>> FROM THE HEART, THERE WERE INNOVATION, MONEY, AND POWER COLLIDE IN SILICON VALLEY AND BEYOND. THIS IS BLOOMBERG TECHNOLOGY WITH CAROLINE HYDE AND ED LUDLOW. ED: LIVE FROM SAN FRANCISCO TO OUR AUDIENCES AROUND THE WORLD, WELCOME TO A SPECIAL EDITION OF BLOOMBERG TECHNOLOGY. IN A FEW MOMENTS, NVIDIA CEO JENSEN HUANG WILL JOIN US FOR A LIVE INTERVIEW FOLLOWING THEIR LATEST EARNINGS REPORT. THE COMPANY POSTING A REVENUE FORECAST THAT BEAT CONSENSUS BUT FELL SHORT OF SOME OF THE MOST OPTIMISTIC ESTIMATES, STOKING CONCERN THAT THE EXPLOSIVE GROWTH IS WANING. THAT'S GET RIGHT TO IAN KING WHO JOINS ME ON SET. LET'S START WITH THE FISCAL

THIRD QUARTER FORECAST AND WHAT WE LEARNED. IAN: THE FORECAST WAS FINE IF YOU COMPARE TO CONSENSUS. IT WAS THERE OR THEREABOUTS. FOR MOST OTHER COMPANIES IN THE WORLD, THAT WOULD BE GREAT AND EVERYONE WOULD BE HAPPY. BUT THIS IS NVIDIA. THIS IS A COMPANY WHICH BEATS PRETTY MUCH EVERY QUARTER BY AN ORDER OF MAGNITUDE, AND IT DID NOT DO THAT AND DID NOT INDICATE IT WOULD DO THAT. ED: THERE ARE MANY STORYLINES. I THINK DEMAND FROM HYPER SCALARS IS INTACT. BUT BLACKWELL WAS EVERYTHING. I WANT TO PLAY A SOUNDBITE OF WHAT JENSEN HUANG SAID ABOUT BLACKWELL. JENSEN: THE CHANGE TO THE MASK IS COMPLETE. THERE WERE NO FUNCTIONAL CHANGES NECESSARY. SO WE ARE SAMPLING FUNCTIONAL SAMPLES OF BLACKWELL, GRACE

BLACKWELL IN A VARIETY OF SYSTEM CONFIGURATIONS AS WE SPEAK. ED: THE MAIN POINT HERE IS THAT THERE IS NOT A DESIGN ISSUE WITH BLACKWELL ITSELF AS HAD BEEN REPORTED, BUT BASED ON WHAT NVIDIA SAID, THIS WAS ABOUT A PRODUCTION MECHANISM. IAN KING, YOU DID A VERY GOOD JOB IN THE TOPLINE BLOG OF EXPLAINING THE GPU MASK. CAN YOU TRY TO GET A SHORT VERSION OF THAT TO OUR AUDIENCE? IAN: THE MASK IS BASICALLY THE BLUEPRINT, WHICH BASICALLY IS USED TO BURN THE CIRCUIT ONTO THE SURFACE OF THE CHIP TO GIVE IT ITS FUNCTION. THAT IS THE FUNDAMENTAL BLUEPRINT. THEY WERE SAYING, WE DID NOT GET IT WRONG, BUT WHEN IT CAME TO MANUFACTURING, IT DID NOT PRODUCE AS MANY GOOD TROOPS AS WE WANTED, SO WE MADE SOME TWEAKS TO THAT, DID NOT HAVE TO

REDESIGN IT. AND THAT IS HELPING US GET A BETTER YIELD. ED: IT IS WORTH NOTING THE STOCK IS DOWN ALMOST 7% IN AFTER-HOURS. BASICALLY BECAUSE WE DIDN'T LEARN ENOUGH ABOUT BLACKWELL. WHAT THEY SAID WAS IN THE FISCAL FOURTH-QUARTER OF FISCAL YEAR 25 THERE WILL BE SEVERAL BILLION DOLLARS OF SALES THROUGH BLACKWELL. WHY DOES THE MARKET WANT MORE AND WHAT DO THEY WANT? IAN: THEY WANTED A REASSURANCE FROM NVIDIA'S MANAGEMENT AND IN THE FORM OF DETAILS. THEY WANTED JENSEN TO PUT HIS ARM AROUND EVERYBODY AND SAY DON'T WORRY, THIS WILL BE FINE. CONSISTENTLY WERE ASKED HOW MANY BILLIONS AND WHEN WILL THEY COME IN.

