AMD Athlon X4 845 Review: A Perfect Budget CPU For Gaming And Multitasking

[4960X]

Have you guys seen this? the X4 845 has pretty good multi thread performance for $60. I don't know if I should get one for my son since the FM2+ socket is dead.

http://www.relaxedtech.com/reviews/amd/athlon-x4-845/1

 

[deasd]

Power consumption is nice, the VID is quite low for 1.048V@3.5-3.8Ghz if it was reading correctly. It's a pity they don't make HEDT with Excavator.

 

[jhu]

Looks fine if you're okay with the inability to upgrade the CPU in the future.

 

[frozentundra123456]

Reads more like advertising copy than an objective review. It *is* probably the best choice for a budget gaming system, unless one plays mostly older lightly threaded games.For general use, a pentium is a better choice because it has an igpu and better single threaded performance.

 

[DrMrLordX]

I question the value of the CPU vs the 860k unless power consumption is that big of a deal (not taking OC into account here). They spent too much time benching it vs the G4400 and not enough time benching it against the 860k and 850. Also that small L2 can hurt sometimes . . .

 

[cbn]

.For general use, a pentium is a better choice because it has an igpu and better single threaded performance.

Low end dGPUs are not expensive:

http://forums.anandtech.com/showpost.php?p=38246424&postcount=183

http://forums.anandtech.com/showpost.php?p=38250482&postcount=185

http://forums.anandtech.com/showpost.php?p=38254969&postcount=186 (An alternative to the second link if you don't mind rebates)

http://forums.anandtech.com/showpost.php?p=38254983&postcount=187

http://forums.anandtech.com/showpost.php?p=38254993&postcount=188



And as far as single thread goes Athlon x 4 845 is not that far behind Pentium G4400 according to Passmark:

Athlon x 4 845: 1775 Single thread CPU marks https://www.cpubenchmark.net/cpu.php?cpu=AMD+Athlon+X4+845&id=2721

Pentium G4400: 1864 Single thread CPU marks http://www.cpubenchmark.net/cpu.php?cpu=Intel+Pentium+G4400+@+3.30GHz

 

[24601]

Most games nowadays are designed explicitly with the minimum recommended target of the stock 2500k.

Getting anything less powerful than a 2500k for a build emphasizing PC gaming is a pure waste of money.

The recommendation nowadays should honestly be to get the 40-60 USD celeron that corresponds to the intel mainstream motherboard you plan on upgrading to the corresponding i5k or i7k in the future.

 

[cbn]

Most games nowadays are designed explicitly with the minimum recommended target of the stock 2500k.

Athlon x 4 845 is not that far off going by passmark:

Athlon x 4 845: 5548 CPU marks https://www.cpubenchmark.net/cpu.php?cpu=AMD+Athlon+X4+845&id=2721

i5 2500: 6223 CPU marks http://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i5-2500+@+3.30GHz

(I included the i5 2500 results rather than the i5 2500K results to remove OC results from boosting the score)

 

[24601]

Athlon x 4 845 is not that far off going by passmark:

Athlon x 4 845: 5548 CPU marks https://www.cpubenchmark.net/cpu.php?cpu=AMD+Athlon+X4+845&id=2721

i5 2500: 6223 CPU marks http://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i5-2500+@+3.30GHz

(I included the i5 2500 results rather than the i5 2500K results to remove OC results from boosting the score)

http://www.anandtech.com/bench/product/1707?vs=288

I find passmark to be too forgiving of bulldozer architecture compared to normal client workloads to be honest.

 

[cbn]

http://www.anandtech.com/bench/product/1707?vs=288

I find passmark to be too forgiving of bulldozer architecture compared to normal client workloads to be honest.

Too bad there aren't any gaming or browsing results in that comparison.

But yes, I can see how construction cores are lagging behind in benchmarks that use a lot of floating point.

 

[24601]

Too bad there aren't any gaming or browsing results in that comparison.

But yes, I can see how construction cores are lagging behind in benchmarks that use a lot of floating point.

Unfortunately, doing gaming tests for CPU comparison in CPU limited areas of games isn't something people like doing for low level cores.

Most tests that proport to be testing CPU limits in PC games are simply done horrendously.

 

[frozentundra123456]

Low end dGPUs are not expensive:

http://forums.anandtech.com/showpost.php?p=38246424&postcount=183

http://forums.anandtech.com/showpost.php?p=38250482&postcount=185

http://forums.anandtech.com/showpost.php?p=38254969&postcount=186 (An alternative to the second link if you don't mind rebates)

http://forums.anandtech.com/showpost.php?p=38254983&postcount=187

http://forums.anandtech.com/showpost.php?p=38254993&postcount=188



And as far as single thread goes Athlon x 4 845 is not that far behind Pentium G4400 according to Passmark:

Athlon x 4 845: 1775 Single thread CPU marks https://www.cpubenchmark.net/cpu.php?cpu=AMD+Athlon+X4+845&id=2721

Pentium G4400: 1864 Single thread CPU marks http://www.cpubenchmark.net/cpu.php?cpu=Intel+Pentium+G4400+@+3.30GHz

Even the cheapest low end video cards make it more expensive than the pentium and basically the same as the A8-7600. And I doubt either the 6450 or 710 are as competent as the A8-7600 igpu. The athlons with no igpu only make sense if pairing them with a decent (R7-250x GT740/750) level dgpu. Otherwise, I still contend you are better off with intel for OK igpu for day to day use plus good cpu, or AMD APU for light gaming.

 

[Dresdenboy]

Reads more like advertising copy than an objective review. It *is* probably the best choice for a budget gaming system, unless one plays mostly older lightly threaded games.For general use, a pentium is a better choice because it has an igpu and better single threaded performance.

I don't know if that iGPU is worth being considered here as on of the last c't magazine issues had some fps graphs showing lots of spikes and valleys for the Pentium and a smooth framerate at maybe 5% avg for a Godavari. I'll try to give more details here later.

 

[24601]

Unfortunately, doing gaming tests for CPU comparison in CPU limited areas of games isn't something people like doing for low level cores.

Most tests that proport to be testing CPU limits in PC games are simply done horrendously.

http://gamegpu.com/тест-gpu/rts-/-стратегии/

Looking at the comparison on the gamegpu tests, it looks like the FX-4100 (3.6 ghz 3.8ghz boost)[the proxy i'm using for 845] is about 60% as performant as the 2500k at stock in tests that aren't GPU limited.

 

[cbn]

Unfortunately, doing gaming tests for CPU comparison in CPU limited areas of games isn't something people like doing for low level cores.

Most tests that proport to be testing CPU limits in PC games are simply done horrendously.

Here are some gaming results from an old thread I have:

In this case, there is a huge difference in favor of the i5 2500K over the Athlon x 4 860K.

 

[24601]

Here are some gaming results from an url=http://forums.anandtech.com/showthread.php?t=2437733&highlight=]old thread[/url] I have:

img]http://pclab.pl/zdjecia/artykuly/chaostheory/2015/04/gtav/charts9/gtav_n_cpu2.png[/img]

img]http://pclab.pl/zdjecia/artykuly/chaostheory/2015/05/w3/charts6/w3n_ultra_cpu.png[/img]

In this case, there is a huge difference in favor of the i5 2500K over the Athlon x 4 860K.

Those results roughly concur with the gamegpu results at ~60% as performant as stock 2500k in CPU limited game tests.

And those were tests in games that are significantly multithreaded (With targets past 4 cores).

In games that aren't as well thread load-balanced the results only go down drastically from there.

 

[cbn]

Those results roughly concur with the gamegpu results at ~60% as performant as stock 2500k in CPU limited game tests.

And those were tests in games that are significantly multithreaded (With targets past 4 cores).

In games that aren't as well thread load-balanced the results only go down drastically from there.

Yep, I agree with you.

