[00:00:02]
[1. Call General Session to Order ]
LIVE TO THE PUBLIC ON ERCOT WEBSITE, I HAVE CONFIRMED THAT A QUORUM IS PRESENT.IN FACT, ALL MEMBERS ARE PRESENT.
UM, FIRST LET ME ASK PUC CHAIR TOM GLEASON, IF HE WOULD LIKE TO CALL AN OPEN MEETING OF THE PUBLIC UTILITY COMMISSION.
THIS MEETING OF THE PUBLIC UTILITY COMMISSION OF TEXAS WILL COME TO ORDER TO CONSIDER MATTERS THAT HAVE BEEN DULY POSTED WITH THE SECRETARY OF STATE FOR JUNE 1ST, 2026.
BEFORE MOVING TO TODAY'S BUSINESS, LET ME REMIND YOU AGAIN OF THE ANTITRUST ADMONITION AND THE EVACUATION MAP THAT IS POSTED IN FRONT OF YOU.
[2. Notice of Public Comment, if Any ]
FIRST ORDER OF BUSINESS ON TODAY'S AGENDA IS ITEM TWO.NOTICE OF PUBLIC COMMENT, IF ANY.
UM, TO DATE, NO ONE HAS EXPRESSED INTEREST IN ACCOMMODATING.
IS THAT CORRECT, BRETT? THAT IS CORRECT.
WITH THAT, WE WILL MOVE ON TO AGENDA ITEM THREE,
[3. April 20, 2026, General Session Meeting Minutes ]
THE MINUTES FROM THE PRIOR MEETING.I'M ASSUMED EVERYONE HAS LOOKED AT IT.
DOES ANYONE WISH TO MAKE A MOTION? UTE? THANK YOU, JULIE.
[4. Increasing Compute Density, Agentic, and Physical Artificial Intelligence - How Flexible AI Factories Address Resiliency, Affordability, and Speed to Power ]
UP IS AGENDA ITEM FOUR.OUR GUEST PRESENTATION THIS MORNING IS INCREASING COMPUTE DENSITY, AGENTIC AND PHYSICAL ARTIFICIAL INTELLIGENCE, HOW FLEXIBLE AI FACTORIES ADDRESS RESILIENCY, AFFORDABILITY, AND SPEED TO POWER.
UM, I'M DELIGHTED TO INTRODUCE OUR TWO GUEST SPEAKERS, MARK SPEER AND ASSIGN, I'M GOING, I'M GONNA MESS THIS UP AS SO,
UM, AND, UH, MARK IS THE SENIOR MANAGING DIRECTOR FOR GLOBAL ENERGY AND NVIDIA, WHERE HE LEADS EFFORTS TO HELP ENERGY COMPANIES ADDRESS HIGH PERFORMANCE COMPUTING, AI, AND ADVANCED ANALYTICS TO ADDRESS CHALLENGES ACROSS THE POWER, UTILITIES AND RENEWABLES AND OIL AND GAS SECTORS.
HIS CAREER HAS BEEN FOCUSED ON THE INTERSECTION OF TECHNOLOGY AND ENERGY, INCLUDING ROLES AT NVIDIA, HALLIBURTON, AND SILICON GRAPHICS.
UM, MARK ALSO SERVES ON THE BOARD OF DIRECTORS OF EPRI, WHICH YOU'LL HEAR MORE ABOUT, AND THE ADVANCED BOARD, UH, ADVISORY BOARD OF GRID FORWARD, WHICH SUPPORTS INNOVATION AND MODERNIZATION ACROSS THE ELECTRIC POWER INDUSTRY.
ASSAN IS THE MANAGING DIRECTOR OF ENERGY AND NVIDIA, WHERE HE DRIVES DIGITAL INNOVATION ACROSS THE ENERGY SECTOR THROUGH AI AND ACCELERATED COMPUTING.
HE WORKS CLOSELY WITH OIL AND GAS POWER, UTILITIES, EPC AND TECHNOLOGY FIRMS TO ADVANCE THE ADOPTION OF AI AND DIGITAL TWINS FOR FASTER AND MORE SUSTAINABLE ENERGY TRANSITION.
OVER THE LAST 20 YEARS, HE'S WORKED AT IN LEADERSHIP ROLES AT SIEMENS, GE, AND BAKER HUGHES, IN, IN ASIA, MIDDLE EAST, AFRICA, UM, EUROPE, AND THE AMERICAS.
I'D LIKE TO THANK OUR GUESTS FOR JOINING US TODAY AND, UM, NOTE THAT THE COMMITTEE MEMBERS WILL HAVE OPPORTUNITIES TO ASK RISK MANAGEMENT RELATED QUESTIONS DURING THE EXECUTIVE SESSION THAT WILL FOLLOW THE MAIN SESSION.
WITH THAT, LEMME TURN IT OVER TO MARK AND ASAN.
AND THANK YOU FOR HAVING US TODAY.
IT'S A GREAT OPPORTUNITY FOR NVIDIA TO COME AND ADDRESS ERCOT AND, AND THEIR STAKEHOLDERS AS WE BELIEVE TEXAS IS A FANTASTIC OPPORTUNITY FOR AI DEVELOPMENT AND GROWTH IN THE NEXT INDUSTRIAL REVOLUTION, WHICH WE SEE AI PLAYING A SIGNIFICANT PART IN.
SO THIS MORNING I WANTED TO TAKE A FEW MINUTES, UH, AS, AS THE CHAIRMAN ACKNOWLEDGED, UH, MARK SPEER, I, I LEAD THE GLOBAL ENERGY BUSINESS, UH, HERE IN, UH, BASED HERE IN HOUSTON, OR DOWN IN HOUSTON, IN TEXAS.
UM, BUT RESPONSIBLE GLOBALLY AND SN
UM, FOR THOSE OF YOU WHO ARE NOT OVERLY AWARE OF WHAT NVIDIA DOES, UM, WHICH IS, UH, YOU KNOW, FIVE YEARS AGO OR SEVEN YEARS AGO WHEN I STARTED AT NVIDIA, MOST PEOPLE DIDN'T KNOW NVIDIA, EXCEPT IF THEY WERE A GAMER, MAYBE A BITCOIN MONITOR.
MAYBE WE'RE LUCKY ENOUGH TO INVEST IN US EARLY, BUT TODAY WE'VE, WE'VE BECOME QUITE A BIT MORE POPULAR.
AND, UH, JENSEN JUST LAST NIGHT INTRODUCED OUR FIRST, UH, PC CHIP, UH, ALONGSIDE MICROSOFT TO BASICALLY POWER AI AT THE, AT THE PC LEVEL.
UM, NVIDIA CONTINUES TO TRANSFORM ITSELF, UM, ALL THE TIME.
IT'S TRANSFORMED ITSELF CONSIDERABLY SINCE I BEGAN SEVEN YEARS AGO WITH THE COMPANY.
AND EVERY DAY WE LOOK TO SOLVE THE WORLD'S MOST CHALLENGING PROBLEMS AND USE ACCELERATED COMPUTING AND AI AND, AND FIGURE OUT HOW WE DO THAT.
WE HAVE A GREAT OPPORTUNITY IN TEXAS TO LEAD THAT, RIGHT? AND SO BEING A TEXAN AND, AND, AND BEING BASED HERE FOR THE LAST 25 YEARS, I STRONGLY BELIEVE THAT THE INTERSECTION OF, OF ENERGY AND AI IS, IS PARAMOUNT RIGHT NOW.
[00:05:01]
AND THE ABILITY TO UNDERSTAND WHAT'S HAPPENING IS, IS GOING TO BE CRITICAL AS WE BRING MORE AND MORE RESOURCES BOTH LOAD, UM, AND GENERATION ONTO THE GRID.AND SOMEBODY ONCE TOLD ME A GOOD CUSTOMER IS NOT THE CUSTOMER WHO USES THE MOST POWER, BUT IT'S THE MOST PREDICTABLE CUSTOMER.
SO WE'RE TRYING HARD TO HELP SOLVE THOSE PROBLEMS FOR THE INDUSTRY.
SO, UM, LET ME GO TO THE NEXT SLIDE HERE.
OH, RIGHT TO LEFT, NOT UP AND DOWN.
SO I, I WANTED TO DESCRIBE WHAT, WHAT WE CALL AS THE FIVE LAYER CAKE.
AND, AND THIS IS HOW WE BASICALLY MAKE THE MOST AMOUNT OF TOKENS PER WATT, RIGHT? AND, AND CREATE THE MOST ENERGY EFFICIENT WAY OF PRODUCING AI.
AND THAT STARTS AT THE FOUNDATIONAL LEVEL WITH ENERGY.
IF ANY OF YOU HAVE EVER BEEN INVOLVED IN THE MANUFACTURING BUSINESS, AND, AND I KNOW FOR, UH, ENERGY COMPANIES, THEY ARE, THEY TAKE FEED STOCKS AND, AND THEY TURN 'EM INTO ELECTRONS AND THEN MOVE 'EM ALONG THE GRID OR ANY MANUFACTURING PROCESS.
WE RUN 'EM THROUGH A FACTORY, WHICH WE CALL AN AI FACTORY, AND WE TURN THEM INTO TOKENS, WHICH ARE UNITS OF, UH, KNOWLEDGE AND THOSE TO TOKENS THAT ARE THEN SOLD TO PEOPLE WHO CAN CONSUME THEM AND TURN THEM INTO EFFICIENCIES AND, AND REVENUE IN THE BUSINESSES THAT THEY RUN.
ON TOP OF THAT ENERGY, YOU HAVE COMPUTER, THE CHIP SETS THAT, THAT WE PRODUCE.
AND THEN THERE'S INFRASTRUCTURE ON TOP OF THAT THAT BASICALLY BUILDS THAT INTO, UH, AN AI FACTORY, OR PEOPLE CALL 'EM DATA CENTERS.
BUT THE DIFFERENCE BETWEEN A DATA CENTER AND AN AI FACTORY IS DATA CENTERS ARE VERY HETEROGENEOUS.
THEY DO WEB SERVICES, EMAIL, S UH, SAP SYSTEMS, ERP SYSTEMS AND THAT.
SO EVERY RACK'S A LITTLE BIT DIFFERENT.
THERE'S LOTS OF DIFFERENT APPLICATIONS RUNNING AT DIFFERENT POWER DEMANDS.
AND AI FACTORY IS, IS VERY HOMOGENEOUS, RIGHT? IT PRODUCES TOKENS, IT DOES ONE THING, AND WE NEED TO MAKE SURE THAT IT DOES THAT ONE THING VERY WELL AND VERY EFFICIENTLY.
AND SO WE HAVE TO WORK WITH AN ECOSYSTEM OF INFRASTRUCTURE PLAYERS IN THE COOLING AND THE POWER SPACE AND, YOU KNOW, LAND POWER CONNECT CONNECTIVITY AND, AND THE SHELL ITSELF TO MAKE IT AS EFFICIENT AS POSSIBLE.
AND THEN WE HAVE ALL THESE AI MODEL BUILDERS ON TOP OF THAT.
AND IT'S ONLY WHEN WE'RE WORKING AT THAT FULL STACK THAT WE CAN REMOVE NOT ONLY THE INEFFICIENCIES IN THE SYSTEM, BUT WE CAN MAKE THE SYSTEM MORE PREDICTABLE, RIGHT? BECAUSE IT'S NOT JUST A MATTER OF THE, THE, UH, SYNCHRONIZATION OF, OF JOBS BEING DONE FOR TRAINING OR INFERENCE AND HOW THEY RAMP UP AND RAMP DOWN.
THAT'S NOT JUST AT THE HARDWARE LEVEL.
IT'S NOT JUST AT THE INFRASTRUCTURE LEVEL WHEN COOLING SYSTEMS KICK ON OR DIFFERENT ENERGY SYSTEMS KICK ON, BUT IT'S ALSO WHEN THE APPLICATIONS ARE DOING DIFFERENT THINGS AND WE NEED TO LOOK AT THAT FULL STACK TO BECOME ENERGY EFFICIENT.
CAN I ASK YOU A QUESTION ABOUT THE STACK? SURE.
IN THE STACK, THE MIDDLE OF BAR IS CALLED INFRASTRUCTURE.
WHAT DO YOU MEAN BY THE TERM INFRASTRUCTURE? INFRASTRUCTURE? THAT, THAT'S THE ENTIRE BUILD OUT OF THE AI FACTORY.
SO THAT COULD BE, YOU KNOW, UM, TRANSFORMERS, RECLOSERS, IT COULD BE PDU AND CDUS AND, AND EVERYTHING ELSE THAT SURROUNDS THE CHIPS THAT ACTUALLY DO IT.
SO IF, IF A LOT OF PEOPLE ARE ASKING, IS AI A A A BUBBLE? AND, AND WE OBVIOUSLY DON'T BELIEVE IT IS, WE, WE BELIEVE THAT WE'RE GONNA START TO TRANSFORM TODAY.
YOU KNOW, WE TALK ABOUT THESE THREE DIFFERENT SCALING LAWS FOR AI WITH LLMS, AND MOST PEOPLE THINK OF AI AS, YOU KNOW, CHAT, GBT AND, AND WHAT WE DO TODAY.
BUT, YOU KNOW, IT STARTED THAT WAY AND IT HAS BECOME EXTREMELY USEFUL FOR PEOPLE IN GETTING ANSWERS AND, AND SOLVING PROBLEMS VERY QUICKLY.
BUT AS WE MOVE INTO AG GENTECH WORKFLOWS WHERE WE ACTUALLY HAVE VIRTUAL HUMANS HELPING US TO SOLVE PROBLEMS, AND AS WE TAKE THE LEAP INTO PHYSICAL AI WHERE WE'RE GONNA HAVE ROBOTS AND CARS AND, UH, MANUFACTURING FACILITIES, AND THEN IT'S EVENTUALLY GONNA LEARN THE LAWS OF, UH, THE, THE PHYSICS AND CHEMISTRY AND BIOLOGY AND OTHERS TO HELP US SOLVE DIFFERENT PROBLEMS. AND THEN, UH, THESE OPEN MODELS THAT ARE BEING CREATED AS SPECIFICALLY DOMAIN SPECIFIC MODELS ARE GONNA HELP US SOLVE PROBLEMS. JENSEN ALWAYS SAID THERE'S MORE PROBLEMS THAN PEOPLE IN THIS WORLD.
AND, YOU KNOW, WE'RE GONNA CONTINUE TO SEE THE EFFICIENCY OF PEOPLE AND THE ABILITY FOR PEOPLE TO SOLVE BIGGER AND BIGGER PROBLEMS. ONE THING TO THINK ABOUT, I'M SURE WHEN CALCULATORS CAME OUT, MOST PEOPLE THOUGHT MATHEMATICIANS WOULD GO AWAY, RIGHT? BUT IT JUST ALLOWED MATHEMATICIANS TO DO MUCH MORE COMPLEX MATH IN THEIR LIFETIME AND SOLVE PROBLEMS THAT THEY COULDN'T HAVE SOLVED IN A REASONABLE AMOUNT OF TIME.
THAT'S WHAT AI IS GONNA DO IN THE FUTURE.
IT'S GONNA HELP PEOPLE SOLVE REALLY
[00:10:01]
HARD PROBLEMS WHERE THEY CAN'T AFFORD TO STAFF THE PEOPLE, THEY CAN'T FIND THE PEOPLE, OR THEY NEED TO HIRE MORE PEOPLE, BUT THEY NEED 'EM TO BE DO THE WORK.INSTEAD OF 30,000 PEOPLE DOING THE JOB, THEY NEED 50,000, AND THEY'RE GONNA CREATE THAT THROUGH EFFICIENCIES USING AI TO INCREASE THEIR WORK, UH, WORK STAFF.
SO IF, IF WE, IF WE LOOK AT THIS CONCEPT OF, OF NOT A BUBBLE, RIGHT? YOU KNOW, IN 2023 CHAT, GBT CHANGED THE WAY WE LOOK AT AI, RIGHT? AND REALLY THE BIG FACTOR THERE WAS IT MADE IT COMMERCIALLY VIABLE, RIGHT? I LOOKED AT SOME OF THE EARLY STATISTICS AND BACK, BACK IN, UH, 2023 AND, AND 2024, IF YOU WERE TO TRAIN A 1.8 TRILLION PARAMETER MODEL, AND THAT'S ALL THE DIFFERENT, UH, DATA THAT GOES INTO IT, IT WOULD'VE TAKEN ABOUT 5,500 GIGAWATT HOURS TO TRAIN THAT MODEL.
SO, TO PUT IT IN PERSPECTIVE, IN ABOUT 2016, IF THAT WOULD'VE BEEN DONE WITH OUR CURRENT GENERATION OF, UH, COMPUTE, IT WOULD'VE TAKEN ABOUT 215,000 RACKS OF, OF GPUS.
FAST FORWARD 10 YEARS TO 2026 WITH BLACKWELL AND ALL THAT, IT TAKES ABOUT THREE GIGAWATT HOURS.
SO WHAT, 2,700 X SPEED UP AND, YOU KNOW, MAYBE 70 RACKS, 75 RACKS.
