Intel Core Ultra 9 285k 'Arrow Lake' Discussion/News ("15th gen") on LGA-1851

[RSR]

It isn't an [totally] unsalvageable situation - I'd assume the issues like compatibility and crashing can be fixed, but they'll have to significantly discount the CPUs to where they are realistically with 1-2 generation old performance and probably dig up another architecture (like some kind of "Meteor Lake++") to flesh out the high end, the platform seems to fully support stuff like Hyper-Threading despite it not being present on Arrow Lake.

What compatibility and crashing issues are these? If you mean the crashing when installing Windows due to the active internal GPU that's been fixed for ages. I've been using my Apex / 285K, and it's been running fine.

I have been comparing it against my 14900KS and 9950X on a daily base, so using it for work and using it for everything else. In terms of power, power pulled from the wall its been pretty impressive as it averaged around ~114w (285K) over the day vs ~145w for the KS. I don't have a whole day's worth of data yet from the 9950X as it has decided to nuke itself. I've ordered a new one which should arrive at some point next week.

 

[Locky]

You're different. You buy everything and try and kill it..

 

[Rroff]

What compatibility and crashing issues are these? If you mean the crashing when installing Windows due to the active internal GPU that's been fixed for ages. I've been using my Apex / 285K, and it's been running fine.

I have been comparing it against my 14900KS and 9950X on a daily base, so using it for work and using it for everything else. In terms of power, power pulled from the wall its been pretty impressive as it averaged around ~114w (285K) over the day vs ~145w for the KS. I don't have a whole day's worth of data yet from the 9950X as it has decided to nuke itself. I've ordered a new one which should arrive at some point next week.

Some people have had crashing or compatibility problems with slightly older production software, some have had intermittent crashing with certain games which doesn't seem to be a configuration issue, etc.

It often struggles to beat the 14700K which uses less power than the 14900 never mind the performance difference to the 14900KS.

 

[RSR]

You're different. You buy everything and try and kill it..

The irony of that statement is my 14900KS has outlasted my 9950X.

 

[RSR]

Some people have had crashing or compatibility problems with slightly older production software, some have had intermittent crashing with certain games which doesn't seem to be a configuration issue, etc.

It often struggles to beat the 14700K which uses less power than the 14900 never mind the performance difference to the 14900KS.

If you can give me specific examples I can go and have a look.

I’ve had no issues with running Adobe PS/LR, Citrix and VMware etc. At the moment it’s just running in stock form with CKD DIMMs.

Yes, I am aware the gaming side is hit and miss. As I am still going through COD / 2042 and CP which my main go to choice.

 

[Vimes]

What compatibility and crashing issues are these? If you mean the crashing when installing Windows due to the active internal GPU that's been fixed for ages. I've been using my Apex / 285K, and it's been running fine.

I have been comparing it against my 14900KS and 9950X on a daily base, so using it for work and using it for everything else. In terms of power, power pulled from the wall its been pretty impressive as it averaged around ~114w (285K) over the day vs ~145w for the KS. I don't have a whole day's worth of data yet from the 9950X as it has decided to nuke itself. I've ordered a new one which should arrive at some point next week.

Apart from being able to test your AMD system with another CPU, to confirm that the CPU has died, how else would you be able to test that if you did not have something to swap it out with..?
I can't ever recall having a CPU die on me in the decades of building PC's. But I have, not often at all, noted a post or three about AMD CPU's dying.
Is the swap out covered directly by AMD in a reasonable time frame..?

Thanks

 

[LtMatt]

Apart from being able to test your AMD system with another CPU, to confirm that the CPU has died, how else would you be able to test that if you did not have something to swap it out with..?
I can't ever recall having a CPU die on me in the decades of building PC's. But I have, not often at all, noted a post or three about AMD CPU's dying.
Is the swap out covered directly by AMD in a reasonable time frame..?

Thanks

His 9950X was delidded so that may have been a factor. That said and touch wood, I’ve never had a CPU die on me yet either.

 

[RSR]

Apart from being able to test your AMD system with another CPU, to confirm that the CPU has died, how else would you be able to test that if you did not have something to swap it out with..?
I can't ever recall having a CPU die on me in the decades of building PC's. But I have, not often at all, noted a post or three about AMD CPU's dying.
Is the swap out covered directly by AMD in a reasonable time frame..?

