Kategorie: News

New AMD Instinct™ MI200 Series Accelerators Bring Leadership HPC and AI Performance to Power Exascale Systems and More

Novem­ber 08, 2021 12:00pm EST

- With new AMD CDNA™ 2 archi­tec­tu­re, AMD Instinct MI200 series acce­le­ra­tors deli­ver ground-brea­king 4.9x advan­ta­ge in HPC per­for­mance1 com­pared to com­pe­ting data cen­ter acce­le­ra­tors, expe­diting sci­ence and discovery -

- MI200 series acce­le­ra­tors are first mul­ti-die GPU, first to sup­port 128GB of HBM2e memo­ry, and deli­ver a sub­stan­ti­al boost for appli­ca­ti­ons cri­ti­cal to the foun­da­ti­on of science -

SANTA CLARA, Calif., Nov. 08, 2021 (GLOBE NEWSWIRE) — AMD (NASDAQ: AMD) today announ­ced the new AMD Instinct™ MI200 series acce­le­ra­tors, the first exas­ca­le-class GPU acce­le­ra­tors. AMD Instinct MI200 series acce­le­ra­tors includes the world’s fas­test high per­for­mance com­pu­ting (HPC) and arti­fi­ci­al intel­li­gence (AI) acce­le­ra­tor,1 the AMD Instinct™ MI250X. (…) Wei­ter­le­sen »

Was zeigt AMD morgen — Trento, Instinct MI250(X), OAM, Milan‑X?

Nach­dem in den letz­ten Wochen bereits ers­te Leaks zu den kom­men­den MI200-Gra­fik­be­schleu­ni­gern ver­öf­fent­licht wur­den, hat AMD vor Kur­zem für den mor­gi­gen 8. Novem­ber ein Event ange­kün­digt, dass sich dem The­ma Daten­zen­tren und CPU- und GPU-Beschleu­ni­gern wid­men soll. Dabei dürf­ten vor allem die bereits für den anste­hen­den Exas­ca­le-Super­com­pu­ter “Fron­tier” aus­ge­lie­fer­ten spe­zi­el­len Pro­duk­te vor­ge­stellt wer­den. Ein Tweet von AMDs Prä­si­den­tin und CEO Dr. Lisa Su gibt aber wohl auch Hin­wei­se dar­auf, dass ein OAM-Beschleu­ni­ger zum Ein­satz kommt. (…) Wei­ter­le­sen »

Neue Downloads der KW 43 und 44

In den letz­ten zwei Wochen wur­de der AMD-Gra­fik­trei­ber drei­mal aktua­li­siert, vor allem um die Per­for­mance bei neu­en Spie­len wie Marvel’s Guar­di­ans of the Gala­xy und Age of Empires IV zu ver­bes­sern. Die ein­zel­nen Ver­bes­se­run­gen kann man in der Ver­si­ons­his­to­rie über­sicht­lich nach­voll­zie­hen. Zusätz­lich gab es noch neue Ver­si­on der PowerT­oys von CapFrameX und auch Nvi­di­as Gra­fik­trei­ber wur­de aktua­li­siert. HWiNFO64 und ande­re Infor­ma­ti­ons­pro­gram­me wie CPU‑Z unter­stüt­zen nun auch Intels Alder Lake Pro­zes­so­ren bzw. haben die­se Unter­stüt­zung verbessert.

Neue Versionen

Neu aufgenommene Programme

  • kei­ne

(…) Wei­ter­le­sen »

Mercury Research: AMD mit Rekordmarktanteil bei Notebook-Prozessoren

Laut neu­es­ten Zah­len von Mer­cu­ry Rese­arch für das drit­te Quar­tal 2021 hat AMD im Bereich Mobi­le (Note­books) einen neu­en Rekord beim Markt­an­teil erreicht. Zule­gen konn­te man eben­falls bei den Ser­ver-Pro­zes­so­ren sowie dank der neu­en Kon­so­len­ge­ne­ra­ti­on von Sony und Micro­soft beim Gesamt­an­teil der x86-Pro­zes­so­ren. Ledig­lich im Seg­ment der Desk­top-Pro­zes­so­ren wur­de ein leich­ter Rück­gang von 17,1 auf 17,0 Pro­zent Markt­an­teil beobachtet.
(…) Wei­ter­le­sen »

AMD Radeon™ PRO V620 GPU Delivers Powerful, Multi-Purpose Data Center Visual Performance for Today’s Demanding Cloud Workloads

SANTA CLARA, Calif., Nov. 04, 2021 — AMD (NASDAQ: AMD) announ­ced the AMD Rade­on PROV620 GPU, built with the latest AMD RDNA™ 2 archi­tec­tu­re which deli­vers high-per­for­mance GPU acce­le­ra­ti­on for today’s deman­ding cloud workloads inclu­ding immersi­ve AAA game expe­ri­en­ces, inten­si­ve 3D workloads and modern office pro­duc­ti­vi­ty appli­ca­ti­ons at sca­le in the cloud.