SO THE PERFORMANCE IMPROVEMENT FROM 10 YEARS AGO TO, TO JUST A FEW YEARS AGO HAS BEEN TREMENDOUS.
AND ON THE INFERENCING SIDE, RIGHT, WHEN YOU ACTUALLY ASK THE QUESTION AND GET THE OUTCOME, IT'S GONE UP A HUNDRED THOUSAND TIMES.
AND, YOU KNOW, WHILE WE GET MORE COMPLEXITY, WE'RE GONNA SEE THIS, THIS PARADOX CALLED JEVONS PARADOX, WHERE AS WE INCREASE PERFORMANCE AND REDUCE COSTS, WE'RE GONNA SEE MORE AND MORE USE BECAUSE MORE WORKLOADS ARE GONNA BECOME COMMERCIALLY ACHIEVABLE, AND WE'RE GONNA SEE A HUGE INCREASE.
AND SO WE ALREADY STARTED TO SEE THE MODELS AND, AND, AND CONTEXT GO UP 10 X AND TOKENS GO UP, UH, 10 X BACK IN 2024.
AND THEN MOVING INTO 25, WE START TO SEE ALL OF THESE AGENTS AND AND REASONING MODELS, AND WE'RE SEEING A HUNDRED X INCREASE.
AND THE GREAT THING IS, IS WE'RE DOING EVERYTHING WE CAN TO MAKE SURE THAT OUR ACCELERATED COMPUTING PLATFORMS KEEP UP WITH THAT PACE, NOT ONLY FROM A HARDWARE INFRASTRUCTURE PERSPECTIVE, BUT REALLY FROM A SOFTWARE PERSPECTIVE.
THESE ARE ALL SOFTWARE DEFINED SOLUTIONS.
SO OUR ABILITY TO WRITE BETTER SOFTWARE MAKES A HUGE DIFFERENCE.
AND JENSEN WILL BE THE FIRST JENSEN AS OUR CEO.
MOST PEOPLE KNOW HIM AS JENSEN.
UM, BUT HE'S, HE'S PUBLICLY TALKED ABOUT HOW MUCH WE SPEND ON AI AGENTS AND AI SOFTWARE DEVELOPMENT FOR OUR EMPLOYEES.
AND, AND IT'S, IT'S, IT'S A LOT.
AND YET WE'RE STILL HIRING DEVELOPERS EVERY MONTH BECAUSE WE CAN'T KEEP UP AND WE'RE MAKING THEM MORE AND MORE EFFICIENT.
SO IT'S NOT GONNA REPLACE PEOPLE, IT'S NOT GONNA REMOVE JOBS.
YOU'RE GONNA BE ABLE TO SOLVE PROBLEMS THAT YOU MIGHT NOT HAVE SOLVED FOR THREE TO FIVE YEARS.
YOU'RE GONNA PULL FORWARD THOSE WORKLOADS BY USING AI.
SO ONE OF THE THINGS I, I WANTED TO TALK ABOUT QUICKLY ALSO WAS MAXIMIZING, UH, AI TOKENS PER WATT.
WHEN YOU LOOK AT THESE LARGE AI FACTORIES, RIGHT? AND THE INEFFICIENCY OF THE FACTORY, AND THIS GOES BACK TO LOOKING AT THINGS AS, AS A COMPLETE MODEL, WHEN YOU LOOK AT THE POWER GOING INTO THE DATA CENTER, AND ACTUALLY THE POWER THAT'S BEING CONSUMED TO GENERATE TOKENS, YOU KNOW, IT, IT'S BASICALLY 60 TO 70% OF THE POWER IS ACTUALLY BEING USED TO GENERATE TOKENS.
AND THAT'S THE VALUE, RIGHT? THAT'S WHAT PEOPLE ARE PAYING FOR THAT PROVIDES THE COMMERCIAL VALUE AND THINGS LIKE THAT.
SO, YOU KNOW, A CERTAIN PORTION GOES TO OVERHEAD AND A CERTAIN PORTION GOES TO, UH, DIFFERENT CAPABILITIES AT, AT THE COOLING AND THE RACK LEVEL.
AND, AND I DON'T WANNA SAY INEFFICIENCIES, YOU HAVE TO DO THESE THINGS, BUT AT THE SAME POINT, EVERYTHING WE CAN DO TO REMOVE INEFFICIENCIES AT THE AI FACTORY LEVEL AND AT THE GRID LEVEL.
SO HOW DO WE WORK TOGETHER WITH GRID PROVIDERS TO MAKE SURE THAT EVERY ELECTRON THAT'S DEVELOPED IS UTILIZED FOR THE OUTCOME THAT IT WAS MEANT TO, OR THAT WE NEEDED TO, AND NOT IN COOLING OR POWER OR OTHER THINGS IS, IS GONNA BE CRITICAL.
AND SO WE'VE DEVELOPED A REFERENCE DESIGN CALLED DSX THAT WE OPEN AND WE PROVIDE, AND WE'VE GOT OVER A HUNDRED PARTNERS WORKING ON THAT, AND I'LL TOUCH BASE ON THAT.
BUT THE WHOLE GOAL IS HOW DO, HOW DOES NVIDIA AND OUR PARTNERS WORK TOGETHER TO CREATE THIS EFFICIENCY THAT REDUCES THE WASTE IN THE SYSTEM? NO DIFFERENT THAN WHAT YOU GUYS DO EVERY DAY.
YOU GUYS ARE, ARE, YOU KNOW, AND THE UTILITIES THAT YOU SUPPORT, BASICALLY WORK VERY CAREFULLY TO MAKE SURE THAT YOU CAN REDUCE WASTE IN THE SYSTEM AND MAKE SURE THAT EVERY ELECTRON
[00:15:01]
THAT YOU GUYS PRODUCE HAS THE BEST CHANCE OF BEING USED AT THE END AND NOT DROPPING OFF A LINE OR NOT BEING WASTED AT THE TRANSFORMER, AT THE SUBSTATION, RIGHT? WE'RE DOING THE SAME THING NOW AT THE DATA CENTER LEVEL, BECAUSE EVEN A ONE OR 2% IMPROVEMENT IN PERFORMANCE, IF YOU THINK ABOUT IT, ONE OR 2% IN PERFORMANCE IMPROVEMENT IN IN POWER CREATES ONE TO 2% MORE TOKENS.IF, IF THESE GIGAWATT DATA CENTERS ARE CREATING A HUNDRED MILLION DOLLARS A DAY IN REVENUE, A DAY IN REVENUE FOR THEIR SHARE, FOR THEIR STAKEHOLDERS, YOU'RE TALKING ABOUT ANOTHER ONE TO $2 MILLION A DAY IN, IN, IN REVENUE, RIGHT? YOU TURN, AND THEN YOU LOOK AT THE TAXES THAT THAT CREATES FOR THE, UH, COMMUNITIES IN WHICH THEY SERVE AND EVERYTHING ELSE, IT'S SIGNIFICANT.
SO ENERGY EFFICIENCY AND CAPABILITIES IS, IS CRITICAL.
SO THIS IS WHAT DSX LOOKS LIKE.
UM, IT IS EXTREMELY ELABORATE, EXTREMELY COMPLEX, AND IT REQUIRES AN ECOSYSTEM TO BUILD, RIGHT? AND SO IT, IT'S NOT JUST ABOUT THE UNDERLYING CHIPS AND THE EFFICIENCY IN THE SILICON, IT'S THE SOFTWARE STACK.
IT IS THE, UH, DIRECT LIQUID COOLING INFRASTRUCTURE.
A LOT OF PEOPLE TALK ABOUT WATER AND STUFF LIKE THAT.
YOU KNOW, OUR NEXT GENERATION VERA RUBIN STACKS WILL BE, UH, DIRECT LIQUID COOLED.
AND WHAT THAT MEANS, IT'S, IT'S A CLOSED LOOP SYSTEM.
SO WE'RE NOT WASTING A BUNCH OF WATER, WE'RE RECYCLING THE WATER.
OUR CHIPS CAN ACTUALLY RUN UP TO 45 DEGREES SEA, RIGHT? WHICH IS VERY WARM.
AND SO WE DON'T NEED THE SIGNIFICANT AMOUNT OF CHILLERS AND EVERYTHING ELSE.
ONCE AGAIN, THE LESS WE PUT INTO, UH, POWERING CHILLERS, THE MORE WE CAN PUT INTO GENERATING TOKENS, RIGHT? SO EVERYTHING WE DESIGN FOR IS TO BECOME AS IF ENERGY AS EFFICIENT AS POSSIBLE.
AND SO YOU CAN SEE HERE, AND I KNOW THESE ICONS ARE PRETTY SMALL, BUT IF YOU LOOK ACROSS THE FOUR DIFFERENT AREAS IN WHICH WE'RE ENABLING DSX AND DSX FLEX, MANY OF YOU MIGHT HAVE HEARD, WE, WE'VE PARTNERED WITH EMERALD AI, AND THIS IS HOW WE COMMUNICATE WITH THE GRID IN ORDER TO BRING FLEXIBILITY BACK SO THAT THE GRID OPERATORS CAN HAVE, UH, UM, A DEMAND RESPONSE WITH THESE LARGE AI FACTORIES THAT SAYS, HEY, YOU KNOW, WE CAN GIVE YOU THE POWER THAT YOU'RE LOOKING FOR, BUT THERE'S GONNA BE 10 HOURS, 20 HOURS A WEEK, A YEAR, 30 HOURS A YEAR, THAT WE'RE GONNA NEED YOU TO BE FLEXIBLE BACK WITH US, AND WE'RE GONNA NEED TO REDUCE YOUR POWER.
I'M, I'M GONNA SHOW YOU SOME EXAMPLES, AND WE CAN ACTUALLY REDUCE THE POWER CONSUMPTION IN ONE OF A FEW WAYS, RIGHT? BY EITHER RAMPING DOWN THE POWER AT THE INDIVIDUAL GPU LEVEL BY REDIRECTING INFERENCING WORKLOADS TO OTHER DATA CENTERS IN DIFFERENT TERRITORIES BY SCHEDULING JOBS DIFFERENT.
BUT THESE ARE COMPLEX SOFTWARE SOLUTIONS THAT THE, THE MARKET HAS, UM, DEVELOPED, RIGHT? BETWEEN DIFFERENT COMPANIES WITH US AND, AND OUR ECOSYSTEM.
AND YOU'LL HEAR US TALK A LOT ABOUT ERI.
UM, WE WORK VERY CLOSELY WITH EPRI, UM, TO, TO TRY TO MAKE SURE THAT THIS IS NOT AN NVIDIA BASED SOLUTION, BUT IT'S AN INDUSTRY BASED SOLUTION.
SO, YOU KNOW, JUST JUST LOOKING AT THE NUMBERS, AND THESE AREN'T OURS, BUT THIS WAS A, A STUDY DONE BY DUKE UNIVERSITY IS IF WE COULD, UM, PROVIDE ABOUT A HALF A PERCENT OF FLEXIBILITY WITHIN THE ERCOT AREA, RIGHT? WE COULD, WE COULD FREE UP AS MUCH AS 10 GIGAWATTS OF, OF POWER, RIGHT? SO IF YOU THINK ABOUT 10 GIGAWATTS OF POWER, UM, YOU KNOW, THAT, THAT SIGNIFICANT BUILD OUT OF INFRASTRUCTURE, SIGNIFICANT TAX REVENUES FOR TEXAS, AND, AND IT REALLY ALLOWS US TO SCALE UP.
AND, AND YOU CAN SEE THERE'S LOTS OF OPPORTUNITIES AND ONCE AGAIN, ARE, ARE THESE EXACT PROOF POINTS, I'M NOT SURE.
BUT I THINK RIGHT NOW WE WOULD SEE THAT WE'RE BUILDING OUT AND WE'RE APPROVING BASED ON A FIVE NINE AVAILABILITY FOR DATA CENTERS AND, AND TRADITIONAL DATA CENTERS, YOU MIGHT NEED THAT BECAUSE IT'S RUNNING CORE BUSINESS THINGS.
THOSE JOBS CAN'T BE SCHEDULED WHEN SOMEBODY NEEDS TO ACCESS DATA.
WITH US, WE CAN ACTUALLY SCHEDULE A LOT OF THE AI WORKLOADS, AND SO WE CAN ACTUALLY MOVE TRAINING JOBS, WE CAN MOVE INFERENCING, WE CAN BE FLEXIBLE, WE CAN RAMP IT DOWN.
IF A TRAINING JOB'S GONNA TAKE SIX MONTHS, IS IT A BIG DEAL? IF IT TAKES SIX MONTHS AND SIX HOURS, WELL, PROBABLY NOT IF I CAN START WORKING IN SIX MONTHS VERSUS IN SIX YEARS WHEN I GET THE INTERCONNECTION, RIGHT? SO THESE ARE THE TYPES OF THINGS THAT WE'RE WORKING CLOSELY TO HELP BOARDS AND, AND, UH, PUCS AND OTHERS TO UNDERSTAND THE BENEFITS OF, OF FLEXIBILITY.
SEN FEEL FREE TO JUMP IN AT ANY TIME.
IF I MISS SOMETHING, HE'S EASE THE BRAINS BEHIND A LOT OF THIS STUFF.
UM, SO IF, IF ANY OF YOU, AND I ASSUME MOST OF YOU ARE FAMILIAR WITH ERI, THE ELECTRIC POWER RESEARCH INSTITUTE.
[00:20:01]
MEMBER ON THERE, AND I WORK VERY CLOSELY WITH AOT AND HIS TEAM, AND, AND THEY STARTED THE D DC FLEX INITIATIVE A COUPLE YEARS AGO AS A WAY OF PULLING TOGETHER THE ECOSYSTEM OF PARTNERS TO BUILD STANDARDS FOR DATA CENTERS AS A FLEXIBLE RESOURCE.SO YOU CAN SEE THAT ONCE AGAIN, THIS, THIS CONSISTS OF TECHNOLOGY COMPANIES AND ISOS AND RTOS AND LOTS OF UTILITIES AND ENGINEERING FIRMS, AND IT CONTINUES TO GROW.
AND, YOU KNOW, THIS INDUSTRY IS TYPICALLY VERY CAUTIOUS AND, AND TENDS TO MOVE AT, AT A CERTAIN PACE, BUT I, I THINK WHAT WE'VE SEEN IS THAT THIS INDUSTRY'S MOVING FASTER THAN IT'S PROBABLY MOVED IN IN, IN MANY DECADES.
AND IT'S A, IT'S AN INCREDIBLE OPPORTUNITY.
I'VE SPENT 25 YEARS IN THE INDUSTRY AND, UM, YOU KNOW, IT'S, IT'S ALWAYS BEEN A COMMODITY THAT'S REALLY IMPORTANT FOR A LOT OF PEOPLE.
BUT NOW IT'S VERY STRATEGIC, RIGHT? AND, AND WHAT I WOULD TELL YOU IS, IS AS DATA CENTERS ARE LOOKING AT WHERE THEY'RE GONNA BUILD AND WHERE THEY'RE GONNA INVEST MONEY AND, AND, AND DEVELOP THE COMMUNITIES IN WHICH THEY SERVE, IT'S NO LONGER ABOUT TAX INCENTIVES.
AND WHO PROVIDES TAX INCENTIVES? IT'S WHO CAN PROVIDE POWER, RIGHT? YOU KNOW, IF SOMEBODY OFFERS A GREAT TAX INCENTIVE, THAT'S GREAT, BUT IF THEY CAN'T GET THE POWER FOR FOUR TO FIVE YEARS, THEY'RE GONNA GO SOMEWHERE WHERE THEY CAN GET POWER, BECAUSE THESE ARE REVENUE GENERATING ASSETS, RIGHT? MOST DATA CENTERS ARE COST CENTERS, RIGHT? AND WHEN WE TALK WITH CIOS AND, AND YOU, YOU GO AND APPROACH YOUR CIO AND SAY, HEY, I NEED MORE DATA CENTER SPACE.
IT, IT'S NOT POPULAR, RIGHT? IT'S A COST.
WHEN YOU'RE RUNNING AI FACTORIES AND YOU SAY, HEY, I NEED MORE, MORE SPACE, IT'S BECAUSE YOU'VE SOLD OUT OF TOKENS AND YOU'RE GENERATING A LOT OF REVENUE, AND SO YOU, YOU NEED MORE, RIGHT? SO THE THE OPPORTUNITY TO CREATE THESE FLEXIBLE ASSETS AND GROW THE ECONOMIES IN WHICH THEY SERVICE TREMENDOUS.
AND SO WE'VE GOT THE ENTIRE INDUSTRY COMING TOGETHER TO BUILD THESE.
UM, MANY OF YOU HAVE HEARD ABOUT EMERALD AI, IT'S A SMALL STARTUP.
UM, YOU KNOW, IT, IT'S A COMPANY WE'VE NOW INVESTED IN, AND A LOT OF TIER ONE UTILITIES AND VENTURE GROUPS WITHIN THE POWER AND UTILITY SPACE, INCLUDING GE, RENOVA AND OTHERS HAVE INVESTED.
