FSP Group Partners with Intel to Launch Energy-efficient and Backup-integrated Next-gen Power Solutions

[Nomad76]

Facing global challenges of energy conservation, carbon reduction, and sustainable development, Intel has been committed to pushing the industry toward higher performance and lower energy consumption product standards, actively leading the PC ecosystem toward a dual track of environmental protection and innovation. FSP Group, as Intel's long-term strategic development partner, shares the same philosophy and has continuously invested in power supply technology innovation with Intel's support, particularly focusing on 12VO architecture design that complies with the latest energy efficiency regulations.

At the Intel Client Ecosystem Symposium, FSP showcased its latest V3 version of TFX 12VO PSU, which not only helps brand customers build compliant PC systems but also contributes to carbon reduction and environmental protection with a design that balances high performance and energy efficiency.



FSP500-10TRC Features

• Intel PSDG ATX12VO V3.0 version
• 500 W Bronze efficiency
• Standard TFX form factor, dimensions (L x W x H): 175 x 85 x 65 mm
• Fixed cable design
• Global voltage input range 100Vac-240Vac
• 10TRC series power options: 300 W / 400 W / 500 W

Additionally, FSP announced two innovative battery backup power supply products (PSU Plus BBU), providing practical and reliable solutions for small computing devices and workstation systems that need to maintain stable operation during momentary power interruptions.

EnerXBar 090-19A BBU: Designed for mini PCs and lightweight systems, with an integrated plug-and-play architecture. Demonstrated with ASUS's 15th generation NUC, it showcases its ability to provide zero transfer time (0 ms) backup during power outages, solving the pain points of data loss and system interruption.

EnerXBar 090-19A BBU Features

• Integrated design
• Plug and play
• Long backup time
• Replaceable DC output interface
• Zero transfer time (0 ms)
• Peak power support (>160%@1 millisecond)
• Long service life
• AC failure detection (optional)

EnerXCube 550-M ATX BBU (Patent Pending): Targeting high-performance applications such as workstations, with built-in USB PD and long-lasting LiFePO4 batteries, it provides more advanced power backup capabilities for systems requiring high stability, enhancing overall system resilience and security.

EnerXCube 550-M ATX BBU Features

• 550 W rated output power
• Integrated design
• Built-in 65 W USB PD
• Zero transfer time (0 ms)
• Extra-long backup time
• Long service life
• LiFePo4 rechargeable battery
• RS-232 communication interface

The three new products showcased by FSP not only demonstrate strong technical R&D capabilities but also reflect its proactive approach to sustainable operations and industry responsibility. Through close collaboration with Intel, FSP continues to break through in product innovation and application value of power solutions, bringing more stable and environmentally friendly next-generation choices to global customers. The three products will also be fully unveiled at Computex Taipei 2025, presenting the future direction and potential of power supply technology.

View at TechPowerUp Main Site | Source

 

[_roman_]

We definitely need more expensive mainboards with more voltage regulators and coils.

Simplify power supply units to outsource the efficiency problems and parts to the mainboard.

RS-232C has to die. We have usb already or ethernet

 

[ncrs]

I do not get the appeal of this PSU+battery combo.
A standalone UPS provides more convenience and elasticity. It can power your monitor at the same time, so you can actually save documents, start to shut down cleanly, and so on instead of just keeping the PC running because there's a battery in the PSU.
Batteries, even LiFePo4 do not like heat:

LiFePO4 batteries can typically operate within a temperature range of -20°C to 60°C (-4°F to 140°F), but optimal performance is achieved between 0°C and 45°C (32°F and 113°F).

They do have longer life span, but still from the render it looks to be not user serviceable.
For business uses UPS batteries usually have strict replacement schedules, and that can be done when UPS is running most of the time. With a PSU-integrated battery there's going to be a lot more manual labor required with disruptions to the system itself.

As for the ATX 12VO PSU having bronze efficiency, really? Dell has been shipping platinums in their custom 12V designs (for example Optiplex) for some time now.

 

[LabRat 891]

We definitely need more expensive mainboards with more voltage regulators and coils.

Simplify power supply units to outsource the efficiency problems and parts to the mainboard.

RS-232C has to die. We have usb already or ethernet

In consumer-facing products, agreed. I think these are more Industrial or Enterprise

I do not get the appeal of this PSU+battery combo.
A standalone UPS provides more convenience and elasticity. It can power your monitor at the same time, so you can actually save documents, start to shut down cleanly, and so on instead of just keeping the PC running because there's a battery in the PSU.

