AMD Athlon X4 860K review @ pcpop

[AtenRa]

How fast do you think the average Athlon x4 860K will go on the stock cooler?

Using the 7700K as example, I believe 4.2GHz could be fine with stock cooler. It was perfectly fine at 4GHz and it had more room to go.

 

[AtenRa]

That's a lot of game and gaming synthetic benchmark hits for Kaveri, though. Such throttling should also occur on and CPU + GPU compute benchmark.

That leaves open the question of how long the 7700k was able to stay at full turbo. I'd still like to see 4.5 ghz 860k vs. 5 ghz 760k, no throttling, no turbo, etc. Estimates based on existing benchmarks indicate that they'd be about the same, which sort-of defeats the purpose of the 860k for overclockers.

Using a dGPU with Kaveri doesnt revert the CPU to lower P-States because the CPU can use all the TDP by itself. So with dGPU the Kaveri maintain the highest base frequency always. That is, A10-7700K in Cinebench can sustain 3.4GHz continuously.

Also i didnt see any throttling at 4.3GHz and the CPU was always working at that frequency(turbo off) during the Cinebench run.

 

[burninatortech4]

Would I get better performance upgrading my 760k to a 860k or my 260k to a 285? What will get me the best bang for the buck.

Gaming at 1080p

Athlon X4 760k
Gigabyte A88x Wifi
Gigabyte 260x
8gb Radeon 2133mhz

 

[Insert_Nickname]

Would I get better performance upgrading my 760k to a 860k or my 260k to a 285? What will get me the best bang for the buck.

Gaming at 1080p

Athlon X4 760k
Gigabyte A88x Wifi
Gigabyte 260x
8gb Radeon 2133mhz

Difficult to say. I'd get the 285 before upgrading the CPU however. You could buy a cheap aftermarket cooler like the Hyper 212. You should be able to get that 760K to 6800K-level (4.1/4.4GHz) easily. The 285 will still be CPU limited but it'll properly be OK for 1080p.

 

[FlanK3r]

we want more 860K reviews . Hope Anandtech can do it .

 

[DrMrLordX]

I would like to see them review the 860k as well, especially as a head-to-head vs. the 760k. They cost very nearly the same on the open market, so why not? Throw a G3258 and an i3 into the mix to round it out a bit.

 

[MiddleOfTheRoad]

Would I get better performance upgrading my 760k to a 860k or my 260k to a 285? What will get me the best bang for the buck.

Gaming at 1080p

Athlon X4 760k
Gigabyte A88x Wifi
Gigabyte 260x
8gb Radeon 2133mhz

I just don't think anyone knows enough at this point. I believe Richlands can overclock higher than Kaveris -- so much of the IPC improvements in Kaveri might wash away if the Richlands are able to clock considerably higher.

 

[Alexvrb]

I would like to see them review the 860k as well, especially as a head-to-head vs. the 760k. They cost very nearly the same on the open market, so why not? Throw a G3258 and an i3 into the mix to round it out a bit.

Just as long as they test with modern titles!

I'd like to see how well they do with only modest (cheap!) air cooling. Possibly even just stock coolers. After all, these are budget chips for budget rigs. I'd bet the 760K doesn't really shine unless you beef up the cooling substantially and crank up the voltage. In that case, the 860k may really eclipse its predecessor.

 

[Abwx]

Of the two retail 7850K APUs bought in January by hardware.fr the worst of the two did 4.4 at 98W and 4.3 at 84W under Prime 95, in games or intensive usual softs thoses wattages are 20% lower, at 4.3 a Kaveri should be as good in games as a Richland at 4.6 but with only 75W TDP.

 

[DrMrLordX]

Just as long as they test with modern titles!

I'd like to see how well they do with only modest (cheap!) air cooling. Possibly even just stock coolers. After all, these are budget chips for budget rigs. I'd bet the 760K doesn't really shine unless you beef up the cooling substantially and crank up the voltage. In that case, the 860k may really eclipse its predecessor.

Agreed, though it would be nice to see Cutress give it a shot just to see how it works out for them.

But yeah, getting an 860k to 4.3-4.4 ghz is probaby a lot easier than hitting 4.7-4.8 ghz on a 760k.

Of the two retail 7850K APUs bought in January by hardware.fr the worst of the two did 4.4 at 98W and 4.3 at 84W under Prime 95, in games or intensive usual softs thoses wattages are 20% lower, at 4.3 a Kaveri should be as good in games as a Richland at 4.6 but with only 75W TDP.

I haven't seen many reports on the 860k yet, but it seems to hit 4.4 ghz with weak cooling.

 

[Alexvrb]

Of the two retail 7850K APUs bought in January by hardware.fr the worst of the two did 4.4 at 98W and 4.3 at 84W under Prime 95, in games or intensive usual softs thoses wattages are 20% lower, at 4.3 a Kaveri should be as good in games as a Richland at 4.6 but with only 75W TDP.

That sounds about right. It's not worth it to upgrade if you've already got a 760K. Maybe if they release a Carrizo-based Athlon in the future, it might give existing FM2+ users a reason to upgrade.

