Sabitlenmiş Tweet
Release of TT-Boltz during the keynote by @DavidBennett__ at Dev Day in SF. What a day!
github.com/moritztng/tt-b…
San Francisco, CA 🇺🇸 English
Moritz Thüning
376 posts
Moritz Thüning
@moritzthuening
Implementing Boltz-2 on Tenstorrent hardware. Engineer @ Tenstorrent. My views.
Munich, Germany Katılım Ağustos 2012
444 Takip Edilen2.2K Takipçiler
For the same price of a very expensive GPU you can have a bunch of Tenstorrent Blackhole cards and much higher Boltz-2 throughput in the end.
English
Moritz Thüning retweetledi
That's a wrap on TT-Deploy!
ICYMI we launched Tenstorrent Galaxy™ Blackhole deployed at scale with industry winning benchmarks including:
- @deepseek_ai 671B at 350+ t/s/u, with a roadmap to 500 t/s/u at $6 per million tokens, extending the low-cost serving curve where GPU economics collapse
- 10x faster than real-time high-quality video gen (720p, 81-frame video in 2.5 seconds) w/ @prodialabs
- 90% pass rate across thousands of random @huggingface models
Links to recap + recordings below ⬇️
English
Moritz Thüning retweetledi
Can’t wait to bring Boltz-2 to the world!!
Moritz Thüning@moritzthuening
A lot has happened between these pics, but in the end, we just switched roles. Even though he’s a slightly larger customer than I ever was and I'm just a small part of an incredible team. @DavidBennett__ is crushing it with ai&
English
Moritz Thüning retweetledi
A lot has happened between these pics, but in the end, we just switched roles. Even though he’s a slightly larger customer than I ever was and I'm just a small part of an incredible team.
@DavidBennett__ is crushing it with ai&
David Bennett@DavidBennett__
At @aiand_ we have a mission to be Japan’s best vertically integrated AI data center and most important AI lab. TT is a key tenant of our heterogeneous infrastructure and what we can accomplish with it. We were proud to be part of @tenstorrent’s TT-Deploy. (47m~) TY @jimkxa
English
Moritz Thüning retweetledi
Moritz Thüning retweetledi
Tune in tomorrow! Run fast video, speech, code all on Tenstorrent Galaxy Blackhole. Powered by our Networked AI architecture with native scale-out. Hear from our partners and customers deploying at scale.
Watch the livestream on May 1st @ 1:30 PM PDT: tenstorrent.com/deploy
English
Moritz Thüning retweetledi
Moritz Thüning retweetledi
Congrats to @Tenstorrent on their Galaxy Blackhole launch! 🎉
As models evolve, general-purpose AI compute is the only foundation that keeps up — and it's a key part of what we're building @aiand_
See you at TT Deploy this week. 🚀
EE Times | Electronic Engineering Times@eetimes
Tenstorrent Unveils Next-Gen Servers For Fast Tokens, No Disaggregation Needed eetimes.com/tenstorrent-un…
English
@beffjezos i first heard of atomic labs and Sam all the way in Munich from @moritzthuening , the work they're doing is amazing and Sam is incredibly cracked - unfortunate but makes sense that they're moving to Texas
English
Immense loss for Califronia.
Sam is a genius and his team is the future of fabs.
Sam Zeloof@szeloof
I’m thankful for 4 years in California . But, the Next Phase is in Texas. We shift our hiring focus to build huge operations in Austin, Texas. Join us ! The team is really great. @atomic_semi New 120,000 square feet HQ for R&D and production:
English
Moritz Thüning retweetledi
At TT-Deploy, see our latest benchmarks and hear from partners and customers scaling on Tenstorrent.
Learn more about Networked AI –– our unified compute, memory, and networking in one scale-out architecture. No proprietary interconnects. No rigid workload declarations.
Watch it live this Friday, May 1st @ 1:30 PM PDT: tenstorrent.com/deploy
English
Very well written. Incredible work by the team. And TT-Boltz runs on it beautifully.
Sally Ward-Foxton@sallywf
.@tenstorrent is launching Galaxy Blackhole servers and clusters for fast LLM inference (DeepSeek-671B at up to 350 t/s/u), among other applications - no disaggregation needed, according to @jimkxa: eetimes.com/tenstorrent-un…
English
Moritz Thüning retweetledi
At TT-Deploy we will be introducing TT-Lang—the new Python DSL to supercharge creation of new operations and frontier models. Join us to learn more!
Tenstorrent@tenstorrent
At TT-Deploy, see our latest benchmarks and hear from partners and customers scaling on Tenstorrent. Learn more about Networked AI –– our unified compute, memory, and networking in one scale-out architecture. No proprietary interconnects. No rigid workload declarations. Watch it live this Friday, May 1st @ 1:30 PM PDT: tenstorrent.com/deploy
English
Gonna be incredible!!
Tenstorrent@tenstorrent
At TT-Deploy, see our latest benchmarks and hear from partners and customers scaling on Tenstorrent. Learn more about Networked AI –– our unified compute, memory, and networking in one scale-out architecture. No proprietary interconnects. No rigid workload declarations. Watch it live this Friday, May 1st @ 1:30 PM PDT: tenstorrent.com/deploy
English
English
I'll be in SF / Santa Clara for the next 10 days. Hit me up if you're interested in @tenstorrent hardware for biotech or anything else.
English
@petrohi and the team are brilliant.
ludwig@ludwigABAP
god thank GOD for tt-lang and not having to write metallium kernels for tenstorrent hardware anymore, I can't believe I didn't know it existed until a couple days ago wtf i am beating my handrolled c++ implementations trivially with just codex + the agent sklils they ship
English