should I abandon the AMD ship? (Updated)

[bononos]

Another crusader in the AMD benchmark cherry picking game.

Actually I don't consider that a cherry picked benchmark since it show the stock 8250 scoring 79.1 and the i5-4670 scoring 100. Did I miss something.

 

[frozentundra123456]

Actually I don't consider that a cherry picked benchmark since it show the stock 8250 scoring 79.1 and the i5-4670 scoring 100. Did I miss something.

I dont know if it is cherry picked or not, although there are plenty of those in this thread from the usual suspects.

In any case, I went back and looked at the actual framerates, and the way they calculated, using the higher Intel number as the baseline minimizes the differences. If you use the AMD 8350 as the baseline and calculate the improvement going to the 4670, the average improvement is about 30%. Either way is mathematically correct, but intuitively if one is looking at the improvement going to intel, then I think the lower FX number should be used as the baseline (ie. the denominator in the ratio).

Edit: in regards to the original question, with a 6300 at 4.5, I dont really see the 8350 as a worthwhile upgrade. It would only be faster in a few heavily multithreaded games (and some productivity apps if the op uses those). I would either go 4790k or stick with the 6300.

 

[Arachnotronic]

@ OP

Figure out what programs you are likely to run and look at the performance of the AMD and the Intel chip in those programs.

If one chip performs significantly better in the applications you want to run, then go with that chip.

 

[Abwx]

Actually I don't consider that a cherry picked benchmark since it show the stock 8250 scoring 79.1 and the i5-4670 scoring 100. Did I miss something.

Since it is the average of all games tested it s hardly a cherry picked number, so if i had used a single game it would have been picked but the average of all games is also cherry picked, i guess that some people are really short of arguments because the numbers do not comply with what they would hope for.

 

[Abwx]

If you use the AMD 8350 as the baseline and calculate the improvement going to the 4670, the average improvement is about 30%.

Let see what it does, using the 8350 as baseline will yield : 100/79.1 = 1.264%, wich say that the i5 is 26.4% faster in this graph.

Then using the i5 as base line : 79.1/100 = 0.791, this imply that the FX is 20.9% slower, you will notice that the latter computation is not as straightforward and that it s easier to use the first method so i dont know where is the problem unless you are afraid that one use the second method that would not yield as impressive apparent numbers.?..

Anyway thank for the math clarification, and of course the rounding of 1.264 to 1.30, i guess that lack of precision is not always that bad after all....

 

[davie jambo]

You'll get about 10 more frames in gaming by changing everything to haswell , for $400 it's not worth it

 

[TeknoBug]

I've stuck to AMD for years since the K6-2 days (had an AMD 486 too but it died) however I feel the same way, why should I continue to use AMD when Intel has the hold on performance.

If you move from an FX 8350 to an i7 3770K and not notice a difference, something's wrong, a stock 3.5GHz 3770K can keep up with a 4.5GHz OC'd 8350 so it's a big deal. And lots of games perform significantly better having higher average framerates with Intel (Planetside 2 for eg strugged a lot on AMD with frames dropping to low teens while Intel holding up around 35-40 frames in the same scale fight).

 

[KingFatty]

It still comes down to whether it's worth $300 for this upgrade. Looking at the above charts, you could be talking like 10-20 FPS difference (like 20 FPS to 35 FPS). I wouldn't spend $300 for that small of an increase, seeing how new CPUs are coming out, and a new motherboard is a big investment - why not go DDR4 too?

The discussion might swerve into that realm of which is better when you are buying a new system. But here the situation is different, because OP already has a system. So would a drop-in CPU upgrade make sense? No. Would a new motherboard+CPU make sense? Based on the gains, I wouldn't do it.

Also everything is being limited by the video card(s). So if you see 1080p benchmarks that show 65 FPS on AMD, or 130 FPS on Intel, it doesn't matter because your GPU may only be able to get you 45 FPS so either CPU would seem just as fast when you actually play the game at eyefinity resolutions.

