/
/ 

What the ASUS ROG Equalizer actually does, and where the 87.5% figure comes from

Three laboratories have now measured ASUS’s $50 ROG Equalizer 12V-2×6 cable. The 87.5% voltage-drop figure traces to one Reddit user, and the balancing hardware measurably worsens per-pin current spread.

01 / 01 Sections Sections

2026-08-04 · Hardware

The ASUS ROG Equalizer is a $49.99 12V-2×6 cable sold as a fix for the melting GPU power connector, and three separate laboratories have now measured it. Their finding is consistent and it is not the one in the marketing: the cable carries a fault better than any standard cable, and in normal operation its own balancing hardware makes per-pin current distribution measurably worse.

Key findings

  • The widely circulated claim that the ROG Equalizer cuts total voltage drop by 87.5% traces to a single Reddit user’s software sensor readings, comparing a new cable against his own visibly discolored one, in a Wccftech article published 4 August 2026 that itself states the old cable was worn out.
  • igor’sLAB measured the same ROG Equalizer cable in both orientations on an MSI GeForce RTX 5090 Suprim at 619 W and found the spread between the strongest and weakest of the six +12V pins fell from 1.98 A to 0.72 A when the cable was reversed so its metal block no longer sat behind the GPU connector.
  • Aris Mpitziopoulos of Hardware Busters loaded three adjacent conductors at 17 A each for ten minutes and recorded a maximum of 93.6 °C with no melting, confirming ASUS’s load-capacity claim under a fault condition that normal operation does not produce.
  • ASUS’s own headline thermal comparison of ~87.8 °C against ~146 °C was produced with the middle four +12V wires deliberately disconnected, forcing 25 A down each of two remaining pins.
  • ASUS’s least-publicized and strongest claim is a ten-day soak, stated in its 28 May 2026 press release: 600 W continuous at 55 °C ambient, with cable temperature settling near 100 °C after the first hour against a 105 °C material limit.
  • The ROG Equalizer works with any power supply carrying a native ATX 3.1 12V-2×6 header from any manufacturer, per ASUS’s own FAQ, with one ASUS-only exception: two thin voltage-sense return lines that function only on a compatible ROG Thor supply.
  • igor’sLAB measured the conductor at approximately 105 strands of 0.13 mm, matching a standard AWG 16 105/36 construction, which is the same gauge the Intel and Molex specifications already call for.

Spec file

Field Value
Product ASUS ROG Equalizer 12V-2×6 PCIe Cable
Model numbers 90YE00BN-B0QA00 (black), 90YE00BP-B0QA00 (white)
Amazon ASIN B0H42Q6936, listed at $49.99, 4.2 stars from 11 ratings, checked 4 August 2026
Announced 9 April 2026
Length, weight 750 mm, 0.2 kg
Connectors PCI-E 16-pin (12V-2×6), male to male
In the box Cable, 3 cable combs, quick start guide
ASUS current claim 17 A per wire short-term peak, against 9.2 A for a standard 12V-2×6 wire
ASUS power recommendation 600 W total
Mechanism Passive busbar. All six +12V conductors join a common aluminum block behind the GPU plug, with a second isolated block for ground
Conductor gauge Not published by ASUS. Measured by igor’sLAB as AWG 16, 105/36 construction, 1.33 mm²
PSU compatibility Any ATX 3.1 supply with a native 12V-2×6 header, any brand
ASUS-only feature Two voltage-sense return lines, functional on compatible ROG Thor supplies
Standards, certification ATX 3.1, PCIe 5.1, UL1581 flame test, UL758
Warranty 3 years
Independent tests to date Hardware Busters (30 April 2026), der8auer (May 2026), igor’sLAB (23 June 2026)

Affiliate disclosure. The Amazon link on this page is an affiliate link. If you buy through it, I earn a commission at no extra cost to you. ASUS did not supply a sample, did not review this page, and has no relationship with this site. I have not bench-tested this cable myself. Every measurement below comes from the three laboratories named, and I have linked their work so you can check it against what I say about it.

What the record shows

Why does the 12V-2×6 connector overheat in the first place?

The 12V-2×6 connector carries up to 600 watts through six +12V contacts and six ground contacts. At 12 volts that is close to 50 amperes total, or about 8.33 amperes per +12V path if the six paths share the load equally.

