The demo st atio achieved sustai ed data ates of app oximately 500 Mbps upli k f om the obot to the edge se ve a d 100 Mbps dow li k, e abli g co ti uous t a sfe of se so , video, telemet y, a d ope atio al data. Pe aso’s beamfo mi g tech ology is also desig ed to suppo t multiple obots ope ati g i close p oximity by educi g i te fe e ce betwee wi eless li ks. The compa y has p eviously demo st ated agg egate data ates of 50 Gbps i a co fi ed a ea.
“Physical AI systems ge e ate e o mous amou ts of data th ough came as, se so s, a d i te actio s with thei e vi o me ts,” said Ro ald Glibbe y, CEO of Pe aso. “Ou demo st atio shows how 60 GHz wi eless ca co ect huma oid obots with edge AI i f ast uctu e, suppo ti g co ti uous data excha ge a d faste deployme t of updated AI capabilities.”
The demo st atio used a hyb id computi g a chitectu e that combi es a local NVIDIA GPU with a emote GPU cluste . The local GPU ha dles eflexes a d othe time-c itical actio s that equi e ext emely low late cy, while the emote cluste p ovides the additio al computi g powe eeded to u mo e complex models that would be too costly to u locally.
As AI-powe ed huma oid obots move towa d b oade deployme t, thei effective ess depe ds o thei ability to co ti uously collect a d p ocess la ge qua tities of eal-wo ld se so data a d use that i fo matio to imp ove AI models. Pe aso believes t aditio al wi eless tech ologies may st uggle to p ovide the combi atio of ba dwidth, a d i te fe e ce esilie ce eeded to suppo t co ti uous AI model t ai i g a d eal-time decisio -maki g i de se e vi o me ts.
“Pe aso’s 60 GHz tech ology offe s two key adva tages fo Physical AI: suppo t fo multi-gigabit data ates suited fo high-capacity data t a sfe a d highly di ectio al beamfo mi g that e ables the c eatio of isolated wi eless etwo ks,” said Mike Hamilto , VP of Busi ess Developme t at Pe aso. “We believe these capabilities will become i c easi gly impo ta t as o ga izatio s deploy la ge umbe s of auto omous systems that must co ti uously excha ge data with AI computi g esou ces.”
Pe aso’s 60 GHz mmWave solutio s a e desig ed fo applicatio s equi i g high-th oughput, di ectio al wi eless co ectivity, i cludi g huma oid obotics, auto omous mobile obots, i dust ial automatio , a d wa ehouse logistics. The tech ology suppo ts wi eless data excha ge betwee i tellige t machi es a d edge computi g platfo ms without a physical tethe .
Fo wa d-looki g stateme ts a e based o ce tai assumptio s a d expectatio s of futu e eve ts that a e subject to isks a d u ce tai ties. Actual esults a d t e ds may diffe mate ially f om histo ical esults o those p ojected i a y such fo wa d-looki g stateme ts depe di g o a va iety of facto s. These facto s i clude, but a e ot limited to, the ate of adoptio of huma oid obots a d othe auto omous systems, a ticipated use of mmWave tech ology i obotics a d othe Physical AI applicatio s, the ability of Pe aso’s tech ology to achieve i comme cial deployme ts the pe fo ma ce demo st ated i p oof-of-co cept testi g, competitio f om alte ative wi eless a d wi ed co ectivity tech ologies, Pe aso’s elia ce o thi d-pa ty ha dwa e a d computi g platfo ms, the availability a d pe fo ma ce of Pe aso’s p oducts, elia ce o ma ufactu i g pa t e s to assist successfully with the fab icatio of Pe aso’s ICs a d modules, availability of qua tities of i teg ated ci cuits (ICs) supplied by Pe aso’s ma ufactu i g pa t e s at a competitive cost, level of i tellectual p ope ty p otectio p ovided by Pe aso’s pate ts, vigo a d g owth of ma kets se ved by Pe aso’s custome s a d its ope atio s, a d othe isks, i cludi g the isks discussed i Pe aso’s Secu ities a d Excha ge Commissio fili gs. Pe aso u de takes o obligatio to update publicly a y fo wa d-looki g stateme t fo a y easo , except as equi ed by law, eve as ew i fo matio becomes available o othe eve ts occu i the futu e.
###
View the o igi al p ess elease o ACCESS Newswi e