 

[Phynaz]

But you have never provided any links, not even one review to back up what you are saying.

In order for you to say the following, it means you have seen that performance and you have the links to provide us with evidence to back up your claims.



If you cant provide evidence for the above, then we should take your comment as bate and thread crapping.

I've provided the links many times over and you know it. Report me if you think I'm thread crapping.

 

[Phynaz]

OP:

http://www.anandtech.com/bench/product/697?vs=836

http://techreport.com/review/23246/inside-the-second-gaming-performance-with-today-cpus

http://techreport.com/review/26996/amd-fx-8370e-processor-reviewed/4

http://techreport.com/review/25293/intel-core-i7-4960x-processor-reviewed/4

 

[Erenhardt]

OP:

http://techreport.com/review/25293/intel-core-i7-4960x-processor-reviewed/4

Thats a nice performance from FX line!
FX6350 better than i5? What a steal for $130

I've seen the fx6300 for 80EUR recently.

 

[Phynaz]

Thats a nice performance from FX line!
FX6350 better than i5? What a steal for $130

Oh look, another cherry picker, who failed to notice that's a LAST GENERATION i5. Wonder why you didn't post any of the other graphs in the article. :thumbsdown:

 

[Erenhardt]

Oh look, another cherry picker, who failed to notice that's a LAST GENERATION i5. Wonder why you didn't post any of the other graphs in the article. :thumbsdown:

Because skyrim is not relevant anymore.
And comparing gaming performance in anandtech bench tool is plain dishonest. The 3 patches and 6 drivers updates between the sampled data makes a difference.

 

[Sweepr]

Because skyrim is not relevant anymore.
And comparing gaming performance in anandtech bench tool is plain dishonest. The 3 patches and 6 drivers updates between the sampled data makes a difference.

Skyrim is a lot more relevant than cherry picked Crysis single-player levels. Also, from the link you posted:

The contrast with some of the recent AMD "APU" chips is kinda brutal. They all seem to do pretty OK with FPS averages, but these other numbers don't look so good. There are some nasty spikes in the plot for that A10-6800K, too, which is kind of the point, I guess. You're probably gonna feel those slowdowns.

 

[MiddleOfTheRoad]

But you have never provided any links, not even one review to back up what you are saying.

In order for you to say the following, it means you have seen that performance and you have the links to provide us with evidence to back up your claims.

If you cant provide evidence for the above, then we should take your comment as bate and thread crapping.

He can't do that, AtenRa -- because what he's saying really isn't true.
Most modern games are now GPU limited, not CPU limited -- so he won't be able to produce meaningful benchmarks to prove what he's saying.

Is a Haswell i7 better than an AMD FX CPU in some games? Sure..... But some people are gaming on a dual core G3258 and that same FX will nearly always provide a better experience than that specific Intel is demanding CPU games. And that's a pretty fair comparison, too -- because a FX-4300 quad core street price is within $15 of a G3258 dual core.

Bottom line, he's just babbling his bias -- Phynaz is not actually discussing anything meaningful. I game on both an i7 3770K and an FX 8320 -- and there isn't a noticeable difference between the two. There isn't a game that runs poorly on either machine. Nowadays, the GPU is nearly always the bigger factor for gaming. With a multiple monitor setup, almost every game is going to be GPU limited.

 

[Erenhardt]

Skyrim is a lot more relevant than cherry picked Crysis single-player levels. Also, from the link you posted:

I will not argue here about which game is more popular...
Yep, min fps and frame times matter a lot:
FX8350:

i7-4770k:

A bit smoother on the fx, but that is probably academic difference not possible to notice with bare eye.

Quite opposite to what many here spread, FX Vishera chips have quite a lot potential. The Fairy tales about fx not being able to run games are... fairy tales...

Metro Redux, Ramestered, Final Edition may run better on fx than on intel, but I guess we will have to wait for that release yet

 

[AtenRa]

I've provided the links many times over and you know it. Report me if you think I'm thread crapping.

Like the ones you just did above ???

and Sapphire 7970 (second in the mail for crossfire).

