ATX 3.x Power Supplies

↪ Notes: FDB fan; Hong Hua

↪ Notes: Fan is of questionable quality. Yate Loon DBB. Loose 12V transient response.

↪ Notes: 5 year warranty, OCP is set way higher than necessary on the minor rails.

↪ Notes: Insane 7 year warranty given the component quality, OCP is set way higher than necessary on the minor rails, very poor 200% excursion response.

↪ Notes: Although OCP is set reasonably (118% on 12V), the platform is at its limit and cannot handle the increased stress, as ripple gets out of control.

↪ Notes: ODM is Dongguan Changhu Electronics Co. Ltd.

↪ Notes: ODM is Dongguan Changhu Electronics Co. Ltd.

↪ Notes: Costdown Core Reactor II

↪ Notes: ATX 3.0 refresh

↪ Notes: Semi-digital

↪ Notes: Totem pole PFC

↪ Notes: Mid quality fan; Same Martech "FDB" fan on the MSI A-GL series.

↪ Notes: Decent electrical performance, although some components are of questionable quality.

↪ Notes: The OCP trggering point for the 5VSB line allows for extremely excessive ripple.

↪ Notes: The OCP trggering point for the 5VSB line allows for extremely excessive ripple. 12V line voltage ripple spikes around the 80-90% load range.

↪ Notes: Read note regarding input range. Mid quality fan; Same Martech "FDB" fan on the MSI A-GL series.

↪ Notes: Questionable fan choice, claimed as FDB, model was noted as rifle in a review of another PSU.

↪ Notes: Sleeve bearing fan. 12V-2x6 is fixed.

↪ Notes: 1000W model has poor transient response on 12V under 200% spike, other models mediocre. 850/1000W avail. in white.

↪ Notes: 12V-2x6 connector (GPU side) has thermistors to act as an "OTP". Stellar ripple suppression, especially on the minor rails.

↪ Notes: OCP is set way higher than necessary on the 12V and 5VSB rails. Good noise profile given its capacity.

↪ Notes: FSP-57AZB. 12V-2x6 connector (GPU side) has thermistors to act as an "OTP".

↪ Notes: Optional "5V Boost" button which increases +5V by 200mV; bridgeless PFC. 12V-2x6 connector has thermistors to act as an "OTP".

↪ Notes: Not to be confused with the Xigmatek Loki.

↪ Notes: 5VSB OCP set high, poor 3.3V transient response in the 200% excursion

↪ Notes: Fan profile is extremely relaxed, OCP is set way too high on the minor rails, no FFP despite being an extremely expensive platform.

↪ Notes: Much better performance than the 1200W model and fewer issues despite using the same platform.

↪ Notes: OPP is set way higher than necessary and isn't reduced even under high operating temperatures. Very poor 200% excursion response.

↪ Notes: Multitude of minor issues leading to low overall performance, does not meet requirements for tier A in Aris' testing. Possible bad sample?

↪ Notes: Ball bearing fan. No warranty period mentioned, detier to B-.

↪ Notes: 2 year warranty; platform is otherwise reasonably capable.

↪ Notes: 2 year warranty; platform is otherwise reasonably capable.

↪ Notes: Not to be confused with the downgraded non-modular Pure Power 12!

↪ Notes: Bridgeless PFC, OCP needs tuning on all rails

↪ Notes: Excellent noise profile.

↪ Notes: Bridgeless PFC, very high component and build quality. OCP needs tuning on all rails. Excellent noise profile.

↪ Notes: Excursion response in the 200% test could be better compared to other units in its class. Excellent noise profile.

↪ Notes: Interleaved PFC; digital operation for primary, APFC and 12V; minor rails are analog.

↪ Notes: 85C bulk caps, excellent noise profile.

↪ Notes: First be quiet! PSU with semi-passive operation.

↪ Notes: Low quality fan, 85C bulk cap.

↪ Notes: Good build and component quality, with good ripple suppression and a decent noise profile. Notably, Aris received a PPG-C sample with a higher capacitance bulk capacitor, different from what other reviews had. PPG-C and PPG-S are identical aside from modularity.

↪ Notes: Good build and component quality, with good ripple suppression and a decent noise profile. Notably, Aris received a PPG-C sample with a higher capacitance bulk capacitor, different from what other reviews had. PPG-C and PPG-S are identical aside from modularity.