THE CFO AND JENSEN HUANG ESSENTIALLY AVOIDED THAT QUESTION AND REFUSED TO GIVE THAT PRECISE REASSURANCE. ED: WE ARE SHOWING AFTER-HOURS REACTION TO NOT JUST NVIDIA ITSELF BUT SOME PEERS, SERVER EQUIPMENT PROVIDERS, AND AMD AND ARM ARE NOTEWORTHY. I WANT TO GO OUT TO CHICAGO AND OUR EQUITIES TEAM. THAT IS THE BROAD SUMMARY FROM ME ABOUT THE NAMES MOVING IN AFTER-HOURS. THERE IS PROBABLY A BIGGER PICTURE AFTER-HOURS AS WELL. START WITH NVIDIA AND WORK OUTWARD FROM THERE. RYAN: ONE THING I WOULD SAY ABOUT NVIDIA'S AFTER-HOURS DECLINE AS IT COMES AFTER A VERY STRONG YEAR-TO-DATE PERFORMANCE. I THINK IT CLOSED UP MORE THAN 150% THIS YEAR. EVEN THOUGH THE FORECAST WAS A LITTLE SHY OF MOST OPTIMISTIC EXPECTATIONS, IT DID

BEAT EXPECTATIONS AND IS COMING OFF SUCH A HUGE GAIN. IT IS NOT NECESSARILY SURPRISING TO SEE A LITTLE BIT OF CONSOLIDATION. EVEN ADA CLIENT IS A LITTLE LESS IN THE OPTIONS MARKET WAS ANTICIPATING. JUST SOME CONTEXT. BUT YOU ARE RIGHT THAT PRETTY MUCH WE ARE SEEING WIDESPREAD WEAKNESS FOLLOWING THE REPORT. PAUL THE MEGACAPS ARE MODESTLY LOWER. CERTAINLY IT SEEMS LIKE THE INITIAL READTHROUGH IS NEGATIVE BUT IT DOES COME AFTER A VERY STRONG START TO THE YEAR. ED: WE MADE A BIG DEAL ABOUT THIS EARNINGS PRINT FOR QUITE A LONG TIME. I LOOK AT SOMETHING LIKE INVESCO QQQ, THE ETF THAT TRACKS THE NASDAQ 100 WE ARE DOWN AROUND ONE PERCENTAGE POINT. IN TRUTH, IN THE MARKET

REACTION WAS THIS THE MACROLEVEL EVENT WE THOUGHT IT WAS GOING TO BE? RYAN: THAT IS A GREAT QUESTION. I WOULD SAY IT IS CLOSE ENOUGH IN LINE WITH EXPECTATIONS EVEN IF IT IS A LITTLE SHY OF THE MOST OPTIMISTIC ONES. I DON'T THINK THIS WILL REALLY CAUSE PEOPLE TO CHANGE HOW THEY ARE ALLOCATING CHANGE. THEIR OPINION ON AI AS THIS FUNDAMENTAL, SECULAR DRIVER. MAYBE WE DO SEE SOME WEAKNESS. SOME STOCKS HAVE BEEN MOVING UP SO MUCH, THERE HAVE BEEN GROWING CALLS ABOUT THEIR VALUATIONS AND CONCERNS ABOUT THAT. I DON'T KNOW IF THIS IS THE ABSOLUTE BLOWOUT THAT'LL CONTINUE CAUSING PEOPLE PILING INTO THE WAY THEY WERE DOING EARLIER THIS YEAR. ED: THERE IS A LOT AND THE NEWS CYCLE INCLUSIVE OF AND OUTSIDE

NVIDIA. ONE OF THE NAMES YOU NOTE IS SUPER MICRO, DOWN SIGNIFICANTLY. THERE IS IN NVIDIA RELATION TO THAT AND THEN THE NEWS IN AND OF THE NAME ITSELF. RYAN: WE SAW YESTERDAY A SHORT REPORT CAME OUT. THEY ARE DELAYING THE FILING OF THE 10K A PRETTY BOTH OF THOSE CAUSED A WEAKNESS IN THE STOCK. A LOT OF REASON TO BE CONCERNED IN GENERAL AND IT IS ONE OF NVIDIA'S BIGGEST CUSTOMERS. THIS IS A REASON FOR PEOPLE WHO MIGHT HAVE BEEN SOURING ON SUPERMICRO. ED: RYAN WITH THE AFTER HOURS OUT OF CHICAGO. LET'S GET REACTION FROM THE SELL SIDE WITH OF THE STOCKS RELATIVE BEARS. HAD A NEUTRAL RATING AT A STREET LOVE $90 PRICE TARGET ON SHARES OF NVIDIA AND JOINS US

