SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Qualcomm Moderated Thread - please read rules before posting
QCOM 181.05-3.5%Oct 28 3:59 PM EDT

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: kech who wrote (196256)10/28/2025 10:19:34 AM
From: Wildbiftek3 Recommendations

Recommended By
Doug M.
JeffreyHF
waitwatchwander

   of 196297
 
There are real issues with nVidia's inferencing performance and efficiency in their initial client / edge implementation w/ Mediatek:

x.com

This product seems like a rather half-baked effort to put a foot in the opening door of edge inferencing solutions which chokes on its own heat. That said, nVidia's main moat-type advantage is its pervasively used software stack so this may still be a preferred solution despite its obvious power limitations.

Qualcomm (and multiple other competitors that are not AMD or nVIdia) are not hamstrung by nVidia's and AMD's graphics first hardware focus which informs their architecture --there are things like texture mapping units and ROPs that don't contribute (so far) to inferencing or training performance and force the data flow to be routed in a specific way as well. Things that vastly improve efficiency (and hence performance within a thermal envelope) like in/near-memory architectures and photonics would take them too far afield from their core competences and may take the current incumbents longer to implement than the relative upstarts...
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext