The stuff Hacker News actually recommends.

GPUs and CPUs · AMD Ryzen

AMD Ryzen 5950X

27owners with an opinion
70%of verdicts positive
13/6positive/negative since 2024
4 yrslongest happy owner
1happy after 3+ years

Praised for performance and reliability.

What owners wrote

Likes itOwned 4 yearsStill going
I got an entire Prebuilt PC when POW ended from a miner. AMD 5950x, 64gb ram, 3090, 2tb SSD, all for $1700. This was 4 years ago when the 3090 was $1500 by itself, and with minor upgrades it's still going strong today. It's wild to think that the system now is worth at least as much as I paid for it then if not much more than that. I saw a similar one going for $2500.
Likes it
i'd say it depends (as with everything) i'm still using an old i7 3770k @ 4.8ghz with 16gb (ddr3) ram running linux for random tasks like executing tests. obviously, power consumption is higher. my main machine is a MBP M1 Max which i use for everything. i also have my main linux desktop workstation that has a 5950x with 128gb (ddr4) ram. i'll probably get 2-5 years out of my MBP, and my AMD workstation will probably be good for another 5-10 years. i'm not a gamer, but i have a 3080. i'm sure my 5950x will still be good for gaming in 10 years if paired with a modern GPU.
Likes it
I'm using qwen 3.6 35B unsloth 4 bit with my 5950x (128 gb memory) and a 3060 12 gb gpu with a self made harness. At 10k context I get about 40 tps generation and 500 tps prefill. At 100k context I get about 25 tps generation and 400 tps prefill. It works, but I often use gpt or claude to make a detailed enumerated plan of what I want to do first, then have qwen follow it. I'm not sure if it is economical or not, but I have solar on the roof so the power use is not really an issue and I already have the hardware. The biggest benefit for me is that it's all done locally, and I know the harness is not uploading anything or sending telemetry to someone else.
Likes it
> My 5950X tower certainly doesn’t That's what I have in my gaming tower, and yeah I feel zero pressing need to upgrade. I did manage to put 64GB of DDR4 in it just before prices went totally bonkers, thankfully. Where I'm falling behind is my GPU I'm still on an nvidia 1660 super, but I just can't justify paying what they cost right now. I would gain pretty much nothing moving to a newer board w/ DDR5.
Likes it
Yeah.. late 90's to early 00's was pretty peak builder era. I learned through a relative fluke how much going a bit over the top on memory and faster drives really helped over even a faster CPU/GPU a lot of the time for general use. My current computer is literally the first I've built in decades that I didn't max out the RAM... I mean, I kind of did as 2x48gb was the most I could get in DDR5@6000 and only use 2 slots (or it would run much slower). 4th gen Core series was the longest I'd held onto a single PC (close to 5 years total for a 4790K). I did a mid-cycle gpu and nvme upgrade and that was it. I bumped to a 3950X/5950X and now 9950X since... AM3 is really the first socket in a long time I'd done an in-place upgrade for any CPU. My daughter's Ryzen 2400 to a 5000 series, and my own build from a 3600 -> 3950X -> 5950X... the 3600 was a placeholder as I couldn't get a 3950X for a few months. I couldn't even name half the CPUs I ran from 1998 to 2005 or so... it was such a blur of upgrades every 6-12 months... I'd upgrade my computer, my wife's, my son's... etc. Then, things just completely stagnated... I mean there's been progress, but it's over the course of years, not seeing 2-3x in under a year.
Likes it
I didn't really understand the performance table until I saw the top ones were 8B models. But 5 seconds / token is quite slow yeah. I guess this is for low ram machines? I'm pretty sure my 5950x with 128 gb ram can run this faster on the CPU with some layers / prefill on the 3060 gpu I have. I also see that they claim the process is compute bound at 2 seconds/token, but that doesn't seem correct with a 3090?
Likes it
In what world is it embarrassing to not buy hardware you don't need? The servers worked fine for years. When there was an actual reason to spend money, they bought something new. Sounds like good stewardship of the donations they receive. I finally just upgraded my 9 year old computer with an i5-6600k to a Ryzen 9 5950x because I wanted to be able to edit home videos. I already rarely even used 1 core on the old CPU, the new one is 7x more powerful, and it's an ebay part from 5 years ago. I don't foresee needing to upgrade again for another decade. I probably would've been good for another 15-20 years if I had upgraded to a DDR5 platform, but RAM prices had already spiked, so I just swapped the motherboard and CPU.
Neutral
This is Black Friday pricing at least, if you're willing to shuck. Seagate drives are still sub-$10/TB which... a single 24-26TB is enough for all my photos (ever), media and some dataset backup for work. I'm planning to backup photos and other "glacier"-tier media like YouTube channels to BluRay (a disk or two per year). It's at the point where I'd rather just pay the money and forget about it for 5-10 years. I built the case from Makerbeam and printed panels, an old Corsair SF600 and a 4 year old ITX system with one of Silverstone's backplanes. They make up to 5 drives in a 3x5-1/4 bay form factor. It's a little overpowered (a 5950X), but I also use it as a generic server at home and run a shared ZFS pool with 2x mirrored vdevs. Even with inefficient space it's more than I need. I put in a 1080ti for transcoding or odd jobs that need a little CUDA (like photo tagging). Runs ResNet50-class models easily enough. I also wondered about treating it as a single-node SLURM server.
