TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

IBM's NorthPole is faster and uses less energy than Nvidia's best AI GPU

30 pointsby danboarderover 1 year ago

3 comments

Ukvover 1 year ago
To my understanding this is a small low-energy chip designed for applications like basic object detection in a doorbell camera, or speech recognition for a home assistant.<p>I think it&#x27;s a bit misleading to say &quot;faster [...] than Nvidia&#x27;s best AI GPU&quot; and &quot;outperofrms[sic] all major architectures&quot; - it specifically means latency and energy efficiency. Its throughput&#x2F;memory are so tiny that it&#x27;s not at all a replacement for Nvidia&#x27;s best AI GPUs (and not intended to be):<p>&gt; NorthPole&#x27;s core array includes 192 MB of flexible memory (768KB of unified memory per core). Assigning 2&#x2F;3rd of this memory to parameters, such as weights and biases, provides 128MB for network storage.<p><a href="https:&#x2F;&#x2F;www.science.org&#x2F;action&#x2F;downloadSupplement?doi=10.1126%2Fscience.adh1174&amp;file=science.adh1174_sm.pdf" rel="nofollow noreferrer">https:&#x2F;&#x2F;www.science.org&#x2F;action&#x2F;downloadSupplement?doi=10.112...</a>
sillywalkover 1 year ago
&quot;It&#x27;s worth clarifying a few things early here. First, NorthPole does nothing to help the energy demand in training a neural network; it&#x27;s purely designed for execution. Second, it is not a general AI processor; it&#x27;s specifically designed for inference-focused neural networks. As noted above, inferences include things like figuring out the contents of an image or audio clip so they have a large range of uses, but this chip may do you any good if your needs include running a large language model because they&#x27;re too large to fit in the hardware.&quot; [0]<p>[0] <a href="https:&#x2F;&#x2F;arstechnica.com&#x2F;science&#x2F;2023&#x2F;10&#x2F;ibm-has-made-a-new-highly-efficient-ai-processor&#x2F;" rel="nofollow noreferrer">https:&#x2F;&#x2F;arstechnica.com&#x2F;science&#x2F;2023&#x2F;10&#x2F;ibm-has-made-a-new-h...</a>
eggyover 1 year ago
IBM was always a powerhouse of R&amp;D, so glad to see this to keep the field competitive between the few players in the game. I love APL and that grew out of Ken Iverson&#x27;s work at Harvard, but then completed at IBM with double the salary [1].<p><pre><code> [1] https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Kenneth_E._Iverson</code></pre>