↪ Notes: Platform is well reviewed although no direct transient response data.

↪ Notes: Fundamentally a Polaris ATX 3.0 refresh. Load regulation is somewhat subpar on the 1050W model.

↪ Notes: Basically G1278 updated to ATX 3.0

↪ Notes: Still essentially G1278 with further updates to ATX 3.1

↪ Notes: Has purple heatsinks for glam. Loose load regulation on 5VSB rail with middling 12V excursion response.

↪ Notes: One review showed ripple on the 12V rail at 62mV for the 1600W, other samples were within 50mV. 5VSB OCP is set high.

↪ Notes: "Semi-digital" - offers digital fan control and monitoring via MasterPlus software; actual components are analog.

↪ Notes: Rated for 850W under 115VAC and advertised as 850W in North America.

↪ Notes: Rated for 1100W under 230VAC, but can still be used under 115VAC for 850W.

↪ Notes: Same platform as the Edge Platinum but with active cooling; full 1300W rating regardless of input range.

↪ Notes: Modified MWE Bronze V3, same as G Gold V2.

↪ Notes: 230V input only

↪ Notes: Very similar to the GX III that superseded it.

↪ Notes: GX III ATX 3.0, but upd. to ATX 3.1. Yate Loon rifle-bearing. Has purple heatsinks for glam.

↪ Notes: GX III w/ upgraded fan (Hong Hua FDB/Yate Loon DBB w/ region specific DBB?) but no inrush protection (NTC Thermistor & bypass relay).

↪ Notes: Modular inputs on side of PSU rather than the rear. Use caution and check case clearances: see Jon Gerow's list of cases (archive)

↪ Notes: 160mm length

↪ Notes: Semi-digital. Good noise profile. 5VSB OCP is set quite high. Look above for case clearance list.

↪ Notes: Class-leading performance, excellent build quality, relatively quiet for the capacity.

↪ Notes: Refresh with included 12V-2x6 cable and listed ATX 3.1 compliance. Actual design appears to be the same internally.

↪ Notes: Refresh with included 12V-2x6 cable and listed ATX 3.1 compliance. Actual design appears to be the same internally.

↪ Notes: Minor changes on the primary side, overall very similar; merely a refresh. Embossed cables instead of flat cables from older generations. (Fancy)

↪ Notes: Refresh with included 12V-2x6 cable and listed ATX 3.1 compliance. Actual design appears to be the same internally.

↪ Notes: Black label; China only

↪ Notes: Poor ripple suppression on the minor rails. Uses semi-digital PFC, 7 year warranty. Can get loud (in excess of 40dbA) under heavy load.

↪ Notes: 650W has 300W 12V-2x6 cable. Poor 3.3V transient response. All use Hong Hua rifle-bearing fans. Y-shaped fan grille perforations.

↪ Notes: 750/850W: 140mm, 1000/1200W: 150mm. 2023 version includes 12VHPWR cable. Early reports of coil whine and premature failures/DOAs, but supposedly fixed.

↪ Notes: The OCP triggering points are set too high on the minor rails, 5VSB regulation exceeds 5% at 5.37%, filtering capacitors are of questionable quality.

↪ Notes:

↪ Notes: The OCP and OPP triggering points are set too high on all rails, especially 5VSB, where it allows for extremely excessive ripple. Good noise profile.

↪ Notes: Ripple suppression, while within the limits for tier A, isn't as good as other units in its class. The OCP trip points need to be reduced on all rails.

↪ Notes: One review had 12V line voltage deviating slightly more than the limit in the 160% ATX 3.1 excursion test. OCP, OPP, and OVP are set too high on all rails.

↪ Notes: >50mV ripple suppression on 12V.

↪ Notes: Very high inrush current, poorly tuned protections on the minor rails; otherwise decent electrical performance.

↪ Notes: Protections need tuning, 5VSB regulation doesn't quite qualify for A tier.

↪ Notes: 850W model can come in white.

↪ Notes: Semi-digital. Outstanding performance per Aris' review.

↪ Notes: Can come in white variants. Mediocre fan, no SCP on -12V rail, which isn't used in modern PCs. Avoid if using very old hardware.

↪ Notes: Can come in white variants. Protections need tuning. ATX 3.0 refresh of the PM-D.