They do not share it equally, and nothing in the connector makes them. The six contacts sit in parallel, and current takes the path of least resistance. Fairness does not enter into it. As Igor Wallossek of igor’sLAB put it in his 23 June 2026 test, small differences in contact resistance, contact pressure, crimping, line resistance, or alignment are enough to unbalance the distribution.

The consequence is disproportionate because heat in a contact follows P = I² × R. A contact carrying 20% more current generates 44% more heat at the same resistance. If its resistance rises at the same time through aging, an unfavorable seating angle, or weakened spring force, the two effects reinforce each other. The contact heats, the material expands, transition resistance climbs, and local dissipation rises again. That feedback loop is the mechanism behind the melted connector photographs.

The 2022 revision from 12VHPWR to 12V-2×6 lengthened the power contacts and shortened the sense contacts, so a card only permits full power draw once the plug is seated far enough. Igor Wallossek’s assessment of that change is that it improved detection of an incompletely inserted plug and removed neither the current density nor the sensitivity to small resistance differences.

Roman Hartung, who publishes as der8auer, established the scale of that imbalance in February 2025. Measuring an RTX 5090 Founders Edition with a third-party Moddiy cable that had previously run on an RTX 4090, he recorded over 22 amperes on a single wire against roughly 6 amperes expected, other wires carrying as little as 2 amperes, and a PSU-side connector temperature of 150 °C.

What is the ROG Equalizer, and what is inside it?

The ASUS ROG Equalizer is a 750 mm 12V-2×6 cable announced on 9 April 2026 and sold at $49.99 in the United States, around €47 in Germany. It is bundled with 2026 ROG Thor III and ROG Strix Platinum power supplies carrying an -E suffix in the model name, and sold separately under model 90YE00BN-B0QA00 in black and 90YE00BP-B0QA00 in white.

ASUS describes the mechanism in its FAQ as “an innovative conductive bridging architecture” and elsewhere as “a physical bridging design” and “a reinforced bridging structure”. The company does not explain further on its own product page.

Aris Mpitziopoulos of Hardware Busters described what that means in his 30 April 2026 evaluation. All six 12V lines are tied to a common low-impedance metal busbar near the GPU connector, with each pin making pressure contact against that shared plate. His summary of the effect: “If one contact attempts to carry more current, its local voltage drops slightly, and current is automatically redistributed through the busbar to the remaining pins.” In his words, “this turns the cable into a small parallel power plane rather than a bundle of independent conductors.”

Igor Wallossek cut one open on 23 June 2026 and reported the contents: tinned copper strands, a coated copper clamp, two massive aluminum bus blocks, two return lines for voltage measurement, a further bridge structure of iron or steel, screw connections, contact springs, and a large volume of black potting compound. A second, electrically separate rail performs the same joining job for the ground conductors.

Two dimensions from that teardown matter more than they look. Roughly 10.6 mm of free cable remains between the rear edge of the connector body and the start of the Equalizer housing, and the housing itself measures about 11.9 mm long and 23.0 mm wide. The twelve conductors therefore have a little over one centimeter in which to move independently before they become a single rigid assembly.

The conductor itself is ordinary. igor’sLAB counted approximately 105 individual strands at 120 to 140 µm, averaging well over 130 µm, which matches the standardized 105/36 construction of AWG 16 at 1.33 mm² to within a percent. Both the Intel guide and the Molex specification already call for 16 AWG on a standard 12V-2×6 power path, with 9.2 A per power contact under a defined thermal test. ASUS publishes no AWG figure or strand construction for the ROG Equalizer.

That last point contradicts a common explanation for the cable’s benefits. The wire is not thicker than a standard 12V-2×6 wire. Material analysis did confirm ASUS’s plating claim: gold over a nickel barrier layer, though igor’sLAB could not determine the layer thickness or the hardness of the gold.

Where does the 87.5% voltage-drop figure come from?

The 87.5% figure comes from one Reddit user’s screenshots, published by Wccftech on 4 August 2026 under the headline “ASUS’ ROG Equalizer Shows 87.5% Reduction In Total Voltage Drop For The RTX 5090”.

The user, posting as sammyranks, had an RTX 5090 with a power supply’s default 12VHPWR cable that showed brown discoloration and a system that was randomly restarting. His 3.3 V rail was falling to 2.968 V under load. After fitting a ROG Equalizer, he reported the 12V rail holding 11.9 V to 12.0 V while running PRAGMATA at 600 W to 635 W, against 11.2 V on the old cable.

That is where the percentage comes from, and the arithmetic is worth doing because nobody printed it. A drop from 12 V to 11.2 V is 0.8 V. A drop from 12 V to 11.9 V is 0.1 V. One tenth divided by eight tenths is 12.5%, so the reduction is 87.5%. The figure is a ratio of two readings from one user’s monitoring software.

Wccftech’s own article states the limitation in its closing paragraph: “It should be noted that the original 12VHPWR cable belonging to sammyranks was subjected to sufficient wear and tear, causing those wild voltage readings. Replacing the worn out cable with a standard 12VHPWR unit would remedy the issue.”

The user also had no way to measure temperature. Wccftech reports his estimate that the old cable was around 80 °C on the basis that it was too hot to touch.

Where does the 10 °C temperature figure come from?

The roughly 10 °C figure traces to a Chiphell forum post relayed by Uniko’s Hardware and reported by TweakTown on 18 April 2026. A user who received a ROG Equalizer bundled with a newer batch of ROG Strix 1200W Platinum supplies compared it against the stock cable of a ROG Thor 1200W II, driving a GIGABYTE GeForce RTX 5090D AORUS Master ICE under a 600 W FurMark load. Under a thermal imager the stock cable read 59.8 °C and the ROG Equalizer read 50.7 °C, a gap of 9.1 °C.

The same post recorded voltage drop from the nominal 12 V of 0.08 V to 0.14 V on the stock cable against 0.005 V to 0.04 V on the ROG Equalizer.

That is a single comparison between two different cables on one system, which leaves conductor tolerances, crimping, and connector seating as uncontrolled variables. It is a data point. It is not a measurement series.

What did ASUS actually test?

ASUS’s own headline comparison, published on its product page, is a fault-injection test, and ASUS says so in the footnote. It does not compare two working cables under normal load.

The company’s stated method: “This extreme test removes the four middle wires of the +12V 16-pin PCIe power cable to evaluate cable temperature performance under maximum current load.” The stated conditions are 25 °C ambient and a distribution of 25 A + 0 + 0 + 0 + 0 + 25 A for 600 W total. Elsewhere on the same page the conditions for the thermal comparison are given as 55 °C ambient at 600 W with the middle four wires disconnected.

Under that condition ASUS reports approximately 87.8 °C for the ROG Equalizer against approximately 146 °C for a standard 12V-2×6 cable.

The pin table ASUS publishes alongside it shows the same setup electrically.

Pin Current per pin, PSU to ROG Equalizer [A] Current per pin, PSU to connector [A]
1 8.31 25.00
2 8.30 0.00
3 8.33 0.00
4 8.37 0.00
5 8.36 0.00
6 8.33 0.00

Read the right-hand column. Four of the six pins carry zero amperes because ASUS disconnected those wires. The comparison demonstrates that a busbar redistributes current when most of the supply lines are missing. It does not describe a cable in working order.

ASUS’s internal documentation, which igor’sLAB was shown, contains a second and more revealing test. In an “Enlarged-Hole Current Distribution Test” run with a GPU heater at 631.2 W and 25 °C ambient, ASUS degraded the contacts at pins 1 and 5 and measured what the busbar did about it.

Measurement Pin 1 Pin 2 Pin 3 Pin 4 Pin 5 Pin 6
Contact resistance [mΩ] 3.70 2.66 2.60 2.52 4.20 2.43
Current [A] 1.68 11.82 11.86 10.66 6.54 10.04

An ideal distribution at that power would put slightly over 8.7 A through each pin. Pin 1 carried 1.68 A. Pins 2 and 3 carried nearly 11.9 A. The busbar did not force equal currents, because a contact with 3.70 mΩ of transition resistance still refuses current and the current still moves to the better paths. What the design did do was keep the thermal situation controlled at over 630 W with two bad contacts.

There is a figure disagreement worth printing here. ASUS’s public product page claims approximately 146 °C for a standard cable in the four-wires-removed test at 55 °C ambient. The internal slide igor’sLAB describes gives 93.1 °C for the standard cable in the equivalent test at 23 °C ambient, with the ROG Equalizer at 87.4 °C, a difference of only 5.7 K at the hottest point. The 32 K gap in ambient temperature accounts for part of the difference between 146 °C and 93.1 °C and does not obviously account for all of it. The internal slide also shows where the design does clearly work: the downstream cable points fell from 75.0 °C and 73.2 °C to 53.6 °C and 50.6 °C, because the busbar spread roughly 25 A per remaining path back out to about 8.3 A per path.

The most substantive number ASUS has published is also the one nobody picked up. Its 28 May 2026 press release describes an endurance test at 55 °C ambient with a constant 600 W draw through the cable over ten days. ASUS reports that cable temperature settled around 100 °C after the first hour and then “hardly wavered, never once exceeding the 105 °C material limit”. That is the claim I would have led with. It describes sustained normal operation with every wire intact, and ten days at temperature is a real endurance figure. It also shows how thin the margin is on this connector standard: 100 °C steady-state against a 105 °C limit leaves 5 K.

What did independent testing find?

Three parties have tested the ROG Equalizer and they reached different conclusions, because they were measuring different properties of it. Hardware Busters found that the busbar carries extreme fault loads well and validated ASUS’s approach. der8auer found that per-pin current distribution in normal operation is worse with the cable’s bridge fitted than without it. igor’sLAB then confirmed der8auer’s direction using the most tightly controlled method of the three, while agreeing with Hardware Busters about the load capacity. Both findings are correct, and the rest of this section is about why they coexist.

Aris Mpitziopoulos, Hardware Busters, 30 April 2026. He connected the cable to a power supply with a native 12V-2×6 connector, started it with a jumper, and ran all load current through a test fixture using a Korad KEL2060 electronic load rated to 2400 W and 170 A. At 600 W, roughly 52 A, he progressively removed conductors until a single conductor carried the full current. His verdict was that the cable “performs well under realistic conditions and clearly improves current distribution thanks to its busbar design” and “can safely handle higher per-conductor loads than standard cables, validating ASUS’s approach”.

His criticism was structural. The busbar exists only on the GPU side. In his words, “reinforcing only one end shifts the weakest point elsewhere, typically to the PSU side”, and under extreme conditions that is exactly what he observed. He judged that a second busbar on the PSU side would have cost little and completed the design, and called the product “effective but expensive”.

In a second test, described by igor’sLAB, Aris Mpitziopoulos loaded three adjacent paths at approximately 17 A each while the other three carried about 1 A, for 54 A total, over ten minutes at 26 °C ambient. The heavily loaded side peaked at 93.6 °C and the lightly loaded side at 69.7 °C. Temperatures stabilized and the connector housing did not melt. That result supports ASUS’s 17 A claim under a deliberately extreme imbalance.

One clarification from Aris Mpitziopoulos himself, answering a reader in his own comments on 1 May 2026, is worth quoting because it changes how the 17 A number should be read: “the cable cannot handle 17A per pin; it is just that the bridge can address issues that could push loads to 17A at some gauges, because the bridge (busbar) equalizes them with the rest, so they get lowered.”

der8auer, May 2026. Roman Hartung measured per-pin current on an RTX 5090 using a Thermal Grizzly WireView Pro II and found the distribution uneven, in some cases worse than conventional 12V-2×6 cables. Reporting on his first video described differences of up to 4 A between the highest and lowest loaded pin, one near 10 A and one near 6 A, with the distribution changing on every disconnect and reconnect. igor’sLAB, summarizing the same work across several cables, gives the range as 2.0 A to 3.6 A between the strongest and weakest channel, against roughly 0.6 A to 0.7 A for conventional comparison cables in the same setup, with one Equalizer channel reaching 9.8 A and triggering a warning on the measuring device.

He then removed the internal bridge and retested on the same RTX 5090. The spread fell to around 1.5 A, with pins running from roughly 7.5 A to just under 9 A. The component ASUS sells as the balancing mechanism was, in his measurements, the thing making balance worse. He also identified a resistance cost, describing the bridge as adding 1 to 2 mΩ on top of the resistance of the connectors and wire.

His teardown raised two further points. The ROG Equalizer uses gold-plated contacts while most GPU-side 12V-2×6 connectors use tin, and gold against tin can degrade over time through oxidation. He found tin residue on the gold-plated contact area. He also measured the actual contact patch at roughly 0.2 to 0.4 mm, much narrower than the full spring width. He stopped short of calling the cable unsafe, noting that the bridge design should still protect both the GPU and PSU sides in a major malfunction.

One disclosure the reader should have: der8auer has had a product development collaboration with Thermal Grizzly since 2015, and the WireView Pro II he measured with is a Thermal Grizzly product that also competes in this problem space. The right response to that is to check his result against an independent one. Igor Wallossek did.

igor’sLAB, 23 June 2026. Igor Wallossek built the cleanest available test of the specific question der8auer raised. His method is worth describing, because it removes the objection that sinks every other comparison on this page.

Comparing two different cables would introduce differences in cross-section, strand count, length, crimping, contacts, plating, and insulation stiffness, and any of those could be blamed for the result. So he reversed the same physical cable. That moved the metal block from the GPU side to the PSU side, so the six conductors arrived at the GPU header individually and mechanically independent, while every other variable stayed identical. He measured on an MSI GeForce RTX 5090 Suprim drawing 617 W, using a WireView Pro II on the GPU side and a PSU-side adapter shorting all six 12V and ground pins to exclude the supply as a source of error, plugging each variant in ten times.

Above 600 W the total current was effectively identical between the two orientations, at 53.34 A against 53.39 A.

Pin Equalizer at GPU, avg [A] Cable reversed, avg [A]
1 8.85 9.08
2 8.94 9.04
3 7.99 8.50
4 8.48 8.60
5 9.11 8.94
6 9.97 9.22
Total 53.34 53.39

With the Equalizer block in its normal position behind the GPU plug, pin 6 averaged 9.97 A and pin 3 averaged 7.99 A, a spread of 1.98 A. With the cable reversed, the spread fell to 0.72 A. Every dispersion measure moved the same way: average maximum-to-minimum spread per measurement point fell from 1.99 A to 0.79 A, standard deviation of the pin averages from 0.606 A to 0.258 A, largest measured spread from 2.81 A to 1.29 A, and the highest single current from 10.48 A to 9.49 A.

Thermally, measured with a calibrated Optris Pi640 at 21 °C over 15 minutes, the reversed cable ran about 1.55 K cooler on average, 69.23 °C against 70.78 °C, and had the lower hotspot, 71.14 °C against 72.69 °C. The Equalizer orientation had the tighter temperature spread across the six points, 3.33 K against 4.81 K, mostly because one point on the reversed cable ran unusually cool. Both stayed far from any thermal limit.

A number disagreement inside igor’sLAB’s own article, which I am flagging and have not resolved. The measurement tables above 600 W give a spread of 1.98 A falling to 0.72 A and a maximum single current of 10.48 A falling to 9.49 A. The article’s summary page instead states a spread of 4.2 A falling to 1.4 A and a maximum single current falling from 10.6 A to 9.3 A. Those are not the same numbers. The summary is internally consistent with itself, because 10.6 A to 9.3 A gives exactly the 23% reduction in I²R dissipation that it quotes, while 10.48 A to 9.49 A would give 17.9%. The summary therefore appears to draw on a different cut of the data. It does not look like a transcription error. The figures I have quoted above are from the measurement tables I read, which are labeled as the above-600-watt series. The direction of the finding is identical in both.

Does the ROG Equalizer work with non-ASUS power supplies?

Yes, with one exception, and ASUS answers this directly in its own FAQ. Asked which models it supports, ASUS replies: “No specific model requirements. The ROG Equalizer is bundled with the 2026 ASUS ROG Thor III and ASUS ROG Strix Platinum power supplies, and is also compatible with power supplies (ATX3.1 with native 12V-2×6 connector) from all leading manufacturers.”

The requirement is about the connector. Brand does not come into it. The ROG Equalizer is a 16-pin to 16-pin cable, male to male, so the power supply needs a native 12V-2×6 header on the unit itself. Any ATX 3.1 supply from Corsair, Seasonic, be quiet!, MSI, Cooler Master, Thermaltake, Super Flower, or anyone else qualifies. An older supply that reaches a 16-pin plug through an 8-pin to 16-pin adapter does not, because there is no native header to plug into. ASUS’s press release puts the swap in one line: “Simply swap out an existing 12V-2×6 cable to add an extra layer of protection.”