With its inno­va­ti­ve GPU-par­ti­tio­ning capa­bi­li­ties, mul­ti-stream hard­ware acce­le­ra­ted enco­ders and 32GB GDDR6 memo­ry, the AMD Rade­on PRO V620 offers dedi­ca­ted GPU resour­ces that sca­le to mul­ti­ple gra­phics users, hel­ping ensu­re cost-effec­ti­ve gra­phics acce­le­ra­ti­on for a ran­ge of workloads1. Built using the same GPU archi­tec­tu­re that powers the latest gene­ra­ti­on game con­so­les and PC game expe­ri­en­ces, the AMD Rade­on PRO V620 GPU is also desi­gned to deve­lop and deli­ver immersi­ve AAA game experiences.

AMD is brin­ging high-per­for­mance AAA gam­ing expe­ri­ence to the cloud with the intro­duc­tion of the AMD Rade­on PRO V620,” said Jeff Con­nell, cor­po­ra­te vice pre­si­dent, Visu­al and Cloud Gam­ing, AMD. “We’re see­ing adop­ti­on of gam­ing in the cloud from cus­to­mers around the world, taking advan­ta­ge of the latest vir­tua­liza­ti­on fea­tures deli­ve­ring effi­ci­ent and low laten­cy con­tent strea­ming to mul­ti­ple simul­ta­neous users. AMD Rade­on PRO V620 is yet ano­ther pro­of point of AMD’s excel­lence in gam­ing – from the PC to the con­so­le and now to the cloud.” (…) Wei­ter­le­sen »

AMD EPYC™ Processors Expand Performance and Security Innovation Across Microsoft Azure Virtual Machines Portfolio

—Micro­soft Azu­re intro­du­ces new Dasv5 and Easv5 VMs based on 3rd Gen AMD EPYC™ pro­ces­sors, and new con­fi­den­ti­al VMs based on AMD SEV-SNP technology —

SANTA CLARA, Calif. 11/04/2021 — Today, AMD announ­ced its con­tin­ued momen­tum and col­la­bo­ra­ti­on with Micro­soft Azu­re,  who is offe­ring the 3rd Gen AMD EPYC™ pro­ces­sor in the latest gene­ra­ti­on of Dasv5 and Easv5 Azu­re Vir­tu­al Machi­nes (VMs). Azu­re is also intro­du­cing new con­fi­den­ti­al VMs, DCasv5 and ECasv5, which use the latest advan­ced secu­ri­ty fea­tures available in 3rd Gen EPYC pro­ces­sors, inclu­ding Secu­re Encrypt­ed Vir­tua­liza­ti­on-Secu­re Nes­ted Paging (SEV-SNP).

The new Azu­re con­fi­den­ti­al VMs, DCasv5 and ECasv5, the first EPYC pro­ces­sor-based con­fi­den­ti­al VMs at Azu­re and the first con­fi­den­ti­al VMs to use SEV-SNP, will enable cus­to­mers to have the data in their secu­ri­ty focu­sed appli­ca­ti­ons encrypt­ed in use, in tran­sit and at rest. The updated Dasv5 VMs, opti­mi­zed for gene­ral pur­po­se workloads, and the Easv5 VMs, opti­mi­zed for memo­ry-based workloads, deli­ver bet­ter pri­ce-per­for­mance for most gene­ral pur­po­se and memo­ry inten­si­ve workloads com­pared to pri­or EPYC pro­ces­sor-based Micro­soft Azu­re VMs.
(…) Wei­ter­le­sen »

Reviewübersicht Intel Alder Lake: Core i9-12900K, i7-12700K und i5-12600K

Nach­dem Intel vor kur­zem bereits die neu­es­te Gene­ra­ti­on von Pro­zes­so­ren mit dem Code­na­men Alder Lake offi­zi­ell vor­ge­stellt hat­te, ist heu­te das Embar­go für die Reviews gefal­len. Im Vor­feld gab es eini­ge Leaks zu Bench­marks und heu­te hat auch das ein oder ande­re Review einen Früh­start hin­ge­legt. Mit der Hybrid-Archi­tek­tur aus gro­ßen Per­for­mance- und klei­nen Effi­zi­enz­ker­nen star­tet Intel aus eige­ner Sicht ein neu­es Zeit­al­ter und unter­stützt sowohl den neu­en DDR5-Spei­cher­stan­dard, als auch — als ers­ter Desk­top-Pro­zes­sor — PCI-Express 5.0.  (…) Wei­ter­le­sen »

