quiz bu lëkkale ak tegtal · Diggante Niveau

Quiz bu am yaatuwaayu mémoire

Saytu HBM stacking, TSVs, interface yu yaatu, yaatuwaayu band ak kàttan, ak HBM4 bi fi nekk ci wetu bi njëkk.

Yooni njiit yu jëm ciMemoire bu yaatu
Laaj 1 de 8

Lan mooy jafe-jafe bi gëna mag bi Memoire Bandwidth bu kawe di saafara ci gaawaayu IA yi?

Xoolaatal laaj bu nekk8 laaj

Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.

  1. Lan mooy jafe-jafe bi gëna mag bi Memoire Bandwidth bu kawe di saafara ci gaawaayu IA yi?
    • calculer cores yu toog inadul def seen liggéey di xaar done yi yegsi
    • GPU yi dañuy tàng lool ndax leeraay bu bari
    • Denc fichier yi ba fàww su kuuraŋ bi dee
    • Keppal GPU bi ak internet bi
  2. Lan moo wuutale HBM ak mémoire GDDR bi?
    • Dañu koy denc ci disk bu dëgër
    • Dafay stack DRAM yu bari dies vertikaal ba noppi boole leen ci vias silicon
    • Dafay jëfandikoo fiil fibre optik ci biir puce bi
    • Amul tranzistor
  3. Lan la HBM gëna yéem ngir mëna am bandwidth bu rëy?
    • Gaawaayu montor bi gëna rëy yeneen mémoire yi
    • Seddal ak azot ndoxe
    • Paralelism bu tar jaaraleko ci interface bu yaatu lool
    • Komprime bépp done ci benn bit
  4. Lan moo waral ñuy def ay pilu HBM ci kaw ab interposer silisium ci wetu GPU bi?
    • Ngir defar kofre bi mu gëna rafet
    • Ndax HBM mënul liggéey lu ëpp benn meetar
    • Ngir may jëfandikukat bi mu dindi ko ci anam wu yomb
    • Ngir fiil yi gëna gàtt, wàññi doole bit bu nekk ak latency
  5. Ci wàllu làkk yu mag, lu tax HBM mëna am solo ni ordinatër bu ñor?
    • Ndax diisaayu model yi dañu leen wara stream saa yu nekk, te loolu dafay laaj kàttan gu bari ak yaatuwaayu band
    • Ndax model yi duñu def benn math
    • Ndax ordinatër bi amul fayda waaye mémoire bi dafay fay juuti
    • Ndax LLM yi CPU bi kese lañuy jëfandikoo
  6. Which HBM generations are part of the progression from HBM2 to currently sampling HBM4 products?
    • HBM2, HBM2E, HBM4E, then HBM3
    • HBM2, HBM2E, HBM3, HBM3E, and HBM4
    • GDDR6, GDDR7, DDR5, and HBM4
    • HBM3E directly followed by HBM1
  7. Why can high HBM bandwidth help an autoregressive language model during token generation?
    • It eliminates the need to store model weights
    • It performs tokenization instead of the GPU
    • It can move model weights and other data to the compute units more quickly
    • It makes every model compute-bound
  8. What is the difference between HBM capacity and HBM bandwidth?
    • They are identical measures of GPU clock speed
    • Capacity measures transfer speed and bandwidth measures data size
    • Neither is related to memory
    • Capacity is how much data can be stored; bandwidth is how quickly data can be transferred