Nvidia ka Server Rack kura bɛna AI Chips baara minnu dilannen don Rivals fɛ
Nvidia ka Server Rack kura bɛna AI Chips baara minnu dilannen don Rivals fɛ Jatigɛwale dɔ kɔnɔ, Nvidia bɛ ka ɲɔgɔndanba min bɛ ka bonya AI puce sugu la, o ɲɛnabɔ, a kɛtɔ ka sèrwɛri rack kura dɔ jira min dabɔra k’a yɛrɛ ka fɛnɲɛnɛmaw dɔrɔn tɛ baara kɛ, nka a ɲɔgɔndankɛlaw ye puwɛntiw fana baara. O wale in bɛ Nvidia bila jɔyɔrɔ la walasa ka tɔnɔ sɔrɔ AI ka fɛnsɔrɔsiraw bɛɛ lajɛlen na, hali n’a y’a sɔrɔ ɲɔgɔndankɛlaw i n’a fɔ AMD ni Intel k’a ka fanga sɔsɔ. Laseli in kɛra GTC ka san o san yiriwalikɛlaw ka lajɛba la, o b’a jira ko fɛnba dɔ bɛna wuli ka bɔ puwɛntiw dɔrɔn feereli la, ka taa ɲɔgɔndan nafamaba di bi kunnafonidilanw ma.
Nvidia ka Pivot Stratégique ka sugu la min bɛ ɲɔgɔn sɔrɔ AI teliya sugu bɛ ka sumaya joona. San caman kɔnɔ, Nvidia ka GPUw ye hakililata camanba in motɛri ye min tɛ sɔsɔ. Nka, jamana cogoya bɛ ka Changé ka da a kan tekinoloji tɔnbabaw ni semiconducteur ɲɔgɔndanw bɛ wari caman bila u yɛrɛ ka custom AI puwɛntiw dilanni na, minnu bɛ wele tuma caman na ko ASICs. O cogoya in, n’a bɛ fɔ a ma ko "silicon diversification", o bɛ Nvidia ka suguya sɔrɔta bila farati la. Baarakɛda minnu bɛ i n’a fɔ Google, Amazon ani Microsoft, olu bɛ ka puwɛntiw dilan minnu bɛ bɛn u ka baara kɛrɛnkɛrɛnnenw ma. O waati kelen na, AMD ni Intel bɛ ka GPU wɛrɛw labɛn minnu bɛ se ka ɲɔgɔn sɔrɔ. Nvidia ka sèrwɛri rack kura ye jaabi tilennen ye nin ɲɔgɔndan degun in na, k’a Jira k’a ka fɛɛrɛbɔ bɛ To cɛmancɛ la hali n’a ka silikon bɛ sèrwɛri kɔnɔ.
K’a Tà Chip Feerekɛla la ka Taa Plateforme Dilanbaga la O fɛɛrɛ in bɛ Nvidia kɔrɔta ka bɔ fɛnw dilanbaga la ka kɛ plateforme dilanbaga nafama ye. Ni a ye rack system dɔ Dabɔ min ye agnostic ye puce feerela fɛ, Nvidia bɛ pari kɛ a ka networking ni software stack kan k’a kɛ a ka moat fɔlɔ ye. Sosiyete ka rezow fɛɛrɛbɔ, kɛrɛnkɛrɛnnenya la a ka InfiniBand ni Spectrum-X Ethernet platforms, o jatera ko a ka fisa ni tɔw bɛɛ ye AI kuluw la minnu bɛ baara kɛ kosɛbɛ. Rack kura in bɛ Nvidia sɛmɛntiya tiɲɛ na ka kɛ AI kunnafonidilan ka "central nervous system" ye. Hali ni kiliyan dɔ bɛ baara Kɛ ni ɲɔgɔndankɛla dɔ ka AI puce ye, u bɛ Se k’u jigi Da Nvidia kan halibi walasa k’a Jira ko o pucew bɛ Se ka kumaɲɔgɔnya Kɛ teliya ni hakɛ la min ka kan ka Kɛ kanko misalibaw degeliw la. O bɛ sɔrɔko barikama dɔ lawuli, min bɛ segin-ka-bɔ, min tɛ se ka kɛ kosɛbɛ puwɛntiw ka ɲɔgɔndan na.
