Robert Friedland stood in Colorado Springs on Monday and told a room of miners to throw out the machines they have used for decades. “What we want to do is take away that SAG mill and that ball mill — bye-bye,” he said. The savings, he said, would be huge, in electricity and in carbon. The technology is his. It sits in a private company, I-Pulse. Its projections, which the news report is careful to label as projections, are up to 80 percent less energy than conventional grinding and about 5 percent more metal recovered. Then he looked at the companies that are not on his shareholder list and handed them a crying towel. “You’re toast. You won’t be able to do this the old-fashioned way.”
This piece has one central idea. A speech that kills the SAG mill is not a mill that has been killed. The number an investor can use is the one Friedland did not give. He did not give the cost of an industrial unit. He did not give the tonnes per hour. He did not give the date a unit would run day and night at a mine. He did not show a result, checked by someone else, against a normal grinding circuit. Until those exist, “you’re toast” is a line on a stage. It is not a cash flow. The public stock a person can buy is Ivanhoe Mines. Ivanhoe is not I-Pulse. A side bet on a private machine is not the same object as the machine.
This is a reading of Henry Lazenby’s September 29, 2026, account of Friedland’s talk at Mining Forum Americas, with the 2023 loan conversion stated from company disclosures of that year. It is not advice. It is not a call to buy or sell Ivanhoe, I-Pulse, or any miner that still owns a mill. Projections can miss. Related-party bets can be brilliant and still be a bad entry. Footage of rock turning to powder is not a flowsheet.
What he wants to kill
A SAG mill is a semi-autogenous grinding mill. A mine blasts rock, crushes it, and then tumbles it in a huge rotating drum so the rock breaks itself, often with steel balls to help. A ball mill takes what comes out and grinds it finer. The point is to free the mineral grain from the waste so a later step can catch the metal. The method is old because it works. It is also a glutton. Steel, gravity, and brute force, applied from the outside, all day, for the life of the mine.
Grinding is one of the mine’s largest power bills. Lazenby’s report cites a U.S. Energy Department study that called grinding the largest energy-using process in U.S. mining. The same sentence does the honest work. Those figures are industry-wide. They are not a measurement of what one mine would save if it unplugged its mill tomorrow. An industry total is not a site budget. Anyone who multiplies “largest power use” by “80 percent less” and calls the product a profit has skipped the only sum that matters, which is at one plant, on one ore, with one power price.
I-ROX, a unit of I-Pulse, says it does the breaking from the inside. The system stores electricity and lets it go in pulses that last from billionths of a second to millionths of a second. Shock waves run through the ore and crack it along the boundaries between minerals, the company says. In theory that frees more of the valuable mineral and makes fewer ultra-fine particles, the slimes that are hard to recover. I-ROX says this can cut energy and lift recovery. The report adds the line that should sit next to the claim in the same size type. The company has not published commercial, mine-scale results that support the headline numbers.
Friedland is 76. Near the end of a talk that also wandered through flying saucers, artificial intelligence, China, and what he called “particularly frightening” military robotic dogs, he played footage. Copper ore fell through an I-ROX test unit and came apart under the pulses. “We’re able to take these large pieces of rock down to a very fine powder, but we do so very efficiently,” he said. The footage may be real. A test unit is a test unit. A mine moves thousands of tonnes, every hour, including the hours when the founder is not in the room. The article’s own test of the idea is that gap. Promising tests are not equipment that can do that job and still beat mills miners already trust.
The two numbers, and who owns them
Hold the 80 and the 5 up to the light. Up to 80 percent less energy. About 5 percent more metal. Both are I-Pulse’s projections for the technology. “Up to” is a ceiling, not a promise. “About 5 percent” is a recovery claim, which in copper is not small if it is real. Five extra units of recovery on a big mine is a lot of metal from rock the mine already paid to dig. That is why the claim is worth a room. It is also why it needs a witness who does not own the machine.