↪ Notes: OCP is set way too high on the minor rails, which led to failure of the unit during Aris' testing. Detier to B- unless verified fixed.

↪ Notes: Subpar component quality, with highly set protections and OOS 3.3V transient response.

↪ Notes: Low component quality, w/ highly set OCP/OPP & OOS 3.3V transient response. Not quite ATX 3.X by failing low load efficiency. 600W 12+2 cable.

↪ Notes: Subpar component quality, with poorly tuned protections, 5V OCP incorrectly set allows for >135mV of ripple. OOS 3.3V transient response.

↪ Notes: Sleeve bearing fan. Meets all other requirements for A tier.

↪ Notes: The OCP and OPP triggering points are set way too high on all rails.

↪ Notes: Good electrical performance for the category, decent component quality, protections are set reasonably.

↪ Notes: ATX 3.0 compliant; uses dual 8-pin to 12VHPWR.

↪ Notes: No performance data, however platform is well reviewed with protections working in similar configs.

↪ Notes: OCP originally set way too high; fixed in later reviews from July 2024. Excellent electrical performance. ATX 3.1 and ATX12VO compatible.

↪ Notes: 122mm depth (extremely compact.)

↪ Notes: 13 year warranty. 5VSB OCP still allows for extremely excessive ripple. Good performance otherwise.

↪ Notes: Transient response doesn't quite make it A tier, no direct protections testing.

↪ Notes: Limited testing; rating could be higher once additional testing exists.

↪ Notes: ATX 3.1 is claimed, but the unit fails to meet the spec due to poor ATX 3.1 excursion response. Treat as ATX 2.x.

↪ Notes: Tier A performance (albeit barely) but only a 3 year warranty since EVGA might go under.

↪ Notes: 850W has excellent noise graph.

↪ Notes: Identical to the ASRock Taichi, however until further testing exists, it will not receive a "+" grade. See teardown here.

↪ Notes: Identical to the PTM X Pro made for ATX 3.0 compliancy.

↪ Notes: Insane performer, extremely quiet, however the 850W model is discontinued.

↪ Notes: Highly likely a similar platform to the PTM Pro ATX 3.0, OTP works in similar configurations (PTM Pro). Superb electrical performance.

↪ Notes: Strong electrical performance (A tier), but with less expensive components and a rifle bearing fan.

↪ Notes: YL rifle bearing fan.

↪ Notes: P1 refreshed for ATX 3.1; Falsely claimed Cybenetics Gold certification, units were never tested.

↪ Notes: Falsely claimed Cybenetics Gold certification, units were never tested.

↪ Notes: Falsely claimed Cybenetics Gold certification, units were never tested.

↪ Notes: Good overall electrical performance and component quality. Version 1.0 is ATX 3.0, Version 1.1 is ATX 3.1 to replace 12VHPWR with 12V-2x6.

↪ Notes: Good overall electrical performance and component quality. Version 1.0 is ATX 3.0, Version 1.1 is ATX 3.1 to replace 12VHPWR with 12V-2x6.

↪ Notes: The OCP triggering point on the 5VSB rail allows for extremely excessive ripple; 5V rail hot triggering point allows for a voltage drop out of spec.

↪ Notes: The OCP triggering point on the 5VSB rail allows for extremely excessive ripple; 5V rail hot triggering point allows for a voltage drop out of spec.

↪ Notes: Various issues including low PF, poor excursion response, and terrible vampire power suppression. OCP is set sky-high on the minor rails.

↪ Notes: Numerous minor issues, failed ATX 3.0 transient response testing, poor build quality.

↪ Notes: Costdown of the Thermaltake Toughpower GF3 ARGB platform; non-modular with Chinese caps and a rifle bearing fan.

↪ Notes: German bulk caps; dual-chamber L-shape design; modular inputs on the main PCB. 1000W and up models come with a USB/fan hub.

↪ Notes: German bulk caps; dual-chamber L-shape design; platform is an EPS-0100 modified to put the modular inputs on the main PCB.

↪ Notes: Good electrical performance and reasonably quiet for the category.

↪ Notes: Rotatable power outlet. 5VSB OCP is high. Includes a USB 2.0 hub.

↪ Notes: Extremely poor electrical performance which leads to severe out of ATX compliance, very abnormal implementation. Refer to here.

