Athlon 200GE - the ultimate great place-holder CPU?

[hojnikb]

The MAX power conmsuption ive ever seen from the 200GE was 75W (CPU-IGP) for the whole system measured at the wall, thats barely twice of a Sempron 2650, im sure ANY PSU can handle that.

The 200GE is really close to match 2650 idle power.

BTW, the 2200G uses up to 160W under the same conditions.

Not really, there are lots of cheap ITX cases, that only come with like 60-90W PSUs. Now if 200GE would be restricted to 35W, that would be fine, but occasional peaks to lets say 50-60W not so much.

 

[Shivansps]

It is still 75W for the full system with a PSU tha ist 80 plus. And the only way to can get anywhere near that is combining full CPU+igp loads.

Anyway, Gigabyte joins the club of 200GE overclokers with the new 1.0.0.6 Agesa bios.

No IGP or memory overclock, it is just like MSI.

 

[Abwx]

It is still 75W for the full system with a PSU tha ist 80 plus.

What is the idle power.?

A 200GE is half a 2400G and at 10% lower frequency, it should barely drain 50% of the big brother s TDP.

 

[Shivansps]

What is the idle power.?

A 200GE is half a 2400G and at 10% lower frequency, it should barely drain 50% of the big brother s TDP.

About 30W max. Idle power while less than the 2200G and 2400G is not half, Full CPU+IGP it is.

 

[VirtualLarry]

https://www.ebay.com/itm/Lenovo-90J...-2200G-with-Radeon-Vega-Graphics/113376148296

With the Ebay 10% off coupon (expires at 3AM PST this morning), that came out to under $300.

I was thinking, I could add an M.2 PCI-E NVMe SSD (EX900 256GB was recently $39.99 on ebay), and replace the 2200G with an A200GE for $59.99 (Newegg), and for $100 extra, minus the value of the 2200G coming out ($100), so for even money, I could put in an NVMe SSD, and then sell/give this rig to a friend of mine.

 

[Insert_Nickname]

It is still 75W for the full system with a PSU tha ist 80 plus. And the only way to can get anywhere near that is combining full CPU+igp loads.

...and this is using a full-size ATX board with an X370 chipset. mATX and ITX boards tend to be a bit more efficient.

Anyway 80+ rating tend not to matter too much below 100W.

I don't think anyone is going to pair up an Athlon 200GE with an X370/X470 mainboard except for testing or as a placeholder. At which I'd question priorities, since you can get a 2200G + A320 board for the same as a X370/X470 board.

 

[Abwx]

About 30W max. Idle power while less than the 2200G and 2400G is not half, Full CPU+IGP it is.

Well, that s about 45W delta at the main, so the chip is within 35W even in CPU + GPU usage.

To get lower power not only the MB is of importance, with a B350 the 2200/2400GE are at 24W at the main provided the RAM is set at 2400MHz, this should be adequate with a 200GE given its 2C and only 3 CUs.

https://www.hardware.fr/articles/973-21/gpu-integre-consommation-efficacite-graphique.html

 

[Shivansps]

Yeah is around those numbers it mostly depends on the PSU, and the margin of error.

idle:
2200G : 35W
200GE : 29W
E3000N(2650 wannabe) : 22W

Youtube:
2200G: 40W
200GE: 33W
E3000N: 29W

CPU:
2200G: 119W
200GE; 55W
E3000N: 31W

Gaming:
2200G: 91W
200GE: 58W
E3000N: 36W

CPU+IGP:
2200G: 143W
200GE: 75W
E3000N: 36W

This using a generic PSU. The idle power and low use power comsuption of the 200GE is really good compared to AM1.

 

[Abwx]

This using a generic PSU. The idle power and low use power comsuption of the 200GE is really good compared to AM1.

AM1 is better than this at idle provided an efficient PSU or 12-18V PSU + DC converter is used, with an Antec ISK 110 case it s about 15W idling and 30-35W loaded using a 2GHz APU.

That being said the 200GE is much better at 356 pts in CB R15 vs 170pts for an AM1 Athlon 5350.

https://www.hardware.fr/articles/921-6/consommation.html

 

[hojnikb]

What is the idle power.?

A 200GE is half a 2400G and at 10% lower frequency, it should barely drain 50% of the big brother s TDP.

It's still based on the same die, just cut down.

 

[Shivansps]

AM1 is better than this at idle provided an efficient PSU or 12-18V PSU + DC converter is used, with an Antec ISK 110 case it s about 15W idling and 30-35W loaded using a 2GHz APU.

That being said the 200GE is much better at 356 pts in CB R15 vs 170pts for an AM1 Athlon 5350.

https://www.hardware.fr/articles/921-6/consommation.html

Im confident that the 200GE can be undervolted and clocked to match AM1 power numbers and still provide better perf. That last thing is really easy.

 

[whm1974]

...and this is using a full-size ATX board with an X370 chipset. mATX and ITX boards tend to be a bit more efficient.

Anyway 80+ rating tend not to matter too much below 100W.

I don't think anyone is going to pair up an Athlon 200GE with an X370/X470 mainboard except for testing or as a placeholder. At which I'd question priorities, since you can get a 2200G + A320 board for the same as a X370/X470 board.

I would get the 2200G anyway unless I needed an APU with 35W TDP. Other then the last bit I just don't see the value of the 200GE.

 

[Abwx]

Apparently overclock on a A320 chipset is possible with the relevant AGESA :

https://www.computerbase.de/2018-12/amd-athlon-200ge-overclocking/

Posted by the reviewer in their forum :

https://www.computerbase.de/forum/attachments/1-1080-1228862015-jpg.734683/

 

[hojnikb]

Wait, wait, so A320 can now overclock too ? Thats _awesome_

 

[naukkis]

And seems that also PCI-express lanes can be unlocked.

