Chii chaitika
Groq announced on August 24 that it will be among the first adopters of Nvidia Groq 3 LPX and plans to deploy the hardware alongside Nvidia Vera Rubin NVL72 in its purpose-built AI cloud. Groq said Dell Technologies will help deploy the systems and that the capacity will eventually be made available to developers, enterprises and AI companies using GroqCloud.
Groq akati ichave pakati pevakatanga kutora Nvidia Groq 3 LPX, iyo yaakatsanangura seanosimudzira accelerator chinangwa chekuwedzera token-chizvarwa kumhanya paNvidia Vera Rubin NVL72 masisitimu. Groq akati iri kushanda neDell Technologies kuendesa zvivakwa muchinangwa chayo-chakavakwa AI Cloud. Chiziviso chinotsanangura kuwedzera kwakarongwa, kwete kutanga kwapera: kana huwandu hwauya pamhepo, kambani yakati ichapa nzira yekutanga yemabhizinesi nemakambani eAI kushandisa Hardware pamabasa ekugadzira.
Makambani akaronga kutenderera kwakadzokororwa, nekukurumidza modhi mhinduro kune yakakura-mamiriro ekushandisa uye agentiki AI, umo software inogona kuburitsa akawanda modhi mafoni paunenge uchiita basa. Groq akataura nezveNvidia-yakadhindwa manhamba ekuita, kusanganisira 3,400 yekubuda tokens pasekondi inomhanya Gemma 4 31B ine 100,000-chiratidzo mamiriro muArtificial Analysis bhenji. Groq akadzokorodzawo zvichemo zvakapetwa kana zvakanyanya kupindirana pane iri padyo imwe chikuva uye akati mamwe mabasa ekukodha anogona kupedzwa mumaminitsi kwete maawa.
Nhamba idzodzo zvirevo zvakaunzwa naGroq uye zvakaverengerwa muchikamu kune zvakaburitswa neNvidia; iyo sosi inopa hapana bvunzo yekumisikidza, kuenzanisa Hardware, latency kugovera, fungidziro yemitengo kana yakazvimirira kudzokorora. Izvo zvakare hazvitaure kana iyo yakadudzwa bhenji inoratidza Groq's deployed kumisikidzwa kana yakafara Vera Rubin chikuva mhedzisiro. Chaiyo-yepasirese performance inoenderana nemhando yekuvaka, yekupinda uye kureba kwekubuda, batching, networking, software optimization uye sevhisi-level zvipingaidzo.
Groq akati vanopfuura mamirioni matanhatu vanogadzira, mabhizinesi eFortune 500 uye zviuru zvemakambani eAI-ekuzvarwa akavakira paGroqCloud, achigadzira matiririyoni ematokeni svondo rega rega munzvimbo dzedata muNorth America, Europe, Middle East neAsia-Pacific. Kwakabva hakungozvimiririra kunyora iwo manhamba kana kuziva vatengi. Inotiwo Groq yakave Nvidia Cloud Partner muna Nyamavhuvhu, ichiibvumira kugadzira, kutumira uye kushandisa yakakurumidza komputa zvichibva paNvidia's referensi yekuvaka uye maitiro ekushanda. Basa raDell rinotsanangurwa sekupa zvivakwa zvakabatanidzwa uye kugona kwepasirese-yekugovera, asi chiziviso chinopa hapana huwandu hwehurongwa, nzvimbo dzezvivakwa kana nguva yekuendesa.
Nei zvichikosha
The announcement points to a growing effort to build infrastructure specifically for : the stage when trained AI models generate responses for users and applications. Faster inference could make long-context systems and AI agents more responsive, but the public evidence here is limited to company and vendor claims. The announcement does not establish when the systems will be available, how much they will cost or how they will perform across workloads beyond the cited .
zvivakwa zvinowedzera kuona kuti vanhu nesoftware vanogona kukurumidza sei kushandisa maAI modhi mushure mekunge iwo mamodheru adzidziswa. Kune yakajairika chatbot, yakaderera mhinduro latency inogona kuita kuti kupindirana kunzwe kwekukurumidza. Kune mumiririri weAI anoronga, anotora ruzivo, anonyora kodhi kana kufonera mamwe masevhisi, kukurumidza chizvarwa kunogona kuderedza nguva yakashandiswa kumirira pamatanho akawanda anoteedzana. Chiziviso chaGroq saka chiri pamusoro peiyo inoshanda yekuendesa layer yeAI, kwete modhi nyowani kana kugona kutsva kunoratidzwa neiyo modhi pachayo.
Kuendeswa kwakarongwa kunoratidzawo kuti AI zvivakwa zviri kuwedzera sei hunyanzvi. Groq iri kumisikidza mayuniti ayo ekugadzirisa mutauro uye mashandiro egore akatenderedza chizvarwa chechiratidzo, nepo Nvidia naDell vanounzwa sevabatsiri vanopa chikuva chekusimudzira uye masisitimu akabatanidzwa. Kana hurongwa uhu huchishanda sekutsanangurwa kwazvinoitwa, vatengi vanogona kuwana mukana wehunyanzvi hwekufungidzira pasina kuvaka uye kushandisa hardware ivo pachavo. Izvo zvingave zvine basa kumakambani ari kutsvaga nguva dzinofanotaurwa dzekupindura sevhisi yevatengi, kusimudzira software kana mamwe maagent-based application.
