this post was submitted on 02 Feb 2026
8 points (55.0% liked)
Technology
79985 readers
4883 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Typically I’ve seen a motherboard supports about 2 generations of gpu before some underlying technology makes it no longer can keep up.
If you are going from a 30 series to a 50 series gpu there is going to be a need for increased pci bandwidth in terms of lanes and pcie- spec for it to be fully utilized.
I just saw this play out with a coworker where he replaced 2x3090 with a 5090. The single card is faster but now the he can’t fully task his storage and gpu at the same time due to pci-lane limits. So it’s a new motherboard, which needs a new cpu which needs new ram.
Basically a 2 generation gpu upgrade needs a whole new system.
Each generation of pcie doubles bandwith so a future 2x pcie-6 gpu will need an 8x pcie 4 worth of bandwidth.
Even then gpu’s and cpu have been getting more power hungry. Unless you over spec your psu there is a reasonable chance once you get past 2 gpu generations you need a bigger Psu. Power supplies are wear items. They continue to function, but may not provide power as cleanly when you get to 5+ years of continuous use.
Sure you can keep the case and psu but literally everything else will run thunderbolt or usb-c without penalties.
At this point why not run storage outside the box for anything sizeable? Anything fast runs on nvme internal.
This doesn’t make any sense, especially the 2x 3090 example. I’ve run my 3090 at PCIe 3.0 over a riser, and there’s only one niche app where it ever made any difference. I’ve seen plenty of benches show PCIe 4.0 is just fine for a 5090:
https://gamersnexus.net/gpus/nvidia-rtx-5090-pcie-50-vs-40-vs-30-x16-scaling-benchmarks
1x 5090 uses the same net bandwidth, and half the PCIe lanes, as 2x 3090.
Storage is, to my knowledge, always on a separate bus than graphics, so that also doesn’t make any sense.
My literally ancient TX750 still worked fine with my 3090, though it was moved. I’m just going to throttle any GPU that uses more than 420W anyway, as that’s ridiculous and past the point of diminishing returns.
And if you are buying a 5090… a newer CPU platform is like a drop in the bucket.
I hate to be critical, and there are potential issues, like severe CPU bottlenecking or even instruction support. But… I don’t really follow where you’re going with the other stuff.
That is the point of the article.
The problem my friends has is that he is rendering video so he has a high performance Sas host adapter on the same PCI bus as the GPU. He upgraded both hoping the 5090 would play nicer with the sas adapter but he can't pull full disk bandwith and render images at the same time. Maybe it's ok for gaming, not for compute and writing to disk.
The thing with power supplies, they continue to provide enough power long after they lose the ability to provide clean power under load. Only when they are really on their last legs will they actually stop providing the rated power. I have seem a persistent networking issue resolved by swapping a power supply. Most of the time you don't test a power supply under load to understand if each rail is staying where it needs to be.