Intel researchers have developed a prototype interconnect that they say will both increase bandwidth and lower power requirements, whether used for simple CPU-to-CPU connections or scaled up to connect tens of thousands of CPUs in a data center. "Based on past trends, every four years the bandwidth requirements for systems is …
Kindly ask your Neological Dept to check that their creations are readily pronounceable before releasing them into the wild. Thanks.
Say the word 'date', then the word 'apocalypse'; then say them one after the other with no gap between them. I have previous experience of helping foreign students with English speech and writing and this is a useful technique. Are there any other words you'd like some help with?
Sorry, it's not helping.
Dappyclippalips. Dottacapalist. Cataplasdopey. Decalsatopyaqp. Decoypapaplast.
G-G-GYURLS!!!! D-D DRINK!!!!
I'll give you the international phonetic alphabet spelling on top of what the other Frank wrote, in case that helps also. You could very well be truly English language impaired, but I doubt it since you used a word like neological.
Anyway the word is a neologism comprised of two words, Data (deɪ-tə) and Apocalypse (ə-ˈpɑ-kə-ˌlɪps), it would be pronounced as intended as Datapocalypse, or deɪ-təˈpɑ-kə-ˌlɪps.
Guess they've never heard of latency!!
Throughput is one thing, latency is another. This is hardly radical or new as it's been applied to processor cores for years. However, it only works for some types of workload, those where latency isn't a big issue. Not really quite sure what the fuss is about here. I don't see they've really been very radical at all and it's been pretty obvious all along. However, turning down I/O has much more of an impact on the processor core than turning down the core itself due to difference in speeds between the core and the I/O. If the core has to wait much longer (relatively) for the I/O, it'll affect it more. Whereas, slowing the core would just move it nearer the speed of the I/O!!