Array
(
[content] =>
[params] => Array
(
[0] => /forum/whats-new/latest-activity?before_id=114785
)
[addOns] => Array
(
[DL6/MLTP] => 13
[Hampel/TimeZoneDebug] => 1000070
[SV/ChangePostDate] => 2010200
[SemiWiki/EmailDomainReplace] => 1000010
[SemiWiki/Newsletter] => 1000010
[SemiWiki/WPMenu] => 1000010
[SemiWiki/XPressExtend] => 1000010
[ThemeHouse/XLink] => 1000970
[ThemeHouse/XPress] => 1010570
[XF] => 2031070
[XFI] => 1060170
)
[wordpress] => /var/www/html
)
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 !
JavaScript is disabled. For a better experience, please enable JavaScript in your browser before proceeding.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an
alternative browser .
. comments on Jalapeño by Richard Ho of OpenAI, ex-Google on TPU
. from scratch, full-stack design matters, used OpenAI models .. from...
Thursday at 11:27 PM
F
https://www.trendforce.com/news/2026/06/26/news-micron-reportedly-acknowledges-chinas-memory-progress-while-noting-output-remains-largely...
Thursday at 11:23 PM
T
It would be super easy for Intel to pick up NAND again. They still remember most of the parameters in their head. DRAM, for leading edge...
Thursday at 10:09 PM
B
Oh it's hardly a piece of cake. Transistor structure is all different. High aspect ratio capacitor. You can only get good margins from...
Thursday at 9:01 PM
F
Oh it's hardly a piece of cake. Transistor structure is all different. High aspect ratio capacitor. You can only get good margins from...
Thursday at 8:30 PM
T
Compared with 18A, this is almost straightforward. The bigger challenge is that Lip-Bu Tan may not be deeply familiar with the process...
Thursday at 8:26 PM
F
exactly
Thursday at 8:23 PM
F
DRAM would require substantially new process development and it is also an expensive process. In their recent collaboration with...
Thursday at 8:16 PM
N
. comments on Jalapeño by Richard Ho of OpenAI, ex-Google on TPU
. from scratch, full-stack design matters, used OpenAI models .. from...
Thursday at 8:09 PM
B
IBM leading-edge chip technology development has no internal or external customers.
But it doesn't seem to be going away, as hist78...
Thursday at 8:04 PM
T
Intel could re-enter the memory business if it really wanted to. Compared with 18A, DRAM is arguably much less complex from an execution...
Thursday at 8:02 PM
O
Yep. That's where my head is as well. I do believe that the AI age will permanently increase the demand for RAM; however, the dilemma...
Thursday at 7:00 PM
O
I chalk this up to the idea that most people working on AI models (including my Son who is just finishing up his PhD) work with Python...
Thursday at 6:54 PM
Yes, that is the private feedback I have been getting. 14A test chips are running but not full chip design. The TSMC inner design circle...
Thursday at 6:29 PM
K
One of these IBM news releases is generating a lot of confusion. AFAIK, no one is able to bond two device wafers, one with PMOS and one...
Thursday at 6:18 PM
Daniel Nenni reacted to NY_Sam2's post in the thread OpenAI and Broadcom unveil LLM-optimized inference chip with
Like.
. comments on Jalapeño by Richard Ho of OpenAI, ex-Google on TPU . from scratch, full-stack design matters, used OpenAI models .. from...