 

[Shivansps]

AM1 is better than this at idle provided an efficient PSU or 12-18V PSU + DC converter is used, with an Antec ISK 110 case it s about 15W idling and 30-35W loaded using a 2GHz APU.

That being said the 200GE is much better at 356 pts in CB R15 vs 170pts for an AM1 Athlon 5350.

https://www.hardware.fr/articles/921-6/consommation.html

Still AMD has a LOT of work ahead in idle power, i just tested a G3930 and the damn thing beats Gigabyte E3000N idle/youtube power, i belive DDR4 vs DDR3 is helping a bit here. But this is done whiout even doing undervolting.

idle
200GE 29W
G3930 20W
E3000N 22W

Youtube
200GE 33W
G3930 27W
E3000N 29W

CPU
200GE 55W
G3930 40W
E3000N 31W

Dirt Rally
200GE 58W
G3930 49W
E3000N 36W

CPU-IGP
200GE 75W
G3930 49W
E3000N 36W

The diference is not big, and the 200GE has better performance to anything ratio but AMD has still a way to go in idle, and idle+media only power.

 

[Shivansps]

Apparently overclock on a A320 chipset is possible with the relevant AGESA :

https://www.computerbase.de/2018-12/amd-athlon-200ge-overclocking/

Posted by the reviewer in their forum :

https://www.computerbase.de/forum/attachments/1-1080-1228862015-jpg.734683/

Hang on, until i actually see it this remains a no-no to me. Neither MSI or Gigabyte A320 motherboards can overclock the 200GE regardless of bios version.

And no motherboard that i had seen so far can overclock the IGP.

 

[Abwx]

Still AMD has a LOT of work ahead in idle power, i just tested a G3930 and the damn thing beats Gigabyte E3000N idle/youtube power, i belive DDR4 vs DDR3 is helping a bit here. But this is done whiout even doing undervolting.

idle
200GE 29W
G3930 20W
E3000N 22W


The diference is not big, and the 200GE has better performance to anything ratio but AMD has still a way to go in idle, and idle+media only power.

Curious that Computerbase has 21W idle power, Golem.de is at the same level, and at stock...

https://www.computerbase.de/2018-09...eistungsaufnahme_mit_integr_und_diskreter_gpu

• Asus Prime B350/B360 Plus, 16 GByte DDR4, Seasonic 520W Titanium Fanless, Win10 x64 v1803, Intel 25.20.100.6323, RS 18.9.3 (Bild: Golem.de)

https://www.golem.de/news/athlon-200ge-im-test-celeron-und-pentium-abgehaengt-1810-136829-2.html

 

[LTC8K6]

System idle power will vary a lot.

Why would anyone use it for comparisons?

 

[Shivansps]

Curious that Computerbase has 21W idle power, Golem.de is at the same level, and at stock...

https://www.computerbase.de/2018-09...eistungsaufnahme_mit_integr_und_diskreter_gpu

• Asus Prime B350/B360 Plus, 16 GByte DDR4, Seasonic 520W Titanium Fanless, Win10 x64 v1803, Intel 25.20.100.6323, RS 18.9.3 (Bild: Golem.de)

https://www.golem.de/news/athlon-200ge-im-test-celeron-und-pentium-abgehaengt-1810-136829-2.html

Im using the exact same components in my comparison
Gigabye A320M-S2H for the 200GE
Gigabyte H110M-H for the G3930
Gigabyte E3000N
Same memory, same PSU same HDD.
On stabilized Win10 install.

Maybe the diference will be smaller with a good psu but it kinda defeats the purpouse.

 

[Abwx]

You have all the detailed numbers below, measured at the main and on the 12V ATX for idle/X264/DOTA, all set ups use the same PSU like in the other reviews, so if you find that AMD is not as good as in these reviews it means that the error is in your set up....

http://www.comptoir-hardware.com/articles/cpu-mobo-ram/37581-test-amd-athlon-200ge.html?start=5

I also linked Hardware.fr in a previous post with 24W at the main for a 2200G/2400G..

System idle power will vary a lot.

Why would anyone use it for comparisons?

It shouldnt vary a lot, main culprit is often an average PSU that has passive power factor correction wich can add 10W or so wasted power at low and very low output powers..

 

[Shivansps]

Well i just retested the 200GE and i got the same results.
My results are petty much in-line with computerbase.de results, its just they are running a more efficient PSU on both the intel and the 200GE. So all their numbers are lower.

You cant compare my results to theirs because of that. This is what happens when people buy cheap PSUs, see it more as a "worse case escenario" since no one test these crap PSUs.

 

[hojnikb]

power readers are also notoriously inaccurate. So unless you're using pro grade hw to measure, its really pointless.

 

[LTC8K6]

How are they measuring the power? Is everyone measuring it the same way? Are the methods accurate and repeatable?

 

[Shivansps]

How are they measuring the power? Is everyone measuring it the same way? Are the methods accurate and repeatable?

A standart plug power meter and yes, so far i can repeat every test result, same PSU, same youtube video, same CPU test, same game benchmark. But even if the test is not accurate compared to a review site it should be accurate enoght to compare my own results.
Belive it or not Windows 10 is the mayor problem since it likes to do stuff in the background.

But i did get some weird results, for example a I5-7400 petty much destroys the 200GE in power numbers in every way. And that was a result i was not expecting.
First i trought i was doing something wrong, then i took a look at this:

The 7400 seems to be very power efficient.