G.SKILL Showcases DDR5-7000 CL40 Extreme Speed Memory

(2 Novem­ber 2021) – G.SKILL Inter­na­tio­nal Enter­pri­se Co., Ltd., the world’s lea­ding manu­fac­tu­rer of extre­me per­for­mance memo­ry and gam­ing peri­pherals, is thril­led to announ­ce the achie­ve­ment of DDR5-7000 CL40-40–40-76 32GB (2x16GB) extre­me speed, pas­sing the Mem­test sta­bi­li­ty test. 7000MT/s memo­ry speed is an exci­ting mile­stone, as it was only seen under liquid nitro­gen sub-zero tem­pe­ra­tu­re coo­ling not long ago in over­clo­cking records. Accom­plished with high-per­for­mance Sam­sung DDR5 com­pon­ents, this extre­me speed memo­ry is tru­ly wort­hy of the G.SKILL flag­ship Trident Z5 fami­ly clas­si­fi­ca­ti­on. (…) Wei­ter­le­sen »

Micron Delivers High-Performance GDDR6 Memory for AMD Radeon RX 6000 Series Graphics Cards

Col­la­bo­ra­ti­on lever­a­ges inno­va­ti­on in Ultra-Band­width Solu­ti­ons to enable high frame rates and enhan­ced 4K reso­lu­ti­on gaming

BOISE, Ida­ho, Nov. 01, 2021 (GLOBE NEWSWIRE) — Micron Tech­no­lo­gy, Inc. (Nasdaq: MU), today announ­ced that its high-per­for­mance 16Gb/16Gbps GDDR6 memo­ry solu­ti­on is now available with AMD Rade­on™ RX 6000 Series gra­phics cards built on the AMD RDNA™ 2 gam­ing archi­tec­tu­re. Using Micron’s advan­ced 1z pro­cess tech­no­lo­gy, this latest ver­si­on of GDDR6 enables up to 512GB/s sys­tem per­for­mance for deman­ding appli­ca­ti­ons like gam­ing and gra­phics. Today’s announce­ment con­ti­nues Micron’s rich histo­ry of inno­va­ti­on and col­la­bo­ra­ti­on with indus­try lea­ders to deli­ver breakth­rough per­for­mance that enables the most advan­ced gam­ing solutions.

As gra­phics and gam­ing appli­ca­ti­ons beco­me more deman­ding, so do the requi­re­ments for high band­width memo­ry and sys­tem per­for­mance. Modern gamers expect high-reso­lu­ti­on, immersi­ve expe­ri­en­ces, and GDDR6 deli­vers with its sup­port of fast frame rates. The result is excel­lent per­for­mance and speed that mini­mi­ze lag time and pro­vi­de gamers with life­li­ke effects.
(…) Wei­ter­le­sen »

Börsengang von GlobalFoundries vollzogen

Auch wenn der Bör­sen­gang von Glo­bal­Found­ries vom Emis­si­ons­vo­lu­men (2,6 Mil­li­ar­den) her der dritt­größ­te in die­sem Jahr bezo­gen auf US-Unter­neh­men war, star­te­te die Aktie am gest­ri­gen Tag sehr ver­hal­ten und schluss­end­lich mit einem leich­ten Minus gegen­über dem Aus­ga­be­preis von 47 US-Dol­lar an der Nasdaq. (…) Wei­ter­le­sen »

AMD Promotes Server SoC Architect Kevin Lepak to Corporate Fellow

Lepak reco­gni­zed at hig­hest tech­ni­cal level for out­stan­ding con­tri­bu­ti­ons to AMD pro­ces­sor design

SANTA CLARA, Calif. 10/28/2021 AMD (NASDAQ: AMD) today announ­ced the appoint­ment of Seni­or Fel­low and Ser­ver SoC (sys­tem on chip) Archi­tect Kevin Lepak to AMD Cor­po­ra­te Fel­low, in reco­gni­ti­on of his cri­ti­cal role in the design of next-gene­ra­ti­on AMD EPYC™ pro­ces­sors. Lepak achie­ved an excel­lent record of over­see­ing the exe­cu­ti­on and deli­very of lea­der­ship data cen­ter pro­ces­sors for AMD, inclu­ding 2nd and 3rd Gen AMD EPYC™ CPUs and next gene­ra­ti­on AMD Instinct™ acce­le­ra­tors. As AMD Cor­po­ra­te Fel­low, Lepak expands on his lea­der­ship in chip design to play a lar­ger role in evol­ving AMD’s pro­ces­sor and sys­tems visi­on. (…) Wei­ter­le­sen »