A recovery gain and an energy cut are different gifts, and they do not have to arrive together. A process can use less power and recover less metal. It can recover more metal and use the same power. It can look brilliant on a selected ore and ordinary on the ore that actually pays the bills. Preferential breaking along mineral boundaries is a geologist’s hope. It is not a law. Some ores break that way. Some do not. The speech did not walk through a bad ore. Speeches rarely do.
The reporter states the commercial prize plainly, and then states the hurdle in the next breath. If the gains show up at real scale, miners could cut a top operating cost and pull more metal from lower-grade rock. Higher recovery can mean more metal from the same tonnes, a longer mine life, and a marginal deposit that starts to clear its cost of capital. Lower power can mean a new mine needs a smaller power plant, which matters as pits go deeper, harder, and poorer. All of that is the prize. The hurdle is the sentence after it. The machine has to process thousands of tonnes around the clock and compete with mills the industry has used for decades. Decades are not a mood. They are spare parts, trained crews, and a banker who has seen the mill before.
What he did not say
The most useful paragraph in the news story is the one headed by absence. Friedland did not say what an industrial I-ROX unit, or a G-Pulse unit, would cost. He did not say how much ore it could treat. He did not say when one might run continuously at a mine. He did not present results, checked by an outside party, comparing I-ROX with a normal grinding circuit.
Those four silences are the investment. Cost tells you whether the crying towel is a threat or a brochure. Tonnes per hour tell you whether the powder in the video can replace a mill that already has a nameplate. A date tells you whether this is a 2028 flowsheet or a 2035 hope. An outside comparison tells you whether the 80 and the 5 survive contact with someone who is not on the cap table. A founder who has the footage and not those four facts has chosen the order. The order is the tell. Showmanship is not a sin at a forum. It is a sin only if a buyer treats the forum as a data room.
There is a smaller fact that makes the silence louder. In 2023 Ivanhoe agreed to turn a loan of about $76 million, owed to it by High Power Exploration, into shares of I-Pulse, HPX’s parent. The loan had started at $50 million in 2019 and grown with interest. Ivanhoe said then that Kamoa-Kakula, its big copper complex in the Democratic Republic of Congo, would be the first of its assets to test I-ROX. It also said the approach had been tried in labs for years and was not yet in commercial mining. Three years later, the 2026 report still says mine-scale commercial results have not been published. A test can be underway and unpublished. A test can have slipped. This piece will not invent which. It will say the clock the company itself started in 2023 has not, in public, struck a commercial result. Investors do not get to pretend the clock started on Monday.
The public stock is not the private machine
Ivanhoe Mines trades in Toronto as IVN and in the United States as IVPAF. Friedland is its founder and executive co-chairman. The article says Ivanhoe has more than his showmanship riding on the technology, and the 2023 conversion is why. The miner owns a piece of the private parent, with rights that included a way to keep its percentage if I-Pulse sold more shares, until a public listing. That is a real link. It is not a tracking stock. If I-Pulse is worth a great deal more someday, Ivanhoe’s piece may be worth more. The market will not tell you how much, because the piece is private, the percentage is easy to dilute, and the value depends on a machine that has not posted the four missing facts.
There is also a related-party stain that a fan letter will skip. The loan was to High Power Exploration, a Friedland-backed firm. It was converted into equity in a Friedland company. Ivanhoe’s public shareholders funded a loan that became a stake in the chairman’s technology house. That can be a smart use of a miner’s balance sheet. It can also be a chairman selling his public company a dream he controls. Both can be true in sequence. The protection, such as it was, is disclosure and a collaboration aimed at a real mine, not a speech. The test of that protection is Kamoa, or whichever mine actually runs the unit and publishes the recovery and the power bill against the mill it replaced. Until then, the public shareholder owns a copper business plus an IOU to a story.