Stacked redundancy or compactness.

Batteries, even LiFePo4 do not like heat:
They do have longer life span, but still from the render it looks to be not user serviceable.
For business uses UPS batteries usually have strict replacement schedules, and that can be done when UPS is running most of the time. With a PSU-integrated battery there's going to be a lot more manual labor required with disruptions to the system itself.

Point, but there's apparently edge use cases that called for it. Off the top of my head: CnC/3Dp control computer? Remote cellular network infrastructure?
LiFePo4 cells are probably the most tolerant of all stresses* amongst the common lithium chemistries. *excepting cold.

As for the ATX 12VO PSU having bronze efficiency, really? Dell has been shipping platinums in their custom 12V designs (for example Optiplex) for some time now.

The PSU has an always-online UPS integrated into it. (0ms transfer time) There's stacked AC-DC-AC-DC conversion going on.
On the plus side, these PSUs should have very clean power output with good isolation from outside interference.

 

[fec32a4de]

I do not get the appeal of this PSU+battery combo.
A standalone UPS provides more convenience and elasticity. It can power your monitor at the same time, so you can actually save documents, start to shut down cleanly, and so on instead of just keeping the PC running because there's a battery in the PSU.
Batteries, even LiFePo4 do not like heat:

They do have longer life span, but still from the render it looks to be not user serviceable.
For business uses UPS batteries usually have strict replacement schedules, and that can be done when UPS is running most of the time. With a PSU-integrated battery there's going to be a lot more manual labor required with disruptions to the system itself.

As for the ATX 12VO PSU having bronze efficiency, really? Dell has been shipping platinums in their custom 12V designs (for example Optiplex) for some time now.

I guess the only appeal of PSU+battery combo if power flickers often but not a full outage. Electricity in my area is shit so I have UPS on anything that needs to have power all the time.

 

[ncrs]

Point, but there's apparently edge use cases that called for it. Off the top of my head: CnC/3Dp control computer? Remote cellular network infrastructure?

All the use cases you mentioned would benefit from a standalone UPS over this PSU+battery combo due to the reasons I mentioned before. Cellular towers have dedicated battery banks and sometimes even diesel generators when placed in remote areas, and most of the time it's running all on DC anyway due to efficiency.

LiFePo4 cells are probably the most tolerant of all stresses* amongst the common lithium chemistries. *excepting cold.

Still the insides of a PSU is less hospitable to them than an external PSU which only gets warm when running DC/AC conversions during charging/discharging.

The PSU has an always-online UPS integrated into it. (0ms transfer time) There's stacked AC-DC-AC-DC conversion going on.

I doubt it's running fully online mode all the time, that would generate a lot of loss (especially at low loads, for example a 900W APC on-line UPS efficiency curve) and heat, and be quite expensive (the APC mentioned is over $1000).
From the datasheet it sounds like they might be using a PR move by utilizing ATX-mandated hold-up time (>=12ms) as an internal buffer, so from the external point of view it's 0ms switch time. But that's just my theory

On the plus side, these PSUs should have very clean power output with good isolation from outside interference.

Edit: from the datasheet it doesn't look like it. ±5% to ±10% on regulation. 120 mV ripple on 12V, when the last TPU PSU review scored up to 24 mV. 120 is the upper limit of the ATX specification, so this looks quite bad.

 

[anonuser57]

We definitely need more expensive mainboards with more voltage regulators and coils.

Simplify power supply units to outsource the efficiency problems and parts to the mainboard.

RS-232C has to die. We have usb already or ethernet

No. With USB you can't have long cables and with Ethernet you need a who built in networking stack when you just want a command console. With Rs232 you don't have any networking involved and you can run the connection to another room in the building if you want. In particular if the device happens to be near an rj45 patch panel you can get a dumb adapter and get the console access at any desk you want without exposing it to the network.

 

[fec32a4de]

No. With USB you can't have long cables and with Ethernet you need a who built in networking stack when you just want a command console. With Rs232 you don't have any networking involved and you can run the connection to another room in the building if you want. In particular if the device happens to be near an rj45 patch panel you can get a dumb adapter and get the console access at any desk you want without exposing it to the network.

With USB you can have powered/amplified extension cables. They're against the USB rules, but they do work.

