Technology

43148 readers

292 users here now

This is the official technology community of Lemmy.ml for all news related to creation and use of technology, and to facilitate civil, meaningful discussion around it.

Ask in DM before posting product reviews or ads. All such posts otherwise are subject to removal.

Rules:

1: All Lemmy rules apply

2: Do not post low effort posts

3: NEVER post naziped*gore stuff

4: Always post article URLs or their archived version URLs as sources, NOT screenshots. Help the blind users.

5: personal rants of Big Tech CEOs like Elon Musk are unwelcome (does not include posts about their companies affecting wide range of people)

6: no advertisement posts unless verified as legitimate and non-exploitative/non-consumerist

7: crypto related posts, unless essential, are disallowed

founded 7 years ago

MODERATORS

MinutePhrase@lemmy.ml

Alibaba's TSMC-Built 5nm RISC-V Chip, XuanTie C950, Now Runs Qwen-3.8 27B Model Natively, Unlocking Massive Vertical Integration Tailwinds (wccftech.com)

submitted 6 days ago by yogthos@lemmy.ml to c/technology@lemmy.ml

9 comments fedilink hide all child comments

you are viewing a single comment's thread
view the rest of the comments

[–] Asafum@lemmy.world 8 points 6 days ago (2 children)

as well as a 500 dollar GPU

(Cries in $1,300 Intel arc fighting tooth and nail to break 25t/s...)

[–] yogthos@lemmy.ml 8 points 6 days ago

haha I keep forgetting how insane prices are now

[–] yxc999@lemmy.ml 3 points 5 days ago (1 children)

Huh, my B70 gets peaks of 37 tok/s TG and 800 tok/s PP, what are your settings?

[–] Asafum@lemmy.world 1 points 5 days ago* (last edited 5 days ago) (1 children)

At first I was mislead by some faulty benchmarking I was doing and thought that llama.cpp + SYCL was worse than Vulkan stock. Vulkan stock couldn't run past 10t/s with MTP for some reason, so after a whole bunch of shenanigans I ended up again back on llama.cpp + SYCL but was lead to believe MTP was hurting me so I didn't retry until last night. I finally got the 35-40t/s with llama.cpp + SYCL and MTP and 800PP with AOT over using JIT. I drop to 25ish after 40-50k context and stay kinda flat 12ish at 160k

I didn't want to move to Linux to try vLLM so this was all just trying to fight windows b.s.

If you're on windows and fighting with the VRAM offload after 70 seconds like I was: HKLM\SYSTEM\CurrentControlSet\Control\GraphicDrivers

EnableRuntimePowerManagement set to 0

[–] yxc999@lemmy.ml 2 points 5 days ago (1 children)

I'm using this on Linux, so the first two lines won't apply to you. Besides that, you are completely right, llama.cpp with the SYCL backend and MTP is the fastest way to run LLMs on Intel Arc GPUs, it's just braindead people and bots on Reddit repeating the same line while they run a NVIDIA GPU and Ollama.

source /opt/intel/oneapi/setvars.sh
/opt/llama.cpp-sycl/bin/llama-server \
  --model /path/to/your/model/Qwen3.8-27B-UD-Q6_K_XL.gguf \
  --device SYCL0 \
  --n-gpu-layers 999 \
  --load-mode none \
  --flash-attn on \
  --jinja \
  --reasoning-preserve \
  --ctx-size 100000 \
  --cache-type-k q8_0 \
  --cache-type-v q8_0 \
  --temp 1.0 \
  --top-p 0.95 \
  --top-k 20 \
  --min-p 0.00 \
  --presence-penalty 0.0 \
  --repeat-penalty 1.0 \
  --spec-type draft-mtp \
  --spec-draft-n-max 2 \
  --port 9931

[–] Asafum@lemmy.world 1 points 5 days ago (1 children)

Thanks for the info! I believe I'm running MTP n-max 4 if I'm not mistaken which seems to be holding out well

[–] yxc999@lemmy.ml 2 points 5 days ago* (last edited 5 days ago)

I feel like you definitely can get away with as much as even 5, but if you actually look at the CLI, it will tell you how many draft tokens actually got accepted. In some situations like writing simple text, over 90% of my two draft tokens get accepted so setting it to five would have made it faster. However if you run a task which requires heavy reasoning, even two draft tokens start approaching close to only 60%. Considering Qwen3.8-27B overthinks more than an anxious teenage girl, you should generally with two. If you take my example with the quant dropped to UD-Q5_K_XL with 262K tokens and reasoning low for a task that isn't actually complex, just long, then go right ahead and let it run with 5.

Edit: excuse the horrendous grammar, I was distracted