[Techspot] Athlon x4 860K vs Pentium G3258 deathmatch

[escrow4]

http://www.techspot.com/review/1017-best-budget-gaming-cpu/page7.html

Verily:

"Having watched the Pentium G3258 and Athlon X4 860K trade blows in no less than 20 games using two mid-range GPUs, the G3258 is undeniably faster and it's the processor you want if you're building our Budget Box with the intentions of gaming."

"For a lot of the games, such as Tomb Raider, Sleeping Dogs, Battlefield Hardline, Watch_Dogs, Hitman: Absolution and Dragon Age: Inquisition, performance was so close it really didn't matter, so you could happily go either way. Sadly though, the Athlon just doesn't deliver the same consistency as the Pentium G3258, being much slower at times."

"The problem for AMD, which we have seen time and time again, is core efficiency. Having more cores available isn't much use if they are slower. Moreover, just because a game can use four cores (and all the games we tested could) doesn't mean a dual-core will be inferior if two cores are working harder than four."

"Touching on that a little more, the Core i3 isn't really an ideal upgrade. Although it does support four threads thanks to Hyper-Threading, the highest clocked model runs at just 3.8GHz with no Turbo boost, so it won't be much of an upgrade from an overclocked G3258, if at all."

***Puts on flamesuit*** :whiste:

 

[cbn]

When I did informal testing of 4.3 Ghz Pentium G3258 vs stock speed Athlon x 4 860K in Batlefield 4 using R7 250X, the OC Pentium was about 10 FPS faster than the Athlon in the leveled building scene in 60-64 player Seige of Shanghai.

http://forums.anandtech.com/showpost.php?p=37322379&postcount=7

In my testing of the following leveled building scene in BF4:

https://www.youtube.com/watch?v=ZGiHqR61O4I&t=278

My overclocked G3258 @ 4.3 Ghz actually beat my stock clocked Athlon x4 860K with 60 to 64 players on the map using the same R7 250X card and detail settings.

I made numerous runs over and over again during the span of several games using both processors. I would say the OC G3258 was probably 10 FPS faster on average in that scene.

And here is the record of the frame rates I got for OC G3258 (for reference):

http://forums.anandtech.com/showpost.php?p=37279088&postcount=173

Regarding the informal BF4 64 player testing I mentioned back in post #164, Here is one of the areas I am using for comparisons:

https://www.youtube.com/watch?v=ZGiHqR61O4I&t=278

Walking thru the leveled building in Seige of Shanghai the Xbox One (which runs the game at 1280 x 720) stays mostly in the 40's/low 50's for FPS, but does drop as low as 30 FPS at one point.

With my G3258 overclocked to 4.3 Ghz, MSI Z97 U3 Plus, 2 x 4GB DDR3 1600 (which is actually faster than a Non-Z board would run the RAM), R7 250X running 1280 x 720 High setting (using Mantle API) I am getting 45 to 65 FPS ~99% of the time in that very same area with 60 to 64 players on the map. This over the span of several games. FPS did drop to a low of 31 FPS, but this was during an explosion and heavy shooting among multiple players.

 

[VirtualLarry]

For a lot of the games, such as Tomb Raider, Sleeping Dogs, Battlefield Hardline, Watch_Dogs, Hitman: Absolution and Dragon Age: Inquisition, performance was so close it really didn't matter, so you could happily go either way. Sadly though, the Athlon just doesn't deliver the same consistency as the Pentium G3258, being much slower at times.

No minimums? No frame rate over time graphs like [H]? Fluff piece.

 

[jj109]

No minimums? No frame rate over time graphs like [H]? Fluff piece.

TechSpot's tech reviews have some issues.

I still remember when they picked a heavy DDOF scripted cutscene in BF4 to do a CPU benchmark. All of the CPUs except for the slowest dual cores were tied... I wonder why?

 

[cbn]

No minimums? No frame rate over time graphs like [H]? Fluff piece.

Back when Pentium G3258 first came out, Tom's did a frame time comparison of stock and OC Athlon x 4 750K vs stock and OC G3258 . OC Pentium actually won most of the games (vs. OC Athlon x4 750K).

http://forums.anandtech.com/showpost.php?p=37175841&postcount=65

No one brings terrible frame times of pentium? Very strange to see people give any weight to average fps numbers these days.