NOW. DOES THAT CHANGE AND YOUR MAIN TAKEAWAY FROM THE ANALYST CALL? >> THE QUARTER WAS A VERY GOOD QUARTER. THEIR ABILITY TO CONTINUE TO GROW DELIVERING TRIPS AT THIS RATE AND THIS SCALE IS FANTASTIC AND UNPRECEDENTED. I DO NOT THINK THERE IS MUCH OF A REVENUE OR GROWTH ISSUE HERE. A LITTLE BIT OF A PUSHBACK IS PROBABLY MORE AROUND THE MARGIN SITUATION. YOU HAD A GOOD CONVERSATION ABOUT BLACKWELL AND SOME MOVING PIECES. PART OF WHAT IS HAPPENING WITH BLACKWELL, IT IS A NEW PRODUCT THAT HAS A FEW HITCHES ALONG THE WAY. THAT PUT PRESSURE ON GROSS MARGINS. OPERATING MARGINS GREW MORE THAN ANTICIPATED. LESS OF THE TOPLINE UPSIDE FLOW TO THE BOTTOM LINE THAT IS

PROBABLY WHY -- WHERE INVESTORS ARE GETTING MORE CAUTIOUS THAN BEFORE. IN THE LAST SEVERAL QUARTERS, THE HUGE UPSIDE TO REVENUE FLOAT ALL THE WAY TO THE BOTTOM LINE, CREATING EACH UPSIDE. IN NVIDIA, IS TO BELIEVE 2026 CALENDAR WILL BE ABOUT $200 BILLION OF REVENUE AND $4.5 TO $5 OF EARNINGS. THAT IS ONLY POSSIBLE IF THEY CONTINUE TO GET THIS KIND OF GROWTH AND MORE LEVERAGE. THE SECOND PART OF THAT EQUATION LOOKS A LITTLE LESS CERTAIN. ED: WHEN YOU SAID THAT $200 BILLION FIGURE, IAN KING, WHO HAS COVERED THIS SINCE 1998, SAT NEXT TO ME STILL, FOR WHAT IS WORTH, GAVE A LITTLE SMILE. IT IS AN ASTONISHING FIGURE. THERE IS A LOT OF EMPHASIS ON

BLACKWELL. HOW DID YOU REACT TO THE NEWS OF SEVERAL BILLION DOLLARS IN FISCAL FOURTH WITH BLACKWELL IN THE EXPLANATION THIS WAS A PRODUCTION ISSUE IMPACTING NOT THE COURT ISSUE ITSELF? >> OUR ESTIMATE IS FAR LESS THAN $200 BILLION FOR 226. IN TERMS OF BLACKWELL, NVIDIA USED TO BE ON A TWO-YEAR CYCLE. THEY DECIDED TO AGGRESSIVELY MOVE TO A ONE-YEAR CYCLE AND THEN PUT A LOT OF NEW FEATURE FUNCTIONALITY AND BUNDLE INTO THE BLACKWELL PLATFORM. IT WAS A VERY AGGRESSIVE AGENDA. IT IS NOT THAT SURPRISING THAT THERE ARE SOME CHALLENGES THEY HAVE TO ENCOUNTER. IF THEY CAN STILL SHIP IN THEIR FISCAL FOURTH-QUARTER, THAT IS A VERY GOOD SIGN. BY THE WAY, THEIR CUSTOMERS DON'T CARE THAT MUCH.

THEIR CUSTOMERS ARE GOING TO CONTINUE TO BUY THE LATEST AND GREATEST. IF THAT IS H-200, THEY WILL CONTINUE TO BUY H-200. MICROSOFT, META, AMAZON, GOOGLE, TESLA, MADE IT ABUNDANTLY CLEAR THEY WILL BUY AS MUCH GP USE AS NVIDIA WILL REDUCE, AT LEAST THIS YEAR. THEY ARE SAYING THE DEMAND SIGNAL IS THERE. THEY WILL BUY WHATEVER NVIDIA IS SELLING. CAN TOWARDS THE END OF THE YEAR A PIECE OF THAT WILL BE BLACKWELL. ED: JENSEN HUANG GAVE A DETAILED ANSWER THAT BASICALLY SUMMARIZED NVIDIA AS A SYSTEMS VENDOR. THE DISCUSSION AROUND CPU'S, EVERYTHING BEYOND GPU. HOW MUCH CREDIT DO YOU ASSIGN THE COMPANY FOR THAT, THE COMPLETE SYSTEMS THAT THEY SELL? >> TREMENDOUS AMOUNT OF CREDIT.