THEY'VE COME UP WITH THIS GREAT CONCEPT OF BEING ABLE TO, YOU KNOW, BASICALLY ORCHESTRATE COMMUNICATION WITH THE GRID OPERATORS AND BE ABLE TO FEED, UM, UH, SUPPLY AND DEMAND SIGNALS BACK TO, UH, AI FACTORIES AND DATA CENTER CAPABILITIES SO THAT WE CAN BASICALLY SEE WHERE THEY'RE PREDICTING NEED AND RESPOND TO IT IN REAL TIME.
AND WE'VE DONE NOW, UH, I THINK IT'S, UH, FIVE OR SIX STUDIES WITH THEM WHERE WE'VE SEEN POSITIVE RESULTS, WHERE WE'VE GOTTEN INDICATION FROM THE UTILITY THAT THEY NEED EXTRA, UM, CAPACITY IN ORDER TO MEET A PEAK DEMAND.
AND WE'VE BEEN A ABLE TO CURTAIL DEMAND RESPONSE IN REAL TIME, SIGNIFICANT LOADS, RIGHT? AND SO YOU SEE THOSE DIFFERENT, UH, CAPABILITIES HERE.
AND ONCE AGAIN, THOSE, THOSE ARE IN PILOT PHASE OR, OR, UH, YOU KNOW, TEST PILOTS AND STUFF LIKE THAT.
THE LAST ONE THAT YOU SEE IN, UH, WHICH WILL TAKE PLACE IN PANASAS IN VIRGINIA, WILL ACTUALLY BE A HUNDRED MEGAWATT.
IT'S, I THINK IT'S 96 MEGAWATT FACILITY THAT WE'RE BUILDING.
AND WE WILL HAVE IT A HUNDRED PERCENT FLEXIBLE TO THE GRID UP IN VIRGINIA AND BASICALLY BE ABLE TO CURTAIL LOAD AS THEY NEED IT.
AND SO THIS WILL BE THE FIRST, UH, ONLINE PRESENCE.
BUT, YOU KNOW, WE'VE ALREADY SEEN THIS WORK, UH, A FEW TIMES.
YOU KNOW, THIS IS AN EXAMPLE THAT WE DID WITH NATIONAL GRID, WITH EMERALD AI, WHERE BASICALLY WHEN, WHEN THE FOLKS IN IN THE UK ARE WATCHING SOCCER GAMES OR FOOTBALL GAMES AS THEY REFER TO 'EM, UH, AT HALFTIME, THEY TURN ON THEIR ELECTRIC TEE KETTLES, AND ALL OF A SUDDEN THERE'S A HUGE POWER SPIKE IN THE, IN THE IN DEMAND.
AND YOU CAN SEE THAT ON THE TOP, UM, ON THE TOP, UH, SET OF LINES, RIGHT? SO ALL OF A SUDDEN, YOU KNOW, AT HALFTIME, THEY SEE A HUGE INCREASE IN POWER.
AND THEN AT TOWARDS THE END OF THE GAME, YOU SEE ANOTHER HUGE INCREASE.
AND WHAT WE WERE ABLE TO DO WAS ACTUALLY FLEX AN AI FACTORY IN ACCORDANCE IN THE SAME AREA TO BASICALLY PROVIDE THE OFFSETTING POWER REQUIREMENTS FOR THOSE INCREASE IN DEMANDS.
AND SO THE GOAL IS, IS HOW DO, HOW DO WE DO THIS ON A REGULAR BASIS? THIS, THIS SHOULDN'T BE AN ALL THE TIME PHENOMENON, BUT WE, WE DEFINITELY KNOW THAT THERE ARE 10 HOURS, 20 HOURS, RIGHT? 50 HOURS A YEAR WHERE PEAK DEMAND IS, IS IS IMPORTANT.
AND THESE GRID ASSETS BECOME HUGE, UM, UM, DEMAND RESPONSE ASSETS MUCH DIFFERENT THAN ASKING PEOPLE TO ADJUST THEIR THERMOSTATS OR TURN DOWN THEIR AC OR OTHERS.
WE ACTUALLY HAVE LARGE POOLS HERE.
SO I WANTED TO TALK FOR A SECOND ALSO ABOUT, SO I'VE TALKED A LOT ABOUT ENERGY FOR AI.
HERE'S THE OPPORTUNITY TO TALK ABOUT AI FOR ENERGY.
WE ARE DOING A LOT TO INCREASE RESILIENCY, RELIABILITY, SAFETY, SUSTAINABILITY
[00:25:01]
ALONG THE GRID.AND, AND, AND AS PART OF THAT, UH, EPRI STARTED THE OPEN POWER AI CONSORTIA WHERE WE'VE GOT, I THINK ALMOST 200 UTILITIES AND PARTNERS WORKING TOGETHER IN A GOAL THAT WE COULD BUILD DOMAIN SPECIFIC MODELS FOR THE INDUSTRY.
THIS INDUSTRY'S NOT REWARDED THE SAME WAY, LET'S SAY, AS, AS OTHER INDUSTRIES WHO CAN MAKE SIGNIFICANT PROFITS OFF OF MAJOR INVESTMENTS AND ALL THAT.
SO THE ONE THING I LOVE ABOUT THIS INDUSTRY IS WE COLLABORATE AND WE, WE COME TOGETHER AND WE CAN BUILD DOMAIN SPECIFIC MODELS THAT REALLY UNDERSTAND GRIDS, UNDERSTAND GENERATION AND CAPABILITIES, THAT'S BEEN TRAINED ON PUBLIC DATA AT FIRST.
AND THEN SECONDLY, POTENTIALLY, UH, PROPRIETARY DATA THAT'S NON-CONFIDENTIAL, RIGHT? YOU KNOW, A LOT OF PEOPLE HAVE DATA THAT'S NOT IN THE PUBLIC, UH, SPHERE THAT HASN'T BEEN TRAINED INTO THE LARGE FOUNDATIONAL MODELS.
THIS GIVES US AN OPPORTUNITY TO TAKE THAT AND TRAIN IT.
THIS WILL HELP THE FUTURE WORKFORCE TO ACCELERATE THE SPEED IN WHICH THEY GET UP TO SPEED, RIGHT? A LOT OF PEOPLE SPEND A LOT OF TIME WORKING.
WE HAVE HIGH TURNOVER AT LOWER LEVELS WITHIN ORGANIZATIONS, LINEMEN AND OTHERS, GIVING THEM THE CAPABILITY TO ASK AI AGENTS AND, AND FOR HELP TO BE ABLE TO TAKE PICTURES, UNDERSTAND WHAT THEY'RE LOOKING AT, AND BE MENTORED BY REALLY ALL THE DATA THAT'S BEEN CREATED IN THIS INDUSTRY WILL HELP THEM GET UP TO SPEED AND PROGRESS IN THEIR FIELDS.
AND WE'VE PARTNERED WITH SOME APPRENTICESHIP COMPANIES TO HELP DEVELOP APPRENTICES, TO HELP BUILD OUT AI FACTORIES AND HELP BUILD OUT, UH, ENERGY INFRASTRUCTURE, LEVERAGING AI, BUT ALSO LEVERAGING THE APPRENTICESHIP ASPECT OF IT TO PUT MORE PEOPLE TO WORK IN THE INDUSTRY.
BUT WE SEE A HUGE OPPORTUNITY FOR AI HELPING SERVE THIS INDUSTRY BECAUSE IT'S AN INDUSTRY THAT'S TYPICALLY UNDERSTAFFED AND, AND THE COMPLEXITY IN WHICH THEY NEED TO WORK IS BECOMING GREATER AND GREATER.
SO THEREFORE, IN ORDER TO ACHIEVE THE AFFORDABILITY GOALS, WE BELIEVE THAT WE CAN HELP THESE, THESE EMPLOYEES BECOME MORE EFFICIENT.
SO I WANT TO LEAVE YOU WITH THIS, AND I, I, I, UH, I GUESS I'LL TURN IT TO ESSEN IF YOU HAVE ANY CLOSING COMMENTS OR THOUGHTS.
UM, BUT REALLY, TEXAS SHOULD BE LEADING THIS, RIGHT? YOU KNOW, WE, WE, WE'VE GOT GREAT RELATIONSHIPS WITH SPP AND PJM AND OTHERS, AND, AND WE'RE, WE'RE WORKING WITH STARTUPS AND WE'RE WORKING WITH THE ECOSYSTEM, BUT ERCOT IS, IS TO ME, HOLDS A SPECIAL PLACE BECAUSE BEING IN TEXAS, WE'VE GOT THE OPPORTUNITY TO PRODUCE NATURAL GAS, TURN IT INTO POWER, MOVE THAT POWER VERY EFFICIENTLY, AND AND SERVICE A MARKET THAT THAT'S GONNA BE IN THE NEED FOR A LOT OF ELECTRICITY.
AND WE ALSO HAVE, UM, YOU KNOW, I, I THINK A, A ROBUST NETWORK OF, OF, UH, TRANSMISSION CAPABILITIES.
AND I THINK WORKING TOGETHER, IF WE CAN ADD MORE PREDICTABILITY AND, AND DEMAND RESPONSE CAPABILITIES AND FLEXIBILITIES, WE CAN ACTUALLY DRIVE A HIGHER UPTAKE OF THE POWER THAT'S EXISTING WITHIN THE ERCOT NETWORK, AND THEREFORE, UM, INCREASE THE DENOMINATOR ON THE FACTOR IN WHICH PEOPLE GET CHARGED AND THEREFORE MAKE THE FIXED ASSETS THAT ARE OUT THERE, UH, BETTER UTILIZED AND OVERALL REDUCE THE COST TO, UH, TEXANS EVERYWHERE AS, AS THEY, UH, LEVERAGE POWER ACROSS, UH, ALL OF THEIR NEEDS.
SO, UH, ESSAM, WHAT DID, DID I MISS ANYTHING SPECIFICALLY? NO, I THINK YOU'VE COVERED IT WELL, I THINK THE, UH, FOR, I MEAN, ONE OF THE IMPORTANT THING IS JUST, JUST TO REEMPHASIZE RIGHT, THAT, UH, THE FIRST SLIDE, WHICH MARKS TALKS ABOUT THE, IN THE FIRST TIME THE TECHNOLOGY COMPANY PUTTING ENERGY AS A TECHNOLOGY STACK.
WE HAVE TAKEN POWER AS, AS, UH, FOR GRANTED FOR AGES RIGHT NOW.
IT IS AN ESSENTIAL PART OF THE TECHNOLOGY.
IT'S NOT A COMMODITY, IT'S A CORE PIECE TO WHAT WE ARE TRYING TO BUILD AND BRING EFFICIENCY ALL THE WAY.
SO THIS IS WHERE THIS MEETING, THIS, UH, UH, OPPORTUNITY IS AN AMAZING THING FOR US.
HOW WE CONTINUE TO WORK TOGETHER, COMMUNICATE, UNDERSTAND.
WE ARE DOING A, A, A ALREADY A LOT WITH ERCOT.
I THINK THAT WE CONTINUE TO DO A LOT MORE AND, UH, TO BRING, MAKE POWER AFFORDABLE AND RELIABLE AND, AND DRIVE THE WHOLE ECONOMIC GROWTH.
ANY QUESTIONS? QUICK QUESTION.
WHEN YOU'RE LOOKING, UM, LIKE WITH YOUR SLIDE WITH THE T KETTLE, UM, ARE YOU THINKING THAT THESE LOADS WOULD REGISTER AS CONTROLLABLE LOADS RESOURCES, OR, OR IS IT MORE AS AN EMERGENCY TYPE COMPONENT? OH, SO WE ARE EXPOSING TELEMETRY DOWN TO THE RACK LEVEL INSIDE THE DATA CENTER.
SO THE UTILITIES WILL HAVE THAT
[00:30:01]
POWER.SO YOU JUST SAY, HEY, I NEED YOU TO SHED A HUNDRED MEGAWATT WITH THOSE RAMPED DOWN RATES BY THIS TIME.
AND, AND THEN THESE SOFTWARES WILL TAKE CARE OF IT.
THEY EITHER, THEY'RE GONNA THROTTLE, THEY'RE GONNA, UH, UH, SHIFT THE LOAD, PAUSE THE LOAD, WE KNOW WHAT NEEDS TO BE DONE AT THE RACK LEVEL, AT THE GPU LEVEL, THAT'S GONNA BE DONE.
SO IT'S EXPOSED YOU EVERYTHING.
SO IN A, EVEN IN DURING THE SHOULDER MONTHS, IF YOU NEED POWER MORE, BECAUSE THERE'S A NOW PEAK DEMAND COMES UP, OR DURING THE, UH, A NORMAL, UH, YOU KNOW, THE USUAL EXTREME WEATHER EVENTS, THESE, THESE TELEMETRIES WILL GONNA BE THERE EXPOSED FROM, FROM A, A DATA CENTER DOWN TO THE RACK LEVEL AND MAKE IT AVAILABLE TO YOU.
I APPRECIATE Y'ALL COMING FOR YOUR PRESENTATION TODAY.
YOU KNOW, NVIDIA HAS BEEN A REMARKABLE COMPANY TO GO FROM GRAPHICS PROCESSORS TO, UH, TO GAMERS TO, UH, AI.
WHAT DO YOU THINK, UM, KEEPS YOUR MANAGEMENT AWAKE AT NIGHT WHEN THEY'RE LOOKING AT THE FUTURE? HA, HAVING SPENT A LOT OF TIME WITH JENSEN E EVERYTHING KEEPS THAT GUY AWAKE,
BUT FOR HIM, I, YOU KNOW, THERE'S, UH, THERE'S DEFINITELY A PARANOIA ABOUT STAYING AHEAD OF THE COMPETITION AND MAKING SURE WE'RE DEVELOPING, UH, BETTER, FASTER, CHEAPER SOLUTION STACKS FOR THE MARKET.
UM, BUT, BUT REALLY, UM, WE'RE A FIRST PRINCIPLES BASED BUSINESS.
AND, AND JENSEN WILL SAY THAT ALL THE TIME, AND HIS THOUGHTS ARE, IF WE CAN SOLVE REALLY HARD PROBLEMS, THEN WE'LL BE OKAY.
RIGHT? AND SO WE, WE'VE GOT A VERY DIFFERENT BUSINESS MODEL, PEOPLE WHO HAVE SPENT TIME WITH NVIDIA.
WE, WE REALLY FOCUS ON OUR PARTNER ECOSYSTEM AND SOLVING PROBLEMS. 80% OF THE PEOPLE AT NVIDIA ARE ENGINEERS, 60% ARE SOFTWARE ENGINEERS.
WE ARE NOT A SALES AND MARKETING ORGANIZATION.
WE'VE BECOME POPULAR BECAUSE OF JENSEN AND BECAUSE OF STOCK PRICE AND BECAUSE OF OTHER THINGS.
BUT, BUT PROBABLY THE BIGGEST REASON IS, IS WE'RE SOLVING PROBLEMS. AND IT'S NOT JUST NVIDIA, IT'S, IT'S OUR PARTNER.
SO IF, IF YOU ASK JENSEN WHAT KEEPS HIM UP AT NIGHT, IT WOULD BE THE SUCCESS OF OUR ECOSYSTEM.
PEOPLE WHO HAVE INVESTED IN THE NVIDIA PLATFORM AND ARE NOT R REACHING ITS OBJECTIVES OR, OR GROWING AS FAST AS THEY WANT TO, HE, HE REALLY THINKS ABOUT HOW DO WE MAKE THAT PARTNER ECOSYSTEM MORE SUCCESSFUL BECAUSE HE KNOWS IF, IF THEY'RE SUCCESSFUL, WILL BY DEFAULT BECOME MORE SUCCESSFUL.
AND SO WE, WE REALLY FOCUS ON THIS ECOSYSTEM, AND THAT'S WHY WE'VE PARTNERED CLOSELY WITH EPRI AND WE'RE PARTNERING CLOSELY WITH YOU GUYS.
AND, AND ULTIMATELY IF, IF WE SOLVE REALLY HARD PROBLEMS, THEN THE CLOUD PROVIDERS AND THE DATA CENTERS AND, AND HP AND DELL WILL SELL STUFF.
BUT, UM, YOU KNOW, IF WE PARTNER CLOSELY WITH GE, VER NOVA, AND, AND IF YOU HAVEN'T SEEN THE NEW SIEMENS ENERGY NORA PLATFORM, IT'S AMAZING.
YOU KNOW, WE'VE WORKED WITH THEM FOR THREE TO FOUR YEARS ON THIS SOFTWARE PLATFORM THAT IS, IS SOLVING THINGS THAT HADN'T BEEN SOLVED BEFORE.
AND THEIR SUCCESS WILL MEAN SUCCESS FOR US.
SO, YOU KNOW, UH, THIS GUY'S PROBABLY LOST A LOT OF SLEEP OVER THAT PROJECT SPECIFICALLY, BUT THOSE ARE THE THINGS THAT WILL KEEP US UP THE MOST.