Server Rack Architecture kura kɔnɔ Soso minnu bɛ porozɛ in dɔn, olu ka fɔ la, sèrwɛri kura in ka kokuradonni jɔnjɔn bɛ a ka rezow jɛlen fini na. O fɛɛrɛ in ye gundo sosɛti ye min bɛ se ka AI puce kɛmɛ caman siri ɲɔgɔn na — a kɛra Nvidia fɛ wo, a kɛra a dilabaga wɛrɛ ye wo — ka kɛ ɔridinatɛriba kelen ye, min bɛ bɛn ɲɔgɔn ma. Rack in dilannen don walasa ka buteliba ɲɛnabɔ AI belebeleba kɔnɔ: pucew ni ɲɔgɔn cɛ kumaɲɔgɔnya. Ka modeli kura dege, o bɛ puwɛnti ba caman de wajibiya walasa u ka baara kɛ ɲɔgɔn fɛ cogo dafalen na. Donanw cili dɔɔnin dɔɔnin walima dannaya tɛ min na puwɛntiw ni ɲɔgɔn cɛ, o bɛ se ka baarakɛcogo ni baarakɛcogo ɲuman tiɲɛ, o bɛ kɛ sababu ye ka rezow layini kɛ fɛnba ye i n’a fɔ jatebɔ siliyɔmu yɛrɛ.
Yɔrɔ kolomaw ni fɛɛrɛ kuraw Rack in bɛ Nvidia ka dɔnniya juguw ta ɔridinatɛri baarakɛcogo ɲuman na. Hali ni fɛn kɛrɛnkɛrɛnnenw tɔgɔ ma fɔ sɔrɔyɔrɔw fɛ, a ka c’a la, o sistɛmu bɛ Nvidia fɛɛrɛ jɔnjɔn caman don a kɔnɔ: High-Bandwidth Networking Switches: Switches customs minnu sinsinnen bɛ Nvidia ka Spectrum walima InfiniBand architecture kan walasa ka kunnafonidilanw bolicogo belebelebaw ɲɛnabɔ. NVLink Teknoloji: A bɛrɛbɛnna walasa ka jɛɲɔgɔnya teliyalenba, tilennen di GPUw ni AI teliyalan wɛrɛw cɛ minnu bɛ se ka kɛ. Universal Chassis Design: Rack form factor standardized min bɛ se ka OEM server tray suguya caman bila minnu kɔnɔ feerekɛlaw ka puce suguya wɛrɛw bɛ sɔrɔ. Unified Management Software: Sisitɛmu porogaramu min bɛ baara kɛtaw labɛn, ka kɛnɛya kɔlɔsi, ani ka baara kɛcogo ɲuman ɲɛnabɔ siliyɔmu ɲagaminen caman kɔnɔ. O fɛɛrɛ in bɛ kunnafonidilanw baarakɛlaw dɛmɛ u ka se ka fɛn caman sɛgɛsɛgɛ minnu ɲɔgɔn ma deli ka kɛ fɔlɔ. U bɛ se ka fɛnɲɛnɛmafagalanw ɲagami ɲɔgɔn na ka da musaka, baarakɛcogo, walima baara kɛrɛnkɛrɛnnenw wajibiyalenw kan, o bɛɛ kɛra k’a sɔrɔ u bɛ fɛnsɔrɔsiraw kelen mara min bɛ ɲɛnabɔ Nvidia ka platform fɛ.
A nɔfɛkow ka bon kosɛbɛ AI baarakɛyɔrɔ la Nvidia ka nin wale in bɛ na ni ripple effects belebele ye fɛɛrɛtigɛ siratigɛ la. A bɛ sɔn ani ka teliya ka taa a fɛ ka ɲɛsin feerekɛla caman ka AI fɛnɲɛnamafagalan sigidaw ma. Baarakɛda kuntigiw fɛ, nin bɛ se ka kɛ kibaru dumanba ye, bawo a bɛ layidu ta ko sugandili caman bɛ kɛ, ka musakaw dɔgɔya, ani ka dɔ bɔ feerekɛlaw ka dadonni na. Nka, a fana bɛ Nvidia fanga sabati stack (dakun) layɛrɛ wɛrɛ la. Ni a yeka rezow ni ɲɔgɔndɛmɛ sira kɔlɔsi, Nvidia bɛ sariyaw sigi baarakɛcogo ni baarakɛɲɔgɔnya siratigɛ la. Sisan, puwɛnti dilannikɛlaw minnu bɛ ɲɔgɔn sɔrɔ, olu ka kan k’a lajɛ k’u ka fɛn dilannenw bɛ baara kɛ cogo la min tɛ fɛn tiɲɛ, fɛnsɔrɔsiraw hukumu kɔnɔ, u ɲɔgɔndanba bɛ se ka min ɲɛfɔ.
