Ikibazo gihuza ibibazo · Hagati Urwego

Ikibazo cya Cerebras

Irasobanura wafer-igipimo cyo kwishyira hamwe, kuri-wafer itumanaho, kurengerwa, hamwe na verisiyo yihariye ya WSE-3 na CS-4 ibisobanuro birambuye.

Inzira zijyanye no kuyoboraSisitemu ya Cerebras
Ikibazo 1 Bya 8

Niki kidasanzwe mubijyanye na moteri ya Cerebras Wafer-Scale Moteri?

Reba buri kibazo8 ibibazo

Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.

  1. Niki kidasanzwe mubijyanye na moteri ya Cerebras Wafer-Scale Moteri?
    • Nibikorwa byose byubatswe kuri wafer imwe ya silicon yuzuye
    • Nibikoresho bya AI bito ku isi
    • Ikora rwose nta mashanyarazi
    • Ikozwe muri plastiki aho kuba silicon
  2. Ni ukubera iki kubika ibice byose kuri wafer bifasha imikorere ya AI?
    • Bituma chip ihendutse kubyohereza
    • Amakuru yimuka hejuru yinsinga za chip aho guhuza buhoro buhoro
    • Bikuraho gukenera software iyo ari yo yose
    • Yemerera chip kureremba mumazi
  3. Hafi ya transistor zingahe WSE-3 irimo?
    • Abagera ku bihumbi 4
    • Abagera kuri miliyoni 4
    • Hafi ya tiriyari 4
    • Abagera kuri magana ane
  4. What is one Cerebras system-level capability advertised for its CS-4 rack-scale system?
    • It uses only a single CPU and no AI accelerator
    • It replaces every model with a database
    • It guarantees the same speedup on every workload
    • It combines multiple wafer-scale processors in a rack-scale platform
  5. Nigute Cerebras ikemura ibibazo byo gukora kuri chip nini?
    • Ikoresha ingirakamaro hamwe ninzira yo gukora hafi yinenge
    • Ijugunya buri wafer ifite inenge iyo ari yo yose
    • Irengagiza inenge n'amato yamenetse
    • Nintoki isana buri tristoriste mukiganza
  6. What are the WSE-3 specifications Cerebras reports in its product materials?
    • Four billion transistors and 900 cores
    • Four trillion transistors and about 900,000 AI-optimized cores
    • One thousand transistors and one million CPU sockets
    • A fixed count of parameters for every model
  7. How can a wafer-scale processor route around manufacturing defects?
    • Assume every fabricated core is perfect
    • Move all computation to a hard drive
    • Use redundant cores and routing so defective regions can be bypassed
    • Rebuild the silicon after deployment
  8. Why can on-wafer memory and communication benefit some AI workloads?
    • They eliminate every computation bottleneck
    • They make all models fit in one core
    • They guarantee lower latency for every task
    • They can reduce data movement between separately packaged processors and memory