AMD Adaptive Computing Technology Enables Next-Generation DENSO LiDAR System, Delivering 20X Improvement in Resolution

── AMD Xilinx Auto­mo­ti­ve Zynq UltraS­ca­le+ MPSoC in DENSO plat­form pro­vi­des hig­hest point-cloud den­si­ty level of any SPAD LiDAR sys­tem in mar­ket today ──

SANTA CLARA, Calif., Jan. 19, 2023 (GLOBE NEWSWIRE) — AMD (NASDAQ: AMD) today announ­ced that its adap­ti­ve com­pu­ting tech­no­lo­gy is powe­ring lea­ding mobi­li­ty sup­pli­er DENSO Corporation’s next-gene­ra­ti­on LiDAR plat­form. The new plat­form will enable over 20X1 impro­ve­ment in reso­lu­ti­on with extre­me­ly low laten­cy for increased pre­cis­i­on in detec­ting pede­stri­ans, vehic­les, free space and more. The DENSO LiDAR plat­form, tar­ge­ted to begin ship­ping in 2025, will levera­ge the AMD Xilinx Auto­mo­ti­ve (XA) Zynq™ UltraS­ca­le+™ adap­ti­ve SoC and its func­tion­al safe­ty suite of deve­lo­per tools to enable ISO 26262 ASIL‑B certification.

DENSO is using the AMD XA Zynq UltraS­ca­le+ mul­ti-pro­ces­sor sys­tem-on-a-chip (MPSoC) plat­form in its Sin­gle-Pho­ton Ava­lan­che Diode (SPAD) LiDAR sys­tem, which gene­ra­tes the hig­hest point-cloud den­si­ty level of any LiDAR sys­tem on the mar­ket today2. Point-cloud den­si­ty descri­bes the num­ber of points within a given area and is ana­log­ous to image reso­lu­ti­on, whe­re richer data ensu­res that cru­cial decis­i­on-making details are captured.

Gene­ral­ly, SPAD-based sys­tems are being adopted by auto­ma­kers becau­se of the space savings that can be achie­ved. The high­ly adap­ta­ble XA Zynq UltraS­ca­le+ MPSoC enables DENSO’s LiDAR sys­tems to redu­ce the size of cur­rent LiDAR imple­men­ta­ti­ons, allo­wing mul­ti­ple LiDARs to work in tan­dem for for­ward view and side views of a vehic­le. One device can be used for mul­ti­ple DENSO LiDAR sys­tems, inclu­ding future gene­ra­ti­ons, which dri­ves down sys­tem cos­ts and helps designs be future-ready.

Cur­rent vehic­les in pro­duc­tion may have just one for­ward-loo­king LiDAR, but next-gene­ra­ti­on vehic­les will have mul­ti­ple sys­tems inclu­ding forward‑, rear facing- and side-view LiDARs. The addi­tio­nal sys­tems are nee­ded to move bey­ond dri­ver assis­tance to full auto­no­my. DENSO LiDAR can also be used for infra­struc­tu­re moni­to­ring, fac­to­ry auto­ma­ti­on and other non-auto­ma­ted dri­ving applications.

We are exci­ted to expand our col­la­bo­ra­ti­on with AMD as we intro­du­ce our next-gene­ra­ti­on LiDAR sys­tem,” said Eii­chi Kur­o­ka­wa, head of Sens­ing Sys­tem busi­ness unit, DENSO Cor­po­ra­ti­on. “AMD high-per­for­mance, high­ly sca­lable, pro­gramma­ble sili­con offers distinct bene­fits for the extre­me­ly com­plex image pro­ces­sing requi­re­ments of our LiDAR sen­sor archi­tec­tu­re. The fle­xi­bi­li­ty and capa­bi­li­ties of the Zynq UltraS­ca­le+ MPSoC plat­form and its abili­ty to meet strin­gent func­tion­al safe­ty requi­re­ments led us to work with AMD.”