 

[cbn]

Low end dGPUs are not expensive:

http://forums.anandtech.com/showpost.php?p=38246424&postcount=183

http://forums.anandtech.com/showpost.php?p=38250482&postcount=185

http://forums.anandtech.com/showpost.php?p=38254969&postcount=186 (An alternative to the second link if you don't mind rebates)

http://forums.anandtech.com/showpost.php?p=38254983&postcount=187

http://forums.anandtech.com/showpost.php?p=38254993&postcount=188



And as far as single thread goes Athlon x 4 845 is not that far behind Pentium G4400 according to Passmark:

Athlon x 4 845: 1775 Single thread CPU marks https://www.cpubenchmark.net/cpu.php?cpu=AMD+Athlon+X4+845&id=2721

Pentium G4400: 1864 Single thread CPU marks http://www.cpubenchmark.net/cpu.php?cpu=Intel+Pentium+G4400+@+3.30GHz

Even the cheapest low end video cards make it more expensive than the pentium and basically the same as the A8-7600. And I doubt either the 6450 or 710 are as competent as the A8-7600 igpu. The athlons with no igpu only make sense if pairing them with a decent (R7-250x GT740/750) level dgpu. Otherwise, I still contend you are better off with intel for OK igpu for day to day use plus good cpu, or AMD APU for light gaming.

When comparing Athlon x 4 chips to desktop APUs we do have to remember the APU's CPU cores will throttle when the iGPU is under full load (2.7 Ghz for a 65W cTDP A8-7600) while with the Athlon x4 dGPU loads do not affect the cores. Under CPU demanding conditions this can result in a good difference in FPS.

Another thing to consider is how the memory on the desktop APU is configured. Single channel, dual channel, DDR3 1600 or something faster? With dGPU this memory config is less important.

So comparing Athlon x 4 plus low end dGPU vs. desktop APU is not so easy to do on a 1:1 basis.

And with low profile dGPU available in increments going from $10 AR Power Color HD5450 (which I use everyday with my E8500) to $20 AR HD6450 to $25AR/$30 GT 710 to $45 AR (+ $3 shipping) GT 730 GDDR5 and $50 AR R7 250 DDR3 a desktop APU can be hard to justify IMO. This, not to mention, the more powerful regular size cards at and above $50 price category.

 

[24601]

Unfortunately, due to etherium eating up all the former bitcoin/litecoin etc mining 7950, 7970, 280, 280x flood that was on the market (They were ~120 USD or less after the first wave), the best performance per dollar is currently ~200 USD R9 290s on ebay.

Everything below 200 USD is basically a gaming video card value dead zone currently.

An overclocked 2500k (or better) also pairs quite nicely with a R9 290, without the R9 290 being bottle-necked massively (Still pretty bottle-necked though), and the recommendation for budget gaming rig has a clear lower limit for value purposes.

http://www.ebay.com/sch/i.html?_odk...0+-"R9+280x"+-"R9-280"+-"R7+240"&_sacat=27386

 

[myocardia]

...

 

[myocardia]

...

 

[myocardia]

...

 

[cbn]

The Athlon 845 is $70, not $60, and it's far and away the worst thing that AMD currently sells. That goes double for gaming. The Athlon 880k is only $20 more expensive, and it will run circles around the 845, along with the 880k being unlocked, and coming with the best heatsink that AMD includes with any of their CPUs.

Athlon x 4 845 is $57.99 plus shipping at NCIX:

https://pcpartpicker.com/parts/cpu/#k=26

A8-7600 is $71.99 plus shipping at the same retailer.

Athlon x4 880K is $81.99 plus shipping from the same retailer.

Athlon x 4 860K is $69.99 plus $1.99 shipping from Newegg.

 

[AtenRa]

BTW, a $120 i3-6100 will utterly destroy the Athlon 880k in pretty much every game ever released.

With a GTX 1080 sure, with the GTX 750Ti it will not even have more than 1-2 fps difference in the vast majority of games.

 

[cbn]

Unfortunately, due to etherium eating up all the former bitcoin/litecoin etc mining 7950, 7970, 280, 280x flood that was on the market (They were ~120 USD or less after the first wave), the best performance per dollar is currently ~200 USD R9 290s on ebay.

Everything below 200 USD is basically a gaming video card value dead zone currently.

An overclocked 2500k (or better) also pairs quite nicely with a R9 290, without the R9 290 being bottle-necked massively (Still pretty bottle-necked though), and the recommendation for budget gaming rig has a clear lower limit for value purposes.

http://www.ebay.com/sch/i.html?_odk...0+-"R9+280x"+-"R9-280"+-"R7+240"&_sacat=27386

Performance per dollar is still good at the low end.

For example, take this new R9 270 (basically half a R9 290) at $104.99 AR free shipping:

http://www.newegg.com/Product/Product.aspx?Item=N82E16814131688