Athlon 3000G - let's all get one for an OC challenge?

[Zerileous]

sounds like fun!

 

[Aleslammer]

I just ordered one so I'm in!

 

[PolRoger]

I noticed that my local Micro Center was showing ~3 units available in stock this morning... I reserved one on-line and picked it up after work this afternoon.




Edit: Wow!... The cooler that comes with this chip is tiny!

AMD Wraith Prism on the left, Athlon 3000G cooler in the center and an old (socket A) Athlon XP Palomino core cooler on the right.




Edit #2:

I've been doing some testing on the 3000G... Since it is a Picasso (Zen+ core) I was kind of hoping for around a ~4.2GHz overclock but my sample seems to clock more like a 1st generation Ryzen.

Maybe my 3000G falls toward the lower end of the silicon bin/lottery? I'll just have to wait for more samples to be posted up to find out for sure.

The Stock 3.5Ghz VID on my sample is (~1.3v) and it seems to run up to ~3.8Ghz with 1.3v. The voltage wall comes up pretty fast and hard for higher clocks... With 3.9Ghz my sample needs ~1.3625v/1.375v vcore and for 4.0Ghz it needs ~1.425v/1.4375v.

I have my doubts about it reaching 4.1? and I think 4.2 is out. The chip seems voltage limited not so much temp limited. The IMC seems okay... I'll be testing that some more in the future.

Cinebench R20 4.0 3333C16:

 

[freakdiablo]

Edit #2:

I've been doing some testing on the 3000G... Since it is a Picasso (Zen+ core) I was kind of hoping for around a ~4.2GHz overclock but my sample seems to clock more like a 1st generation Ryzen.

Maybe my 3000G falls toward the lower end of the silicon bin/lottery? I'll just have to wait for more samples to be posted up to find out for sure.

The Stock 3.5Ghz VID on my sample is (~1.3v) and it seems to run up to ~3.8Ghz with 1.3v. The voltage wall comes up pretty fast and hard for higher clocks... With 3.9Ghz my sample needs ~1.3625v/1.375v vcore and for 4.0Ghz it needs ~1.425v/1.4375v.

I have my doubts about it reaching 4.1? and I think 4.2 is out. The chip seems voltage limited not so much temp limited. The IMC seems okay... I'll be testing that some more in the future.

Cinebench R20 4.0 3333C16:

Were the 2200G/2400Gs clocking to 4.2ish? Sorry, it's almost 11pm, taking almost all my brain power just to type this out and can't work the googles. I looked at them briefly last year for my HTPC, and my understanding is that with the G cpus, since the onboard GPU cores share a power rail with the CPU cores, they really don't like to boost to their rated upper clock under real world cases due to the power constraints - would make sense if they don't OC as well either.

 

[trents]

Yeah, I was only getting about 3.8 on my sample but I was using a very low end mATX motherboard with no heatsinks. I was hoping those with a better board would see better results. I was disappointed. I just put it back to stock and will use it in a customer build.

 

[notarat]

Never mind the new threadrippers, we have this baby! Maybe it is just me, but I think it is cheap enough for many to buy assuming they can drop it into an existing AM4 system, and we can have an OC comp around it on an equal-CPU footing. Anyone interested?

I dunno about getting one to overclock but I've been looking into building a larger NAS than my QNAP (Maybe a FreeNAS build) and this is an intriguing CPU. It seems like it would be far better for this type of use than my backup system and I'm sure I could pick up a good board to pair it with.

 

[Acidburn]

Every stinking time! Do you have any idea why this is the case, ED? It's as if AMD won't let anybody outside of the CPU team see the design until it's released, THEN the APU designers can start their work. Ideally it would happen semi-concurrently, as it does with Intel, but I don't believe AMD has ever displayed this in the history of their integrated GPU/CPU designs.

It's frustrating when spec'ing out an inexpensive system for a customer. Going AMD always means you're a generation behind on CPU.

Have you been living under a rock since 2005?