Xilinx Reports Record Revenue in Fiscal Second Quarter 2022

Xilinx Reports Record Reve­nue in Fis­cal Second Quar­ter 2022

  • Record reve­nue of $936 mil­li­on, repre­sen­ting 7% sequen­ti­al growth and 22% annu­al growth amidst con­ti­nuing indus­try-wide sup­p­ly chain challenges
  • Aero­space & Defen­se, Indus­tri­al and Test, Mea­su­re­ment & Emu­la­ti­on (AIT) reve­nue increased 20% sequen­ti­al­ly, with strong per­for­mance in all sub-mar­kets led by ano­ther record per­for­mance in the Indus­tri­al end mar­ket and impro­ve­ment in Aero­space & Defen­se business
  • Auto­mo­ti­ve, Broad­cast and Con­su­mer (ABC) reve­nue in the quar­ter increased 19% sequen­ti­al­ly, with record quar­ters in all sub-mar­kets, led by the Auto­mo­ti­ve end market
  • Wired and Wire­less Group (WWG) reve­nue increased 9% sequen­ti­al­ly and 42% year-over-year as robust glo­bal 5G deploy­ments con­ti­nue and strength from the Wired business
  • Data Cen­ter Group (DCG) reve­nue decli­ned mode­st­ly, down 3% quar­ter-over-quar­ter, as Net­wor­king strength was off­set by a decli­ne in Compute
  • Plat­form trans­for­ma­ti­on con­ti­nues with total Adap­ti­ve SoC reve­nue, which includes Zynq and Ver­sal plat­forms, up 9% sequen­ti­al­ly and 56% year-over-year, and repre­sen­ting 29% of total revenue

SAN JOSE, Calif.–(BUSINESS WIRE)–Oct. 27, 2021– Xilinx, Inc. (Nasdaq: XLNX), the lea­der in adap­ti­ve com­pu­ting, today announ­ced record reve­nues of $936 mil­li­on for the fis­cal second quar­ter, up 7% over the pre­vious quarter.

GAAP net inco­me for the fis­cal second quar­ter was $235 mil­li­on, or $0.94 per diluted share. Non-GAAP net inco­me for the quar­ter was $266 mil­li­on, or $1.06 per diluted share.

As per­mit­ted as of Octo­ber 27, 2021 under the terms of the Mer­ger Agree­ment bet­ween Xilinx and Advan­ced Micro Devices, Inc. (AMD), the Xilinx Board of Direc­tors voted unani­mously to decla­re a cash divi­dend of $0.37 per out­stan­ding share of com­mon stock paya­ble on Novem­ber 15, 2021 to all stock­hol­ders of record at the clo­se of busi­ness on Novem­ber 8, 2021. The divi­dend is con­di­tio­ned upon and will only be paya­ble if the mer­ger has not clo­sed on or befo­re the record date for such divi­dend. (…) Wei­ter­le­sen »

Xilinx and Leading Broadcast and AV System and IP Integrators Deliver Complete, Production-Ready Multimedia Streaming End-Point Solutions

Oct 26, 2021 

Powered by Xilinx adap­ti­ve SoCs, high­ly inte­gra­ted strea­ming solu­ti­ons are rea­dy-to-ship or cus­to­mi­ze by cus­to­mers for a simp­ler, fas­ter time-to-mar­ket and deployment

SAN JOSE, Calif.–(BUSINESS WIRE)– Xilinx, Inc. (NASDAQ: XLNX), the lea­der in adap­ti­ve com­pu­ting, today announ­ced it is working with its IP and sys­tem inte­gra­tor eco­sys­tem to pro­vi­de the industry’s first and only pro­duc­tion-rea­dy mul­ti­me­dia strea­ming end­point solu­ti­ons for broad­cast and pro­fes­sio­nal audio/video (AV) appli­ca­ti­ons. The­se com­ple­te solu­ti­ons fea­ture powerful Xilinx® Zynq® UltraS­ca­le+™ EV Mul­ti-Pro­ces­sor (MP) sys­tem-on-a-chip (SoC) and Zynq-7000 SoC devices, with inte­gra­tors con­tri­bu­ting FPGA IP, media frame­work soft­ware and pro­duc­tion-rea­dy pro­ducts. The high­ly inte­gra­ted solu­ti­ons are rea­dy-to-ship, or rea­dy to cus­to­mi­ze, making it signi­fi­cant­ly fas­ter and easier for cus­to­mers to bring broad­cast and pro­fes­sio­nal AV pro­ducts to market.

Mul­ti­me­dia strea­ming requi­res fle­xi­ble AV inter­faces, real-time video pro­ces­sing, sup­port for dif­fe­rent pro­fes­sio­nal-gra­de video codecs, and a relia­ble high-band­width Ether­net trans­port lay­er for AV-over-IP, all of which are pro­vi­ded by Xilinx Zynq devices. AV-over-IP is at the heart of broad­cast end­points and infra­struc­tu­re, key­board-video-mou­se tech­no­lo­gies, strea­ming end­points, AV rou­ters and swit­chers, video wall con­trol­lers and col­la­bo­ra­ti­on equip­ment. (…) Wei­ter­le­sen »