DENSO’s SPAD LiDAR can gene­ra­te over three mil­li­on points-per-second at 10 frames per-second3. This midd­le-ran­ge LiDAR sys­tem uses XA Zynq UltraS­ca­le+ MPSoCs for sys­tem moni­tor func­tion­a­li­ty to help enable the tem­pe­ra­tu­re and over­all sys­tem to func­tion cor­rect­ly. Becau­se the sys­tem uses time-to-digi­tal con­ver­si­on ins­tead of ana­log-to-digi­tal con­ver­ters, the over­all sys­tem size and cost can be opti­mi­zed while still deli­ve­ring high-per­for­mance high-den­si­ty data.

DENSO has deve­lo­ped an excep­tio­nal­ly pre­cise LiDAR sys­tem. With LiDAR con­ti­nuing to evol­ve, the­re are new tech­no­lo­gy requi­re­ments, dri­ving the need for impro­ve­ments in sen­si­ti­vi­ty, den­si­ty and per­for­mance,” said Mark Wad­ling­ton, seni­or vice pre­si­dent and gene­ral mana­ger, Core Mar­kets Group, AMD. “Through AMD adap­ti­ve com­pu­ting tech­no­lo­gy we’re hel­ping to enable a reduc­tion in sys­tem size and space, while also impro­ving reso­lu­ti­on for increased pre­cis­i­on in object detec­tion, all at very low latency.”

AMD in Automotive
As the pace of inno­va­ti­on con­ti­nues to acce­le­ra­te in the auto­mo­ti­ve indus­try, the need for high-per­for­mance com­pu­te, com­pu­te acce­le­ra­ti­on and gra­phics tech­no­lo­gies is incre­asing. AMD is a lea­der at this inflec­tion point, with a broad line of high-per­for­mance CPUs, GPUs, FPGAs and Adap­ti­ve SoCs. From powe­ring in-vehic­le info­tain­ment sys­tems to advan­ced dri­ver-assis­tance sys­tems, auto­no­mous dri­ving and net­wor­king appli­ca­ti­ons whe­re func­tion­al safe­ty is of para­mount importance, AMD pro­vi­des car­ma­kers with a one-stop shop for sili­con and soft­ware solu­ti­ons. For more infor­ma­ti­on, visit AMD’s Auto­mo­ti­ve web­site.

Sup­port­ing Resources:

About AMD
For more than 50 years AMD has dri­ven inno­va­ti­on in high-per­for­mance com­pu­ting, gra­phics and visua­liza­ti­on tech­no­lo­gies. Bil­li­ons of peo­p­le, lea­ding For­tu­ne 500 busi­nesses and cut­ting-edge sci­en­ti­fic rese­arch insti­tu­ti­ons around the world rely on AMD tech­no­lo­gy dai­ly to impro­ve how they live, work and play. AMD employees are focu­sed on buil­ding lea­der­ship high-per­for­mance and adap­ti­ve pro­ducts that push the boun­da­ries of what is pos­si­ble. For more infor­ma­ti­on about how AMD is enab­ling today and inspi­ring tomor­row, visit the AMD (NASDAQ: AMDweb­siteblogLin­ke­dIn and Twit­ter pages. 

©2023 Advan­ced Micro Devices, Inc. All rights reser­ved. AMD, the AMD Arrow logo, Zynq, UltraS­ca­le+, Xilinx, and com­bi­na­ti­ons the­reof are trade­marks of Advan­ced Micro Devices, Inc. Other names are for infor­ma­tio­nal pur­po­ses only and may be trade­marks of their respec­ti­ve owners.

All per­for­mance and cost savings claims are pro­vi­ded by DENSO and have not been inde­pendent­ly veri­fied by AMD. Per­for­mance and cost bene­fits are impac­ted by a varie­ty of varia­bles. Results her­ein are spe­ci­fic to DENSO and may not be typi­cal. GD-181

———————————————————

1 Reso­lu­ti­on impro­ve­ment vali­da­ted by DENSO during inter­nal evaluation
2 Achie­ve­ment con­firm­ed by DENSO during in-lab testing
3 DENSO obser­ved LiDAR out­put with 2D echo