LET'S START WITH THE BEGINNING. WHEN HISTORY IS WRITTEN, WE WILL TALK ABOUT HOW JENSEN HUANG AND NVIDIA CREATED THE CAPABILITIES THAT MAKE GENERATIVE AI POSSIBLE TODAY. IN A SIMILAR WAY, ELON MUSK IS RESPONSIBLE FOR THE ELECTRIC VEHICLE. PART OF THE GENIUS WAS NOT JUST SEEING THIS WAS COMING. AI CAPABILITIES WERE COMING. IT'S ALSO REALIZING IN ORDER TO CREATE A MOAT AROUND THE CHIP, THEY NEED TO DO A COUPLE THINGS. ONE IS TO OWN THE PROPRIETARY SOFTWARE, BUT THE OTHER THING IS TO ENHANCE THE BUNDLE AS MUCH AS POSSIBLE. THESE GPU'S WORK IN A RISE. WHEN THEY WORK TOGETHER, SOMETIME TENS OF THOUSANDS TOGETHER. WHEN YOU BOX THOSE UP TOGETHER IN A BOX WITH MULTIPLE GP USE, CPUS, WIRING, CABLE, FIRMWARE,

YOU MAKE IT HARDER FOR YOUR COMPETITORS TO CATCH UP, AND THEY HAVE DONE A REMARKABLE JOB OF THAT. ED: THE STORY THUS FAR HAS BEEN ABOUT HYPER SCALARS AND META FIVE NAMES ESSENTIALLY BEING THE CORE CUSTOMER OF NVIDIA. WE GOT SOME DETAIL IN LOW DOUBLE-DIGIT BILLION SALES THIS YEAR IN SOVEREIGN AIP. DO YOU SEE EVIDENCE THE MARKET IS BROADENING OUT JUST BEYOND CLOUD PROVIDERS? >> NOT REALLY. IT SEEMS THOSE FIVE ARE STILL A SUBSTANTIAL AMOUNT OF THE REVENUE. IN THEIR MOST RECENT CALLS, THEY HAVE SAID RECENT THINGS. THEY HAVE CONSISTENTLY SAID THEY ARE OVER INVESTING. THEY SAID THEY DEFINITELY WANT TO INVEST MORE, AND WE ARE SEEING THAT IN RESULTS TODAY. THEY ALSO USE THE TERM OVER

INVESTING. THAT IS SOMETHING YOU DO BEFORE YOU DO THE OTHER THING. THEY MADE IT CLEAR THEY ARE ONLY GOING TO DO SO WHEN THEY HAVE THE DEMAND SIGNALS. IF FOR WHATEVER REASON DEMAND FOR AI CAPACITY GOES DOWN, FOR INSTANCE, BECAUSE COMPUTE IS NOT CONTINUING TO DRIVE PERFORMANCE OF FOUNDATIONAL MODELS, YOU ARE GOING TO SEE THOSE SAME FIVE ACTORS PULL BACK AND SAY MAYBE WE HAVE ENOUGH DATA CENTER CAPACITY OR MAYBE WHAT WE HAVE IN THE PIPELINE IS ENOUGH. THAT IS THE CONCERN FOR 2025 AND 2026. IS THAT THOSE VERY COMPANIES COULD DO THAT. BEYOND THAT, THESE ARE THE SAME COMPANIES DEVELOPING THEIR OWN CHIPS. THEY ARE AT VARIOUS STAGES FOR THIS BUT GOOGLE'S TPU'S ARE PROBABLY AS GOOD AS NVIDIA'S. GOOGLE HAS ALREADY CAUGHT UP.

AMAZON IS CATCHING UP HERE. META AND MICROSOFT ARE IN EARLIER STAGES. THESE SAME COMPANIES BUYING EVERY GPU FROM NVIDIA THEY CAN WILL BE THE ONES WHO CAN BUY POSSIBLY LESS NEXT YEAR AND THE YEAR AFTER. ED: I WANT TO REFLECT, THERE ARE PEOPLE TUNED IN AROUND THE WORLD, MY UNDERSTANDING IS THERE ARE WATCH PARTIES IN CITIES FROM NEW YORK TO LONDON FOR THESE EARNINGS. THIS IS PROBABLY THE MOST UNIQUE EARNINGS PRINT I HAVE COVERED IN MY CAREER. WHAT ABOUT FOR YOU IN THE SEMICONDUCTOR SPACE? WHY IS THIS SO DIFFERENT? >> FIRST OF ALL, I TRIED TO PARTICIPATE BY WEARING NVIDIA GREEN TODAY. BUT IMPORTANTLY, NVIDIA IS IMPORTANT BECAUSE OF HOW MUCH THEY HAVE ADDED TO THE CUMULATIVE EARNINGS AND PERFORMANCE.