Other large miners have put money in, which is not nothing and not proof. Over recent years the names attached, in company statements and in accounts of this week’s speech, have included BHP, Rio Tinto, Newmont, Teck, Anglo American, and Codelco, plus, on the I-ROX side in earlier reports, a European fund tied to Bill Gates’ Breakthrough Energy. Treat the list as reported, not as an audited cap table from this desk. A strategic check from Codelco or BHP is a better signal than a keynote. It is still a signal about interest, not about tonnes. Big miners buy options on technologies all the time. Most options expire. The crying-towel line treats the list as a wall between the saved and the doomed. A wall made of minority stakes in a private firm is a thin wall. The stakes say “we might want this.” They do not say “we have thrown out the mill.”
Toast, and a five-million-dollar hole in the ground
The geothermal aside is the scale check the copper boast does not want. Friedland said the core technology needed to crush and grind rock, and to drill, is the same technology needed to go after geothermal heat. Lazenby’s account says the plan is to validate G-Pulse at the Millungera Basin project in northwestern Queensland. Under the partnership described, I-Pulse would take operational control and earn an initial 80 percent interest by investing at least $5 million. The stake would become 65 percent when the joint venture is formed.
Read $5 million next to “you’re toast.” Five million dollars is a drill program. It is not the replacement of the world’s grinding fleet. An 80 percent earn-in that later steps down to 65 percent is a normal farm-in, not a revolution already won. Geothermal and mine grinding can share a pulse of electricity and still be different businesses, with different rocks, different regulators, and different ways to fail. The honest link is the one he stated. The same idea wants to crack rock, whether the goal is a mineral grain or a path for heat. The dishonest link would be to treat a Queensland earn-in as evidence that a Chilean SAG mill is already obsolete. It is evidence that the idea is still being pointed at the ground.
This is useful for the investor who likes the man and distrusts the adjective. Friedland’s pattern, for decades, has been to see a shortage early and to be unbearable about it in public. Sometimes the rock was there. Sometimes the timing was not the timing in the talk. The unbearable line this week is that outsiders will not be able to mine the old way. The humble facts in the same story are a test unit, unpublished mine-scale results, no cost, no throughput, no date, and a geothermal foothold priced like a single exploration season. Humble facts do not make him wrong. They make “toast” a forecast with a long fuse. Long fuses are fine if you know you own the fuse and not the blast.
Eight Escondidas does not require this machine
He did not leave the copper scare off the stage. The report says he put a fresh spin on a familiar warning. The industry faces a projected 2040 supply gap equal to almost eight Escondidas, and better processing could recover more metal from ore miners already handle. Escondida, in Chile, is the giant against which copper people measure other giants. “Almost eight” of them is a way of saying the gap is enormous. This piece will not pretend to audit that 2040 figure from a conference line. It will use the logic he attached to it, because the logic is the part an investor can test without a model of the year 2040.
The logic is that recovery is a mine you do not have to permit. If a plant gets 5 percent more of the metal out of tonnes it already mines, that metal does not need a new pit, a new road, or a new decade of hearings. On a strained copper balance, that is a serious sentence. It is also a sentence that does not belong only to I-ROX. Any improvement in recovery, from a better reagent, a better grind size, or a better circuit, is the same kind of hidden mine. His machine is one candidate. The gap he describes would still be a gap if the machine fails. The gap is not evidence that this design works. It is evidence that a design that worked would be valuable. Value if it works is the definition of a venture bet. It is not the definition of a sure replacement for the mill.
There is a second, quieter point. Lower-grade ore is why grinding’s power bill hurts more every year. You move more rock for the same metal, so the mill works harder to free a smaller prize. A technology that breaks that spiral is worth more in 2035 than it was in 2005, if it is real. “If” is still doing the work. The spiral is real whether or not his pulse is. Investors who buy the spiral are buying a problem. Investors who buy I-Pulse’s answer are buying one proposed fix. Do not pay the problem price and tell yourself you bought the fix, or the reverse. Ivanhoe, as a miner, is exposed to the problem and, through a private stake and a hoped-for test, to one fix. Those exposures are not equal. The mines are the business. The fix is the option.
What an investor can actually use
This is a filter, not a ticket.