Pentium OC G3258 actually won 4 out of 7 games (in terms of frame time variance) against OC Athlon x4 750K in the Tom's test:

http://forums.anandtech.com/showpost.php?p=36517046&postcount=17

http://forums.anandtech.com/showpost.php?p=36517050&postcount=18

http://forums.anandtech.com/showpost.php?p=36517071&postcount=19 (<---Here is where I tallied results)

,...but one of the newer games Thief there was a large discrepancy in frame time variance favoring OC Athlon x4 750K.

http://forums.anandtech.com/showpost.php?p=36873247&postcount=30

Unfortunately Thief was very new at the time (it was the newest game in the test I believe). My conclusion (after comparing Thief to the oldest game to scale quad core in the test, Metro Last Light) was that is was probably a driver issue.

http://forums.anandtech.com/showpost.php?p=36875563&postcount=49

However, these days with Athlon x4 860K being more powerful than Athlon X4 750K and newer games using more cores (or not optimized for dual core) the situation could be very different (depending on the games tested).

 

[mikeymikec]

No minimums? No frame rate over time graphs like [H]? Fluff piece.

Yeah, I found it amusing when they said that 1FPS more may as well be a tie, but 2FPS more constitutes a lead.

 

[TheELF]

No minimums? No frame rate over time graphs like [H]? Fluff piece.

Doesn't matter if your 10% (or whatever % ) faster,if you want to have the same smoothness as the quad you can slow down by 10% ,just do a fps limit and you are set.
But if you are ~50% slower on the quad you can't do anything to gain speed.

 

[mikeymikec]

Doesn't matter if your 10% (or whatever % ) faster,if you want to have the same smoothness as the quad you can slow down by 10% ,just do a fps limit and you are set.
But if you are ~50% slower on the quad you can't do anything to gain speed.

I think VL's point was that their results would likely be affected by say a single frame rate spike or more importantly that product x may give a far better average with fewer spikes while product y gives performance spikes but then flounders as soon as anything of note appears on the screen.

 

[TheELF]

product y gives performance spikes but then flounders as soon as anything of note appears on the screen.

I think that this becomes very clear by the different benchmarks,you can see that as soon as something,an effect or scene, that requires single speed will come up the athlon will fall far far behind.
Avg fps are ok for benchmarks as long as enough different games are shown,then you can tell how a cpu will behave in multi heavy situations as well as in single heavy situations,even within the same game.

 

[NTMBK]

No measurement of frame time consistency, article is useless. Ignore the metric which looks worst for the Pentium, and surprisingly the Pentium looks good.

 

[Shehriazad]

Frame time consistency especially in titles like GTAV and Witcher 3 (which are...you know...very popular new games) would be really nice to see.

Not to mention that I keep seeing that while the Pentium has higher averages on benchmarks for those 2 games...their lows are lower.

I'm not fighting the fact that the Pentium is better in IPC or games that only really fire up 2 cores for the work....but anyone seriously thinking they should be buying a dual core in 2015 if they want to play new games without frustration...is clinically insane to me.

Not to mention there have been a multitude of games already where people had to wait for patch or crack fixes so the games actually just properly work with dual cores...but you don't see such a fact being mentioned at all...ever.

Yes yes...all is well in dual core land.


Also...what the heck are they doing in those benchmarks. I happened to have JUST played Metro and Metro Last Light redux(steam sales ahoy)...neither one would drop below 30 fps....EVER with all the CPU intensive settings cranked up.
Meanwhile for Witcher 3 if you search for benchmarks you can also see that something is terribly off here.

The "highs" of the Pentium are known to be a tad higher...but the lows go down all the way to like 25fps ish....which for most gamers is unacceptable...meanwhile the 860K never goes below 30.

Yet the article happily proclaims that the pentium got the better consistency? Like hell it does.

Yes...the pentium will manage higher max fps...it will also have worse frame time, lower lows in new titles and you have to deal with occasional waiting times while dual cores not working properly has to be fixed. Not to mention since you HAVE to OC the pentium to really make it worth your time...your choice of boards in the lower sector is rather limited as not all will let you OC....meanwhile camp AMD just doesn't give a damn. OC all you want on that 30$ board. (Not about the price, it's about that you can use ANY board).