I WOULD NOT NECESSARILY EXTRAPOLATE THEIR SUCCESS OR SOMETIMES SETBACKS FOR THE REST OF THE MARKET. BECAUSE WHAT THEY DO IS VERY SPECIFIC. GROWTH AND DATA CENTER CONSTRUCTION AND DATA CENTER GPU FULFILLMENT INTO THOSE DATA CENTERS IS A VERY SMALL SET OF COMPANIES. NVIDIA IS SUCKING THE OXYGEN FROM THE ROOM FOR A LOT OF OTHER TECHNOLOGY COMPANIES. WE CELEBRATE FROM NVIDIA BECAUSE THEY HAVE ADDED SO MUCH VALUE. I WOULD NOT NECESSARILY READ THROUGH TO THE REST OF TECHNOLOGY. ED: THE RELATIVELY MOST BEARISH OR PERHAPS LEAST BULLISH NAME ON THE THREE, THANK YOU. WELCOME TO THIS SPECIAL EDITION OF BLOOMBERG TECHNOLOGY. JOINING ME NOW IS JENSEN HUANG, CEO OF NVIDIA, STRAIGHT FROM

THE ANALYST CALL. GOOD EVENING, GOOD AFTERNOON TO YOU. I THINK THE MARKET WANTED MORE AROUND BLACKWELL. THEY WANTED MORE SPECIFICS. I AM TRYING TO GO THROUGH THE CALL AND THE TRANSCRIPT. IT SEEMS LIKE VERY CLEARLY THIS WAS A PRODUCTION ISSUE AND NOT A FUNDAMENTAL DESIGN ISSUE WITH BLACKWELL. BUT THE DEPLOYMENT IN THE REAL WORLD, WHAT DOES THAT LOOK LIKE TANGIBLY, AND IS THERE A DELAY IN THE TIMELINE OF THAT DEPLOYMENT AND THUS REVENUE FROM THAT PRODUCT? JENSEN: UH. THE FACT THAT I WAS SO CLEAR AND IT WAS NOT CLEAR ENOUGH KIND OF TRIPPED ME UP RIGHT AWAY. WE MADE A MASS CHANGE TO IMPROVE THE YIELD. FUNCTIONALITY OF BLACKWELL IS WONDERFUL. WE ARE SAMPLING BLACKWELL ALL

OVER THE WORLD TODAY. WE SHOW PEOPLE, I AM GIVING TOURS TO PEOPLE OF THE BLACKWELL SYSTEMS WE HAVE UP AND RUNNING. YOU CAN FIND PICTURES OF BLACKWELL SYSTEMS ALL OVER THE WEB. WE HAVE STARTED VOLUME PRODUCTION. VOLUME PRODUCTION WILL SHIP IN Q4. Q4, WE WILL HAVE BILLIONS OF DOLLARS OF BLACKWELL REVENUES. AND WE WILL RAMP FROM THERE. THE DEMAND FOR BLACKWELL FAR EXCEEDS ITS SUPPLY, OF COURSE IN THE BEGINNING, BECAUSE THE DEMAND IS SO GREAT. BUT WE'RE GOING TO HAVE LOTS AND LOTS OF SUPPLY, AND WE WILL BE ABLE TO RAMP, STARTING IN Q4. WE HAVE BILLIONS OF DOLLARS IN REVENUES AND WE WILL RAMP FROM THERE. WE WILL HAVE A GREAT NEXT YEAR AS WELL. ED: WHAT IS THE DEMAND FOR

ACCELERATED COMPUTING BEHIND THE HYPER SCALARS AND META? JENSEN: HYPER SCALARS REPRESENT ABOUT 45% OF OUR TOTAL DATA CENTER BUSINESS. WE ARE RELATIVELY DIVERSIFIED TODAY. WE HAVE HYPER SCALARS. WE HAVE INTERNET DATA SERVICE PROVIDERS. WE HAVE SOVEREIGN AI'S. WE HAVE INDUSTRIES, ENTERPRISES. SO IT IS FAIRLY DIVERSIFIED. OUTSIDE OF HYPERSCALERS IS THE OTHER 55%. THE APPLICATION USE OUTSIDE THE DATA CENTER STARTS WITH ACCELERATED COMPUTING. ACCELERATED COMPUTING DOES EVERYTHING OF COURSE FROM -- WELL, THE THINGS WE KNOW ABOUT WHICH IS GENERATIVE AI, AND THAT GETS MOST OF THE ATTENTION. BUT AT THE CORE WE ALSO DO DATABASE PROCESSING, PRE-AND POST PROCESSING OF DATA BEFORE YOU USE IT FOR GENERATIVE AI.