Two practical constraints come with it. ASUS states that the cable must not be disassembled or attached to other cables: “users are strongly advised against disassembling the ROG Equalizer or connecting it to other cables.” ASUS also asks for at least 4 cm of clearance between the chassis side panel and the graphics card’s power connector, so the cable can form a natural curve. Both der8auer and igor’sLAB independently noted the rigid block makes the cable stiffer near the plug than a conventional cable.

There is one ASUS-only element, and it is not on the compatibility page. igor’sLAB found two thin return lines running to a compatible ROG Thor supply, one carrying the 12 V potential measured at the Equalizer and one the corresponding ground. This is remote voltage sensing: the supply reads voltage closer to the graphics card than its own output terminals and can compensate for drop across the cable. On any other manufacturer’s supply those lines do nothing.

That feature is also the most likely explanation for the voltage-drop results people are reporting, and it comes with a limit that igor’sLAB states plainly. The return path measures only the common potential at the busbar’s summing point. Whether the current beyond that point is distributed evenly or very unevenly across the six GPU-side contacts is invisible to it. A single contact can have elevated resistance while the measured voltage looks unremarkable. igor’sLAB adds that in the worst case the regulation can slightly increase total current, because the supply raises output to hold the target voltage at the measuring point.

The graphics card side has no such requirement. The busbar is passive and brand-agnostic, so the cable behaves the same on an NVIDIA, AMD, or Intel card with a 12V-2×6 socket.

The card does decide whether you get the second layer ASUS advertises, and here ASUS is admirably clear about the limits. Per-pin monitoring runs through the Power Detector+ feature of the GPU Tweak III app, and ASUS’s 28 May 2026 press release names exactly which cards support it: the ROG Astral GeForce RTX 50 Series, the ROG Matrix GeForce RTX 5090, and the ROG Matrix GeForce RTX 4090 Platinum. On one of those, GPU Tweak III tracks amperage on each pin of the 12V-2×6 connector and raises a popup if any pin drops to 0 amperes or exceeds a safe upper limit. ASUS states plainly that “such support is not necessary to take advantage of the ROG Equalizer”. A separate “ROG Equalizer mode” toggle inside Power Detector+ was still marked “Coming Soon” on the product page when I checked it on 4 August 2026.

Worth noticing what that means. The per-pin monitoring that would actually tell you whether the cable is balancing anything lives in the graphics card, not in the cable, and only on three ASUS product lines.

Was there a melted ROG Equalizer?

A photograph of a melted ROG Equalizer connector appeared on the Chiphell forums, spread through Reddit, and reached Wccftech on 12 June 2026 under the headline “ASUS’s $50 ROG Equalizer Cable Was Sold as the Fix for 16-Pin GPU Burn-Outs, Yet It Just Melted Too”. Eleven days later igor’sLAB examined the image and concluded it is a fake, generated by an image AI.

The reasoning is specific and checkable. Three contact chambers in the upper row, the first, fifth, and sixth from the left, carry black areas that lack any thermal transition. The violet plastic immediately around them retains almost unchanged color saturation, surface texture, and gloss. Real thermal damage to that plastic produces a gradient: yellowing, then browning, loss of gloss, softening, shrinkage, blistering, and only then charring, spreading along the heat flow. The rightmost chamber shows a dark raised element with no corresponding material displacement in the adjacent walls, where melted thermoplastic would have to be missing in one place and piled up in another.

igor’sLAB named Wccftech and VideoCardz as the outlets that amplified it. Note that the Wccftech article carrying the 87.5% claim links to that melting story as established fact.

Cite as: Max Avery, “What the ASUS ROG Equalizer actually does, and where the 87.5% figure comes from”, maxavery.org, 4 August 2026. Measurements are attributed in the text to their sources: Hardware Busters (Aris Mpitziopoulos, 30 April 2026), der8auer (May 2026), and igor’sLAB (Igor Wallossek, 23 June 2026). Manufacturer figures are from the ASUS ROG Equalizer product page and the ASUS press release of 28 May 2026.