↪ Notes: Excellent performance and noise profile. OPP and OCP on the minor rails need tuning. They do work but are set way higher than necessary.

↪ Notes: Protections need tuning and mediocre ATX 3.0 transient response; overall performance is dwarfed by the CWT-built Titan Gold.

↪ Notes: OCP is set high on 3.3V and 5VSB, 5VSB voltage regulation >3% at 4.87%, mediocre excursion response in the 200% test. High inrush current.

↪ Notes: ATX 3.0 refresh

↪ Notes: Semi-digital.

↪ Notes: Only one 8-pin PCIe cable! Dual 12VHPWR. Good performance in pre-production sample, but needs fine-tuning. Semi-digital.

↪ Notes: Digital primary & 12V; analog minor rails & standby.

↪ Notes: Minor rails' OCP is set lower compared to the CWT-built high-wattage models. 650W is noisier than the 750W on account of using a DBB fan.

↪ Notes: Extremely quiet operation; however OCP is set high on the minor rails.

↪ Notes: Interleaved PFC, semi-digital

↪ Notes: Solid electrical performance, let down by poorly tuned protections, mediocre transient response, and a low quality fan.

↪ Notes: Yate Loon rifle bearing fan. The OCP triggering point for the 5VSB line allows for extremely excessive ripple.

↪ Notes: 5V OCP malfunctioned under high temps; CSZ costdown with Chinese caps and inexpensive bridge rectifiers per Aris' testing.

↪ Notes: Yate Loon rifle bearing fan. The OCP triggering point for the 5VSB line allows for extremely excessive ripple.

↪ Notes: Excellent noise profile.

↪ Notes: Available in black and white

↪ Notes: Semi-digital, interleaved PFC, magnetic levitation (maglev) bearing fan

↪ Notes: Modular board protudes from the unit

↪ Notes: Overall good performance, uses a YL rifle-bearing fan.

↪ Notes: Some degree of false advertising going on. Claims to be LLC resonant with an FDB fan and 105C caps. Internal shots show otherwise. Recommended to proceed with caution. Even though the platform is safe, it's not as good as advertised.

↪ Notes: Some degree of false advertising going on. Claims to be LLC resonant with an FDB fan and 105C caps. Internal shots show otherwise. Recommended to proceed with caution. Even though the platform is safe, it's not as good as advertised.

↪ Notes: OTP is set over >200C (doesn't function as intended). 850/1000W might not be similar to the 1200W.

↪ Notes: Globe Fan FDB fan. Assumed to be quite noisy (>48dBA) based on other units on the same platform. OPP is set way higher than necessary.

↪ Notes: Internal shots show the unit using the same "ICB" (in reality rifle bearing) fan as the SilverStone Triton Rx.

↪ Notes: The OCP and OPP triggering points need tuning. Very good ripple suppression, and decent component quality. Excellent noise profile.

↪ Notes: Mediocre load regulation on the minor rails, and poor excursion response in the 200% test. The OCP and OPP triggering points are set too high on all rails. 1000W model has high inrush currents. Good performance otherwise.

↪ Notes: Mediocre load regulation on the minor rails, and poor excursion response in the 200% test. The OCP and OPP triggering points are set too high on all rails. 1000W model has high inrush currents. Good performance otherwise.

↪ Notes: Mediocre load regulation on the minor rails, and poor excursion response in the 200% test. The OCP and OPP triggering points are set too high on all rails. 1000W model has high inrush currents. Good performance otherwise.

↪ Notes: Black Diamond with a PCIe 5.0 interface.

↪ Notes: 122mm depth; same platform as Enermax Revolution D.F. 12, but with cheaper components.

↪ Notes: Electrical performance is mediocre and OCP triggering points are sky high. Recommended only for low-end hardware.

↪ Notes: Failed JonnyGuru's burn-in test (gave up the ghost), protections are poorly misconfigured on this platform.

↪ Notes: Based on the same platform as the Black Hole; OTP untested; some parts remain unidentifiable; white variant available.

↪ Notes: ATX 3.0 transient response could be better. 5VSB OCP is stupidly high.

↪ Notes: Black model has white line accents, compared to V3. Stupidly high inrush current, mediocre transient response, minor rails' OCP need tuning.