Joined June 2023
- Tweets 4,266
- Following 189
- Followers 74,960
- Likes 772
560 Photos and videos
Pinned Tweet
May 12
All products are now available to order! $45k for tinybox green with 5090s, and we have two variants of the tinybox pro v2, one with 5090s and one with RTX6000 Blackwells (that's a whopping 768 GB of RAM)
24
8
337
101,085
Jun 13
Also might I mention, a great time to buy a tinybox.
14
5
228
7,220
Jun 12
Up and benchmarking itself on 4xMI300X (this is with vLLM). Memory bandwidth says we should be able to go so much faster.
Jun 12
π Kimi-K2.7-Code, our latest coding model, is now released and open-sourced!
π· Improved coding & agent performance over K2.6: 21.8% on Kimi Code Bench v2, 11.0% on Program Bench, and 31.5% on MLS Bench Lite.
π· Reasoning efficiency: Less overthinking, with 30% lower reasoning-token usage compared to K2.6.
π· Long-horizon coding: Improved instruction following, higher end-to-end coding task success rates.
β‘οΈ 6x High-Speed Mode coming soon!
π Available today via Kimi API and Kimi Code.
π Kimi Code: kimi.com/code
π API: platform.moonshot.ai
10
8
281
21,110
Jun 12
It's really nice to have a local Kimi, this is as good as the best models just 6 months ago, and you know nobody can take it away from you or stop you from finetuning so it'll never refuse your requests. With this cranked to 1000 tok/s it might even be the best experience.
1
2
134
4,551
Jun 10
A reminder to look past the hype and look at the numbers. Google is the largest owner of compute in the world. AI is not a race, it's a decentralized revolution that will take decades to play out.
Jun 10
Model weights available on Hugging Face under Apache 2.0 license, read more here: blog.google/innovation-and-aβ¦
20
19
387
22,682
the tiny corp retweeted
Jun 9
i look forward to our chinese brothers liberating the knowledge from within fable-5 and selling it to me at 5% the cost & 2x the speed
319
1,586
24,639
1,057,101
the tiny corp retweeted
Jun 10
Concentration of power, capabilities and economic wealth is the biggest risk in AI. We need open science and open-source more than ever!
111
477
3,085
159,765
Jun 10
This makes me not want to waste any time using it. Who knows if it's silently sandbagging me. Is tinygrad close enough to frontier LLM for it to? Just makes the model completely untrustworthy.
Jun 9
mythos will be bad ON PURPOSE on ai "frontier llm research" tasks, this is very very sad for the research community
also the fact that this is un purpose not visible to the user is crazy
20
36
935
39,106
Jun 3
Because it's the full stack from Tensors to MMIO, the ceiling on speed in tinygrad is higher than in any other framework.
Iβm grinding away making inference faster on AMD using tinygrad and itβs super gratifying, also an incredible way to learn gpu programming
5
3
139
20,719
May 29
tinygrad wants to make it as easy as possible to answer three questions about Tensor compute. What happens? Where does it happen? And when does it happen?
5
7
209
15,835
May 28
tinygrad will write that C for you. Our new driver compiles all interaction with the GPU to C, so once it's running the CPU does next to nothing.
SpaceX has almost finished writing V1.0 of an in-house AI training stack in C that exact-maps to 220k GB300s with 800G NICs, making heavy use of pipeline parallelism and getting as close to bare metal as possible.
The potential speed improvement vs JAX for large training runs is over an order of magnitude.
36
66
1,477
293,213
May 28
1 of 8 NVIDIA RTX PRO 6000 Blackwell being torn down for tinybox pro install. Don't worry, it's only $10,000 if you shear one of the ribbon cables.
22
16
329
29,861
May 21
Our supplier is raising the price of Blackwell cards $2200 each. As a result, we'll have to further raise the price of the tinybox green v2 blackwell. This is the last week to get your order in at the old price.
12
7
260
43,209
May 25
Prices updated. For those people who placed orders, you have 5 days to land payment or they will be cancelled and you'll have to reorder at the new price. I wish we could make these machines cheaper for you, but we're actually making even lower margins with these new prices.
14
6,355
May 21
.@tenstorrent when you are ready, we'll get you on MLPerf for $10M. Ground up stack, one 1MB pip install, zero C 20 (pure Python). From how many people I see on X trying to rewrite it, your current software approach isn't working.
15
12
448
34,287
May 19
Our first assembly backend, DEV=CPU:X86 is merged! Instruction selection register allocation added, and it's all visible with VIZ=1. This is the path to kernels that outperform everyone playing at the LLVM/PTX layer.
10
18
373
20,685
May 19
New $1,000 bounty: "RDNA3 assembly backend reusing the X86 assembly infrastructure passing all tests and competitive on speed"
We already have all the RDNA3 instruction encodings and emulator. And now with the CPU:X86 backend, we have a register allocator and isel template.
2
2
36
8,549
May 18
To everyone who wants to invest in tiny, preorder an exabox. At launch, it will be the cheapest compute you can buy. Deploy it and run it and make returns!
We don't want VCs who invest other people's pensions with only upside potential for them. We want people in the trenches.
21
11
440
29,201
May 18
This is who we want as tiny corp customers.
1
3
86
16,673