TRANSCODING, SCIENTIFIC SIMULATIONS, COMPUTER GRAPHICS, COMPUTER PROCESSING OF COURSE. SO, THERE ARE TONS OF APPLICATIONS THAT PEOPLE USE OUR ACCELERATED COMPUTING FOR. ONE OF THEM IS GENERATIVE AI. WHAT ELSE CAN I SAY? THAT COVERS IT. ED: LET ME PUT A PIN IN SOVEREIGN AI. IT WAS SO INTERESTING TO HEAR SOMETHING BEHIND IT, THAT IN THIS FISCAL YEAR, THERE WILL BE LOW DOUBLE-DIGIT BILLIONS OF DOLLARS IN SOVEREIGN AI SALES. TO THE LAYPERSON, WHAT DOES THAT MEAN? IT MEANS DEALS WITH SPECIFIC GOVERNMENTS. IF SO, WHERE? JENSEN: IT IS NOT NECESSARILY. SOMETIMES IT IS DEALS WITH PARTICULAR REGIONAL SERVICE PROVIDER THAT HAS BEEN FUNDED BY THE GOVERNMENT, AND OFTENTIMES THAT IS THE CASE.

IN THE CASE OF JAPAN, FOR EXAMPLE, THE JAPANESE GOVERNMENT CAME OUT AND OFFERED SUBSIDIES OF A COUPLE BILLION DOLLARS, I THINK, FOR SEVERAL DIFFERENT INTERNET COMPANIES AND TELCOS TO BE ABLE TO FUND THEIR AI INFRASTRUCTURE. INDIA HAS A SOVEREIGN AI INITIATIVE DOING, AND THEY ARE BUILDING THEIR AI INFRASTRUCTURE. CANADA, THE U.K., FRANCE, ITALY, I AM MISSING SOMEBODY, SINGAPORE, MALAYSIA. A LARGE NUMBER OF COUNTRIES ARE SUBSIDIZING THEIR REGIONAL DATA CENTERS SO THAT THEY CAN BECOME ABLE TO BUILD OUT THEIR AI INFRASTRUCTURE. THEY RECOGNIZE THAT THEIR COUNTRY'S KNOWLEDGE AND DIGITAL DATA IS ALSO THEIR NATURAL RESOURCE. NOT JUST THE LAND THEY ARE SITTING ON, NOT JUST THE AIR.

ABOVE THEM, BUT THEY REALIZE NOW THAT THEIR DIGITAL KNOWLEDGE IS PART OF THEIR NATURAL AND NATIONAL RESOURCE. AND THEY OUGHT TO HARVEST THAT AND TRANSFORM THAT INTO THEIR NATIONAL DIGITAL INTELLIGENCE. AND SO, THIS IS WHAT WE CALL SOVEREIGN AI PRE-YOU CAN IMAGINE ALMOST EVERY SINGLE COUNTRY IN THE WORLD WILL EVENTUALLY RECOGNIZE THIS AND BUILD OUT THEIR AI INFRASTRUCTURE. ED: YOU USED THE WORD RESOURCE, AND THAT MAKES ME THINK ABOUT THE ENERGY REQUIREMENTS HERE. ON THE CALL, YOU TALKED ABOUT HOW THE NEXT GENERATION MODELS WILL HAVE MANY ORDERS OF MAGNITUDE GREATER COMPUTE NEEDS. HOW WILL THE ENERGY NEEDS INCREASE, AND WHAT IS THE ADVANTAGE YOU FEEL NVIDIA HAS IN THAT SENSE? JENSEN: THE MOST IMPORTANT THING THAT WE DO IS INCREASE THE PERFORMANCE AND EFFICIENCY OF