HOW ABOUT YOUR SUPPLY CHAIN ON THE CHIP SIDE, WHERE, WHAT'S THAT LOOK LIKE? UM, BOY, NOT MY AREA OF EXPERTISE, BUT AS, AS YOU KNOW, WE'VE, WE'VE COMMITTED PUBLICLY TO A LOT OF BIG SPEND WITH TSMC IN TAIWAN, AND OBVIOUSLY THEY'RE BUILDING MORE AND MORE MANUFACTURING HERE.
WE'RE ACTUALLY BUILDING, UH, IN, IN CONJUNCTION WITH FOXCONN, A HUGE MANUFACTURING FACILITY IN HOUSTON, RIGHT? SO WE'LL BE POWERING THAT WITH ERCOT AND, AND CENTERPOINT AND, AND REALLY WE, WE CONTINUE TO TRY TO GROW OUR AMERICAN SUPPLY CHAIN, BUT WE'RE, WE'RE A VERY GLOBAL ORGANIZATION.
AND THE, THE THING THAT I RECOGNIZE THE MOST OVER MY LAST SEVEN YEARS IS JENSEN IS EXTREMELY LOYAL.
AND SO EVEN DURING COVID AND TIMES WHERE PEOPLE WEREN'T MANUFACTURING AS MUCH AND WEREN'T BUYING AS MUCH, JENSEN WORKED WITH OUR SUPPLY CHAIN TO MAKE SURE THAT THEY WERE WELL TAKEN CARE OF, WOULD BUY IN ADVANCE, WOULD HELP THEM FILL THEIR LINES, WOULD DO THE RIGHT THING TO KEEP THEM SUPPORTED.
AND, AND THEREFORE WE'VE BUILT THESE LONG TRIED AND TRUE RELATIONSHIPS, RIGHT? AND, AND WE CONTINUE TO INVEST AND WE HAVE LONG-TERM RELATIONSHIPS WITH OUR SUPPLIERS, BUT THE GOAL IS TO OBVIOUSLY RAMP UP IN THE US QUITE A BIT.
THAT, THAT RIGHT NOW IS, IS A BIG PUSH FROM THE ADMINISTRATION, AND JENSEN HAS, UH, RESPONDED TO THAT VERY STRONGLY.
WE'RE ALSO LOOKING AT OTHER PLACES AROUND THE WORLD IN WHICH,
[00:35:01]
WHICH IT MAKES SENSE TO INCREASE OUR SUPPLY CHAIN.SO WHEN WE'RE ASSESSING LARGE DATA CENTER LOADS THAT ARE APPLYING TO COME ONTO THE TEXAS, THE ERCOT GRID, UM, WHAT QUESTION SHOULD WE BE ASKING THEM ABOUT THEIR ENERGY FLEXIBILITY? UM, AND I'LL PROBABLY DEFER TO ESSENCE TO CLARIFY THAT, BUT IT, IT'S WHAT LEVEL OF FLEXIBILITY DO YOU HAVE, RIGHT? ARE YOU BUILDING A FLEXIBLE DATA CENTER? ARE YOU BUILDING IT WITH BATTERIES AND, AND ARE YOU BUILDING IT WITH THE CAPABILITIES OF BEING ABLE TO REDIRECT LOADS? AND HOW MUCH FLEXIBILITY CAN YOU PROVIDE TO US AS, AS A GRID OPERATOR, RIGHT? BECAUSE I, I THINK IN THE PAST THERE'S, THERE'S BEEN LITTLE FLEXIBILITY BECAUSE OF FIVE NINE REQUIREMENTS WITH THEIR CUSTOMERS, AND THEY SAID, NO, WE CAN'T BE FLEXIBLE.
I THINK TODAY, DEPENDING ON THE TYPE OF, UH, FACILITY THEY'RE DEVELOPING, THEY CAN BE EXTREMELY FLEXIBLE.
AND EPRI'S ACTUALLY DEVELOPED A NEW, YOU KNOW, THOUGHT PROCESS ALONG THIS DC FLEX CALLED, UH, MOSAIC, WHERE THEY'RE ACTUALLY TRYING TO CLASSIFY THE FOUR OR FIVE DIFFERENT TYPES OF FLEXIBILITY SO THAT PEOPLE CAN ACTUALLY SAY, OKAY, I'M, I'M BLUE, OR I'M GREEN, OR I'M THIS, AND THAT IDENTIFIES THE FLEXIBILITY OF THE FACILITY THAT THEY'RE TRYING TO BUILD, GIVING YOU AN INDICATION OF WHAT KIND OF SLA YOU CAN HAVE WITH THEM AND, AND FEEL GOOD ABOUT.
YEAH, SO JUST ONE THING, RIGHT? THE AI FACTORIES ARE INHERENTLY FLEXIBLE, 50% FLEXIBLE, INHERENTLY, THEY DON'T RUN ON BASE LOAD.
SO IF THIS IS SOMETHING, A TECHNOLOGY WHICH IS TRULY FLEXIBLE, WHY NOT LEVERAGE THAT? YOU'RE PROVIDING THEM A FLEXIBLE LOAD TO THE GRID, WHICH BECOMES AN ASSET.
SO WHERE WE ARE, THERE'S SO MANY NEW ENTRANTS IN THE MARKET, WE KNOW HYPERSCALERS KNOW HOW TO BUILD THOSE, UH, LARGE DATA CENTERS IN OUR FACTORIES.
OUR CONTRIBUTION, OUR EFFORT IS TO REALLY HELP THESE NEW ENTRANTS, UH, TO BUILD THOSE THINGS.
BEING THE BEST CITIZEN TO THE GRID, THAT IS WHAT WE ARE TRYING TO WORK WITH.
THE WORK WE ARE DOING WITH THE AIR CODE, UH, SYSTEM PLANNINGS TEAM IS, UH, TAKING THEIR REQUIREMENT BACK TO OUR DATA CENTER REFERENCE DESIGN, YOU KNOW, THIS IS THE DSX IS BASICALLY, WE ARE OPEN SOURCING THE WHOLE MINDSET WE DID FOR THE SOFTWARE INDUSTRY, OPEN SOURCING, YOU KNOW, HOW WE OPEN SOURCE, THE DATA CENTER, REFERENCE DESIGN SO PEOPLE CONFORM TO AT THE, AT, AT THE CORE MINIMUM THAT THESE ARE THE FLEXIBILITY, THESE ARE THE THINGS THEY WILL CONFORM TO.
SO WE ARE STANDARDIZING FROM OURSELVES, FROM OUR SIDE, UH, PUTTING THOSE WHITE PAPERS.
EVERYBODY WHO'S COMING TO US SAY, HEY, YOU NEED TO CONFORM TO THOSE THINGS, AND THIS IS WHERE THE WORK IS.
AIR CODE IS SO IMPORTANT THAT WE TAKE WHAT WE ARE DOING, UH, AIR CODE AND THE UTILITIES UNDERSTAND FULLY SAY, OKAY, HEY, THERE'S NO BLIND SPOT HERE, AND THIS IS A CONTINUOUS COMMUNICATION COLLABORATION, WHICH IS NOW NEEDED FOR FOREVER.
UH, I REALLY WAS FOCUSING TODAY ON, UH, ONE OF YOUR METRICS, WHICH WAS TOKENS PER MEGAWATT.
SO YOU TALKED A LOT ABOUT THE GRID SIDE OF THAT METRIC.
COULD YOU TALK A LITTLE BIT ABOUT THE ROOT CAUSE OF THAT METRIC, WHICH IS THE POWER DISSIPATION FROM BOTH THE EFFICIENCY OF THE CODE AND THE POWER DISSIPATION OF THE SEMICONDUCTOR HARDWARE, THE PROCESSORS, WHAT, WHAT'S BEING DONE INSIDE OF NVIDIA TO DRIVE DOWN A MEGAWATTS DISSIPATED PER TOKEN? RIGHT? SO A LOT ACTUALLY, UH, THIS IS HAPPENING, RIGHT? THE FIRST THING WHICH WE DID CURRENTLY, AS YOU KNOW, A LOT OF THOSE, UH, RACKS WORK ON A 400 VOLT, YOU KNOW, AND WE SAW THERE'S A LOT OF AC DC CONVERSIONS HAPPENING DURING THE POWER LOSSES, UH, HAPPENING ACROSS THE WHOLE, UH, SUPPLY CHAIN OF POWER INSIDE THE DATA CENTER.
UH, WE HAVE GONE TO 800 WORD DC AND THIS WAS ALL DC D DC UH, REALLY MINIMIZING, UH, THE POWER LOSSES ALL THE WAY GOING INTO THE CHIP.
OKAY? SO, SO, UH, THE NEW STANDARDS WHICH WE ARE BUILDING, UH, HOW WE ARE PUTTING RECTIFIER AND CAPACITORS, UH, ACROSS, UH, UH, THE PDU AND TO THE RECS LEVEL, UH, BUILDING THESE NEW PRODUCT LINES WITH THE SUPPLY CHAINS LIKE SCHNEIDER'S OF THE WORLD, UH, SIMMONS AND GS OF THE WORLD, MAKING SURE THEY HAVE THOSE THINGS.
UH, UH, AND SO THIS IS ONE PART, RIGHT, REALLY GOING, UH,
[00:40:01]
BRINGING NEW STANDARD COMPLETELY DC DC RIGHT? UH, AND ADDITIONALLY, UH, WHAT WE ARE DOING IS, UH, MAKING SURE ALL THESE GPUS DOWN TO THE CORE, THEY ARE ALL SOFTWARE DEFINED, RIGHT? AND WHEN WE WAS TALKING ABOUT THE EMERAL OF THE WORLD DOING, WE ARE EXPOSING ALL THAT SOFTWARE DEFINED HOW WE ARE, UH, MANAGING THOSE CORES IN THOSE GPUS EXPOSING TO THE WORLD.NOW YOU CAN COMMUNICATE WITH THOSE TO MANAGE THOSE, RIGHT? SO REALLY WHEN THERE IS NO LOAD IS WORKING, SWITCH OFF THOSE CORES.
SO THIS IS THE, THAT'S THE, UH, GRANULARITY, UH, WHERE NVIDIA IS GOING, MAKING SURE THERE'S NO POWER WASTAGE WHEN WE ARE DOING THOSE COMPUTE.
SO FROM CORE TO AN, UH, SINGLE SERVER NODE, WE CALL IT TO A RACK, THEN UP ALL THE WAY TO THE DATA CENTER.
SO LET ME, AND LET ME TOUCH BASE REALLY QUICK ON THE SILICON, BECAUSE I'M GUESSING THAT MIGHT BE THE ROOT OF OF YOUR QUESTION AS WELL.
AND THE, THE SILICON CONTINUES TO GET FASTER AND FASTER.
NOW, ON A, ON A PER CHIP, OR ON A PER GPU BASIS, WE SEE THE POWER ENVELOPE CONTINUING TO GO UP FROM 700 MEGAWATTS OR 700 KILOWATTS TO 700 WATTS TO 800 WATTS AND SO ON.
UM, HOWEVER, IF YOU LOOK AT THE JUMP FROM HOPPER TO BLACKWELL AND BLACKWELL TO VERA RUBIN, WE'VE INCREASED 10 X IN THE AMOUNT OF TOKENS, UM, THAT CAN BE PRODUCED WITH EACH OF THOSE, RIGHT? SO AS YOU CONTINUE TO INCREASE, YOU MIGHT INCREASE 10% IN POWER, BUT YOU'VE INCREASED 10 X IN WATT IN TOKENS.
AND SO OUR PERFORMANCE PER WATT AT THE SILICON LEVEL CONTINUES TO GROW EXPONENTIALLY.
AND THEN FROM THERE, I WOULD SAY 30 40% OF THAT IS, IS SILICON BASED.
THE OTHER IS, UH, SOFTWARE BASED.
AND ONCE AGAIN, THE GREAT THING IS, IS WITH, WITH OUR SILICON BEING SOFTWARE DEFINED, I, AND I OFTEN COMPARE US TO TESLA OR OTHER AUTONOMOUS VEHICLES, IS WHEN YOU GET AN UPDATE IN YOUR SOFTWARE, YOU MIGHT GET 5% MORE EFFICIENCY OR YOU GET MORE SAFETY FEATURES.
WE UPLOAD NEW VERSIONS OF OUR SOFTWARE ALMOST MONTHLY, AND PEOPLE CAN TAKE ADVANTAGE OF THAT, AND THAT'S WHERE THEY CAN GET AN ADDITIONAL CAPACITY AND ADDITIONAL SAVINGS.
IN SOME APPLICATIONS, YOU MIGHT SEE THAT YOUR PERFORMANCE DOUBLES OR TRIPLES OVER THE FIRST THREE TO FOUR YEARS.
YOU HAVE THAT INFRASTRUCTURE, EVEN THOUGH IT'S THE SAME INFRASTRUCTURE, YOU DIDN'T CHANGE THE HARDWARE.
WHAT YOU'VE CHANGED IS THE SOFTWARE, UH, LIBRARIES AND DRIVERS AND FIRMWARE THAT WE PROVIDE.
AND BY DOING THAT, YOU CONTINUE TO GET MORE TOKENS PER WATT THAN WHEN YOU STARTED.
SO WE'LL ALWAYS TELL YOU THAT OUR, YOUR RELATIONSHIP CONTINUES TO GET BETTER AFTER YOU BUY THE EQUIPMENT THAN, THAN UM, THAT YOUR FIRST DAY HAVING IT.
AND THE SILICON TAKES TREMENDOUS STEPS UP EVERY GENERATION, BUT THE SOFTWARE CONTINUES TO DRIVE A LOT OF THE VALUE.
WE'LL HAVE YOU BACK IN THE EXECUTIVE SESSION.
[5.1 Projects and Technology Update]
AND WE'RE GONNA GO INTO COMMITTEE BRIEFS AND WE'RE GONNA START WITH A PROJECT AND TECHNOLOGY UPDATE WITH, UH, MANDY BALD AND TROY ANDERSON.JUST I, SO I DON'T SEE THEM ON THERE.
UM, THIS MORNING I'VE GOT MANDY BALD.
I'M THE SENIOR DIRECTOR FOR ERCOT PROJECT MANAGEMENT OFFICE.
I'M ALSO PARTNERED UP WITH TROY ANDERSON TODAY.
HE'S THE DIRECTOR OF OUR PORTFOLIO MANAGEMENT OFFICE.
WE SERVE KIND OF TWO DIFFERENT FUNCTIONS.
I REALLY KIND OF FOCUS ON THE ACTIVE PROJECTS WHERE HE'S LOOKING AT THE OVERALL PORTFOLIO PLANNING.
AS WE MOVE AHEAD, I'LL PROVIDE KIND OF THE STANDARD VIEW AS TO WHAT'S HAPPENING IN SOME ACTIVE PROJECTS.
AND TROY WILL GIVE YOU SOME INSIGHTS INTO HOW THAT RELATES TO, UM, REVISION REQUESTS AS THEY COME THROUGH THE MARKET PROCESS.
THIS CHART IS, LOOKS A LITTLE SIMILAR TO LAST MONTH'S IN TERMS OF FRAMEWORK, HOWEVER YOU'VE PROBABLY NOTICED.
IN TERMS OF ACTIVE PROJECTS, WE HAVE JUMPED UP QUITE A BIT.
SO LAST TIME I REPORTED TO YOU, WE WERE AROUND, UH, 70 ACTIVE PROJECTS.
THIS REPORT SHOWS 80 AT THE TIME I PULLED IT, WE'RE ACTUALLY AT 82 TODAY.
I MEAN, IT, IT IS A KIND OF A STEADY RISE IN PROJECTS.
UM, AS WE'VE BEEN MOVING INTO THIS YEAR, WHICH IS WHAT WE EXPECTED, WE ARE EXPECTING SOME CLOSINGS AS WE MOVE FORWARD.
[00:45:01]
WE ARE EXPECTING TO GET INTO A FAIR AMOUNT OF CLOSINGS IN THE NEXT FEW MONTHS.UM, AND AS WE'RE MOVING INTO THESE, TROY AND HIS TEAM WORK WITH THE RESOURCE MANAGERS AND MY TEAM, AND THERE'S A, A HEAVIER FOCUS ON MAKING SURE WE'VE GOT THE RIGHT RESOURCES AVAILABLE.
SO WE ARE AT, WE ARE AT SOME HIGH DEMAND LEVELS, BUT WE CONTINUE TO STAY FOCUSED ON DO WE HAVE THE RESOURCES AVAILABLE TO GET INTO THIS WORK? UM, AND LOOKING KIND OF AT THE PROFILE OF ACTIVE PROJECTS THAT ARE RUNNING TO MAKE THOSE DECISIONS.
AND, YOU KNOW, WE'LL SEE IF WE KINDA STABILIZE AT THIS LEVEL NOW THAT WE'VE GOT, UM, THAT ADDITIONAL CAPACITY THIS YEAR AVAILABLE TO US AS RTC ROLLED OFF.