However, I think if you're building a new budget FM2+ system, the 860K is the way to go. It's easier to get an 860K up to low 4's and match a mid-4's Richland with lower power/cooling requirements.

 

[Kalessian]

I would really like to see a full review of this chip, too. Would the IPC improvement of steamroller make it faster clock-for-clock than the L3-cache equipped fx-4300?

I'm thinking at the least, the lower power requirement of steamroller would make overclocking the 860k a cheaper operation than overclocking a 4300 or 6300?

 

[Insert_Nickname]

I'm thinking at the least, the lower power requirement of steamroller would make overclocking the 860k a cheaper operation than overclocking a 4300 or 6300?

Undoubtedly. You get a not insignificant saving too from the FM2+ platform itself since it has no northbridge (integrated into the APU). Plus you get a much more modern platform. AM3(+) is pushing 5 years after all.

 

[MiddleOfTheRoad]

I would really like to see a full review of this chip, too. Would the IPC improvement of steamroller make it faster clock-for-clock than the L3-cache equipped fx-4300?

That is really what I'd like to see. My guess is that older games will still be faster due to the L3 on the FX, but newer multithread games should be faster on the 860k.

 

[Abwx]

At stock a A10 7850 is on par in games and 4% faster in applications than a stock FX4300, obviously the cache has a few % influence according to this difference, the turbo is negligible for the 7850 in games so a 860K should have no problem outperforming a FX4300, not counting a lower TDP.

 

[SPBHM]

considering richland vs vishera, for gaming the 4300 should be faster or around the same at the same clock as Kaveri for most games,

if you look here http://pclab.pl/art57691-3.html stock 4300 can keep up with 4.5GHz Richland easily.

fm2+ platform is nicer, but it lacks upgrade options for now, anyway, I think in this comparison (860K vs 4300) price would be the most important thing.

 

[Abwx]

considering richland vs vishera, for gaming the 4300 should be faster or around the same at the same clock as Kaveri for most games,

if you look here http://pclab.pl/art57691-3.html stock 4300 can keep up with 4.5GHz Richland easily.

fm2+ platform is nicer, but it lacks upgrade options for now, anyway, I think in this comparison (860K vs 4300) price would be the most important thing.

Kaveri is not Richland, it is 5.8% faster in games clock/clock according to Hfr, the comparison i posted is right from their benches comparing Kaveri to FXs.

 

[SPBHM]

Kaveri is not Richland, it is 5.8% faster in games clock/clock according to Hfr, the comparison i posted is right from their benches comparing Kaveri to FXs.

finding a comparison between richland and kaveri is easy, and the clock difference (4300 vs 4.5GHz richland) is much higher than 6%, so...

 

[pw257008]

finding a comparison between richland and kaveri is easy, and the clock difference (4300 vs 4.5GHz richland) is much higher than 6%, so...

?? 4.3 vs 4.5 isn't more than a 6% difference

 

[Kalessian]

He means the FX-4300, which is 3.8ghz->4.0 turbo, not 4.3ghz.

And the pclab and hfr results are not consistent with one another. Hfr has 6800k beating the 4300, pclab has it the other way around (4300 beats 4.5ghz 760k). Could the pclab 4.5ghz results have some throttling going on? Or slower ram maybe, idk.

It's really too bad 860k won't do 4.8+ ghz

 

[pw257008]

He means the FX-4300, which is 3.8ghz->4.0 turbo, not 4.3ghz.

And the pclab and hfr results are not consistent with one another. Hfr has 6800k beating the 4300, pclab has it the other way around. Could the pclab 4.5ghz results have some throttling going on? Or slower ram maybe, idk.

It's really too bad 860k won't do ~4.8+ ghz

whoops, figured it was a reference to Kaveri vs Richland frequencies.

 

[Lyfer]

WTB an Anand review on this.

 

[chrisjames61]

Difficult to say. I'd get the 285 before upgrading the CPU however. You could buy a cheap aftermarket cooler like the Hyper 212. You should be able to get that 760K to 6800K-level (4.1/4.4GHz) easily. The 285 will still be CPU limited but it'll properly be OK for 1080p.

I run my A10-5800K all day at 4.6 GHz with Prime95 running in the background. Thermal margin stays at 25C according to AMD overdrive. My case is a C70 with 5 fans. My cooler is low end. A Hyper 212 EVO. My board is a Crossblade Ranger. The air coming out the rear exhaust is coolish. I think the 750K, 760K and 860K have very little differentiation between them. If I were to buy one it would be the 860K. But I actually can get A10-7850's for $90 open box at all the Microcenters in NYC.

 

[Kalessian]

http://www.overclock3d.net/reviews/cpu_mainboard/amd_athlon_860k_black_edition_cpu_review/1

Sadly no power consumption or overclocked 760k

 

[DrMrLordX]

Wait. A stock vcore of 1.5v? That board had to be overvolting out of the box. There is no way in the Nine Hells that an 860k has a stock vcore of 1.5v.