I was running almost 50ft of USB to monitor my compact medicine fridge that is on a 1500VA/1000w UPS.

Although USB2.0 (UPSs tend to use 2.0) assumes data is lost if it doesn't arrive in some milliseconds (I forget) I guess the active repeaters passed on info of incoming data or something. With USB3.0+ there's no such restriction so active USB3.0 works extremely well over long distances. Because I was bored or something, I connected a SATA SSD to the 50ft USB3.0 repeaters and got just over 400MB/s sequential read and write.

Although I do agree with you on the simplicity of RS232

 

[_roman_]

I think RS485 is the proper product for such scenarios. I do prefer Ethernet

 

[EsaT]

As for the ATX 12VO PSU having bronze efficiency, really?

So much for ATX 12VO allowing higher efficiency because of simpler design...

But that's really wet dream of MBA scammers:
Components must be about dirt cheap and as new different type of product they can give it high price sticker.
So win-win.

Who cares (definitely not Intel) that more power gets wasted as heat causing also need for more cooling...

https://silentpcreview.com/wp-content/uploads/2012/12/Figure-4-Seasonic.png

That's amount of wasted power from 80+ Bronze to Platinum PSU...

Power Lost: A Better Way to Compare PSU Efficiency

CA_Steve, an active member and a moderator in the SPCR forums, examined the 80 PLUS definition of efficiency as a % figure at different power levels, also defined as a % of the rated power of a PSU. Steve decided that this is too misleading and shows why power lost (as heat) is a better

silentpcreview.com

From the datasheet it sounds like they might be using a PR move by utilizing ATX-mandated hold-up time (>=12ms)

Edit: from the datasheet it doesn't look like it. ±5% to ±10% on regulation. 120 mV ripple on 12V...

Another "upgrade" by MBA and marketing scammers:
Hold-up time requirement used to be 17ms.
Guess ATX 3.2 goes to below half wave of 50Hz like 9ms...

Maybe FSP digged up old Epsilon platform.
PSUs using that routinely exceeded max allowed ripple in reviews.
That should also fit to inefficiency rating.

 

[trsttte]

The PSU has an always-online UPS integrated into it. (0ms transfer time) There's stacked AC-DC-AC-DC conversion going on.

Why!? If they have the battery inside the unit there's no reason for it to have an AC side, it can be just AC-DC with the battery on the DC side, no need to stack anything.

So much for ATX 12VO allowing higher efficiency because of simpler design...

It does allow for higher efficiency and simpler designs. A psu has no information of what devices downstream might need or not, so 3.3v and 5v rails have to be kept active even if they're not needed for anything. If that's shifted to the motherboard, you can power up and down things as necessary because the motherboard does know if it has things using those rails or not.

But just because it can doesn't mean it will produce better designs, the manufacturers can of course just pocket the savings if they want, that's where ratings come in and having an 80 bronze rating should mean an instant no buy.

 

[R-T-B]

So much for ATX 12VO allowing higher efficiency because of simpler design...

I mean it's efficiency considering the built in always online UPS is actually rather incredible, tbh.

 

[EsaT]

It does allow for higher efficiency and simpler designs. A psu has no information of what devices downstream might need or not, so 3.3v and 5v rails have to be kept active even if they're not needed for anything. If that's shifted to the motherboard, you can power up and down things as necessary because the motherboard does know if it has things using those rails or not.

Those small step down switchers used by every even remotely modern PSU for 3.3V and 5V are highly efficient.

And motherboard wouldn't have any more idea of something needing 3.3V/5V than PSU.
Without those for example of electronics of HDD/SSD wouldn't start at all.
Pretty sure also most expansion cards would be pretty mute.

And 5V is anyway needed for the USB, so you can't avoid generating it somewhere.
Or are we going to trash also that just for having change for the sake of change?

In highly controlled/specified environment these type changes can work.
But in general PC world they would just create more new issues than solve.

 

[trsttte]

And motherboard wouldn't have any more idea of something needing 3.3V/5V than PSU.

Of course it does, the motherboard knows if there are devices connected to PCIe, to SATA, to usb, to whatever, and it can also probe for them - "let me supply some power, is there a device there starting up? oh I guess I need to keep this bus active then." The PSU can't really do that.

On a fully loaded machine it's simply moving the problem from one place to another, but most machines are not that so ATX12VO ends up making a lot of sense, so much so it was already becoming a thing in oem machines before the standard was even published.