doesn't look good for them. 

Do you know they didn't? The L3 cache fix from MS was released as a preview on 21st, less than a week ago. The AMD chipset driver seemed to be released the same day.

 

I don't know if it is the case, but maybe they have done informal testing since, did not see it materially affect the message, and continued with the old results rather than try to shoehorn in new testing.

 

2 minutes ago, Elisis said:

You severely overestimate how much bad rep this will generate.

Many here do not trust 1st party results anyway. If there is a major discrepancy, it will be turned up in reviewer testing which should be ongoing now.

[Dracarris]

1 hour ago, RejZoR said:

Also this whole unboxing thing and other useless stuff almost a week before actual launch is the dumbest s**t ever. No one cares about theoretical would be, might be stuff. We just want to see hard numbers, not this worthless PR fluff.

Yeah seriously what's this BS, all the tech influencer get to show it today, Roman breaks and delids it but nobody is allowed to turn the freakin thing on on camera - WHY??

Are they expecting a significant amount of people to pre-order/buy before actual reviews are out?

[cj09beira]

22 minutes ago, Dracarris said:

Yeah seriously what's this BS, all the tech influencer get to show it today, Roman breaks and delids it but nobody is allowed to turn the freakin thing on on camera - WHY??

Are they expecting a significant amount of people to pre-order/buy before actual reviews are out?

because this way they can have their own "benchmark" results marinating in people's heads, so that by the time they see the reviews they have already decided they like the cpu

[Dracarris]

2 minutes ago, cj09beira said:

because this way they can have their own "benchmark" results marinating in people's heads, so that by the time they see the reviews they have already decided they like the cpu

I really hope not many of their buyers are actually this stupid

[leadeater]

6 minutes ago, cj09beira said:

because this way they can have their own "benchmark" results marinating in people's heads, so that by the time they see the reviews they have already decided they like the cpu

@Dracarris

It also mostly has to do with Youtube reviewers as well, few years (product cycles ago) some of them pushed the boat out and got a bit cheeky and did unboxing videos and showing off their review samples etc before the NDA date for reviews expired. This was just to get views etc etc

 

Since then Intel, AMD, Nvidia etc have basically accepted this as a thing and now there is even an NDA date for unboxings separate from reviews.

 

In the past all we would have gotten today is the news announcement of an official release date and pricing, that's it. If you can pre-order (hell even buy) then that is actually outside what Intel allows and is not an official thing.

[Blademaster91]

4 minutes ago, cj09beira said:

because this way they can have their own "benchmark" results marinating in people's heads, so that by the time they see the reviews they have already decided they like the cpu

Their benchmarks also gets people to pre-order, pre-ordering a CPU is even worse than pre-ordering games IMO.  I don't really see any excuse that Intel couldn't run the tests again with the W11 patches,  and I mean it's kinda ironic too as Intel says benchmarks don't matter.

[leadeater]

2 minutes ago, Blademaster91 said:

Their benchmarks also gets people to pre-order, pre-ordering a CPU is even worse than pre-ordering games IMO.  I don't really see any excuse that Intel couldn't run the tests again with the W11 patches,  and I mean it's kinda ironic too as Intel says benchmarks don't matter.

Intel doesn't do "Pre-Orders", 4th of November is the first "official" time you can purchase the CPUs and when the reviews will go up.

[Blademaster91]

16 minutes ago, leadeater said:

Intel doesn't do "Pre-Orders", 4th of November is the first "official" time you can purchase the CPUs and when the reviews will go up.

Not sure what you mean but pre-orders are up on Amazon at least, and I've seen Z690 boards listed on Newegg.

https://www.techpowerup.com/287990/intel-12th-gen-core-availability-from-november-4-pre-orders-begin-october-27-report

[leadeater]

2 minutes ago, Blademaster91 said:

Not sure what you mean but pre-orders are up on Amazon at least, and I've seen Z690 boards listed on Newegg.

https://www.techpowerup.com/287990/intel-12th-gen-core-availability-from-november-4-pre-orders-begin-october-27-report

Yes 3rd party retailers are selling them, some for quite a few days, this is not sanctioned or technically allowed by Intel right now.

 

Edit:

And you'll note if you do go on to Amazon and search "12900k" you won't get any hits.

[igormp]

1 hour ago, leadeater said:

DDR5-4800 is 38.4 GB/s per channel (both 32 bit channels aka kinda per DIMM/64bit). DDR4-3600 is 28.8 GB/s per channel.

But 4800 is the best case scenario for the spec, Intel ran the sticks @ 4400. I don't think we can expect XMP to go much over that with this first batch of DDR5, not to mention the latencies.

[leadeater]

Here's a list of all the places that list the CPU but none of them allow you to buy it or pre-order aka pay any money. Nonfiction list only.

https://www.amazon.com/dp/B09FXDLX95/?tag=6890-20

https://www.bhphotovideo.com/c/search?Ntt=i9-12900K&N=0&InitialSearch=yes&sts=ma&SID=2322

https://www.newegg.com/intel-core-i9-12900k-core-i9-12th-gen/p/N82E16819118339?nm_mc=AFC-RAN-COM&cm_mmc=AFC-RAN-COM&utm_medium=affiliates&utm_source=afc-GearInStock&AFFID=3456265&AFFNAME=GearInStock&ACRID=1&ASUBID=&ASID=&ranMID=44583&ranEAID=3456265&ranSiteID=10snU7HFqKs-SMG5aih7xz1OvzXD15CGWQ

https://www.bestbuy.com/site/searchpage.jsp?st=i9-12900k&_dyncharset=UTF-8&_dynSessConf=&id=pcat17071&type=page&sc=Global&cp=1&nrp=&sp=&qp=&list=n&af=true&iht=y&usc=All+Categories&ks=960&keys=keys&irclickid=yIs1RxzH0xyIWDAV4GR1JSPoUkBS6eyBexF5TU0&irgwc=1&ref=198&loc=qcy&acampID=0&mpid=60117&intl=nosplash

 

Any place that lets you buy it or pre-order it, yea that's not Intel "allowing" it. Not everyone plays by the rules, not every company goes with the iron fist of destruction like Nvidia does either. So it is what it is.

[leadeater]

3 minutes ago, igormp said:

But 4800 is the best case scenario for the spec, Intel ran the sticks @ 4400. I don't think we can expect XMP to go much over that with this first batch of DDR5, not to mention the latencies.

Sure but it'll still be around the 35 GB/s range. Not saying either are good though in raw bandwidth terms, just a nice to know because if you lookup DDR5 bandwidth you'll often see only 51.2 GB/s mentioned which is for DDR5-6400 aka non existent.