A New Era of Coopetition (Jɛɲɔgɔnya waati kura). O ko in bɛ "coopetition" dinamikɛ gɛlɛn dɔ Dabɔ. AMD ɲɔgɔnna ɲɔgɔndanw bɛna ɲɔgɔn Sɔrɔ kosɛbɛ ni Nvidia ye walasa k’u ka MI300X walima AI puce nataw feere. O bɛɛ n’a ta, u bɛ Se k’a Sɔrɔ fana k’u ka puwɛntiw bɛ Lase Nvidia ka rack systems kɔnɔ. O bɛ fɛɛrɛw jira minnu bɛ ye tekinoloji seko ni dɔnko siratigɛ wɛrɛw la, yɔrɔ min na, tɔnw bɛ ɲɔgɔn sɔrɔ fɛn kelen kan, k’a sɔrɔ u bɛ jɛ ka baara kɛ ɲɔgɔn fɛ, min b’a to a bɛɛ bɛ baara kɛ. Baarakɛla labanw fɛ, nafa laban ka kan ka kɛ kokuradonni teliya ye ani ɔridinatɛri AI barikama, min bɛ baara kɛ ka ɲɛ. Instagram SEO nafama bɛ ɲini cogo min na ka platform in ka algorisimuw faamuya walasa ka sɔrɔ, ka ɲɛtaa sɔrɔ AI infrastructure la, o bɛna a ɲini ka faamuyali sɔrɔ cogo min na ka ɲɛsin Nvidia ka rezow paradigm kura ma, a mana kɛ cogo o cogo, puce architecture min bɛ a kɔnɔ.
Kuncɛli: A Masterstroke in Ecosystem Control Nvidia ka sèrwɛri rack kura ye masterstroke ye stratégique positionnement la. Sani a ka kɛlɛ lafasali kɛ AI puce kura donbaga bɛɛ kama, a ka baarakɛda bɛ ka siraba in jɔ min nafa ka bon kosɛbɛ, AI siraba bɛɛ ka kan ka taa min kan. O b’a to a nafa ka bon ani a ka sɔrɔ bonya hali siliyɔmu sugu tilalen na. Nin fɛɛrɛ in ɲɛtaa bɛna bɔ izini caman tali la ani ni ɲɔgɔndankɛlaw sɔnna Nvidia ka platform ma i n’a fɔ sariya min tɛ mɔgɔ si ta ye. Fɛn kelen jɛlen don : AI ka fanga kɛlɛli tɛ kɛ dɔrɔn tugun ko jɔn de ka puce teliya ka bon, nka jɔn bɛ kunnafonidilan ka ɲɔgɔndan fini mara. Jagokɛlaw minnu b’a fɛ ka AI nafa, ka to ɲɛfɛ, o kɔrɔ ye ka nin fɛnsɔrɔko jiginni ninnu faamuya. Walasa ka hakilina caman sɔrɔ ka ɲɛsin ɲɔgɔndan tekinoloji yɔrɔw taamacogo ma, aw k’an ka sɛgɛsɛgɛliw sɛgɛsɛgɛ yecogo sɔrɔli kan sigidaw la minnu bɛ ka wuli. Yala i labɛnnen don k’i ka nizɛri fɛɛrɛ tigɛ ka ɲɛ wa? A to Seemless k’i dɛmɛ ka labɛn dɔ jɔ min bɛ se ka siniko ɲɛnabɔ. Aw ye kuma an ka jɛkulu fɛ bi walasa ka lajɛ kɛ.