Kukosha kwechido cheruzhinji kunogumira pane zvinotaurwa nemusoro wenyaya. Iyo sosi hairatidze kuti kukurumidza kugadzirwa kwetokeni kunoburitsa mhinduro dziri nani, vamiririri vakachengeteka kana kudzikisa mitengo yese. Kumhanya kwekubuda chikamu chimwe chete chebasa reAI. Vashandisi vanodawo kukwana kwakakwana, kubata mamiriro, nguva, chengetedzo, data-kutonga kutonga uye mitengo inofanotaurwa. Masisitimu anokurumidza anogona kutowedzera kudiwa kwezvivakwa kana vakaita kuti ive nemari yekumhanyisa mafoni emhando kana kufambidzana kwenguva refu.
Chiziviso ichi chakakoshawo kumakwikwi muAI computing. Iyo yekuchengetedza yemukati inosanganisira mimwe mishumo ichangoburwa nezve Nvidia masisitimu, kusanganisira chirevo chakasiyana chekuti Indian AI data-center femu yakaraira Vera Rubin masisitimu. Icho chiitiko chakasiyana, kwete humbowo hwekuti kutumirwa kwaGroq kwatanga. Kutorwa pamwe chete, zviitiko zvakadaro zvinoratidza kufarira kwemusika mune nyowani AI zvivakwa, asi sosi iyi yega haigone kumisa musika wekugovanisa, kuwanikwa kwekupa kana kuti Groq ichawana mukana wezvinhu pamusoro pevamwe vanopa makore.
Kune vatengi, mubvunzo unoshanda hausi wekungoti hardware inogona kuburitsa yakakwira bhenji nhamba. Izvo ndezvekuti vagadziri vanogona kuiwana kuburikidza neyakagadzika APIs, pamutengo unotsigira mitoro yavo yebasa, mumatunhu mavanoshanda, nezvisungo zvebasa zvavanogona kuvimba nazvo. Hapana kana imwe yemamiriro aya inotsanangurwa muchiziviso. Iyo sosi zvakare haiburitse mashandisiro emagetsi, kutonhora zvinodiwa, zvirambidzo zvemodhi kana chengetedzo uye zvirevo zvekuvanzika zvinotonga kutumirwa kwebhizinesi.
Interactive Mechanism: Iyo Inonyatsoshanda
Ongorora ari pasi tekinoroji kuseri kwekusimudzira uku uchipindirana.
Which component of an AI application is the machine-learning model itself?
Zvekutarisa zvinotevera
The key test will be whether Groq turns the announced hardware partnership into broadly accessible production capacity. Watch for a deployment schedule, pricing, service regions, customer access, independent benchmarks and evidence from real workloads. It also remains unclear how the new systems will compare with competing accelerators on total cost, energy use, reliability and the range of models they can support.
Chekutanga, tarisa humbowo hwekuti chiziviso chasimuka kubva mumutauro wekudyidzana kuenda pakushanda. Groq haina kupa zuva rekuwanikwa kwekutanga, nhamba yehurongwa hwainotarisira kuendesa, nzvimbo dzenzvimbo dzakakodzera data kana chikamu cheGroqCloud traffic iyo ichashandisa iyo nyowani hardware. Chiziviso chenguva yakazotevera, zvinyorwa zvebasa kana kuburitswa kwevatengi kunojekesa mamiriro eprojekiti.
Chechipiri, tarisa kuongororwa kwakazvimirira kwezvichemo zvekuita. Kuenzanisa kunobatsira kunoshuma kupera-kusvika-kumagumo latency, nguva yekutanga tokeni, yakaramba ichibuda nekukurumidza, kubuda pasi pekuda panguva imwe chete, kuita mukureba kwakasiyana uye mhedzisiro mumamodeli akati wandei. Vanofanirawo kutaura chikuva chinokwikwidza, software stack uye fungidziro dzemitengo. Iyo sosi ye3,400-token nhamba uye ina-kanguva yekudyidzana kudai haigone kupindura iyo mibvunzo yega.
Chechitatu, vatengi nevaongorori vanofanirwa kuongorora hupfumi. Chiziviso chinoti chikuva chakagadzirirwa scalable, yakaderera-mutengo agent AI, ichitora Nvidia mukuru, asi haipe mutengo pa tokeni, zvibvumirano zvechibvumirano, fungidziro dzekushandisa kana mutengo wakakwana wevaridzi. Kuburitswa kweramangwana kunogona kuratidza kana nyanzvi hardware inodzikisa mutengo mumitoro yebasa rakajairwa kana kuti mabhenefiti ayo akaiswa mune mamwe mamodheru uye maitiro emigwagwa.
Chechina, tarisa kushanda uye tekinoroji miganho yekutumirwa. Groq anoti ichashandisa Nvidia's reference architecture uye Dell's integrated infrastructure, asi sosi yacho haitsananguri kuenderana kwesoftware, zvinodiwa zvekutama, tsigiro yemhando, kuregererwa, kugara kwedata kana hurongwa hwekupindura chiitiko. Iwo mameseji anozoona kana mabhizinesi achigona kushandisa iyo inonzwisisika kana bhizinesi-yakakosha maapplication.
Chekupedzisira, tarisa kana iyo nyowani nyowani ichichinja magadzirirwo anovaka maAI vamiririri. Kana kunonoka kunonoka kunowanikwa nguva dzose, vagadziri vanogona kushandisa mamiriro akareba, maturusi akawanda ekufona kana mamwe matanho ekufunga anodzokorora. Izvo zvinogona kuvandudza kuterera mune zvimwe zvikumbiro uku uchiwedzera mashandisirwo etokeni uye kudiwa kwezvivakwa. Chiziviso ichi hachipi humbowo parizvino nezve iwo ezasi mhedzisiro, saka ivo vanoramba vari mikana yekuyedza pane yakagadziriswa mhedzisiro.