Neutral
Well, I just updated my 5950x with 128 GB to AMD Strix Halo with the same amount of much faster RAM. It is noticeable faster, but what's better: the whole computer is tiny and sips energy. I'm very happy I ordered this in the summer, framework delivered it to me early this month. I wonder will these machine just be out of stock now or the price goes up a lot...
Doesn’t like it
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 these issues. Sometimes it’s something completely different like the PSU. After the last issues, I decided to buy a prebuilt (ThinkStation) with on-site service. The cooling is a bit worse, etc., but if issues come up, I don’t have to spend a lot of time debugging them. Random other comment: when comparing CPUs, a sad observation was that even a passively cooled M4 is faster than a lot of desktop CPUs (typically single-threaded, sometimes also multi-threaded).
Doesn’t like it
I've got a 5950x that I can reliably crater with a very specific .NET 8 console app when it would otherwise be stable 24/7/365, even under some pretty crazy workloads like Unity. I think a lot of it boils down to load profile and power delivery. My 2500VA double conversion UPS seems to have difficulty keeping up with the volatility in load when running that console app. I can tell because its fans ramp up and my lights on the same circuit begin to flicker very perceptibly. It also creates audible PWM noise in the PC which is crazy to me because up til recently I've only ever heard that from a heavily loaded GPU.
Likes it
I wonder if cooling/power is really the key here. I've got a 5950x that ended up getting the water loop I'd intended for my next threadripper - only to find they were not selling the blasted things to anyone but a few companies. With the cooling sized for almost twice what the 5950x could put out, it has been a very stable machine for some crazy workloads. That old dog will likely keep the setup when a zen 5 TR gets swapped in. For a long time, my Achille's heel was my Bride's vacuum. Her Dyson pulled enough amps that the UPS would start singing and trigger the auto shutdown sequence for the half rack. Took way too long to figure out as I was usually not around when she did it.
Neutral
I think this is the best path if your problem can support it. I use a 5950X for running genetic programming and neuroevolution experiments and about once every 100 hours the machine will just not like the state/load it is experiencing and will restart. My approach is to checkpoint as often as possible. I restart the program the next morning and it deserializes the last snapshot from disk. Worst case, I lose 5 minutes of work. This also helps with Windows updates, power outages, and EM/cosmic radiation.
Doesn’t like it
The criteria for purchase for anybody trying to use it is "run slowly but acceptably" vs. "run so slow as to be unusable". My memory is wrong, it was the 32b. I'm running the 70b against a similar prompt and the 5950X is probably going to take over an hour for what the M4 managed in about 7 minutes. edit: an hour later and the 5950 isn't even done thinking yet. Token generation is generously around 1 token/s. edit edit: final statistics. M4 Pro managing 4 tokens/s prompt eval, 4.8 tokens/s token generation. 5950X managing 150 tokens/s prompt eval, and 1 token/s generation. Perceptually I can live with the M4's performance. It's a set prompt, do something else, come back sort of thing. The 5950/RTX3080's is too slow to be even remotely usable with the 70b parameter model.
Likes it
On the subject of under-powered hardware, this will be a minority opinion, but I don't even bother self-hosting on Pis anymore. I started using more powerful systems like NUCs and that was good, but I still kept running into hardware limits with things like Jellyfin. What I do now, and IMHO what the holy grail is (for some with a finite budget) is to build a great desktop PC and do all the self hosting on that. I built a beastly "gaming" rig that ran me about $3800 (though $2k of that was the damn GPU because it was early 2021 at the time). With a Ryzen 9 5950X 16-Core Processor and 128 GB of RAM, and a RX 6900 XT, it easily handles everything I throw at it, and occasionally when there's an intensive job like a long re-encode, it blazes through it in record time. Currently I could rebuild a similar system for about <$2k. I put Fedora on there, and I then use this rig for: 1. A normal desktop experience with multi-monitor for work during the day. 2. A remote dev workstation. I use neovim & tmux so can do development completely headless, so I SSH into this rig for doing dev work. It compiles and runs tests blazingly fast and with up to 32 threads so I rarely have to wait for compilations. It's also great to have something that's "always on" so I can SSH in from anywhere and instantly be dropped right into the last state. 3. Running all my self-hosted stuff like HA, Jellyfin, Audiobookshelf, ArchiveBox, and many more. Some run in containers on the host while others r…
Neutral
