new diamond from FlanK3r - AMD FX-8120 first part

[muziqaz]

Leeghoofd, I don't think this is a comparison. Its just for reference. We all know that FX is quite a dog of a CPU. So I don't understand all the rage from you two. If you want some comparisons from all angles from all the CPUs in the world don't look at single thread in single forum from a guy who spares his 20minutes of free time to run some tests. His emphasis is on bd performance not on some Intel cpu which is on a market for some 9 months.

[FlanK3r]

right, 2600k 3.8 GHz in graphs is only for interest, because I had done before some results with it.

[liberato87]

@ flanker

what max frequency prime95 stable on air?
what vcore?
and what about temps?

I read you talking about 65c on noctua nh d 14 and Leeghoofd talking about 90c under amd stock liquid cooling at full rpm.

what about power consumption? you dont have a watt meter to test? thank you

[Leeghoofd]

Take note that I only have the early sample CPU and test Prime 95 with custom set 12K-12K... which seems pretty hot... 4.7ghz is 10-12°C colder depending on cooling used...

Recalibrated the CPU temps with an offset to let it idle around 28-30°C iso the 14°C coretemp read out. (this to be on a similar readout by the temp probe on the side of the CPU )

[FlanK3r]

4840 MHz 1.5V stable (but I wrote, I had no time, so about 20 minutes), Noctua NHD14, ambients about 20C (i dont like "hot" in room , max 22C ), temps about 65 C Coretemp (so about 75C real?).
90C under Antec Kuhler 920 is not possible...Namegt tested some chips 8150 under AIR in PRIME and had about 70-77C (I think, he used Frio)

[BeepBeep2]

4840 MHz 1.5V stable (but I wrote, I had no time, so about 20 minutes), Noctua NHD14, ambients about 20C (i dont like "hot" in room , max 22C ), temps about 65 C Coretemp (so about 75C real?).
90C under Antec Kuhler 920 is not possible...Namegt tested some chips 8150 under AIR in PRIME and had about 70-77C (I think, he used Frio)

About 80c real. Thats a bit hot for my tastes, and temps will continue to rise over time (probably to 68-70c in coretemp) so you will be around 85c real if you stress it long enough.

You feel safe using those kinds of volts?

[FlanK3r]

1.5V is safe for aircooling, it is in BD OC quide or SImon can tell you. Of course, all is based at temps etc...It can not be said simply 1.5V is safe for all FX pieces....

temps from Namegt -11x FX 8150 with the same voltage, 30mins PRIME with Cooler Master 612
eb5fa3b19ac2de531d1496c7910c57d8.jpg

[Leeghoofd]

And you honestly believe 15°C idle, temps below ambient The CPUs can handle 1.5volts, dunno if the boards can do this over a longer period...

[FlanK3r]

coretemp....but temps are as my..I had about 65C in Coretemp (so +10C in real).

[BeepBeep2]

I dont have a CPU to test with, but motherboard socket temp should be near correct if not 2-3c low.

I don't even know why you are recording CoreTemp numbers when the motherboard socket temp is accurate. My K-Type probe (Extech EasyView) on the side of IHS reads 3-4c lower than motherboard temp with Thuban, so its pretty accurate.

My Thuban was 13-16c off...

[FlanK3r]

as my Thuban..... 20C idle in Coretemp, max load 48 Coretemp with Ultra Kaze (so, real under 60). I think, the diference is between 10-12C. Maybe with new version of Coretemp could be fixed?

[konlaos]

@ flanker
what max memory freq. LinX maxmem stable ?
Thank you

[FlanK3r]

I dont tested RAM in LINX, because LINX dont work now with FX corectly. But seems, with good RAM kits u can get about 2300-2400 MHz stable (example 9-10-9)

[liberato87]

thanks for the reply flanK3r.
thank you for sharing these results.

[Zeus]

Is it me or do the 8120's oc at least as good or even better than the 8150's?