You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
x3d? - "Compute dies are placed on top of Intel 3 base dies using 3D stacking. Base dies host the chip’s mesh interconnect and L3 slices. Placing the L3 on separate base tiles gives Intel the die area necessary to implement 8 MB L3 slices, which gives a chip 576 MB of last level cache."
Those are the current statistics. He said they want the figure to be below 30%. The reason for this is to maintain flexibility in terms of capital investments.
No, I think people should refer to the original source instead of a secondhand source that may contain added biases. The link I shared is my personal recording to keep track of Intel's communications. I've turned off all ads on such videos, and I'm not making any profit from them. The video also includes Youtube automated transcription to assist people with hearing difficulties.
By studying each event and keeping a record of past communications, it is possible to answer questions such as those related to IFS on mobile handsets.
Ideally, each event should be uploaded to YouTube by Intel's Investor Relations team, so that all stakeholders can review them. Once they do that, I won’t need to spend my time on it.
Another point is that people should refer to the actual voice recording instead of the transcript, as the tone of the conversation also conveys important information.
Another point is that people should refer to the actual voice recording instead of the transcript, as the tone of the conversation also conveys important information.
Another point is that people should refer to the actual voice recording instead of the transcript, as the tone of the conversation also conveys important information.
Unless you’re not a real person, you should have noticed that the first line of the first post in this thread links to real Q&A conversations audio replay, which your YouTube channel copied. If you’re telling us not to trust that, then how are we supposed to trust yours?
By the way, in the recording, I also normalize the audio levels in DaVinci Resolve according to YouTube standards. As a result, I can clearly hear the conversation even in my car. Personally, I always listen this way rather than through their website.
That’s why I said Intel should publish those events on YouTube with better audio.
By the way, in the recording, I also normalize the audio levels in DaVinci Resolve according to YouTube standards. As a result, I can clearly hear the conversation even in my car. Personally, I always listen this way rather than through their website.
That’s why I said Intel should publish those events on YouTube with better audio.
Unless you’re not a real person, you should have noticed that the first line of the first post in this thread links to real Q&A conversations, which your YouTube channel copied. If you’re telling us not to trust that, then how are we supposed to trust yours?”
... What you can do:
1. Download the Youtube Video
2. Stripe the audio
3. Apply Whisper
4. Apply diff to your transcription
Anyway, in the video, I didn’t include any extra personal or opinionated comments. It’s meant to help people have a better experience consuming relevant information about Intel.
I am not changing my reasoning; I am simply explaining why I am doing it. It is evident from the statistics that people prefer consuming materials on YouTube. I have to agree because I prefer that approach too.
Once I recorded it in real-time, I enhanced the audio and then uploaded it to YouTube to leverage their live transcription. This way, I can carefully study the material. I then relistened it again while driving to work.
... What you can do:
1. Download the Youtube Video
2. Stripe the audio
3. Apply Whisper
4. Apply diff to your transcription
Anyway, in the video, I didn’t include any extra personal or opinionated comments. It’s meant to help people have a better experience consuming relevant information about Intel.
Or you could vibe-code the four points I listed to verify the content — I guess it could work. I recently uploaded a video demonstrating vibe coding for educational purposes:
By the way, I also highlight relevant Intel products that can be used for LLM inferencing in that video.
chipsandcheese.com