AND THEN LAST TIME THERE WAS A QUESTION, UM, AND SOME INTEREST IN LOOKING AT KIND OF THE DISTRIBUTION OF OUR PROJECTS.
SO, UH, I'D LIKE TO START BY JUST KIND OF DESCRIBING HOW I ADDRESSED INTERNAL AND EXTERNAL.
SO WHEN I REFER TO INTERNAL HERE, I'M TALKING ABOUT PROJECTS THAT WE'RE RUNNING FOR THE SAKE OF KIND OF ERCOT INC.
SO THIS COULD BE OUR INFRASTRUCTURE, IT PROJECTS, UM, ERP PROJECTS, THINGS THAT WE NEED TO DO TO RUN OUR BUSINESS TO THEN SUPPORT THE EXTERNAL MARKET AND THE STAKEHOLDERS AND ALL OF THAT.
BUT KIND OF THE INNER WORKINGS.
UM, THE EXTERNAL IS THE LABEL I USED FOR ANYTHING THAT REALLY IS DIRECTLY SUPPORTING GRID MARKETS, COMMERCIAL OPERATIONS, ALL OF THE THINGS WE DO FOR OUR EXTERNAL STAKEHOLDERS.
UM, AND WHAT'S INTERESTING IS, BY AND LARGE, IT'S, IT'S, IT'S RELATIVELY BALANCED.
SO WE LEAN TOWARDS THE EXTERNALLY FOCUSED EFFORTS AND THOSE FUNCTIONS IN TERMS OF VOLUME OF PROJECTS WE'RE RUNNING AND OUR TOTAL LABOR.
AND THIS IS LOOKING AT ACTIVE PROJECTS RIGHT NOW.
AND THIS IS NOT HISTORICAL OVER TIME, THIS IS, UH, GIVEN THE MID-MAY ACTIVE PROJECTS.
UM, SO WE'RE LEANING A LITTLE TO THE EXTERNAL FOCUSED PROJECTS IN TERMS OF VOLUME AND LABOR, BUT OUR SPEND IS A BIT HIGHER ON OUR INTERNAL PROJECTS.
AND THAT MAKES SENSE WHEN WE CONSIDER THE HEAVY INFRASTRUCTURE COSTS THAT WE MIGHT SEE ON SOME WORK.
WE'VE GOT DC SIX AS A PROGRAM IN HERE.
WE'VE GOT THE IDENTITY AND ACCESS MANAGEMENT PROGRAM INCLUDED IN THIS, AND THOSE ARE MORE EXPENSIVE INITIATIVES.
UM, WHAT I DON'T HAVE SPLIT OUT HERE, BUT I DO WANT TO CALL OUT, IS EMBEDDED IN THAT ERCOT KIND OF INTERNAL NUMBER.
IT ARE SECURITY DRIVEN EFFORTS.
AND WITHIN THAT VOLUME THAT REPRESENTS EXTERNAL, THAT'S ABOUT A THIRD OF THOSE PROJECTS.
SO, UM, JUST WANTED TO CALL THAT OUT 'CAUSE THAT, THAT, UM, THAT IS PRETTY SIGNIFICANT AS WELL IN TERMS OF KEEPING THAT FOCUS ON SECURITY.
SO THAT CLOSES OUT MY PORTION, BUT I WANT TO PAUSE AND SEE IF THERE'S ANY QUESTIONS OR DISCUSSIONS BEFORE I HAND IT OFF TO TROY.
ANY QUESTIONS FOR MANDY? OKAY.
THANK YOU MANDY FOR THE INTRODUCTION.
I'M TROY ANDERSON WITH ERCOT PORTFOLIO MANAGEMENT.
I'M HERE TODAY TO REPORT ON THE STATUS OF REVISION REQUESTS THAT ARE PENDING PROJECT IMPLEMENTATION.
UH, THIS IS A TOPIC THAT HASN'T REALLY BEEN DELVED DEEPLY INTO AT THE BOARD, SO I'LL KIND OF GET, I'LL DISCUSS SOME PROCESS POINTS FIRST, THEN WE'LL TALK ABOUT THE SLIDES YOU SEE HERE.
THEN I'M GONNA ADDRESS A COUPLE QUESTIONS THAT I THINK WILL BE AT THE FRONT OF YOUR MIND AFTER WE HAVE THESE NEXT FEW MINUTES OF DISCUSSION, THEN WE CAN SEE WHAT OTHER QUESTIONS YOU HAVE AFTER THAT.
I'VE WORKED IN THE, UH, PROJECT PRIORITIZATION REALM WITH ERCOT FOR 20 PLUS YEARS.
SO I'VE, UH, BACK TO THE PRE NODAL DAYS, I'VE KIND OF SEEN THE EBBS AND FLOWS OF THIS AND HAVE A LOT OF THE HISTORY BEHIND IT.
SO I'M HAPPY TO SHARE SOME OF THAT TODAY.
BUT BEFORE WE GET INTO THE SLIDE, I WANT TO SET THE STAGE A LITTLE BIT.
SO AS MOST OF YOU KNOW, AT EVERY BOARD MEETING, YOU ARE ASKED TO CONSIDER A GROUP OF REVISION REQUESTS THAT COME FROM TAC THROUGH THE STAKEHOLDER PROCESS.
AND PR NPRS GERS, YOU'LL HAVE A BIG PIGGER TOMORROW.
UM, SOME AREN'T CONSENT, 16 ARE ON CONSENT TOMORROW, SEVEN OF THEM ARE NOT.
AND YOU'LL TALK ABOUT THOSE IN DETAIL.
EVERYONE HAS AN IMPACT ANALYSIS TIED TO IT THAT WILL TELL YOU IS THERE A SYSTEM PROJECT RELATED TO THIS OR NOT? IF THERE IS, IT'LL ESTIMATE THE COST AND DURATION.
THE ITEMS THAT DO HAVE A RELATED PROJECT ARE WHAT ARE PART OF MY TRACKING THAT I WE'LL BE TALKING ABOUT TODAY.
ABOUT A THIRD OF THE REVISION REQUESTS THAT YOU'LL SEE FOR APPROVAL HAVE A RELATED PROJECT.
THE OTHER TWO THIRDS ARE MORE LANGUAGE CHANGES THAT CAN GO IN EFFECT MUCH SOONER.
THEY DON'T REQUIRE SYSTEM CHANGES.
EVERY ITEM THAT DOES HAVE A RELATED PROJECT IS GIVEN A PRIORITY.
THAT PRIORITY NUMBER THAT YOU'LL SEE IS REALLY WHAT'S THE TARGET YEAR? WE WANT TO START THIS WORK.
THAT'S ALWAYS A TARGET, IT'S NOT A GUARANTEE.
[00:50:01]
ARRIVES, WE ALWAYS HAVE TO LOOK AT, YOU KNOW, WHAT ARE THE EXISTING PRIORITIES, WHO ARE THE RESOURCES THAT WE NEED, ET CETERA, TO DETERMINE WHEN WE CAN START THE, THE WORK.SO THAT, UH, ACTUAL START YEAR IS SUBJECT TO CHANGE.
NOW, OVER THE YEARS, UM, WE'VE ALWAYS HAD SOME REVISION REQUESTS IN OUR QUEUE WAITING TO BE DONE.
UM, POST NODAL GO LIVE, THAT WAS A PRETTY SHORT LIST, LIKE IT WAS LESS THAN 10 IN AT MANY POINTS IN TIME.
BUT BACK IN 2020, CONCERNS STARTED TO GROW THAT THE BACKLOG CONTINUED ITEMS THAT WEREN'T GETTING SUFFICIENT TRACTION.
SO WE NEEDED TO HAVE A STAKEHOLDER DISCUSSION ABOUT WHAT WERE THESE OLDER ITEMS ON THE LIST.
WE STARTED CALLING THEM THE AGING ITEMS, WHAT'S ON THE LIST, WHAT'S THE CURRENT PRIORITY BASED ON THE LATEST, UH, MARKET CONDITIONS.
AND THAT FIRST DISCUSSION WAS IN AUGUST OF 2020, WHICH IS WHY THE FIRST BAR OR FIRST COLUMN YOU SEE ON THE CHART, IT STARTS IN AUGUST, 2020.
UH, AT PRS IN THAT MONTH, WE TALKED ABOUT THE ITEMS THERE.
THAT'S THE, THE DARKER ORANGE.
A HANDFUL OF THEM WERE ON HOLD AND SOME WERE IN FLIGHT.
THAT WAS THE POPULATION AT THAT TIME.
NOW FROM LEFT, RIGHT AS YOU GO ACROSS THE CHART, THAT'S THE, THE LATEST COUNT BY THESE VARIOUS CATEGORIES.
FOR EV AT THE START OF EVERY MONTH, SINCE AUGUST OF 2020, AS THE HEIGHT OF EACH COLUMN INCREASES, THAT MEANS NEW THINGS WERE APPROVED THROUGH THE BOARD AND PUCT ADDED TO OUR QUEUE THAT WE'LL BE MANAGING GOING FORWARD.
SO THAT'S WHY THE, THE COLUMNS WILL ONLY GO UP BECAUSE THIS IS AN ADDITIVE POPULATION STARTING WITH THAT INITIAL POPULATION IN 2020.
THIS IS NOT A COUNT OF PROJECTS.
SO MANDY TALKED ABOUT HOW MANY PROJECTS ARE IN THE QUEUE.
THIS IS A COUNT OF REVISION REQUESTS THAT ARE BEING DELIVERED BY PROJECTS.
SO WHEN YOU THINK ABOUT RTC PLUS B, FOR EXAMPLE, THERE WAS A SLIDE BACK IN YOUR JUNE REPORT OF LAST YEAR THAT HAS LIKE 30 REVISION REQUESTS THAT WERE BEING DELIVERED BY THAT.
SO THAT'S GOING TO GET A, A BIG CHUNK OF THE COUNT.
THE DARK ORANGE AT THE BOTTOM ARE THE NOT STARTED ITEMS. OBVIOUSLY A FOCUS WHEN WE'RE TALKING ABOUT WHAT'S THE PRIORITY TO BE DONE NEXT.
THE LIGHTER ORANGE ARE THINGS THAT ARE ON HOLD, TYPICALLY PROJECTS THAT WE STARTED.
AND THEN WE RAN INTO ISSUES WITH OTHER HIGHER PRIORITY ITEMS THAT WERE IN THE QUEUE.
SO WE HAD TO STOP THAT WORK TEMPORARILY.
THE BLUE OR THINGS THAT ARE IN FLIGHT, THE PROJECT IS IN MOTION TO DELIVER SOME, UH, SYSTEM CHANGES.
AND OF COURSE, THE GREENER THINGS THAT WE HAVE IMPLEMENTED, WHICH FOR THE PURPOSES OF THIS CHART, THAT'S THINGS THAT HAVE GONE LIVE.
SO THAT'S WHY ON DECEMBER 1ST YOU SEE THAT THE, THE BLUE IS 40.
'CAUSE ON DECEMBER 1ST OF LAST YEAR, RTC WAS STILL HAD NOT DELIVERED YET IT DELIVERED ON THE FIFTH.
AND SO A WHOLE CHUNK OF THE BLUE MOVED TO GREEN FOR THE JAN SUBSEQUENT MONTH, JANUARY, 2026.
NOW THE SOME WAYS TO INTERPRET THIS, THE, UH, SOME OF THIS IS A FUNCTION OF TIME.
YOU KNOW, THE HEIGHT OF THE COLUMN AND THE SIZE OF THE GREEN SECTION.
IT'S GONNA CONTINUE TO GROW AS I CONTINUE TO BUILD THIS OUT INTO THE FUTURE.
BUT THE, THE ORANGE AND BLUE IS, THAT'S THE, THE INDICATION OF, YOU KNOW, WHAT, HOW MUCH IS IN OUR QUEUE THAT WE'RE TRYING TO MANAGE, THE BLUE SHOWS, HOW MUCH OF, HOW MUCH WORK IS MANDY'S GROUP ACTIVELY WORKING ON TO DELIVER? AND IT CAN, IT, IT SHOWS THAT WE HAVE BEEN ABLE TO MAINTAIN OR EVEN KIND OF DRAW DOWN ON SOME OF THAT, THAT THOSE AGING NUMBERS.
UM, FOR EXAMPLE, NOT STARTED WAS 38, BACK WHEN WE HAD THE FIRST MEETING IN 2020.
IT PEAKED AT 70 IN EARLY 2023, AND NOW WE'RE AT 48.
SO IT'S, IT'S A NUMBER THAT IS CERTAINLY THERE'S WORK TO BE DONE, BUT AT LEAST IT'S NOT GROWING UNCONTROLLED IN THE BACKGROUND.
NOW DURING THIS WHOLE PERIOD OF TIME, THERE WAS OBVIOUSLY A LOT OF THINGS GOING ON AND WE HAD RTC PLUS B THAT, THAT STARTED IN MID 23, DELIVERED LATE LAST YEAR, BACK IN 2022.
WE DELIVERED, UH, FAST FREQ FAST FREQUENCY RESPONSE.
ALSO IN 2022, FIRM FUEL SUPPLY NPRR 1120.
THEN IN 2023, ECRS, ERCOT CONTINGENCY RESERVE SERVICE WENT LIVE.
AND WE WERE DOING SECURITIZATION THROUGH MUCH OF THIS TIME AS WELL.
IT HAD RELEASES IN 22, 23 AND 2024.
ALSO IN THE BACKGROUND, MAJOR TECH HEALTH WORK, EMS, UM, UPGRADE, M-M-S-O-S, TECH REFRESH DATA CENTER FIVE, YOU KNOW, REFRESHING DATA CENTER HARDWARE.
NOW ALSO IN THE BACKGROUND, WE ARE DOING PROJECTS THAT,
[00:55:01]
THAT FIT IN WHERE RESOURCES ARE AVAILABLE.FOR EXAMPLE, CRR HAS ITS OWN QUEUE.
USUALLY A CRR CHANGE DOESN'T CONFLICT WITH MMS OR OTHER AREAS.
SO THAT QUEUE CAN BE MANAGED ON ITS OWN.
NOW, ONE QUESTION I THINK YOU MIGHT HAVE IS WHAT DO WE, WHAT HAVE WE BEEN DOING TO MANAGE THE BACKLOG? WELL, AS I SAID, STARTING IN 2020, WE WERE ACTIVELY LOOKING AT IT EVERY YEAR OR 2 CENTS.
PRS AND TAC HAVE HAD A SPECIFIC DISCUSSION ABOUT THESE WHERE WE WOULD REPRIORITIZE THINGS IN THE QUEUE, DETERMINE, YOU KNOW, WHAT NEEDS TO HIT THE TOP NEXT.
SO THE LAST SUCH MEETING WAS APRIL OF 2025.
WE KNEW THAT RTC WAS GOING LIVE.
WE WANTED TO SET THE STAGE FOR WHAT IS NEXT IN THE QUEUE.
AND OUT OF THAT MEETING, THERE WERE 10 REVISION REQUESTS IDENTIFIED AS TOP PRIORITY.
I'M GONNA JUMP INTO THE APPENDIX, SLIDE 11.
OKAY, SO IN PRS AND SUBSEQUENTLY ATTACK THESE 10 ITEMS WERE DEEMED HIGH PRIORITY.
AND SO WE WERE LOOKING TO FIND A PLACE FOR THEM IN OUR PROJECT PLAN HERE IN 2026.
OUT OF THOSE 10, TWO OF THEM ARE ALREADY COMPLETE, FIVE OF THEM ARE IN FLIGHT WITH A PROJECT RIGHT NOW.
UH, TWO OF THEM ARE PENDING A START.
ONE WILL BE STARTING NEXT MONTH.
THE OTHER ONE, WE HAVE SOME OTHER ANALYSIS TO DO FIRST.
THE NPRR 1216, WE HAVE A WORKAROUND IN PLACE THAT'S ACCEPTABLE TO ERCOT.
SO WE DON'T PLAN ON WORKING ON THAT PROJECT IN THE NEAR TERM BECAUSE OF THE WORKAROUND.
NOW, ALSO, IN ORDER TO KEEP DRIVING DOWN THAT NUMBER OF AGING ITEMS, WE HAVE TO WORK ON THINGS THAT AREN'T HIGH PRIORITY, AS LONG AS THEY DON'T CONFLICT WITH HIGH PRIORITY THINGS.
AND THAT THIS IS A SUBSET OF THAT.
I'M NOT GONNA WALK THROUGH ALL OF THIS, BUT THESE ARE OTHER REVISION REQUESTS THAT ARE INF FLIGHT IN ONE WAY OR ANOTHER, OR PART OF OUR PLANS THAT ARE GONNA BE GOING LIVE, UH, OVER SOME TIMELINE IN THE, IN THE COMING MONTHS.
ONE OF THOSE IS NPRR 1323, WHICH I SEE THE COMMISSION APPROVED LAST FRIDAY.
AND SO THAT'S GONNA PAIR UP WITH 1290.
THAT'S SOME POST RTC TWEAKS THAT WERE DESIRABLE BY THE MARKET.
