Array
(
[content] =>
[params] => Array
(
[0] => /forum/whats-new/latest-activity?before_id=114853
)
[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 .
B
Its brutal
Reportedly thy are cutting even positions that would be building infrastructure, which is interesting in my opinion...
Yesterday at 6:52 PM
What is the metric here for 50% -- cores, sockets, $$ value of CPUs, total racks? is it deployed or last quarters worth?
Yesterday at 6:25 PM
So HBM4E was claimed to be custom logic base die optimized for each customer.
Is that what is actually happening.
I am surprised they...
Yesterday at 6:22 PM
Why this is interesting:
HBM4E is 20% faster than 4 (16Gbps)
Both HBM4 and 4E were introduced this year (2026)
12 layers (48GB) is a...
Yesterday at 6:21 PM
100% agreed. Though if Rapidus is successful at 2nm, I think that could also be considered a success for IBM.
Yesterday at 6:21 PM
"When it comes to IBM, I firmly believe they’re in the camp of saying Moore’s Law is still alive.".
IBM has labs but no fabs. Moore's...
Yesterday at 6:19 PM
B
Are the layoffs due the ridiculous AI financial outlays or due to AI itself? I have a feeling CEO's can now say AI and lay-off with...
Yesterday at 5:41 PM
- Early testing shows that the first-generation accelerator will deliver performance per watt substantially better than current...
Yesterday at 4:38 PM
B
I think the custom logic version is now called C-HBM4E.
Yesterday at 3:38 PM
M
Integrated photonics may be approaching a spark point.
The recent frontier work in ultra-fast modulators, photodetectors, plasmonic...
Yesterday at 12:51 PM
"When it comes to IBM, I firmly believe they’re in the camp of saying Moore’s Law is still alive.".
IBM has labs but no fabs. Moore's...
Yesterday at 12:42 PM
S
14A is less of risk vs A16 cause 18A has derisked lots of the problems for 14A can't say that about A16.
Yesterday at 11:41 AM
Why this is interesting:
HBM4E is 20% faster than 4 (16Gbps)
Both HBM4 and 4E were introduced this year (2026)
12 layers (48GB) is a...
Yesterday at 11:34 AM
April 2026 | Press Coverage
Enterprise Ireland Benelux has highlighted yieldHUB's continued expansion across Belgium and the...
Yesterday at 11:34 AM
yieldHUB has been featured by Enterprise Ireland Korea following the launch of its new website and the introduction of yieldHUB Live...
Yesterday at 11:33 AM
Wow.
Its brutal Reportedly thy are cutting even positions that would be building infrastructure, which is interesting in my opinion...
Xebec replied to the thread SoftBank chipmaker Arm hits 50% share in top AI data centers.What is the metric here for 50% -- cores, sockets, $$ value of CPUs, total racks? is it deployed or last quarters worth?
Like.
So HBM4E was claimed to be custom logic base die optimized for each customer. Is that what is actually happening. I am surprised they...
hist78 reacted to Daniel Nenni's post in the thread OpenAI and Broadcom unveil LLM-optimized inference chip with
AmandaK posted the thread Enterprise Ireland Benelux Highlights yieldHUB's Growth Across the Regional Semiconductor Ecosystem in yieldHUB.April 2026 | Press Coverage Enterprise Ireland Benelux has highlighted yieldHUB's continued expansion across Belgium and the...