OUR NEXT GENERATION. SO, BLACKWELL'S MANY MORE TIMES PERFORMANCE THAN HOPPER AT THE SAME LEVEL OF POWER USED. SO THAT IS ENERGY EFFICIENCY. MORE PERFORMANCE WITH THE SAME AMOUNT OF POWER, OR SAME PERFORMANCE AT A LOWER POWER. THAT IS NUMBER ONE. THE SECOND IS USING LOCAL COOLING. WE SUPPORT AIR COOLING, LIQUID COOLING. LIQUID COOLING IS A LOT MORE ENERGY EFFICIENT. THE COMBINATION OF ALL OF THAT, YOU ARE GOING TO GET A PRETTY LARGE STEP UP. THE IMPORTANT THING TO ALSO REALIZE IS AI DOES NOT CARE WHERE IT GOES TO SCHOOL. SO INCREASINGLY WE ARE GOING TO SEE AIB TRAINED SOMEWHERE ELSE, HAVE THAT MODEL COME BACK AND BE USED NEAR THE POPULATION OR EVEN RUNNING ON YOUR PC OR YOUR PHONE. SO WE ARE GOING TO TRAIN LARGE

MODELS, BUT THE GOAL IS NOT TO RUN THE LARGE MODELS NECESSARILY ALL THE TIME. YOU CAN SURELY DO THAT FOR SOME OF THE PREMIUM SERVICES AND THE VERY HIGH VALUE AI'S. BUT IT IS VERY LIKELY THESE LARGE MODELS WILL THEY'LL -- WILL THEN HELP TO TRAIN AND TEACH SMALLER MODELS AND THEN WE WILL HAVE A FEW LARGE MODELS THAT ARE ABLE TO TRAIN A WHOLE BUNCH OF SMALL MODELS AND THEY RUN EVERYWHERE. ED: YOU EXPLAINED CLEARLY THAT DEMAND TO BUILD GENERATIVE AI PRODUCT ON MODELS OR EVEN AT THE GPU LEVEL IS GREATER THAN CURRENT SUPPLY. IN BLACKWELL'S CASE IN PARTICULAR, EXPLAIN THE SUPPLY DYNAMICS FOR ME FOR YOUR PRODUCTS AND WHETHER YOU SEE AN IMPROVEMENT SEQUENTIALLY QUARTER ON QUARTER OR AT SOME POINT BEYOND THIS FISCAL YEAR.

INTO NEXT YEAR. JENSEN: THE FACT WE ARE GROWING WOULD SUGGEST OUR SUPPLY IS IMPROVING. OUR SUPPLY CHAIN IS QUITE LARGE. ONE OF THE LARGEST SUPPLY CHAINS IN THE WORLD. WE HAVE INCREDIBLE PARTNERS. AND THEY ARE DOING A GREAT JOB SUPPORTING US IN OUR GROWTH. AS YOU KNOW, WE ARE ONE OF THE FASTEST-GROWING TECHNOLOGY COMPANIES IN HISTORY. NONE OF THAT WOULD HAVE BEEN POSSIBLE WITHOUT VERY STRONG DEMAND, ALSO VERY STRONG SUPPLY. WE ARE EXPECTING Q3 TO MAKE MORE SUPPLY THAN Q2. I THINK OUR SUPPLY CONDITION GOING INTO NEXT YEAR WILL BE A LARGE IMPROVEMENT OVER THIS LAST YEAR. WITH RESPECT TO DEMAND, BLACKWELL IS SUCH A LEAP. THERE ARE SEVERAL THINGS HAPPENING. JUST THE FOUNDATION MODEL MAKERS THEMSELVES, THE SIZE OF

THE FOUNDATION MODELS ARE GROWING FROM HUNDREDS OF BILLIONS OF PARAMETERS TO TRILLIONS OF PARAMETERS. THEY ARE ALSO LEARNING MORE LANGUAGES. INSTEAD OF JUST LEARNING HUMAN LANGUAGE, THEY ARE LEARNING THE LANGUAGE OF IMAGES AND SOUNDS AND VIDEOS, AND EVEN LEARNING THE LANGUAGE OF 3D GRAPHICS. WHENEVER THEY ARE ABLE TO LEARN THESE LANGUAGES, THEY CAN UNDERSTAND WHAT THEY SEE, BUT THEY CAN ALSO GENERATE WHAT THEY ARE ASKED TO GENERATE. SO, THEY ARE LEARNING THE LANGUAGE OF PROTEINS AND CHEMICALS AND PHYSICS, CAN BE FLUID AND PARTICLE PHYSICS. SO, THEY ARE LEARNING ALL KINDS OF DIFFERENT LANGUAGES. LEARN THE MEANING OF WHAT WE CALL MODALITY, BUT BASICALLY LEARNING LANGUAGES. SO, THESE MODELS ARE GROWING IN