SO WE'RE EXCITED TO GET THOSE DELIVERED, UH, SOMETIME BEFORE Q4, I BELIEVE IS OUR, OUR TARGET.
NOW ANOTHER, I'LL JUMP BACK NOW, ANOTHER QUESTION THAT I THOUGHT YOU WOULD HAVE IS, WELL, WHAT'S THE STATE OF THE 48 THAT ARE CURRENTLY NOT STARTED? HOW DO WE CHARACTERIZE THOSE? THERE'S A COUPLE WAYS.
UH, MY, THE FIRST LINE THERE IN THE YELLOW BOX, 23 OF THEM WERE APPROVED SINCE RTC STARTED.
AND SO THERE'S SOME LOGIC THERE THAT WE WERE TRYING TO KEEP THE, THE FLIGHT PATH FOR RTC PLUS B CLEAN TRYING TO, YOU KNOW, LIMIT SCOPE CREEP SO THAT THAT COULD DELIVER EARLY IF POSSIBLE, WHICH IT DID.
SO 23 OF THOSE ARE FAIRLY RECENT, WHICH MEANS 25 OF THEM WERE PRE MID 23 IN THE QUEUE.
ANOTHER WAY TO LOOK AT IT IS WHAT, HOW WERE THESE CHARACTERIZED IN LAST YEAR'S PRIORITIZATION DISCUSSION? SO 29 OF THE 48 WERE DEEMED MEDIUM PRIORITY.
SO THINGS LIKE REPORTING CHANGES, UH, UH, THINGS THAT, YOU KNOW, JUST WEREN'T RISING TO A LEVEL OF HIGH PRIORITY.
SO A BIG CHUNK WERE MEDIUM, 10 OF THEM WERE DEEMED NO ACTION AT THIS TIME, MEANING WE WEREN'T READY TO WORK ON IT.
THE MARKET DOESN'T THINK WE SHOULD BE READY TO WORK ON IT.
THINGS LIKE, THERE'S THREE REVISION REQUESTS RELATING TO THE SOUTHERN SPIRIT AND UNTIL THE DAY THE THE DAY ARRIVES, WE'RE READY TO DO THAT.
WE DON'T NEED TO BE STARTING UNNECESSARILY EARLY, MAYBE NEVER.
THERE WERE ALSO A FEW OTHERS THAT ARE RECENT ADDITIONS TO THE LIST.
AND THEN I MENTIONED THE TWO HIGH PRIORITY ONES THAT ARE STILL BEING WORKED INTO THE PLAN.
NOW THE NEXT PRIORITIZATION DISCUSSION WILL BE AT PRS ON JUNE 9TH OF THIS MONTH.
SO NINE DAYS FROM NOW, WE WILL BE TALKING ABOUT WHAT THE NEXT WAVE SHOULD BE.
AND OF COURSE THAT'S ALWAYS DEPENDENT ON LIMITING CONFLICTS WITH HIGH PRIORITY THINGS IN THE QUEUE, WHICH WILL BE DRRS AND FIRM FUEL OVER THE NEXT YEAR PLUS.
UH, BUT ANYTHING THAT WE CAN FIT AROUND THOSE AND GET THEM OFF THE AGING LIST, WE WILL BE SEEKING TO DO.
ONE FINAL COMMENT AND THEN I'LL ASK FOR QUESTIONS.
UM, I DO WANNA THANK MARKER PARTICIPANTS IN THE STAKEHOLDER PROCESS THAT PARTNER WITH US SO CLOSELY ON THIS PROCESS AND THE GIVE AND TAKE WITH THAT, UH, PORTFOLIO MANAGEMENT REALLY DEMANDS BECAUSE THERE, THERE'S SO MUCH GOING ON.
THERE ARE PRIORITIES WE HAVE TO ADDRESS.
THEY LET ME KNOW THEIR PRIORITIES, WHICH IS A GREAT THING, BUT THEY'RE ALSO UNDERSTANDING WHEN WE HAVE TO SHIFT THINGS AROUND BECAUSE OF, UH,
[01:00:01]
'CAUSE OF SITUATIONS WITH RESOURCES OR NEW PRIORITIES IN THE QUEUE, THINGS LIKE THAT.ARE THERE ANY THAT ARE IN THE NOT STARTED THAT WE BEGAN FIVE YEARS AGO AND WE'VE RETHOUGHT AND THINK, YOU KNOW, SAID WE DON'T REALLY WANNA DO THOSE AT ALL, OR THE SCOPE HAS CHANGED? GREAT QUESTION.
AND I THINK YOU'RE POINTING OUT THE LINE IN THE YELLOW BOX THAT I, I FAILED TO MENTION.
SOME HAVE BEEN FLAGGED AS CANDIDATES FOR REMOVAL.
THEY ARE, NO, THERE ARE THINGS ON OUR AGING LIST THAT WE DON'T THINK WE SHOULD DO.
I THINK THE MARKET TENDS TO AGREE.
AND SO SOMEONE JUST HAS TO GO TO THE EFFORT TO FILE A NEW NPRR TO STRIKE THAT LANGUAGE.
SO THAT'S ONE GROUP AND THEY'RE, IT'S, IT'S, YOU KNOW, IT'S MAYBE THREE TO FIVE AS WE LOOK AT THE GROUP TODAY.
UM, THOSE ARE CANDIDATES FOR REMOVAL THAT WE WOULDN'T DO AT ALL.
UM, OUTSIDE OF THAT, IF THINGS NEED CHANGED, IF, IF CHANGES ARE NECESSARY, A NEW NPR WOULD HAVE TO BE FILED, THAT CHANGES THE LANGUAGE, GET THAT APPROVED THROUGH THE NORMAL PROCESS, THEN WE COULD TAKE ON THE WORK.
SO IF WE IDENTIFY SOMETHING IN THE QUEUE THAT WE SHOULDN'T IMPLEMENT AS WRITTEN, WE'RE NOT GONNA START ON IT.
WE'LL LEAVE IT NO ACTION UNTIL SUBSEQUENT LANGUAGE CAN CHEW THAT UP TO THE CURRENT REALITY THAT IT NEEDS TO BE.
ANY OTHER QUESTIONS, ROY? BILL? SO A LOT OF INFORMATION HERE, YOU KNOW IT BETTER THAN WE DO.
IS THERE ANYTHING THAT, FROM YOUR PERSPECTIVE, THAT YOU'RE VERY CONCERNED ABOUT OR THAT WE SHOULD BE AWARE OF OR THAT WE'RE NOT MOVING FAST ENOUGH ON OR YOU DON'T HAVE THE RESOURCES FOR? I'M KIND OF LOOKING FOR, WHAT'S THE PUNCHLINE HERE? I MEAN, OBVIOUSLY YOU GUYS ARE DOING GREAT WORK, BUT WHAT, WHAT SHOULD THE BOARD KIND OF TAKE AWAY HERE? EVERYTHING UNDER CONTROL GOING FINE OR ARE THERE AREAS OF SPECIFIC CONCERN? THE THING I WORRY ABOUT MOST IS THE VOLUME OF PROJECTS THAT WE HAVE IN FLIGHT.
MANDY TALKS ABOUT RECORD SETTING PROJECT COUNTS, AND WHILE THE RESOURCE ANALYSIS THAT I DO SAY THAT, HEY, THIS SHOULD FIT WITHIN OUR CAPACITY, THE CONTEXT SWITCHING OF PEOPLE THAT ARE ON SIX OR EIGHT PROJECTS IS DIFFICULT TO QUANTIFY AND, YOU KNOW, IN, IN THE ANALYSIS.
SO I'M, I'M CONCERNED THERE, BUT I DO, I DO FEEL LIKE THAT OUR, OUR TWO YEAR BUDGET CYCLE IS PROVIDING MORE AND MORE RESOURCE TOWARD THE PROJECT WORK THAT WE NEED.
WE HAVE A SECOND EYE TEST ENVIRONMENT THAT'S PART OF OUR IN-FLIGHT PROJECT QUEUE THAT'S GONNA GIVE US MORE CAPACITY FOR PROJECT WORK TO SOME DEGREE.
SO A LOT OF THINGS I THINK ARE POINTING IN THE RIGHT DIRECTION TO HELP US KEEP WORKING THIS.
BUT EVERY PROJECT IS A NEW ANALYSIS ON WHETHER WE HAVE THE RESOURCES TO DO IT.
SO, BUT NOTHING STANDS OUT RIGHT NOW FROM YOUR PERSPECTIVE AS HIGH URGENCY THAT YOU NEED SUPPORT ON OR THAT TYPE OF THING? YOU FEEL LIKE IN GENERAL YOU'RE ON TRACK? I THINK.
AS LONG AS SOMETHING DOESN'T EMERGE FROM THAT 48 THAT, YOU KNOW, WE HAVE TO PUSH TO THE TOP OF THE LIST.
I UNDERSTAND THERE'S ALWAYS THAT KIND OF A THING.
[5.2.1 ERCOT Website Redesign ]
THE NEXT TOPIC IS GOING TO BE LOOKING AT OUR WEBSITE REDESIGN AND CHAD AND UM, YEP, SEAN ARE GONNA, ARE GONNA DO THAT.WE'LL TRY TO CATCH UP ON SOME TIME HERE.
UM, SO YEAH, WE WANNA TALK ABOUT THE ERCOT RE UH, WEBSITE REDESIGN.
THIS IS A 2026 OBJECTIVE AND KEY RESULT FOR ERCOT.
I AM THE EXECUTIVE SPONSOR FOR THIS BECAUSE IT FALLS UNDER COMMUNICATIONS AND ULTIMATELY AT THE END OF THE DAY, WHAT WE'RE TRYING TO DO THIS YEAR IS CREATE A COUPLE OF SANDBOX ENVIRONMENTS SO THAT WE CAN HAVE SOME USER TESTING FROM OUR STAKEHOLDERS AS WE CONTINUE TO IMPROVE THE USABILITY AND SEARCHABILITY OF THE WEBSITE.
UH, ULTIMATELY THE WEBSITE WILL, WILL HAVE A NEW DEPLOYMENT IN Q1 2027.
SO, AND I'LL SEE IF SEAN CAN DIRECT ME TO THE, TO THE NEXT SLIDE.
IT'S A LITTLE BIT OF HISTORY BEHIND, UH, THE CURRENT WEBSITE IS THAT IT'S BEEN KIND OF THIS STATUS QUO, KIND OF SAME FUNCTIONALITY CAME FRAMEWORK FOR THE LAST 16 YEARS.
UH, REALLY IN 23, 24 WE ENGAGED A, UH, VENDOR TO COME ON BOARD AND START TO CONDUCT INTERVIEWS WITH STAKEHOLDERS, WHICH INCLUDED PARTICIPANTS, UH, LEGISLATURES, THE GENERAL PUBLIC TO GET FEEDBACK ON THE USABILITY OF IT.
AND THE KEY MAIN ISSUES ARE THE ULTIMATE USABILITY AND THE SEARCHABILITY OF THE INFORMATION ON THE WEBSITE.
I THINK EVERYBODY KNOWS HOW CRITICALLY IMPORTANT THE WEBSITE IS FROM A CRITICAL COMMUNICATION PATH TO THE GENERAL PUBLIC.
WE'VE SEEN A LOT OF THAT ON OUR HOMEPAGE WITH OUR DASHBOARDS
[01:05:01]
BECOMING, UH, HIGHLY IMPACTED IN, UH, BOTH LOCAL MEDIA BUT ALSO NATIONAL MEDIA.BUT AS YOU GET FURTHER INTO THE WEBSITE, AS FAR AS THE SEARCHABILITY, IT GETS REALLY CONFUSING AND THE LAYERING OF THE PAGES AND WHERE THE CONTENT IS.
SO THAT WAS PROBABLY ONE OF THE MAIN FEEDBACK POINTS THAT WE GOT AS WE WENT THROUGH, UH, THESE SURVEYS IS THE USABILITY AND SEARCHABILITY OF THAT.
UH, WE TOOK THAT INFORMATION FROM THAT THIRD PARTY TO ROLL IT INTO OUR DESIGN ENGAGEMENT.
SO IT'S A KEY FOCUS AREA AND I KNOW SEAN WILL KIND OF WALK THROUGH SOME OF THE INITIAL TEMPLATES AND HOW THAT IMPROVES THE USABILITY AND THE SEARCHABILITY OF THAT.
AGAIN, UM, LATER THIS YEAR, WE'LL HAVE TWO SANDBOX ENVIRONMENTS IN AUGUST AND TOWARD THE LATER THE YEAR THAT ALLOW USER ACCEPTANCE OF THAT.
AND THEN WITH THE GOAL OF GOING LIVE IN, IN APRIL OF 2027.
ON THE NEXT SLIDE, YOU'LL JUST SEE REALLY SOME OF THE FEEDBACK THAT WE GOT ACROSS THOSE PEER GROUPS.
UH, BOARD MEMBERS, UH, PROVIDED SOME INPUT AS WELL AS FAR AS THE NAVIGATION AND THE COMPLEXITY OF IT.
AGAIN, THE SEARCHABILITY WAS A CRITICAL THEME THAT WE HEARD FROM MARK PARTICIPANTS AND THE GENERAL PUBLIC.
UM, THE SITE IS NOT VERY INTUITIVE WAS ANOTHER THEME.
UH, AND THEN ULTIMATELY CONSIDERING OUR BROAD ROLE IN COMMUNICATIONS, UH, WE REALLY WANNA MAKE SURE THAT THE CONTENT IS CLEAR AND ACCESSIBLE FOR THE PUBLIC.
AND SO AS WE GOT FEEDBACK FROM THE PEC UH, STAKEHOLDERS AND, AND THE GENERAL PUBLIC, THAT'S ALL BEING INCORPORATED INTO OUR ULTIMATE FRAMEWORK TO PRO REALLY MOVE FORWARD WITH FINDABILITY AND CLARITY IN IN THE WEBSITE REDESIGN.
SO WITH THAT, I'M GONNA TURN IT OVER TO SEAN, WHO'S ON THE IT SIDE, UH, HELPING TO LEAD THIS PROJECT FORWARD TO TALK ABOUT SOME OF THE MORE SPECIFIC SLIDES, INCLUDING OUR ENGAGEMENT WITH STAKEHOLDERS.
SO AS HE MENTIONED, MY NAME IS SEAN SCOTT.
I'M THE SENIOR MANAGER OVER IT DIGITAL SERVICES.
WE ARE RESPONSIBLE FOR OUR EXTERNAL PRESENCE.
UH, AS CHAD MENTIONED, WE GOT SOME OF THE FEEDBACK AND I JUST WANNA HIGHLIGHT SOME OF THE DESIGN ASPECTS FOR US.
SO WHAT WE HAVE IS OUR, THAT ONE'S NOT UP.
WE HAVE OUR NEW LANDING PAGE THAT, UH, IT'S CLEANER ON IT.
UM, THE WHOLE ASPECT OF THE FEEDBACK WE GOT WAS THERE WAS TOO MANY PLACES THAT PEOPLE HAD TO GO TO FIND SOMETHING.
SO WE NOW HAVE TWO NAVIGATIONS.
WE'VE GOT AN UPPER NAVIGATION AND A LOWER NAVIGATION.
THE UPPER NAVIGATION IS MORE FOCUSED AROUND, YOU KNOW, HOW DO YOU DEAL WITH ERCOT.
AND THEN THE SECOND ONE IS MORE FOCUSED AROUND, UM, YOU KNOW, OUR MARKET PARTICIPANTS AND GETTING THEM TO WHERE THEY NEED TO BE WITHIN THE PRIMARY NAVIGATION.
ONE OF THE FEEDBACK ASPECTS THAT WE GOT FROM OUR, OUR MARKET PARTICIPANTS IS THEY WERE HAVING TO CLICK THROUGH DIFFERENT LEVELS OF THE NAVIGATORS TO GET TO WHERE THEY WERE.
SO WHAT WE DID IS WE TOOK THEIR FEEDBACK, HOW THEY WORK, HOW THEY DO THEIR BUSINESS, AND WE PUT IT UNDER A MARKET PARTICIPANT TOOL.
SO ALL OF THEIR RESOURCES ARE NOW CONSOLIDATED UNDER ONE PIECE.
TWO OF THE MORE SIGNIFICANT CHANGES THAT YOU'LL SEE IS WITHIN OUR ERCOT NODAL PROTOCOLS AND OUR MARKET RULES.
UM, ALSO MARKET RULES IN A SECOND, BUT FOR THE NODAL PROTOCOL, WHAT YOU'LL SEE IS A COMPLETELY DIGITIZED PAGE GONE IS THE ASPECT OF JUST A LIST OF PROTOCOLS.
AND YOU CLICK IN THERE AND YOU THEN HAVE TO GO THROUGH AND SEE WHAT YOU HAVE.
NOW, WHEN YOU CLICK ON A PROTOCOL OVER TO THE LEFT HAND SIDE, ACTUALLY DISPLAY WITHIN THE CENTER PAGES, YOU CAN SEE THE SECTIONS AND THE DOWNLOAD LATEST VERSIONS OF THE ERCOT NODAL PROTOCOLS.
