Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I am giving up on Intel and have bought an AMD Ryzen 9950X3D

michael.stapelberg.ch

Post reply on HN
Print view View original

341–350 of 370 posts

Re: I am giving up on Intel and have bought an AMD Ryzen 9950X3D

#341

Post by jonbiggums22 »

IAmGraydon wrote: post #265
microtonal wrote: post #2

I feel like both Intel and AMD are not doing great in the desktop CPU stability department. I made a machine with a Ryzen 9900X a while back and it had the issue that it would freeze when idling. A few years before I had a 5950X that would regularly crash under load (luckily it was a prebuilt, so it was ultimately fixed). When you do not have a bunch of components ready to swap out it is also really hard to debug the…

9900x here and zero crashes since I built it 9 months ago. A lot of the stability comes down to choosing the right RAM with the right timing for Ryzen CPUs.

I haven't moved on from AM4 yet but the way XMP is advertised you'd think it was guaranteed instead of overclocking that technically voids your warranty.

Re: I am giving up on Intel and have bought an AMD Ryzen 9950X3D

#342

Post by coldtea »

cma wrote: post #307
coldtea wrote: post #267

Earlier quoted context omitted.

You take backups and do checksums for the valuable data. For the rest, who cares...

Let's say you corrupted one bit in a blender asset 200 revisions ago and it was unnoticeable and still upgraded through five blender upgrades, but now on the sixth upgrade it fails with a corruption error and doesn't upgrade. Without knowing how to fix that error you've lost 200 revisions of work. You can go back and find which revision had the problem, go before that, and upgrade it to the latest blender, but all yo…

>Let's say you corrupted one bit in a blender asset 200 revisions ago and it was unnoticeable and still upgraded through five blender upgrades, but now on the sixth upgrade it fails with a corruption error and doesn't upgrade.

And let's say you have archived copies of it with checksums like I suggested, going back to all revisions ago.

What's the issue again now, that ECC would have solved? Not to mention that ECC wouldn't help at all with corruption at the disk level anyway.

Re: I am giving up on Intel and have bought an AMD Ryzen 9950X3D

#343

Post by enronmusk »

secure wrote: post #336
enronmusk wrote: post #119

If OP's CPU cooler (Noctua NH-D15 G2) wasn't able to cool down his CPU below 100C, he must have been (intentionally or unintentionally with Asus multi core enhancement) overclocked his CPU. Or he didn't apply thermal paste properly or didn't remove the cooler plastic sticker? I have followed his blog for years and hold him in high respect so I am surprised he has done that and expected stability at 100C regardless of…

> If OP's CPU cooler (Noctua NH-D15 G2) wasn't able to cool down his CPU below 100C, he must have been (intentionally or unintentionally with Asus multi core enhancement) overclocked his CPU. Or he didn't apply thermal paste properly or didn't remove the cooler plastic sticker? I did not overclock this CPU. I pay attention to what I change in the BIOS/UEFI firmware, and I never select any overclocking options. Also,…

Thank you for responding. How do you explain your CPU hitting 100C in that case? That should not have happened.

https://www.techpowerup.com/review/intel-core-ultra-9-285k/2... lists maximum temperature as 88.2C with the previous gen NH-D15 cooler.

Re: I am giving up on Intel and have bought an AMD Ryzen 9950X3D

#344

Post by jonbiggums22 »

itvision wrote: post #304
rkrisztian wrote: post #296

Earlier quoted context omitted.

It's the 3D cache, as I wrote in my other response. It has to be powered on at all times, so it affects even the idle power usage.

It has little to nothing to do with the 3D cache. The IOD (die) is extremely inefficient for all desktop Zen CPUs as it never truly idles.

Their APUs don't have the problem from the reviews I've seen, but yes the I/O die has been the bane of the Zen platform when it comes to idle power consumption.

To make matters worse, the x570 chipset basically runs this I/O die upside down as a chipset and sucks twice as much power at idle as the x470 chipset it replaced. I expected them to replace this hack of a product used for the high end when Asmedia's efforts were delayed but all that platform got was B550. It was pretty clear they weren't chasing this part of the market during AM4's heydey, no real idea where they are at now with chipsets on AM5. But given few people talked about how crappy that chipset was in this respect I guess they might be right it wasn't important to most people.

Re: I am giving up on Intel and have bought an AMD Ryzen 9950X3D

#345

Post by fluoridation »

cma wrote: post #307
coldtea wrote: post #267

Earlier quoted context omitted.

You take backups and do checksums for the valuable data. For the rest, who cares...

Let's say you corrupted one bit in a blender asset 200 revisions ago and it was unnoticeable and still upgraded through five blender upgrades, but now on the sixth upgrade it fails with a corruption error and doesn't upgrade. Without knowing how to fix that error you've lost 200 revisions of work. You can go back and find which revision had the problem, go before that, and upgrade it to the latest blender, but all yo…

So don't upgrade it. Export it to an agnostic format and re-import it in the new version. Since it's failing to upgrade, it must be a metadata issue, not a data issue, so removing the Blender-specific bits will fix it.

