𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠 (@IanCutress) / X

𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠

46.8K posts

𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠

@IanCutress

Consultant, Chief Analyst, Influencer. Substack: more-moore.com (@MoreThanMoore2x) Youtube: techtechpotato.com (@TechTechPotato)

London

Joined June 2011

Pinned
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Apr 18, 2024
I was in Oregon to take a look at @Intel's High-NA EUV machine - a $350m monster of lithography. It's almost complete now, going into calibration. Then it'll run test-wafers to prep for R&D into 14A. Here are some details on the machine, plus my thoughts. youtu.be/8i9rs4LNSlI
185K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Aug 16, 2023
Jesus fucking christ. Thoughts go out to Madison, the sheer toxicity is insane, I hope it's at least a small weight lifted from her. Why are so many people such arseholes to women in tech?
Madison | Suop
@suuuoppp
Aug 16, 2023
To stop the speculation and DM's I am receiving. I chose to quit my role at LTT because it, and the working environment I was facing, were ruining my mental health. The number of daily items the Social media role at the time was expected to fill was incredibly high.
309K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Aug 16, 2023
In case anyone missed it, Billet don't want the monoblock back, and have asked for the previously agreed monetary value. reddit.com/r/LinusTechTip…
259K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Aug 16, 2023
This is the correct, professional move. You could argue it's later than it should have been, but the second best time is today.
Linus Tech Tips
@LinusTech
Aug 16, 2023
We are taking Madison’s allegations seriously. We are conducting an internal assessment alongside bringing in a third-party investigator to look into the allegations. We are committed to publishing the findings and implementing any corrective actions that may arise due to this.
146K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Jan 16, 2024
Finished installing my @nvidia rtx 5090
84K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Jan 28, 2025
People saying 'Deepseek used PTX, it breaks the CUDA moat!' are fundamentally misunderstanding what's at play here, and what CUDA actually is. CUDA is a high level language. It makes code libraries and interfacing with NVIDIA GPU hardware easier. It allows for fast iteration.
232K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Nov 5, 2023
This is the warning you get if you try to install windows on an SSD smaller than your total DRAM. SSD: 500GB DRAM: 2TB
143K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Aug 16, 2023
If you've not met Terren before, this is it. I've chatted to him. He knew then that he'd have to define the path for the next gen of LMG. I don't think he knew to what degree. He does now.
189K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Aug 23, 2025
Amazon requires new sellers to have a company name of at least 7 some letters to avoid manual checks. Hence Chinese white label resellers just check some alphabetti spaghetti and come up with CUKTECH and NICGIGA or KPKKQUE or FUXK etc
Jeff Geerling
@geerlingguy
Aug 22, 2025
I can't handle the names of some of these Chinese tech manufacturers...
00:00
61K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Aug 15, 2023
It turns out I'm just a short bus journey away from Billet Labs. I just emailed them (at 10:45pm) asking if I can come by tomorrow for an interview and to look at what they're building next.
91K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Apr 28, 2022
Welcome to the future
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Feb 25, 2025
This looks amazing! @FrameworkPuter desktop with @AMD Strix Halo. ➡️ Top model with 395+, 16 cores, 8060S GPU ➡️ 128GB of DRAM for $1999. ➡️ Two M.2, 5GbE, WiFi 7 ➡️ Connect four together by Eth for AI ➡️ Shipping Q3, fits in LTX backpack ➡️ Motherboard available on its own
101K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Jan 25, 2025
Started watching the Rossman video. Thought I'd make mental notes. Stopped making notes at 10 mins. Realised at 20 minutes it's not worth the mental effort. Not what I expected. Nuance doesn't seem to matter. Facts don't seem to matter. Seemingly instantly confused over important
81K
𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠
@IanCutress
Dec 6, 2024
Intel test chip yield on 18A is 99%. 🤷‍♂️
Dan Nystedt
@dnystedt
Dec 6, 2024
TSMC's 2nm process has achieved over 60% yield in trial production, surpassing expectations, media report citing unnamed supply chain sources. Mass production is on schedule for next year. The manufacturing technology is being developed in TSMC’s Hsinchu (Baoshan) fab, and when
139K