Uploaded June 2026 | Updated September 2026, 2 weeks ago
Run DFlash with SGLang's new Spec V2 overlap scheduler on Qwen3.6-27B and hit 160 tok/s on a single H100 -- hands-on install, benchmark, and real numbers.
🔥 Get 50% Discount on any A6000 or A5000 GPU rental, use following link and coupon:
bit.ly/fahd-mirza
Coupon code: FahdMirza
🔥 Buy Me a Coffee to support the channel: ko-fi.com/fahdmirza
#dflash #lucedflash #lucespark #kvflash #sglang #speculativedecoding
PLEASE FOLLOW ME:
▶ LinkedIn: linkedin.com/in/fahdmirza
▶ YouTube: youtube.com/@fahdmirza
▶ Blog: fahdmirza.com
RESOURCES:
▶ github.com/z-lab/dflash
All rights reserved © Fahd Mirza
Run DFlash with SGLang's new Spec V2 overlap scheduler on Qwen3.6-27B and hit 160 tok/s on a single H100 -- hands-on install, benchmark, and real numbers.
🔥 Get 50% Discount on any A6000 or A5000 GPU rental, use following link and coupon:
bit.ly/fahd-mirza
Coupon code: FahdMirza
🔥 Buy Me a Coffee to support the channel: ko-fi.com/fahdmirza
#dflash #lucedflash #lucespark #kvflash #sglang #speculativedecoding
PLEASE FOLLOW ME:
▶ LinkedIn: linkedin.com/in/fahdmirza
▶ YouTube: youtube.com/@fahdmirza
▶ Blog: fahdmirza.com
RESOURCES:
▶ github.com/z-lab/dflash
All rights reserved © Fahd Mirza