Separate the man, the miner, and the machine. The man can be early and loud. The miner, Ivanhoe, has operating copper and other assets whose value does not require a SAG mill to die on Monday. The machine is private, unproven at the scale that pays, and controlled by the same man. If you own the miner because the copper is good, say that, and treat I-Pulse as a warrant you cannot price. If you own the miner because the mill is doomed and only his friends will thrive, you own the speech. The speech did not include a cost or a throughput. You cannot model a speech. You can only clap.
Ask for the four missing lines before you let “toast” into a model. What does a unit cost, installed, against the mill it claims to retire? How many tonnes per hour, on the ore this mine actually has, not on the ore in the video? What date does continuous duty start, and who pays for the power and the spare parts? Where is the side-by-side, signed by someone who does not work for I-Pulse, showing energy per tonne and recovery against the existing circuit? A company that answers those in a filing has left the forum. A company that answers them with another video is still at the forum.
Watch the test that was named in 2023, not the adjective that was used in 2026. Kamoa-Kakula was the first site Ivanhoe itself pointed at. A published trial there, or a published trial anywhere at mine scale, with recovery and kilowatt-hours and a period long enough to include a breakdown, would change the weight of this story. Silence does not prove failure. It does prove that a buyer who acts now is underwriting the silence. Underwriting silence is allowed. Call it that. Do not call it due diligence.
Discount the shareholder list by what a list can mean. A check from a major is a reason to keep reading. It is not a reason to assume the major has scrapped its own mills. Majors buy options so they are not the ones holding the crying towel if the option hits. They also keep the old mill running, because the old mill ships metal this quarter. If a major later signs a commercial install, with tonnes and a penalty if the recovery misses, that is a different document. That document does not exist in this week’s story.
Size the boast against the $5 million. The geothermal step is priced. The mill-killing step is not. When a founder prices one use of his technology at a few million dollars and describes the other use as the end of an era, believe the priced use more than the era. Eras get announced. Earn-ins get signed. The signed paper is the tell of how early the era still is.
Three ways this can age
These are paths, not predictions.
One. The powder scales. A mine, perhaps one he already named, runs a unit long enough to publish energy and recovery against its own mill. The 80 percent comes in lower, because “up to” almost always comes in lower, but the power cut is still large. The recovery gain is real on that ore. A major signs more than a minority check. In this path the speech was early, not empty. Ivanhoe’s private stake becomes easier to argue about, and the public mines may gain a cost edge if they get the machines first. Outsiders are not toast on a Monday in 2026. They are behind, which is a different and more ordinary word. Behind can be fixed with a purchase order. Toast cannot. The path that makes him look right still has to pass through a purchase order.
Two. The lab stays a lab. The video remains the best evidence. Mine-scale duty eats the savings in wear, power spikes, or a recovery gain that fades on the wrong ore. The $5 million geothermal work proceeds or stalls on its own merits. I-Pulse stays private and interesting. Ivanhoe remains a miner whose chairman has a technology house. In this path nothing catastrophic was said that was false on the day, because the day did not claim a commercial result. The only people hurt are the ones who paid a commercial price for a test-unit story. The SAG mill does not attend their funeral. It keeps turning.
Three. The technology works and the structure does not. The pulses do what he said, and the value sits in the private company, diluted before any public holder of Ivanhoe sees a clean share of it. Or the public miner spends years as the test site, taking the operating risk, while the private parent keeps the rent. Related-party stories often end this way even when the science is good. In this path “you’re toast” was a sales line aimed at future customers, and the public shareholder was the audience, not the landlord. Good science and a bad split of the rent is still a bad investment if you bought the wrong paper.
What this does not mean
It does not mean the idea is silly. Breaking rock along mineral boundaries with a very short pulse is a serious physical proposal. Grinding really is a huge power load. A 5 percent recovery gain, if it survived a year of plant data, would be a large amount of metal. Serious is not the same word as proven. The article you started from is careful on that line. This one will not be less careful because the footage was striking.