YOU CAN ACTUALLY DOWNLOAD THEM.
AND PROBABLY THE BIGGEST ASPECT OF THIS IS NOW A GLOSSARY THAT IS SITTING RIGHT THERE ON THE FRONT PAGE OF IT.
I HAVE TO SAY, THIS IS PROBABLY ONE OF MY FAVORITE ASPECTS OF IT BECAUSE IT CHANGES THE ASPECT OF MARKET RULES SO THAT YOU NOW HAVE WHAT ARE CALLED CHIP FILTERS.
SO SITTING IN THE REQUEST AND ISSUES IN THE CENTER OF THE PAGE, YOU CAN ACTUALLY FILTER NOW DOWN AT THAT LEVEL, YOU CAN SEARCH AND SEE WHAT IS INSIDE EACH MARKET RULE, NOT ONLY AT A SUMMARY LEVEL.
SO YOU CAN SEE OVER TO THE RIGHT HAND SIDE, UM, YOU KNOW, WE'VE GOT THE, THE NPRR 9 5 7 AS AN EXAMPLE.
AND YOU CAN SEE A SUMMARY, YOU CAN SEE THE STATUS, YOU CAN SEE WHEN IT WAS APPROVED.
AND SO YOU GET A, A VERY, UH, QUICK VIEW AT A GLANCE OF WHERE WE ARE SEARCH, UH, CHAD, UH, THIS IS CHAD'S BIGGEST TOPIC.
UM, SO WE HAVE TAKEN GREAT CARE IN IMPROVING OUR SEARCH, NOT ONLY THROUGH A, AN IMPROVED PLATFORM, BUT WE WILL EMBED AI CAPABILITIES INTO OUR SEARCH.
SO YOU'LL BE ABLE TO SEARCH FOR, SHOW ME ALL OF THE PROTOCOLS RELATED TO CLIENT CERTIFICATES, PLAIN LANGUAGE SEARCHABILITY.
[01:10:01]
FEWER ZERO RESULTS.SO WHEN YOU'RE LOOKING FOR SOMETHING, YOU'LL ACTUALLY HAVE A, UH, SEMANTIC RANK THAT UNDERSTANDS THE SYNONYMS. IT'LL UNDERSTAND THE INTENT, GIVING YOU A BETTER, UH, RANKING WITHIN YOUR SEARCH CAPABILITIES.
DIFFERENTIALS WILL UNDERSTAND THE DOCUMENTS VERSUS CONTENT WITHIN A PAGE.
AND THEN OVERALL YOU GET A BETTER USER EXPERIENCE.
AND SO WITH ALL OF THIS, THERE'S BEEN A LOT OF CARE TIME EFFORT TAKEN INTO THIS.
SO WHAT DO WE DO WHEN WE'RE DONE? SO FIRST THING IS WE KNOW WE LISTENED.
AND SO WE'VE TAKEN THAT, WE'VE GONE THROUGH WITH A DESIGN PARTNER, WE'VE CREATED NEW DESIGNS.
I'VE SHOWN YOU JUST A HANDFUL OF THEM.
UM, BUT WHEN WE'RE DONE, HOW DO WE KNOW THAT WE MET THE MARK? THIS IS WHERE CONTINUED FEEDBACK COMES IN.
ONE OF THE KEY PIECES THAT WE NEED TO CONTINUE TO DO IS NOT WAIT 16 PLUS YEARS TO GET FEEDBACK FROM OUR MARKET PARTICIPANTS, OUR STAKEHOLDERS, OUR GENERAL POPULATIONS.
SO WE WILL CONTINUE TO SEEK FEEDBACK.
ONE OF THE THINGS THAT WE ARE GOING TO DO IS IMPLEMENT A, NOT A CONTINUAL SURVEY, BUT WE'RE GONNA DO SURVEYS THROUGHOUT THE, UH, WEBSITE.
SO, YOU KNOW, RANDOM PEOPLE WILL SAY, HEY, DO YOU WANNA PROVIDE SOME FEEDBACK? UH, ONE OF THE OTHER PIECES IS BEHAVIOR MONITORING, NOT, UH, NOT INTRUSIVE BEHAVIOR MONITORING.
IT'S MORE ABOUT WHERE IS YOUR, WHERE ARE YOU SPENDING YOUR TIME ON THE PAGE? WHERE IS YOUR MOUSE GOING? WE WANNA KNOW WHAT PAGES ARE THE MOST FREQUENT, WHICH PAGES DON'T GET VISITED AT ALL.
UM, GIVES US AN OPPORTUNITY TO STREAMLINE WITH 20,000 VISITORS TO ERCOT.COM EVERY DAY WITH SPIKES IN THE 160,000.
WHEN WE HAVE WEATHER EVENTS, WE NEED TO MAKE SURE ERCOT.COM STAYS IRRELEVANT.
WE WANT TO BE THE FOREFRONT OF WHAT PEOPLE SEE WHEN THEY COME TO OUR SITE AND THEY'RE IMPRESSED AND THEY CAN FIND THE INFORMATION THAT THEY WANT.
NOW I JUST WANT TO COMPLIMENT YOU BECAUSE WE ARE A SERVICE ORGANIZATION.
AND SO IT'S GOOD TO MEASURE OUR SERVICE LEVEL.
SO WITH THE SANDBOXES, WHICH IS A GREAT IDEA, WHAT'S THE FORUM? HOW DO YOU GET THE FEEDBACK? HOW DO YOU VET THAT? I MEAN, WHAT, JUST TALK A LITTLE BIT MORE ABOUT THAT, WOULD YOU PLEASE? SO WE'RE, WE HAVE TWO SANDBOX.
UM, THEY'RE GEARED TOWARDS, UM, INTERNAL AS WELL AS EXTERNAL, UM, PARTICIPANTS.
WE ARE GOING TO INVITE PEOPLE INTO THOSE SANDBOXES AND, UH, YOU KNOW, ASK FOR THEIR FEEDBACK.
PROBABLY THE SAME WAY THAT WE DID THE ORIGINAL SURVEYS ON IT.
WE WILL PUT A SURVEY ON THERE, WHAT WORKS, WHAT DOESN'T, UM, AND JUST KIND OF TAKE THE, TAKE THAT FEEDBACK IF YOU WILL.
SO IS THAT LIKE A SMALL USER GROUP? IT IS THEN THAT YOU KIND OF GET BACK YES.
AND YOU TAKE IT AND THEN YOU COME BACK WITH WHAT YOU'RE PROPOSING TO CHANGE OR IMPROVE? YEAH.
IF IT MAY, IF AND WHERE IT MAKES SENSE SURE.
BUT TO MAKE TO LEVEL SET THE FOLKS SO THEY KNOW WHAT IS CHANGING AND WHAT'S NOT.
YEAH, BILL, YOU CAN SEE IN THAT MIDDLE PART OF THIS SLIDE, IT'S A LITTLE BIT MORE SPECIFIC WHEN YOU GET DOWN INTO THE GENERAL PUBLIC ABOUT HOW WE'RE GONNA DO A SURVEY, THE MARK PARTICIPANTS, WHICH GROUPS WE'RE GONNA BE REACHING OUT TO SO WE CAN GET A WIDE SAMPLE OF DIVERSITY OF MARKET PARTICIPANTS.
AND THEN, UH, AS SEAN INDICATED, WE'LL REACH OUT TO SOME OF THE LEGISLATURE MEDIA CONTACTS THAT WE HAVE SO THAT WE CAN ALLOW THEM TO OPT IN AND TEST THAT FUNCTIONALITY THROUGH THE SANDBOX AND CONTINUE TO GET FEEDBACK.
HISTOR, I'VE RUN INTO CHALLENGES BEFORE.
DO YOU DO IT? BUT THEN THEY, THEY, THEY DON'T KNOW KIND OF WHAT THE END GAME IS, WHAT WAS INCLUDED, WHAT'S NOT INCLUDED.
SO CLOSING THAT LOOP I THINK IS IMPORTANT AS YOU WORK THROUGH IT.
YEAH, WE'RE GONNA BE ASKING THEM TO PERFORM SOME ASPECTS THROUGH THE PAGES AS WE GO THROUGH AS WELL.
SO WE'LL PROVIDE A LITTLE BIT OF A SCRIPT SO IT'S NOT GONNA JUST BE A FREEFORM.
UH, WE DO WANT TO HAVE SOME STRUCTURE AROUND WHAT WE'RE GETTING BACK.
ANY OTHER QUESTIONS? HEARING NONE.
UM, LET'S MOVE, LET'S MOVE ON TO OUR NEXT ITEM, UM, WHICH IS ITEM 5.5, 5.2
[5.2.2 New Market Facing Functionalities – Large Load Curtailment Manager and Real Time Operation Communications Tool ]
0.2.UM, NEW MARKET FACING FUNCTIONALITY, LARGE LOAD CURTAILMENT MANAGER AND REAL-TIME OPERATION.
AND JIMMY HARTMAN WAS SUPPOSED TO COME IN FOR THIS, BUT HE'S HAD A FAMILY EMERGENCY.
SO DAN, I THINK YOU ARE GOING TO SORT OF ACT IN HIS PLACE.
UM, AND, AND, UH, SERENA VA BADRI IS UP AT THE PODIUM AND HE'S GONNA LEAD US THROUGH THE CHARTS.
I AM, UH, SERENA VA BADRI, SENIOR DIRECTOR OF GRAND MARKET SOLUTIONS, UM, TEAM.
UM, WE ARE PART OF
UM, I'M HERE TODAY TO PRESENT, UM, AND GIVE A OVERVIEW OF, UH, TWO NEW, UH, MARKET FACING FUNCTIONALITIES, UH, THAT WE ARE WORKING,
[01:15:01]
UM, THIS YEAR.UH, ONE IS, UM, RELATED TO LARGE ROAD CUTTING MANAGER WE CALL, UM, LLCM.
OTHER ONE IS, UM, REAL TIME OPERATION, UH, COMMUNICATION TOOL WE CALL RTOC.
UM, SO DAN WILL BE PROVIDING, UH, THE UH, UH, BUSINESS USE CASE AND THE BACKGROUND THAT DRIVING THESE TWO EFFORTS.
I'LL BE FOCUSING MORE ON IMPLEMENTATION DETAILS AND TIMELINES.
'CAUSE WE CAN'T JUST TELL 'EM ALL AT THE TURN OFF AT THE SAME TIME.
IT'S GOTTA BE DONE IN A MORE ORGANIZED MANNER REGARDLESS OF WHO THOSE INSTRUCTIONS ARE GOING OUT TO.
SO THIS IS REALLY THAT TOOL THAT WILL ALLOW US TO, TO DO BOTH THIS LOOK AHEAD AND, UH, UM, UH, ACTUALLY FIGURE OUT WHICH, WHICH LOADS ARE GONNA BE INTERRUPTED AND WHAT A ORDER AND SO FORTH, LIKE WE TALKED ABOUT AT PREVIOUS MEETINGS.
AND SO THE, UM, THE, UH, FIRST PHASE OF THIS, WE'RE DOING THIS IN PHASES BECAUSE, UM, SOME OF THE RULES HAVEN'T BEEN WRITTEN YET ON SOME OF THE SUB SUBSEQUENT PHASES.
BUT FOR THE, WE, WE'VE GOTTA BE READY BY THIS SUMMER FOR THE, UH, NET METERED FACILITIES.
SO THE ONES THAT ARE CO-LOCATING WITH EXISTING GENERATORS, AND SOME OF THOSE HAVE STARTED TO BE APPROVED BY THE COMMISSION.
AND SO WE'VE GOTTA BE READY TO ISSUE THOSE INSTRUCTIONS.
AND THAT'S KIND OF WHAT THE, WHAT'S IN THIS FIRST PHASE.
SO I'LL GO TO, UH, REAL TIME OFFER COMMUNICATION TOOL AS WELL.
SO DAN WOULD BE PROVIDING THE BACKGROUND AND KEY BENEFITS.
SO THIS, THE SECOND TOOL IS ONE THAT WE'VE BEEN WORKING ON REALLY SINCE, UH, URI.
IF WE, OUR, UH, TRADITIONAL WAY OF ISSUING OPERATING INSTRUCTIONS OUT TO EITHER THE TRANSMISSION OWNERS OR THE TRANSMISSION OPERATORS OR THE, UH, QSE IS TO DO A HOTLINE CALL, WHICH IS A, WE WE CALL EVERYBODY HAS TO PICK UP THE PHONE THAT'S IN THAT CATEGORY.
UH, AND THEN WE ISSUE AN OPERATING INSTRUCTION AND THERE'S REPEAT BACKS THAT HAVE TO HAPPEN AND, AND SO FORTH TO MEET THE NERC REQUIREMENTS.
AND SO, UH, WHAT THAT WORKS WELL WHEN WE'VE GOT A LIMITED AMOUNT OF OPERATING INSTRUCTIONS GOING ON, BUT WHEN WE HAVE A LARGE VOLUME AT ONE TIME, WHICH TENDS TO BE THE CASE IN ISSUE, UH, IN AN EMERGENCY TYPE SITUATION, UH, THE TECHNOLOGY HAS CHANGED SUCH THAT MAYBE THE PHONE CALL ISN'T THE MOST EFFICIENT WAY TO DO THAT.
AND SO, BUT YET IT NEEDS, BECAUSE IT HAS TO MEET ALL THE THREE PART COMMUNICATION REQUIREMENTS, THERE NEEDS TO BE, UM, A, UM, UM, GOOD MECHANISM FOR MAKING SURE THAT IT DOES MEET THOSE NERC OBLIGATIONS.
AND SO WE'VE, UM, THE, THIS WAS ACTUALLY, UH, AS A RESULT, THIS, THIS TOOL WAS AS THE RESULT OF A, SOME OF THE MARKET PARTICIPANTS ACTUALLY SUBMITTING A, A REQUEST FOR US TO DEVELOP THE TOOL.
AND WE WORKED WITH THEM ON TRYING TO FIGURE OUT HOW TO DO THAT.
AND, UM, WE'RE NEARING COMPLETION ON THAT NOW.
SO DOES IT, IT AUTOMATES THE REQUEST, THEY HAVE TO CONFIRM IT, COME BACK, AND THEN YOU HAVE TO COME, AND THEN PHONE IS THE LAST RESORT.
IF YOU DON'T GET A RESPONSE, THAT'S ALWAYS THE BACKUP.
AND, AND THEY, UM, WILL ISSUE IT OUT TO CLASSES AT A TIME.
THERE'LL BE, UH, PAGES WHERE, WHERE THERE'S, UH, KIND OF EXISTING SCRIPTS SO THEY KIND OF KNOW WHAT THE SCRIPT IS THAT WE'RE TALKING TO, BUT THERE'S ALSO THE ABILITY TO, IN UNIQUE CIRCUMSTANCES, IMPLEMENT DIFFERENT SCRIPTS SO THEY KNOW EXACTLY WHAT THE INSTRUCTION'S GONNA BE.
THEY KNOW WHICH ONE IT IS AND THOSE KIND OF THINGS.
UM, I WILL, UH, GO BACK TO, UM, LARGER
I WILL, UH, START, UM, UPDATES ON IMPLEMENTATION DETAILS AND TIMELINES LIKE DAN MENTIONED.
SO THE FIRST PHASE, UM, OF THIS EFFORT, WE'LL BE FOCUSING ON DELIVERING, UM, SB SIX, UH, NET METERING USE CASE, UH, WHICH IS, UH, UM, LARGE DOORS CO-LOCATED WITH THE EXISTING GENERATING UNITS.
WE PLAN TO DELIVER THIS CHANGE BY END OF, UM, THIS MONTH.
AND THEN FUTURE PHASE INCLUDES THE REST OF THE SCOPE RELATED TO, UH, SB SIX, BACKUP GENERATION AND 2 0 2 C BACKUP GENERATION.
AND, UH, IN
SO, UH, THIS IS THE LLCM TOOL, UH, THAT WE ARE BUILDING TO ISSUE THE, UH, CURTAILMENT INSTRUCTIONS TO
[01:20:01]
LARGE LOADS.UM, UH, WE ARE BUILDING THIS TOOL IN, UH, MARKET MANAGEMENT SYSTEM, MMS, UH, THIS, THE LOOK AND FEEL OF THIS DISPLAY SIMILAR TO, UM, THE OTHER DEPLOYMENT TOOLS THAT CONTROL ROOM HAS CURRENTLY, UM, UH, NON SPIN, UM, E-R-S-R-R-S AND, AND ECRS.
SO, UH, IT'S NOT REALLY ENTIRELY NEW DISPLAY, UH, TOOL FOR THEM.
THEY'RE ALREADY AWARE OF THE ACTION THAT NORMALLY THEY HAVE TO TAKE, UH, WHEN THEY'RE USING THOSE TOOLS.
SO IF YOU, UH, LOOK AT, THERE ARE FEW, UM, OPERATOR ACTIONS, UM, UH, THEY HAVE TO TAKE, UH, FIRST MANUAL FETCH.