What I think is going on

The ROG Equalizer is a good piece of hardware doing a different job from the one its name advertises, and the gap between those two things is where every argument about it comes from.

Strip the marketing and the mechanism is a busbar. Six wires arrive, join a lump of aluminum, and six contacts leave. That structure is genuinely excellent at one thing: surviving a fault that has already happened upstream. If two of your supply wires stop conducting, a conventional cable dumps everything through the remaining contacts on the card, and the busbar spreads it back out across all six. Aris Mpitziopoulos pushed over 50 amperes through one conductor into that block and the cable held. Three adjacent contacts at 17 amperes for ten minutes topped out at 93.6 °C without melting. That is real engineering and ASUS deserves the credit for it.

The trouble is that fault tolerance is not what “Equalizer” means to a buyer, and it is not what the product page shows you. What the page shows you is a cable with four of its six wires cut, which is a scenario that essentially never occurs. Igor Wallossek’s line about it is the right one: this is the case where lightning finds your garden shed.

Meanwhile the thing that actually melts connectors is a slightly bad contact on an otherwise intact cable, and on that question the measurements point the other way. The most convincing evidence is igor’sLAB’s reversal test, because it is the only one that isolates the variable. Same cable, same card, same load, same ten plug cycles, and the only thing that changes is which end the block sits on. Spread drops from 1.98 A to 0.72 A when the block is not behind the GPU plug. You cannot explain that away with manufacturing tolerances, because it is the same cable in both runs.

The reason is mechanical, and it is the most interesting finding in any of this work. Contacts in a 12V-2×6 connector need to move. The header’s square pins are rigid, and they are not all perfectly centered or square to the connector plane, so each spring contact has to accommodate whatever geometry it meets. The Equalizer gives them about 10.6 mm of free wire and then locks them into a potted block. They adapt less, so a bad seating angle stays bad.

That also explains the odd pairing igor’sLAB reports. With the block fitted, the distribution is highly reproducible from one insertion to the next, and reproducibly uneven. Stiffness buys you consistency, and consistency is only a virtue when the thing being repeated is good.

Then there is the asymmetry that both Aris Mpitziopoulos and Igor Wallossek land on independently, which tells you it is the real structural criticism. There is a busbar at the card and nothing at the supply. If several conductors fail, the block spreads the residual current nicely across the card’s contacts, and the surviving conductors and their PSU-side contacts still have to carry that whole increased load alone. The weak point moves. It does not disappear. Aris Mpitziopoulos says a second busbar would have cost almost nothing at this price. He is right, and its absence is the most telling thing about the product.

Now the two numbers that brought most readers here. The 87.5% is a real ratio computed from real readings, and it is measuring a worn-out cable against a new one. That is a comparison in which literally any new cable wins. The user’s own account describes a browning connector and a 3.3 V rail sagging to 2.968 V, which is what a failing cable looks like, and Wccftech says so in its own last paragraph while running the number in its headline. The 9 °C is better evidence, from a thermal imager under a controlled 600 W FurMark load, and it is still one cable against one other cable on one system with the conductor tolerances uncontrolled. Neither number is fabricated. Neither one supports the conclusion attached to it.

What I keep coming back to is Igor Wallossek’s closing point, which nobody in this argument is acting on. Almost every product in this category, including this one, treats the cable as the thing to fix. The common factor across his measurements was that pin 3 gave the best result regardless of which end of the cable he plugged in. That points at the header on the card. Nothing in the cable explains it.

Twelve rigid square pins that are not reliably centered or square are the shared root cause here, and nobody is selling a fix, because fixing it means the card makers tightening a manufacturing tolerance and there is no accessory to sell you at the end of that.

What actually reduces your exposure

Start with the part of 12V-2×6 connector safety that costs nothing. The most common preventable cause of a hot connector is a plug that is not fully seated, and the 12V-2×6 revision only helps if you push until it latches. Give the cable a gentle tug to confirm, and leave the clearance the manufacturer asks for, which ASUS puts at 4 cm between the side panel and the connector, so the cable is not levering the plug sideways. Lateral pull on a plug changes contact geometry, and contact geometry is the whole problem.

