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

Benchmarking 15 “E-Waste” GPUs with Modern Workloads

esologic.com

Post reply on HN
Print view View original

21–30 of 68 posts

Re: Benchmarking 15 “E-Waste” GPUs with Modern Workloads

#21

Post by latchkey »

riedel wrote: post #20
latchkey wrote: post #3

Darn, I was hoping to see bc-250's (aka PS5 chips) in there. They've recently become popular for inference and they are only about $200 on ebay. They hold a special place in my heart because I deployed 20k of them and I'm glad to see they are finding a purpose now and not just e-waste.

Cool stuff. Just read: https://github.com/akandr/bc250

Yea, I'm bummed I didn't know about the 40-CU unlock, although it probably wouldn't have had much impact on mining performance. It still would have been neat to test. I did build a whole automated solution for auto-tuning each individual board. It would start at the "best" settings and then downgrade every time there was a crash. If it wasn't crashing, then those were the new "best" settings for that individual chip.

Re: Benchmarking 15 “E-Waste” GPUs with Modern Workloads

#22

Post by 0xbadcafebee »

loeg wrote: post #18
jkonline wrote: post #13

[flagged]

"Don't post generated text" - https://news.ycombinator.com/newsguidelines.html

The rules for comments are now so lengthy that probably 90% of comments would violate one. It's like how police can pull you over for any reason and justify it by picking a law you've unwittingly violated

Re: Benchmarking 15 “E-Waste” GPUs with Modern Workloads

#24

Post by chromadon »

SillyUsername wrote: post #17

No mention of the venerable Tesla P4. 75W peak, 8GB VRAM, about $80 (£60). I have 6x P4s, a Xeon E5 2696v3 (36 threads, 3.8ghz peak but all core turbo unlocked, so 6 cores at 3.8Ghz - about 8 cores at 3.5ghz, or all cores at 3.1ghz), 48GB DDR4, all fit into a micro atx case running on a 650W MSI psu. This gives me a virtual 48GB GPU (llama.cpp ftw) to backup that 48GB of RAM. I typically see scores of at least 7-12t/…

How the hell did you fit 6 P4s in a mATX case?

Re: Benchmarking 15 “E-Waste” GPUs with Modern Workloads

#25

Post by SwellJoe »

When I wanted to tinker with self-hosted models, I bought a couple of Radeon Pro V620 GPUs, because they're 32GB, still supported by current ROCm releases, and a few years newer than the similar-priced 32GB Nvidia cards (which are all EOL). They're a little faster than the old Tesla stuff, as well. 64GB is enough to run Gemma 4 31b 4-bit QAT with pretty big context at a respectable interactive speed (30+ tokens per second sustained).

That said, even the old Radeon Pro stuff has gotten more expensive on eBay, so I'm not necessarily recommending cheap old server cards that need custom-printed fan shrouds to operate in a consumer PC. Probably better to buy the Radeon AI Pro R9700 for $1400, which will be faster, supported for many years, and has a fan already. Or, maybe even the Intel ARC B70 for $1000.

Re: Benchmarking 15 “E-Waste” GPUs with Modern Workloads

#26

Post by eso_logic »

nbf_1995 wrote: post #23

This site does not like being on the front page of HN. ~7MB for pictures of graphs that probably should have html or svg. This is an interesting article though. Bookmarking since my dual e5-v4 system is unplugged until summer is over.

I'm trying bruh fuck!

Re: Benchmarking 15 “E-Waste” GPUs with Modern Workloads

#27

Post by SoftTalker »

A few years ago we got rid of a bunch of K80s at work, they were not only obsolete but had gotten glitchy as hell. I suspect this is from the many heat/cool cycles they went through. When they were running flat out the exhaust air felt like a hair dryer.

Re: Benchmarking 15 “E-Waste” GPUs with Modern Workloads

#28

Post by Joel_Mckay »

Depends on the use case, as for hardware h265 codecs a rtx 5070 Ti works just as well as the rtx 6000 gpu. Legacy GPU don't support modern codecs, but modern Intel chips have h265 HDR hardware support. Lower https://www.pugetsystems.com/pugetbench/creators/davinci-res...

https://www.pugetsystems.com/pugetbench/creators/premiere-pr...

In some cases it is better to have lower passmark scores:

https://www.videocardbenchmark.net/gpu.php?gpu=RTX+PRO+6000+...

Blender is heavily bottle-necked by ray-tracing and de-noising operations:

https://opendata.blender.org/benchmarks/query/?compute_type=...

One metric that isn't considered is VRAM, as some rendering pipelines still rely on composited baked-scenes to reduce each areas memory requirements.

In general, the $/performance unit will depend on what you are doing, but there is 1 more thing to consider... Old GPU use mystery binary BLOB drivers no longer maintained on modern kernels. You might get the software to work with a legacy Windows GPU driver, but the key takeaway concept here is "might". =3

Re: Benchmarking 15 “E-Waste” GPUs with Modern Workloads

#29

Post by Octoth0rpe »

SoftTalker wrote: post #27

A few years ago we got rid of a bunch of K80s at work, they were not only obsolete but had gotten glitchy as hell. I suspect this is from the many heat/cool cycles they went through. When they were running flat out the exhaust air felt like a hair dryer.

Do you have any #s on how old they were at decomm time? There's some suspicion that part of the AI bubble is companies playing games with depreciation, eg assuming that H100/H200s will survive for 5 years.

Re: Benchmarking 15 “E-Waste” GPUs with Modern Workloads

#30

Post by wlesieutre »

eso_logic wrote: post #26
nbf_1995 wrote: post #23

This site does not like being on the front page of HN. ~7MB for pictures of graphs that probably should have html or svg. This is an interesting article though. Bookmarking since my dual e5-v4 system is unplugged until summer is over.

I'm trying bruh fuck!

For an easier change than HTML or SVG, try running them though pngcrush to make the graph images much smaller. Won't give you the lossless vector quality, but you should be able to keep these image files visually indistinguishable at much lower size.
Post reply on HN