Microsoft Takes On Other Clouds With “Braga” Maia 200 AI Compute Engines
… As before, we think that nine of those 56 lanes are used in the all-to-all links to make a Braga quad system board. …
Search
Related topics:
Link Aggregation
Tracked topic
Link Aggregation
Arista Modular Switches Aim At Scale Across Networks, Hit Scale Out, Too
… It has 304 SerDes running at 100 Gb/sec with PAM-4 encoding, with 144 SerDes being used for downlinks and 160 SerDes being used to uplink to 51.2 Tb/sec Ramon 3 spine fabric switches. …
With TPU 8, Google Makes GenAI Systems Much Better, Not Just Bigger
… Given the Zebrafish 8i chip’s focus on inference, Google came up with a new hierarchical network topology for the interchip interconnect ICI that it has used to link machines at the board level and sometimes at the rack level in its TPU systems. …
Rebellions AI Rings Up The Money To Rack Up AI Inference Systems
… This picture of the RebelRacks shows them at a higher density: With this picture of the RebelPod, you can't see inside the racks, so it is hard to say how many machines are in each rack: RebelPod takes a bunch of RebelRacks and interlinks them with an 800 GB/sec Ethernet back-end network, with a 25… …
OpenAI, Microsoft And Friends Build A Better, More Scalable Ethernet
… If you have enough links, as it turns out, and the right protocol, you an heal around link failures and while the AI training job slows down, there are enough ways to reroute traffic that the network can heal around the link failure. …
Driving Down The AI System Roadmap With Nvidia
… Decades ago, Cray supercomputers had copper backplanes in the rack and optical links coming off their routers to interlink the racks. …
Bechtolsheim & Friends Breathe Life Into Pluggable Optics One Last Time
… Let’s say that you want to have a row of compute and networking based on Ethernet that links 512 XPUs together and that you want to use switches with top-of-the-line 1.6 Tb/sec OSFP ports. …
Why Hyperscale Networks Have Been Decades In The Making
… It’s either a local data center in the region, or we are partnering with the local service providers.” Building networks over years and years But if the benefits of an intimate understanding of technology or regional business links might be clear, others might be less obvious. “You need to make cer… …
Nvidia Finally Admits Why It Shelled Out $20 Billion For Groq
… But look at the complexity of the GPU, which is directly proportional to its costs – and most of the bill of materials for the R200 will cover the cost of the HBM4 stacked memory and the interposer it requires to link it to the GPU. …
Arm Comes Full Circle With Homegrown, AI-Tuned Server CPU
… The hyperscalers and cloud builders all want to pair their Arm CPUs to Nvidia GPUs, AMD GPUs, and homegrown XPUs, and in some cases they will use licensed NVLink Fusion ports and in others they will use UALink or ESUN over Ethernet. …
To show you the most relevant results, we’ve omitted some entries very similar to those already shown. Repeat the search with the omitted results included.