Thanks

I have a spare 8500G, which I have tested the X870 Hero with and it works fine.

The bench has been on the side for a few weeks, with a bios default running as I don't leave profiles on them in case of swapping hardware in and out. So between now and then the 9950X has given up the ghost which seeming looks like the IO chip has gone.

Yes, it was delidded but it's been running fine delidded since I first got it around September with no issues till now.

 

[RSR]

His 9950X was delidded so that may have been a factor.

I'm not convinced by that tbh.

As it has been working fine in a delidded form since September and since the TG Block has been mounted its not been touched. The only major things its done lately is update the bios on the board and take it off the side to use it.

I've looked at the die and there are no cracks on the die.

 

[RSR]

Also to add my 14900KS is also delidded and that's been motoring along fine.

 

[Vimes]

I'm not convinced by that tbh.

As it has been working fine in a delidded form since September and since the TG Block has been mounted its not been touched. The only major things its done lately is update the bios on the board and take it off the side to use it.

I've looked at the die and there are no cracks on the die.

I suppose, at the least, delidding it has voided the warranty on the CPU...? Tough break there, the situation and the chip..!

Have you made it into a keyring pendant for a Crimbo gift..?

 

[RSR]

I suppose, at the least, delidding it has voided the warranty on the CPU...? Tough break there, the situation and the chip..!

Have you made it into a keyring pendant for a Crimbo gift..?

Yes, it does. It is what it is, unfortunately. Na, it will just get lobbed in the bin. However, I've ordered a replacement so not all is lost.

I'd have been more impressed if I had killed it in some way tbh, but no I get some random failure on it.

 

[Vimes]

Yes, it does. It is what it is, unfortunately. Na, it will just get lobbed in the bin. However, I've ordered a replacement so not all is lost.

I'd have been more impressed if I had killed it in some way tbh, but no I get some random failure on it.

I did not know if it was similar to those who replace the std cooler on a GPU with a water block, and then be able to return to the std, even replacing warranty stickers at times, if there needs to be a warranty claim.
A pity that if there is absolutely no physical damage from the die to be seen then it could not be reattached, the lid, and claimed.

Still, it seems you missed out on a niche Crimbo gift. I would have liked that. Probably the closest I would have got to such a chip, working or not.

 

[LtMatt]

Yes, it does. It is what it is, unfortunately. Na, it will just get lobbed in the bin. However, I've ordered a replacement so not all is lost.

I'd have been more impressed if I had killed it in some way tbh, but no I get some random failure on it.

Dom's had a couple die like yours, his were delidded too. With multiple packages under the processor lid, I think they can be quite sensitive once the hood is ripped off and then become more prone to failure.

 

[RSR]

Dom's had a couple die like yours, his were delidded too. With multiple packages under the processor lid, I think they can be quite sensitive once the hood is ripped off and then become more prone to failure.

Yet my 7950X3D and 7800X3D were delidded and ran fine for ages.

 

[LtMatt]

Yet my 7950X3D and 7800X3D were delidded and ran fine for ages.

You only have to compare Zen 4/5 die shots to see they are not identical under the hood, aside from using a different process node.

 

[RSR]

You only have to compare Zen 4/5 die shots to see they are not identical under the hood, aside from using a different process node.

I was replying in terms of "With multiple packages under the processor lid, I think they can be quite sensitive once the hood is ripped off and then become more prone to failure."

We know they are different dies, as they would be called the same product.

However, what you are suggesting is the new node is more sensitive than the old one.

 

[LtMatt]

I was replying in terms of "With multiple packages under the processor lid, I think they can be quite sensitive once the hood is ripped off and then become more prone to failure."

We know they are different dies, as they would be called the same product.

However, what you are suggesting is the new node is more sensitive than the old one.

That’s exactly what I’m suggesting, based purely off the failures reported by users that have delidded them. Think Dom’s had 2x9950X and a 9800X3D fail in similar fashion to yours once delidded. Curious to see how your replacement goes down the line.

 

[RSR]

That’s exactly what I’m suggesting, based purely off the failures reported by users that have delidded them. Think Dom’s had 2x9950X and a 9800X3D fail in similar fashion to yours once delidded. Curious to see how your replacement goes down the line.

Did he not crack the die one of those?

 

[LtMatt]

Did he not crack the die one of those?

Not the one I saw. Couple of other users on OCN had similar too.