It does not mean every miner without a stake is doomed. He said that. The evidence for it would be a machine in service, a cost, and a refusal by the vendor to sell to outsiders. What exists is a warning from a founder to a room. Founders talk that way when they want the room to feel late. Late is a sales emotion. Sales emotions are a bad input to a discount rate.
It does not mean Ivanhoe’s mines are a call option with no other value. The company is a miner. Copper, zinc, and the platinum-group work have their own prices, their own costs, and their own politics. A bad week for a private technology does not, by itself, empty those pits. A good week for the technology does not, by itself, fill them. If you cannot describe the mines without the mill story, you are not analyzing the miner. You are renting the chairman’s microphone.
It does not mean the 2040 gap is settled because he said “eight Escondidas.” Supply gaps are forecasts. Forecasts of 2040 copper have been confident before and wrong on the date. Use the gap as a reason recovery matters. Do not use it as proof that this pulse is the way the gap gets filled. Many technologies can fit inside a large gap. A large gap is a market. It is not a purchase order with his name on it.
It does not mean you should buy the stock, avoid it, or try to guess a private valuation from a conference stage. The stage did not contain the valuation. It contained a bye-bye, two projections, a video, a towel, and a set of facts the speaker chose not to add. The adult response is to keep the projections in the column marked “his,” and to keep your money in the column marked “shown.” Shown, this week, is a test. Shown is not a retired mill.
The close
Friedland wants the SAG mill and the ball mill gone. He told Mining Forum Americas that I-Pulse can do the job with up to 80 percent less energy and about 5 percent more recovery, and that both figures are the company’s projections. I-ROX would crack the ore from the inside with pulses measured in millionths of a second and less. The footage shows large rock becoming fine powder. The U.S. Energy Department work he is leaning on says grinding is mining’s biggest energy use in that country, and those figures are industry-wide, not a savings ticket for one pit. He did not give a cost, a throughput, a start date, or an independent comparison. Commercial mine-scale results are still unpublished, three years after Ivanhoe turned a $76 million related-party loan into I-Pulse shares and pointed at Kamoa-Kakula as the first test. The geothermal cousin of the idea is attached to a Queensland earn-in of at least $5 million. The copper cousin is attached to a line about eight Escondidas and a crying towel.
The central idea does not need a winner. A speech that kills the SAG mill is not a mill that has been killed. The number that matters is the one he left out. Cost, tonnes, time, and a check by someone who does not own the pulse. Ivanhoe’s public shares are a miner with a private side bet, not a claim on every mill in the world. Toast is what you are if you confuse the two, and then discover that the old mill, which he said goodbye to, is still the machine shipping the metal.
Important information
This article is for information and education only. It is not investment advice and not a recommendation to buy or sell Ivanhoe Mines, any other miner, or any interest in I-Pulse, I-ROX, or G-Pulse. Mining stocks can fall if metal prices fall, if a project slips, or if a related-party technology stake is worth less than the story. Private companies can dilute public partners. Projections of energy savings and recovery can miss by a wide mark. Past success finding mines does not mean a new machine will work at plant scale.
The account of the September 28, 2026, remarks in Colorado Springs, the 80 percent and 5 percent projections, the quotes, the absence of cost, throughput, timing, and independent comparisons, the Energy Department caveat, the unpublished mine-scale results, the Queensland G-Pulse terms, and the eight-Escondidas line follows Henry Lazenby’s September 29, 2026, report “Mining Forum: Friedland wants to kill the SAG mill,” as carried by Mining.com and The Northern Miner. The 2023 conversion of a loan of about $76 million, originally $50 million in 2019, from High Power Exploration into I-Pulse shares, the investor rights, and the plan to test at Kamoa-Kakula are from Ivanhoe’s disclosures as reported in June 2023. Names of other strategic investors are from company statements and contemporaneous accounts over 2023–2026, not from a cap table audited here. Percentages, dates, and test status can change. This article does not consider any person’s goals or finances.