SIZE, LEARNING FROM MORE DATA, AND THERE ARE MORE MODEL MAKERS THAN THERE WAS A YEAR AGO. AND SO THE NUMBER OF MODEL MAKERS HAS GROWN SUBSTANTIALLY BECAUSE OF ALL THESE DIFFERENT MODALITIES. THAT IS JUST ONE, THE FRONTIER MODEL, THE FOUNDATION MODEL MAKERS THEMSELVES HAVE GROWN TREMENDOUSLY. AND THEN THE GENERATIVE AI MARKET HAS REALLY DIVERSIFIED BEYOND THE INTERNET SERVICE MAKERS TO STARTUPS AND NOW ENTERPRISES ARE JUMPING IN. DIFFERENT COUNTRIES ARE JUMPING IN, SO THE DEMAND IS REALLY GROWING. ED: I AM SORRY TO CUT YOU OFF, I WILL LOSE YOUR TIME SOON. YOU HAVE ALSO DIVERSIFIED. WHEN I SAID TO OUR AUDIENCE YOU ARE COMING ON I GOT SO MANY QUESTIONS. PROBABLY THE MOST COMMON ONE

IS, WHAT IS NVIDIA? WE TALKED ABOUT YOU AS A SYSTEMS VENDOR, BUT SO MANY POINTS ON NVIDIA GPU CLOUD. I WANT TO ASK FINALLY, DO YOU HAVE PLANS TO BECOME LITERALLY A CLOUD COMPUTING PROVIDER? JENSEN: NO. OUR GPU CLOUD WAS DESIGNED TO BE THE BEST VERSION OF NVIDIA CLOUD THAT'S BUILT WITHIN EACH CLOUD. NVIDIA DGX CLOUD IS BUILT WITHIN GPC, INSIDE AZURE, INSIDE OCI. SO WE BUILT OUR CLOUDS WITHIN THERE'S SO WE CAN IMPLEMENT THE BEST VERSION OF OUR CLOUD, WORK WITH THEM TO MAKE THAT AI INFRASTRUCTURE, THE NVIDIA INFRASTRUCTURE AS GREAT AS POSSIBLE. SO THAT STRATEGY HAS WORKED INCREDIBLY WELL. OF COURSE, WE ARE LARGE CONSUMERS OF IT BECAUSE WE CREATE A LOT OF AI OURSELVES,

OUR CHIPS ARE NOT POSSIBLE TO DESIGN WITHOUT AI. SO WE USE IT OURSELVES, A TREMENDOUS AMOUNT OF IT. SELF-DRIVING CARS, THE OMNIVOROUS WORK WE ARE DOING. WE ARE USING THE DGX CLOUD FOR OURSELVES. WE ALSO USE IT FOR NEI FOUNDRY. WE MAKE AI MODELS FOR COMPANIES THAT WOULD LIKE TO HAVE EXPERTISE IN DOING SO. AND SO WE ARE A FOUNDRY FOR AI LIKE TSMC IS A FOUNDRY FOR OUR CHIPS. SO THERE ARE THREE FUNDAMENTAL REASONS WHY WE DO IT. ONE IS TO HAVE THE BEST VERSION OF NVIDIA INSIDE ALL THE CLOUDS. TWO, BECAUSE WE ARE LARGE CONSUMER OURSELVES. THIRD, BECAUSE WE USE IT TO HELP EVERY OTHER COMPANY. ED: JENSEN HUANG, CEO OF NVIDIA, I

WANT TO THANK YOU FOR YOUR TIME. THE EXTENDED CONVERSATION STRAIGHT OFF THE EARNINGS CALL, THANK YOU. GOOD TO SEE YOU. FISCAL SECOND QUARTER. I POINT OUT IN AFTER HOURS THE STOCK IS DOWN ALMOST 7%. THERE WILL BE A LOT OF ANALYSIS TO BE DONE OVERNIGHT BY THE SELL SIDE AND THE BUY SIDE AND WE WILL BRING YOU THE BEST FROM BLOOMBERG. FROM SAN FRANCISCO, THIS IS BLOOMBERG TECHNOLOGY. ♪

Nvidia CEO Jensen Huang | Bloomberg Technology

Welcome to the Bloomberg Technology YouTube channel. Subscribe for original series and mini-docs on game-changing technologies, profiles on innovators, live conversations with experts, and breaking technology news & analysis from Bloomberg TV.