EDIT:
Just look at this:

https://www.youtube.com/watch?v=EjE0BU5egFc (I chose this one because it's a similar setup with a GTX 960)

We aren't even talking ultra here....in the first 2 minutes he already manages to go down all the way to 21 fps lows!!!! Granted, it's at stock speed...but without stock he's still likely to fall down to 25, like professional benchmarks have shown...with stuttering galore. Then he adjusts to medium and while fps is now more stable...you can still see how it's a freakin' stutter fest.

 

[mikeymikec]

I think that this becomes very clear by the different benchmarks,you can see that as soon as something,an effect or scene, that requires single speed will come up the athlon will fall far far behind.

Every gaming engine has its quirks and is reasonably likely to be designed with a different purpose in mind to another, so I don't see how you can make such a claim, it's almost as unhelpful as saying that one can surmise the quality of bicycle tyres by the top speeds being reported. However, in most games there is a scenario that causes frame rate spikes so it's much easier to cheat prospective customers into thinking that your product provides them with the performance they want (which I think I can safely say that most people want consistent FPS above all, followed by high FPS).

Avg fps are ok for benchmarks as long as enough different games are shown,then you can tell how a cpu will behave in multi heavy situations as well as in single heavy situations,even within the same game.

"Testing with as many different scenarios/games" is important in almost every product review (when talking about PC hardware performance).

 

[SPBHM]

EDIT:
Just look at this:

https://www.youtube.com/watch?v=EjE0BU5egFc (I chose this one because it's a similar setup with a GTX 960)

We aren't even talking ultra here....in the first 2 minutes he already manages to go down all the way to 21 fps lows!!!! Granted, it's at stock speed...but without stock he's still likely to fall down to 25, like professional benchmarks have shown...with stuttering galore. Then he adjusts to medium and while fps is now more stable...you can still see how it's a freakin' stutter fest.

you have to consider if the video capture is not affecting the result, even with shadowplay, with only 2 threads at 100% load it could be having an effect

a big portion of the video is showing the GPU limiting performance because of the high settings used,
also strangely he left hairworks on even for medium...
and I think there are considerably more demanding parts of this game.

in any case, this video shows having this kind of CPU for a $200 VGA can deliver some pretty bad results, it's best to balance more the PC specs

even an i3 or 6300 would be a big upgrade for witcher 3

 

[wright.pwa]

EDIT:
Just look at this:

https://www.youtube.com/watch?v=EjE0BU5egFc (I chose this one because it's a similar setup with a GTX 960)

We aren't even talking ultra here....in the first 2 minutes he already manages to go down all the way to 21 fps lows!!!! Granted, it's at stock speed...but without stock he's still likely to fall down to 25, like professional benchmarks have shown...with stuttering galore. Then he adjusts to medium and while fps is now more stable...you can still see how it's a freakin' stutter fest.

Except the fact that he is using a dual-core processor has nothing to do with those slow downs, that's all hair works my friend.

TechSpot's tech reviews have some issues.

I still remember when they picked a heavy DDOF scripted cutscene in BF4 to do a CPU benchmark. All of the CPUs except for the slowest dual cores were tied... I wonder why?

Yeah that's a serious issue LOL. If I recall correctly that was primarily a GPU performance article...

 

[sm625]

Why do they keep publishing articles like these without any data?

 

[Shehriazad]

Except the fact that he is using a dual-core processor has nothing to do with those slow downs, that's all hair works my friend.

Yea, right...the CPU sitting at 100% nearly all the time must mean that it's the GPU totally slowing it down *cough*



Either way.... Like I said...the 860K is far from a perfect piece of hardware...but anyone recommending a dual core without hyperthreading in 2015 is insane....2 thread support will slowly drop out of games...we already see it happening. (Especially with console hardware being pretty much just PC type multicores...and a big number of games being a port of that)

No matter how awesome DX12 is gonna be and all that stuff..I just don't understand how ANYONE would recommend a 2 thread CPU for anything gaming related while looking at today and the future...unless of course you like having to deal with more and more games popping up that DON'T support your hardware right out of the box and having to wait for fixes.


I'll recommend that pentium to anyone that just games casual stuff and stuff that is from the past (Like Skyrim...the Pentium will be tons better there)...but anyone else I would recommend to go at least i3 if they have the extra cash.