What a silly hypothetical. There's a myriad freak occurrences that could make you have to redo work that you don't worry about. Now, I'm not saying single-bit errors don't happen. They just typically don't result in the sort of cascading failure you're describing.

Re: I am giving up on Intel and have bought an AMD Ryzen 9950X3D

#346

Post by fluoridation »

coldtea wrote: post #342
cma wrote: post #307

Earlier quoted context omitted.

Let's say you corrupted one bit in a blender asset 200 revisions ago and it was unnoticeable and still upgraded through five blender upgrades, but now on the sixth upgrade it fails with a corruption error and doesn't upgrade. Without knowing how to fix that error you've lost 200 revisions of work. You can go back and find which revision had the problem, go before that, and upgrade it to the latest blender, but all yo…

> Let's say you corrupted one bit in a blender asset 200 revisions ago and it was unnoticeable and still upgraded through five blender upgrades, but now on the sixth upgrade it fails with a corruption error and doesn't upgrade. And let's say you have archived copies of it with checksums like I suggested, going back to all revisions ago. What's the issue again now, that ECC would have solved? Not to mention that ECC w…

>What's the issue again now, that ECC would have solved?

If the bit flip happened in RAM, the checksum would be of the corrupted data. ECC corrects single bit errors of data on RAM.

>Not to mention that ECC wouldn't help at all with corruption at the disk level anyway.

Yes, using ECC without ZFS, btrfs, ReFS, or checksummed file formats is pretty pointless (unless your application never touches storage).

Re: I am giving up on Intel and have bought an AMD Ryzen 9950X3D

#347

Post by jonbiggums22 »

Fr0styMatt88 wrote: post #251

Alright so two CPUs failing in the same system has gotta be strange; mobo issue? Secondly, what BIOS settings should I be using to run safely? Is XMP/whatever the AMD equivalent is safe? If I don't run XMP then my RAM runs at way below spec (for the stick) default speeds. Anyone know of a good guide for this stuff?

XMP is technically overclocking, nothing inherently safe about it. I've had new dual channel kits fail memtest at XMP settings on Ryzen, it seemed to depend almost entirely on what the individual CPUs memory controller was capable of.

Maybe the situation is better on DDR5 platforms.

Re: I am giving up on Intel and have bought an AMD Ryzen 9950X3D

#348

Post by fluoridation »

adrian_b wrote: post #217
consp wrote: post #78

Earlier quoted context omitted.

Isn't it mostly an ease of mind thing? I've never seen a ECC error on my home server which has plenty of memory in use and runs longer than my desktop. Maybe it's more common with higher clocked, near the limit, desktop PC's. Also: DDR5 has some false ecc marketing due to the memory standard having an error correction scheme build in. Don't fall for it.

Whether you will see ECC errors depends a lot on how much memory you have and how old it is. A computer with 64 GB of memory is 4 times more likely to encounter memory errors than one with 16 GB of memory. When DIMMs are new, at the usual amounts of memory for desktops, you will see at most a few errors per year, sometimes only an error after a few years. With old DIMMs, some of them will start to have frequent error…

>A computer with 64 GB of memory is 4 times more likely to encounter memory errors than one with 16 GB of memory.

No. Or well, not exactly. More bits will flip randomly, but if between the two systems only the total installed memory changed, both systems will see the same amount of memory errors, because bit flips on the additional 48 GB will not result in errors, because they will not be used. Memory errors scale with memory used not with memory installed.

Re: I am giving up on Intel and have bought an AMD Ryzen 9950X3D

#349

Post by dahauns »

stavros wrote: post #280
dahauns wrote: post #271

Earlier quoted context omitted.

How about you "overclock" (overvolt, unlock TDP etc.)? This is about sane, stable defaults. If you want the extra performance far beyond the CPUs sweet-spot it should be made explicit you're forfeiting the stability headrooms.

Good thing that's not the point I was making, then!

Maybe not intentionally.

Re: I am giving up on Intel and have bought an AMD Ryzen 9950X3D

#350

Post by vid »

827a wrote: post #340
vid wrote: post #329

Earlier quoted context omitted.

People are running GPT OSS 120b at 46 tokens per second on Strix Halo systems, which is quite usable and a fraction of the cost of a 128GB NVidia or Apple system. Apple's GPU isn't that strong, so real competition to Apple and NVidia can be created.

Exactly yeah, my point is that there's a lot more to running these models than just the raw memory bandwidth and GPU-available memory size, and the difference between a $6000 M4 Ultra Mac Studio and a $2000 AI Max 395+ isn't actually as big as the raw numbers would suggest. On the flip-side, though: Running GPT-OSS-120b locally is "cool", but have people found useful, productivity enhancing use-cases which justify do…

I use local AI fairly often for innocuous queries (health, history, etc) I don't want to feed the spy machines plus I like the hands on aspect, I would use it more if I had more time and while I hear the 120b is pretty good (I mostly use qwen 30b), I would use it a lot more if I could run some of the really great models. Hopefully Medusa Halo will be all that.
Post reply on HN