UH, SO THAT ACTUALLY, UH, BRINGS THE, UH, LARGE LOAD DETAILS INTO THIS TOOL.
AND, UM, AND THEY HAVE TO SELECT THE LARGE LOAD GROUPS NEXT.
SO THEY WANT TO CURTAIL, SAY, UH, 500 MEGAWATTS OF, UH, LARGE LOADS.
THEY ENTER THE TARGET MEGAWATT HERE AS 500 MEGAWATTS.
AND, UM, WITH THE JUNE 1ST DELIVERY, UH, THEY WILL BE ABLE TO SELECT SB SIX, UH, CO-LOCATED, UH, LARGER CATEGORY, AND THEY HAVE TO ENTER THE DEPLOYMENT END TIME.
AND, UM, SORRY, I THINK, OH, I MESSED UP.
SO WITHIN A CLASS OF, OF THE LARGE LOADS, WHEN YOU DECIDE YOU NEED TO CURTAIL, HOW, HOW DOES THAT, YOU KNOW, AFTER YOU GO THROUGH IT, HOW DOES THAT GET, DOES IT GET REPRIORITIZED SO YOU'RE NOT CURTAILING THE TOP SAME PEOPLE AT THE TOP OF THE STACK EVERY TIME AS YOU GO THROUGH? OR HOW DOES THAT WORK? YEAH, I THINK THAT'S THE GENERAL IDEAS THAT WE'LL, AND, AND WE MAY NOT, UH, CURTAIL ALL OF A PARTICULAR SITE CORRECT.
AND SO WE'RE TRYING TO DO IT IN A FAIR WAY ACROSS, YOU KNOW, WHATEVER, ACROSS MULTIPLE EVENTS.
RIGHT? SO YOU KIND OF BRING SOMEBODY ELSE TO THE TOP AFTER SOMEBODY'S BEEN THERE, SO WE'RE NOT KIND OF HITTING THE SAME PEOPLE.
BUT IT ALSO DEPENDS WHAT NODE YOU'RE ON ISN'T MEANS LOCATION.
SO FOR SOME THINGS THERE MAY BE, UH, IF IT'S, IF REASON THERE'S SOME OF THE OTHER THAT WON'T BE IMPLEMENTED LEVEL OF PHASE ONE, BUT IT DEPENDS ON THE, UM, GEO GEOGRAPHY AND WHETHER IT'S A CONGESTION PROBLEM OR NOT, THEN WE WOULD HAVE TO DO THAT, YOU KNOW, OUT OF THIS ORDER.
BUT FROM, IF IT'S A SYSTEM WIDE THING, IT WOULD BE UNDERSTOOD.
AND I, SINCE WE'RE ON THIS TOPIC, ROUGHLY HOW MANY TIMES A YEAR DO YOU EXPECT TO USE THIS? THAT'S A, UH, UH, CHRISTIE QUESTION.
I MEAN, I THINK THAT'S SOMEWHAT INDICATIVE OF WHEN WE GET THE, UH, UH, THE RELIABILITY STANDARD STUDIES DONE.
YOU KNOW, AT THAT POINT WE'LL HAVE BETTER, UH, IMPLEMENT, YOU KNOW, KNOWLEDGE.
WELL, WE'RE GONNA HAVE TO USE IT BEFORE THE RELIABILITY STANDARDS THAT HE GETS IN.
'CAUSE THAT KEEPS MOVING OUT AND OUT AND OUT, RIGHT.
SO I THINK TO, TO DAN'S POINT, WHEN WE DO THE RELIABILITY ASSESSMENT LATER THIS YEAR, WHAT WE'LL BE LOOKING AT IS HOW MANY TIMES ARE WE HAVING TO LEAN ON THOSE LARGER LOADS TO CURTAIL BECAUSE OF, UH, GRID CONDITIONS? AND WE'LL MAKE SURE THAT WE'RE REPORTING THOSE PERCENTAGES AND NUMBERS AS WELL WHEN WE, UH, PRODUCE THE RELIABILITY STANDARD.
WELL, YOU KNOW, I LIKE NUMBERS.
SO CAN I USE THE MORE REPORT AND JUST, AND ASSUME WHEN THE, THE MOST RISKY HOUR ON THE HOTTEST DAY IN THE SUMMER AND THE COLDEST NIGHT IN THE WINTER ARE, IS WHEN WE WOULD USE THIS TOOL? YES.
AND WE, AND WE MAY BE DOING THE, ISSUING THESE INSTRUCTIONS IN ORDER TO AVOID GOING INTO EEA.
SO IT WOULD BE A FEW MORE TIMES THAN WHAT'S INDICATED BY THE, THAT, THAT COLUMN AND THE MOURA WITH THE, THE EEA.
SO, SO THEY, THEY ENTER THE T SELECT THE LARGE LOADS UNDER THAT SB SIX, UM, CO-LOCATED, UM, CATEGORY, AND THEY ENTER THE, UH, LEAD TIME AND THEN, UH, THEY WILL SELECT, UH, THE, UH, THE LARGE LOADS AND THE TOOL AUTOMATICALLY, UH, SELLS THE LARGE LOADS BASED ON TARGET MEGAWATT THEY ENTER, UH, AND, UM, WHEN THEY SELECT, UH, SO THEY CAN SEE ALL THE, UH, INFORMATION IN THE BOTTOM DISPLAY HERE.
UM, SO THEY CAN VERIFY THE NEW DEPLOYMENT THAT THEY'RE SENDING OUT, UH, BASED ON THE TARGET THEY ENTERED, AND THEY'LL CLICK THAT DEPLOY BUTTON.
THE INSTRUCTIONS WILL BE SENT OUT TO LARGE DOORS, UM, TO NEXT SLIDE.
[01:25:02]
SO QUICK QUESTION.WE'VE HEARD FROM, UM, SOME OF THE FOLKS WITH CRITICAL INFRASTRUCTURE, PARTICULARLY LIKE IN FAR WEST TEXAS, THAT WANT TO, UM, YOU KNOW, KIND OF WORK WITH OUR CO IN TERMS OF PRIORITIZING WHAT COMES OFFLINE.
HOW WILL THAT WORK IN CONJUNCTION WITH THIS? SO, SO AT, AT THIS POINT, THIS IS INTENDED TO DO, UH, THE VOLUNTARY CURTAILMENT PROGRAM THAT WAS PASSED IN.
I'LL GET THE NPR NUMBER WRONG, BUT, UH, THE, THE VOLUNTARY
AND THEN THE SB SIX STUFF, THIS VOLUNTARY, UH, KIND OF OUTSIDE OF THE, UH, THAT THE, SOME OF THE FOLKS IN THE PERMIAN HAVE BEEN TALKING ABOUT, I THINK IT WOULD BE A DIFFERENT THING AND PROBABLY BE ISSUED KIND OF MORE MANUALLY IN THAT WE WOULD GIVE THEM SOME KIND OF INSTRUCTION IN ADVANCE, UH, OR NOT INSTRUCTION, SOME KIND OF INFORMATION IN ADVANCE BECAUSE IT'S REALLY, TRULY IS JUST A VOLUNTARY LOAD REDUCTION IS WHAT THEY'RE TALKING ABOUT.
THIS COMPREHENDS THEM BEING ABLE TO UTILIZE THAT Y IT WOULD, IT WOULD IN THEORY WOULD'VE ALREADY BE HAPPENING BEFORE WE WOULD ISSUE ANY OF THE INSTRUCTIONS UNDER THIS.
SO THIS IS THE DISPLAY THAT WE BUILT IN EMS. UM, THIS IS MAINLY TO PROVIDE THE, UM, VISIBILITY TO CONTROL ROOM OPERATORS, UM, ON LARGER PERFORMANCE MONITORING IN REAL TIME.
SO ONCE, UH, UH, CONTROL ROOM OPERATORS DEPLOYS, UM, TO ISSUE THE MENT MEGAWATTS TO LARGE LOADS, THEY LOOK AT THIS DISPLAY, UH, TO MONITOR THEIR PERFORMANCE.
IF YOU LOOK AT THE BOTTOM, UH, UH, DISPLAY THERE, SO, UH, IT SHOWS, UM, THE, UH, LLD MEGAWATT, UH, WHICH IS THE, UM, LARGE LOAD CURRENT CONSUMPTION AND SAY CURRENT CONSUMPTION IS 25 MEGAWATTS AND THE DEPLOYED MEGAWATT IS, SAY, 25 MEGAWATTS OR THIS LEAD TIME.
UH, BY END OF THIS, UH, DEPLOYMENT TARGET TIME, THE CURRENT CONSUMPTION IS A LARGE LOAD.
UM, MEGAWATT SHOULD BE, SHOULD BE ZERO.
AND THIS DISPLAY WILL BE EXTREMELY USEFUL FOR CONTROL OPERATORS TO MONITOR THE LARGE LOADS IN REAL TIME.
UM, SO, UH, GOING TO R-T-O-C-A TOOL, UM, DANIEL WILL PROVIDE THE BACKGROUND AND THEN KEEP BENEFITS.
I WILL JUMP INTO IMPLEMENTATION DETAILS.
SO THESE ARE THE TIMELINES THAT, UH, SUMMARIZES, UM, THE EVOLUTION OF THIS TOC TOOL OVER LAST TWO YEARS.
UM, AND THEN THE AC THE MILESTONE THAT WE ACCOMPLISHED AND ALSO THE MILESTONES AHEAD OF US.
UH, SO I WANT TO HIGHLIGHT TWO KEY POINTS HERE.
UM, AT THE BEGINNING OF THIS EFFORT BACK IN 2024, WE WORKED WITH, UH, MARKET PARTICIPANTS MAINLY VOS, AND GOT THE FEEDBACK FROM THEM BY BUILDING A, A PROOF OF CONCEPT TOOL.
AND WE INCORPORATED THAT FEEDBACK INTO OUR REQUIREMENTS AND DESIGN.
UH, AND EVEN DURING THE IMPLEMENTATION PHASE, WE CONSTANTLY COMMUNICATED THE IMPLEMENTATION DETAILS TO MARKET PARTICIPANTS AND GOT THE FEEDBACK AS WELL.
CURRENTLY WE ARE IN, UM, MARKET TESTING PHASE.
UM, WE ONBOARDED, UH, ALL THE VOS, THEY'RE ACCURATE TESTING AND, UH, WE ARE ONBOARDING QSCS RIGHT NOW.
IN PARALLEL TO THIS, UM, WE ARE DEVELOPING THEIR DRAFT NPRR TO UPDATE THE PROTOCOL SECTIONS, UH, THAT ALLOWS CONTROL ROOM OPERATORS TO, UH, USE THIS TOOL.
IN ADDITION TO CURRENT, UH, HOTLINE CALLS, DEPENDING ON WHEN THIS NPRR GETS APPROVED, UH, WE WILL BE STARTING THE PARALLEL OPERATIONS WITH MARKET PARTICIPANTS AND THEN WE WILL BE GOING LIVE WITH THIS TOOL.
UM, I, UH, I TOOK A SCREENSHOT OF THIS TOOL.
UM, UH, I WILL TAKE A EXAMPLE OF A A, A GRID EVENT, UM, EA THREE LOAD SHED EVENT AS AN EXAMPLE, UH, TO DEMONSTRATE THE CAPABILITY OF, UH, THIS TOOL.
SO WHEN AIR CUT ENTERS THE EEA THREE AND IT REMINDS THAT FORM LOAD SHED IS NEEDED, UM, CONTROL ROOM OPERATORS, UH, USING THIS TOOL, THEY GO TO THE COMMUNICATION INTERFACE AND BRINGS UP, UH, THIS DISPLAY HERE.
AND, UM, UH, THEY, THEY'LL PICK, PICK THE TEMPLATE EEA THREE, UH, FORM LOAD SHARED TEMPLATE, AND IT AUTOMATICALLY POPULATES ALL THE INFORMATION.
THEY WILL, UH, QUICKLY VERIFY THE INFORMATION, UH, AND IF NEEDED, THEY MAKE ADJUSTMENTS AND THEY WILL, UH, HIT THE SEND BUTTON.
UH, THIS IS THE TVOS VIEW OF THIS TOOL.
[01:30:01]
UH, THE EA THREE LOAD, SHE MESSAGE SHOWED UP IN THEIR INBOX AT THE, AT THE TOP OF THEIR INBOX IN THE RED COLOR, BECAUSE THIS IS THE EMERGENCY AND CRITICAL PRIORITY.IT SHOWS UP IN THE RED COLOR AND ALSO IT SENDS THE NOTIFICATIONS, UH, ON THEIR COMPUTERS.
UH, SO THEY GET, UH, THEY HEAR THE AUDIBLE ALARM TO GET THEIR ATTENTION, UH, TO THIS MESSAGE AND THEY'LL, THEY CANNOT ACKNOW IMMEDIATELY.
AND ONE THING I WANT TO HIGHLIGHT HERE IS, UM, THE, THE DESIGN OF THIS, UH, MESSAGING ASSEMBLED THE OUTLOOK STYLE EMAIL MESSAGING.
YOU CAN SEE THE INBOX HERE AND THEY CAN REPLY, REPLY ALL FORWARD, THE, UH, MESSAGE.
IT HAS A SORTING AND, UM, SUCH CAPABILITIES AS WELL.
THEY CAN SAVE THE MESSAGE AS WELL.
THIS IS ONE OF THE, THE COOL FEATURES OF THIS TOOL.
THERE ARE SEVERAL OTHER NEW FEATURES, BUT TODAY WE ARE FOCUSING ONLY COMMUNICATION INTERFACE ON THIS TOOL ONCE, UM, UH, TVOS ACKNOWLEDGE THIS MESSAGE.
UM, SO IT SHOWS UP, UH, IN TVOS VIEW OF THIS TOOL AS WELL AS, UH, AIR CUT, UM, OVERVIEW OF THIS TOOL.
WE'VE GOT TO KEEP MOVING HERE.
UM, THE SECOND TO LAST AND LAST ITEMS ARE BOTH WITH JP AND, UH, WELL WITH, WITH WITH JP AND THEN JUST ASKING IF THERE'S ANY OTHER BUSINESS.
[6. Future Agenda Items]
THIS ITEM IS FUTURE AGENDA ITEMS. UM, JP YEAH, I KNOW WE TRIED SOMETHING NEW THIS TIME WITH, UH, SHOWING SOME OF OUR IMPLEMENTATION DETAILS.I'M SORRY, JIM, JIMMY HARTMAN COULDN'T BE HERE, BUT HE'S MY ULTIMATE CUSTOMER.
SO IT'S SOMETHING DIFFERENT WHERE I KNOW USUALLY YOU'RE ON THE APPROVAL SIDE OF THE NRSS, BUT THIS ON THE OTHER SIDE, WHAT GETS IMPLEMENTED AND DELIVERED AND WHERE THE DOLLARS ARE SPENT.
SO HOPEFULLY YOU GOT TO SEE SOME OF THAT.
UM, IN TERMS OF, YOU KNOW, UPCOMING, UM, GUEST SPEAKERS, I THINK WE HAVE SOME MORE IDEAS AROUND DATA CENTER AND, UH, WE ARE STILL WORKING ON, UM, MAYBE OTHER IDEAS, BUT PLEASE LET ME KNOW THROUGH JOHN IF THERE'S ANYTHING SPECIFIC THAT YOU'D LIKE TO HEAR.
JUST SO THAT I, I THOUGHT THAT WAS VERY HELPFUL TO KIND OF GO INTO HOW THE SYSTEM PRACTICALLY OPERATES.
AS FAR AS FUTURE TOPICS, I THOUGHT THE NVIDIA PRESENTATION WAS EXCELLENT.
IT WOULD ALSO BE NICE TO TALK TO, UH, DATA CENTER FOLKS WHO WANT TO PUT THEIR OWN GENERATION IN BEHIND THE METER AND GET THEIR PERSPECTIVES ON THAT.
THAT'S NOT THE SAME KNOW STRATEGY THAT THEY'RE USING, BUT I THINK FOR US IT WOULD BE VERY HELPFUL TO UNDERSTAND THAT AND GET THEIR PERSPECTIVE ON IT.
I CAN PROBABLY GET THE FERMI GUYS TO COME IN HERE.
THESE ARE THE GUYS UP IN AMARILLO WHO ARE DOING THAT PROJECT WITH TEXAS TECH, WHICH YEP.
HAS BEEN SUBJECT TO A LITTLE BIT OF CONTROVERSY, BUT, UH, I HAVE SOME GOOD C GOOD CONTEXT THERE, SO THANK YOU.
SO LET ME KNOW IF THERE ARE ANY QUESTIONS.
WE'RE GONNA ADJOURN THE GENERAL SESSION AT THIS POINT
[Convene Executive Session ]
AND QUICKLY SWITCH OVER INTO EXEC SESSION SO WE CAN EXCUSE OUR GUESTS.EVERYONE TAKE A QUICK BREAK, BUT, UH, WE'RE RUNNING ABOUT 15 MINUTES LATE, SO SENSITIVE FOR THE FACT THAT I LOSE SOME OF MY PARTICIPANTS.