 

[TeknoBug]

An OC'd G3258 has one of the fastest IPC's you can find, but it's dual core and only dual core so no HT like i3 and it'll have its limitations. A G3258 is totally a CPU I'd be using when building a PC for my mother or grandmother or for a guest PC in my house, or such otherwise an i3 would be my choice for lightweight gaming and office use. The Athlon X4... I just don't even consider.

 

[AtenRa]

Assuming they used the in-game benchmarks i got the following with my Core i3 4330/A10-7850K and HD7950 1GHz (cat 15.5beta), win 8.1 64bit

BIOSHOCK Infinity : 1080p Ultra DOF
Core i3 4330 min= 30,02fps / Avg= 80,02fps
A10-7850K min= 25,30fps / 73,62fps

Metro Last Light Redux : 1080p Very High AF 16x Tessellation High
Core i3 4330 min= 26,17ps / Avg= 53,12fps
A10-7850K min= 14,18 / 52,08fps

Thief (Mantle) : 1080p High AF 16x
Core i3 4330 min= 42,5ps / Avg= 66,5fps
A10-7850K min= 41,7 /58,4fps

Dragon Age Inquisition : 1080p Ultra 2x MSAA HBAO (Mantle)
Core i3 4330 min= 33,1ps / Avg= 39,1fps
A10-7850K min= 34,1 /38,9fps

Ill post a full review soon.

 

[sm625]

Metro Last Light Redux : 1080p Very High AF 16x Tessellation High
Core i3 4330 min= 26,17ps / Avg= 53,12fps
A10-7850K min= 14,18 / 52,08fps

Ouch that is brutal. I sincerely hope AMD is profiling the heck out of that game for their Zen development, because those numbers are atrocious. It is hard to imagine how anything in the cpu could possibly introduce the 30+ milliseconds of delay it takes to cause the framerate to drop like that relative to the intel chip. Does that not indicate that caches are being completely flushed out multiple times when they should not be? Like the core is thrashing between multiple threads not being scheduled properly...

 

[B-Riz]

This "review" gave me a good LOL.

If someone I knew was wanting to get into PC Gaming cheap, I would hit the used market first, then round out with a new vid card and ssd.

Just for arguments sake, i5 2500k on CL for $120 OBO.
http://toledo.craigslist.org/sop/5063601694.html

Looked for FX-8's, nothing listed currently on CL.

Granted, this works for anywhere with a good used market.

Do people forget about used parts that still kick butt?

 

[DrMrLordX]

Ouch that is brutal. I sincerely hope AMD is profiling the heck out of that game for their Zen development, because those numbers are atrocious. It is hard to imagine how anything in the cpu could possibly introduce the 30+ milliseconds of delay it takes to cause the framerate to drop like that relative to the intel chip. Does that not indicate that caches are being completely flushed out multiple times when they should not be? Like the core is thrashing between multiple threads not being scheduled properly...

It is possible that there are cache flushes happening on both CPUs. Intel chips seem to handle such events with relative aplomb. The same thing ruins AMD performance, at least on anything Construction-core based.

Intel CPUs with HT seem to handle such events even better.

 

[Flapdrol1337]

Pentium has run every game I tried very well.

If you're not gpu limited you have to sometimes use an fps cap though. Otherwise the high priority game threads hog both cores until the background/low priority threads can't be delayed anymore, and you get noticable stutters, or even have the game stall for a second.

Anyway, I have a gtx 670 so I don't often have to do that.

 

[VirtualLarry]

Pentium has run every game I tried very well.

If you're not gpu limited you have to sometimes use an fps cap though. Otherwise the high priority game threads hog both cores until the background/low priority threads can't be delayed anymore, and you get noticable stutters, or even have the game stall for a second.

That's interesting. That's actually because of how the NT kernel scheduler works.

 

[Fox5]

It is possible that there are cache flushes happening on both CPUs. Intel chips seem to handle such events with relative aplomb. The same thing ruins AMD performance, at least on anything Construction-core based.

Intel CPUs with HT seem to handle such events even better.

This is true in general for AMD vs Intel, cache flushes are more penalizing on AMD.
Just look at any JIT languages, like Javascript and Java. AMD can massively (even more so than usual) lose to Intel in those situations.

 

[scannall]

No measurement of frame time consistency, article is useless. Ignore the metric which looks worst for the Pentium, and surprisingly the Pentium looks good.

Overclock the Pentium, and leaving the 860k at stock seems rather biased as well. It's a decent overclocker after all.