> any reasonable CPU from the past 15 years can handle software decoding of high resolution video just fine. 4k VP9 from youtube takes my 5950x around 20-25% CPU usage to handle with hardware acceleration disabled. The fastest consumer CPU available 15 years ago could not handle that. Hell, even CPUs from 10 years ago couldn't do that. Add power & thermal limitations of a laptop CPU? Not a chance. And that's just VP9! HEVC or AV1 would really put the hurt on.
Doesn’t like it
How many CPUs have you owned? Compare this to Rolling a die in a tabletop game. How often do you roll a natural one or snake eyes? On a 20-sided die a natural one only comes up 5% of the time. I don't think I'm quite at having owned 40 (desktop) CPUs, but I've had failed CPUs a number of times. On an old Athlon XP 3000+ it had bad CPU cache and was able to change it from crashing repeatedly to working by disabling a large swath of the L2 cache in a bio setting. I am just now retiring a workstation that had an AMD Ryzen 5950x that was generally unstable, that would pass memtest and all the diagnostics I know how to run would pass, but about once a month would print a message on all the consoles I had open about an MCE kernel exception detected on some CPU number at random. One time I had one of those budget TriCore CPUs that had the 4th core turned on in the BIOS by accident, and that generally caused the ton of issues when I figured out it was being detected as a quad core I went in and disabled that last core and it went back to working just fine. I'm sure at least one or two of my Intel machines failed similarly, I've had a number of dead CPUs including a few Pentium 4s back when I used to fool around with overclocking settings, and at least a few dead server chips. And I know that I've had a few systems that simply wouldn't boot even when all the parts were fresh out of the box but would when I'd swap out one part or another, and sometimes that failed part was a CPU whethe…
Doesn’t like it
I experienced problems with 5950X as well and nobody (including myself) seemed to believe it was the CPU before I got another 5950X which just worked with the same setup. The issue wasn't easy to reproduce, all standard checks and torture tests ran just fine but much more random workload crashed the unit when all cores were maxed for few seconds at a time instead (eg. during compiling). Sometimes it happened twice a day sometimes once a week.
Doesn’t like it
I got a buggy 1800X, shipped for RMA to a logistics center in IIRC the Netherlands from Germany. RMA took about 2 weeks altogether and it was worth it. Since then, I've upgraded to the new top of the Ryzen line everytime there was a significant (>40%) increase in compilation performance. The 5950X was a notable dud in that regard. Recently, setting power limits to get 90-95% of the performance at 60% of the TDP became a thing.
Neutral
Yup, so I've adopted a strategy for my past few desktop builds like this: - Every time a new ToTL GPU comes out for a new family, buy it at retail price as soon as it launches (so, the first-available ToTL models that were big gains in perf: GTX 1080 Ti, RTX 2080 Ti, RTX 3090, RTX 4090) - Every other release cycle, upgrade CPU to the ToTL consumer chip (eg on a 12900KS right now, HEDT like ThreadRipper is super expensive and not usually better for gaming or normal dev stuff). I was with Ryzen since 1800x -> 3950x -> 5950x but Intel is better for the particular game I play 90% of the time. - Every time you upgrade, sell the stuff you've upgraded ASAP. If you do this right and never pay above MSRP for parts, you can usually keep running very high-end hardware for minimal TCO. - Buy a great case, ToTL >1000w PSU (Seasonic or be quiet!), and ToTL cooling system (currently on half a dozen 140mm Noctua fans and a Corsair 420mm AIO). This should last at least 3 generations of upgrading the other stuff. - Storage moves more slowly than the rest, and I've had cycles where I've re-used RAM as well, so again here go for the good stuff to maximize perf, but older SSDs work great for home servers or whatever else. - Monitor and other peripherals are outside of the scope of this but should hopefully last at least 3 upgrade generations. I bit when OLED TVs supported 4K 120hz G-Sync, so I've got a 55" LG G1 that I'm still quite happy with and not wanting to immediately upg…
Neutral
I have 70B at q4_0 running at 20 tokens/s inference using 2 Founders Edition RTX3090s with NVLink. It's an older AMD 5950x with a motherboard with 4-slot spacing between the PCI-e (iirc, the motherboard is an ASRock x570 Taichi). For comparison, I tried the same rig with a single RTX 4090, but it generated an abysmal 2.4 tokens/s, which I would consider nigh unusable. The difference between all GPU VRAM is significant.
Neutral
The baseline draw is still much higher than lower spec alternatives, and it means you would need to cater for the high end scenario in your battery estimations if you intend to actually use it during power outages. I switched from a 5950X + 3080, to a 5700G APU, to finally an M1 MBP + Steam Deck last year for this exact reason. Far cheaper to have a 250wh battery that can handle those two for the ±2h outages every day.
Likes it
With a Noctua NH-D15 and about 30 minutes spent tweaking the motherboard fan curves I was able to get my 5950x to not thermal throttle without producing any noticeable fan noise. This seems extremely over-engineered and sounds like it could have been solved by using Noctua or similar quiet fans.