Replace a cable that has visible discoloration, browning, or heat marking, and do it regardless of what you replace it with. This is the actual lesson of the 87.5% story. That user’s problem was a degraded cable, and a $20 replacement of the correct specification would have resolved it. Do not read a large improvement over a damaged part as evidence about a product.

If you want per-pin visibility, buy an instrument. The Thermal Grizzly WireView Pro II measures current on each pin individually and will tell you if one contact is carrying 11 A while its neighbor carries 6 A, which is the condition that precedes a failure. Every finding in this article that concerns real-world balance was produced with one. It costs more than the cable and it answers a question the cable cannot. If you own an ASUS ROG Astral RTX 50 Series, ROG Matrix RTX 5090, or ROG Matrix RTX 4090 Platinum, you already have this capability through GPU Tweak III Power Detector+ and it costs you nothing.

If you want automatic protection, look at what Corsair shipped in April 2026. The ThermalProtect PCIe 5.1 600 W 12V-2×6 cable puts a passive thermal switch inside a cable comb about 30 mm from the connector, and when it detects the cable above 65 °C it signals through the sense pins to make the card stop drawing power. It lists at $24.99, works with any native 12V-2×6 supply and any 12V-2×6 card regardless of brand, and needs no configuration. That is half the ROG Equalizer’s price, and it goes at the failure mode directly by interrupting the thermal runaway once it starts. Preventing the imbalance underneath is not something it attempts, so it will do nothing at all for your current distribution, and a shutdown mid-game is a real cost. The two products address different halves of the problem and neither one covers both.

One caution on ASUS’s characterization of the competition. Its press release says some competitor solutions “employ an alarm-based system as their first line of defense, requiring users to manually shut down their PC and reseat the cable when a power anomaly is detected”. Corsair does not name that as how ThermalProtect works. By Corsair’s own account the trip signals the card through the sense pins and power draw stops without the user doing anything. The description fits a monitoring instrument like WireView, which alerts the user and leaves the intervening to them.

The strongest available combination is per-pin overcurrent protection in the power supply paired with something that improves fault tolerance in the cable. That is Aris Mpitziopoulos’s own recommendation, and it is worth quoting his framing: the cable improves load balancing on the GPU side while per-pin protection ensures no individual conductor is overstressed. More supply makers are adding per-pin OCP. If you are buying a new power supply for a 575 W card, that feature is worth more than any cable on this page.

So, who should buy the ROG Equalizer? On the evidence, a narrow group. If you already own a 2026 ROG Thor III or ROG Strix Platinum supply with the -E suffix, the cable came in the box, the voltage-sense return lines actually function, and there is nothing to decide. If you bought one of those supplies before the bundling started, ASUS is running an upgrade program until 28 November 2026 where you enter the power supply’s serial number to check whether it qualifies for a discount code, which is a better price than the one on this page. If you run a card near 600 W on a native ATX 3.1 supply and you want the largest fault reserve available in a passive cable, this is the product with the most independent load testing behind it, and Hardware Busters’ stress results are reassuring about what it survives.

If you are buying it because you read that it cuts voltage drop by 87.5% or evens out your pin currents, buy something else. Three laboratories have now looked, and the balancing claim is the one part that does not hold up. Igor Wallossek’s own verdict is the fairest summary anyone has written. He calls it “technically interesting and exceptionally robust”, and adds that “its name only incompletely describes its actual function.”

Check the current price of the ASUS ROG Equalizer on Amazon (affiliate link)

And a limitation that applies to everything above, including the recommendation. None of these products address the root cause. A cable can add margin, monitor temperature, or report current. None of them can make twelve rigid square pins in a graphics card header meet twelve spring contacts at a reproducible angle, and until somebody fixes that, every option here is managing a symptom.

Sources

Corrections

None yet.

Last updated: 2026-08-04

Disclosure. Max Avery is affiliated with Digital Ascension Group (DAG). Investment advisory services are offered through DAG Wealth, an SEC-registered investment adviser (CRD No. 328627). Registration does not imply a certain level of skill or training. DAG is not a law firm and does not provide legal or tax advice. Custody arrangements with third-party independent qualified custodians reduce certain risks but do not eliminate them. Nothing here is investment, legal, or tax advice, or a recommendation to buy or sell any asset. This article describes matters of public record; charges are allegations and defendants are presumed innocent unless and until proven guilty.

More from the journal

Continue reading.