Earlier quoted context omitted.
It seems to be paywalled, though.
There is a DOI at the top/right of the paywalled version - just plug that into SciHub (I just verified it's there)
Today I learned:
DOI = 'Digital Object Identifier'
After 50 years of effort, researchers made silicon emit light
31–40 of 108 posts
Re: After 50 years of effort, researchers made silicon emit light
#32Is the breakthrough here about emission or transmission? Or is the physics for these two connected? It's not clear to me, between the title and the article.
Secondly, could using a "photonic" memory bus bring RAM access speeds close to cache speeds, or is the transmission distance/time not the main issue there?
Re: After 50 years of effort, researchers made silicon emit light
#33Re: After 50 years of effort, researchers made silicon emit light
#34On a scale of zero to really big deal. How much of a big deal is this?
Photonic can only be up to 20% faster... And I'm talking about a fully photonic cpu, not an hybrid one that has a light/electricity translation cost.
I don't think the win here will be speed, particularly chip level as much as enabling lower power and higher throughout architectures at the system level. As in this is mote likely to be used for medium range links where infiniband/pcie/nvlink are used today.
Re: After 50 years of effort, researchers made silicon emit light
#35Re: After 50 years of effort, researchers made silicon emit light
#36Is the breakthrough here about emission or transmission? Or is the physics for these two connected? It's not clear to me, between the title and the article. Secondly, could using a "photonic" memory bus bring RAM access speeds close to cache speeds, or is the transmission distance/time not the main issue there?
They were able to measure a strong photoluminescence signal from the silicon germanium nanowires. Photoluminescence is a good proxy for how efficiently it will light up - carriers are generated optically with a light source more energetic than the bandgap of the material. This is much easier than fabricating a full device with electrical contacts.
Silicon is typically a really lousy photon emitter because it’s an indirect bandgap material. Turning an electron/hole pair into a photon requires an interaction with a phonon. It seems by getting the silicon to grow in a hexagonal orientation, it becomes a direct bandgap material leading to much higher emission efficiency.
Re: After 50 years of effort, researchers made silicon emit light
#37Is the breakthrough here about emission or transmission? Or is the physics for these two connected? It's not clear to me, between the title and the article. Secondly, could using a "photonic" memory bus bring RAM access speeds close to cache speeds, or is the transmission distance/time not the main issue there?
Transmission time isn’t really the main issue, it’s more about the work required to get a memory request through the levels of the hierarchy to DRAM and back. Probing each level of cache, propagating through the miss queues, translation (maybe with TLB miss), waiting for the DRAM controller, etc.
Re: After 50 years of effort, researchers made silicon emit light
#38It's kind of fitting that this was done in Eindhoven, a city that formed around the Philips light bulb factory. It's a place that celebrates a history of making stuff glow.
I graduated from TU/e and in my experience their applied physics department, specifically nanomaterials like this are very well funded and attract a lot of international talent.
Eindhoven likes light! And....making bridges out of beer crates: https://www.ed.nl/default/tu-e-studenten-vestigen-nieuw-reco...
Re: After 50 years of effort, researchers made silicon emit light
#39Is the breakthrough here about emission or transmission? Or is the physics for these two connected? It's not clear to me, between the title and the article. Secondly, could using a "photonic" memory bus bring RAM access speeds close to cache speeds, or is the transmission distance/time not the main issue there?
Transmission time isn’t really the main issue, it’s more about the work required to get a memory request through the levels of the hierarchy to DRAM and back. Probing each level of cache, propagating through the miss queues, translation (maybe with TLB miss), waiting for the DRAM controller, etc.
Interesting - if that's the case I would imagine it make sense for some systems to feature an architecture which skips the idea of multilevel cache entirely and has only RAM connected over a photonic bus. No probing, no cache misses.
Re: After 50 years of effort, researchers made silicon emit light
#40Is the breakthrough here about emission or transmission? Or is the physics for these two connected? It's not clear to me, between the title and the article. Secondly, could using a "photonic" memory bus bring RAM access speeds close to cache speeds, or is the transmission distance/time not the main issue there?
They were able to measure a strong photoluminescence signal from the silicon germanium nanowires. Photoluminescence is a good proxy for how efficiently it will light up - carriers are generated optically with a light source more energetic than the bandgap of the material. This is much easier than fabricating a full device with electrical contacts. Silicon is typically a really lousy photon emitter because it’s an ind…
Thank you. There is a lot here I need to read more about.