Gaming
Bf4
Thief
Titanfall
Csgo
Dota2
Diablo3
Dayz

I also use 3 monitors, if that matters.

The OP will use two GPUs at CF and 3x Monitor setup and you link single GPU 1080p benchmarks. And then you talk about cherry picking on my CrossFire 1440p AT review ???

Does anyone have any Multi GPU Multi Monitor reviews to show the performance difference between FX6300@4.6GHz and Core i5 ??? Throwing 1080p single GPU benchmarks all the time are irrelevant to this thread and they dont help the OP or anyone else.

 

[SlowSpyder]

The OP is using a three monitor setup. If 1080P monitors, that 6.2 megapixel (that's almost 70% more pixels than 1440P). Why all the benchmarks at 720P, 1080P, and 1440P? They are not at all comparable. With two 7970 level cards, I don't think the CPU matters that much. He is likely to be GPU limited the vast majority of the time. If he was building from the ground up, a very good argument for Intel could be made. But not when he already has AM3+ and has a FX 6300. Just my $.02.

 

[davie jambo]

The OP is using a three monitor setup. If 1080P monitors, that 6.2 megapixel (that's almost 70% more pixels than 1440P). Why all the benchmarks at 720P, 1080P, and 1440P? They are not at all comparable. With two 7970 level cards, I don't think the CPU matters that much. He is likely to be GPU limited the vast majority of the time. If he was building from the ground up, a very good argument for Intel could be made. But not when he already has AM3+ and has a FX 6300. Just my $.02.

Yeah this , it's not worth the bother upgrading to intel at the moment. He would see an increase in frames but it won't be very much. Ten or less I should expect

Intel folk seem to think their cpus are vastly superior but mostly there is not that much in it at all. Not enough to spend $400 on if you ask me

 

[Dribble]

I don't understand why the AMD defense squad don't just drop it sometimes. The new i5's are clearly superior gaming cpu's. It's only in another universe where you think otherwise, and as for all games being gpu limited, that's rubbish. They tend to be most gpu limited in the sort of fake automated single player runs that reviewers use, and then you have to only look at average fps. Play on a large multiplayer server and look at min fps which is what you get when the action is hottest and the differences are much more pronounced. Equally there are still a number of huge single threaded games out there (e.g world of tanks basically needs a modern intel cpu to break 45fps).

Op just get an i5K and be done with it, then you know your cpu will always be fast.

 

[kawi6rr]

Hello

I am wanting to update my primary rig. AMD FX 6300 @ 4.6ghz, gigabyte 990FXA-ud3, 8gig ddr3 1600, 2x840 evo ssds, Corsair TX750 PSU, and Sapphire 7970 (second in the mail for crossfire).

Basically I am looking at getting a 8350 for 150$ or i5\Z97 combo for 285$ at micro center. My big issue is, with having a solid board, is the 8350 the smart upgrade or would the Intel path be better. No issues with cost, more focused on which purchase would be wiser of the two. I have been looking and watching prices for awhile, I can not decide.

Gaming
Bf4
Thief
Titanfall
Csgo
Dota2
Diablo3
Dayz

I also use 3 monitors, if that matters.

I took my computer over to my buddy’s house to do some comparisons both running Win 7 64bit professional.

My setup

• FX 8350
• R9 280x
• 16gigs or ram

His setup

• I5 3570k
• GTX 770
• 16gigs or ram

Benchmarks show his system being faster, pretty graphs and that’s it. Side by side running Bf4, Thief, and Titanfall you could not see any difference period. They looked identical to the eye in gameplay. You’re going to have many AMD haters telling you to jump ship but if you don’t have any issues with your current choices other than upgrading graphics then why spend the extra money on a whole new computer.

 

[SlowSpyder]

I don't understand why the AMD defense squad don't just drop it sometimes. The new i5's are clearly superior gaming cpu's. It's only in another universe where you think otherwise, and as for all games being gpu limited, that's rubbish. They tend to be most gpu limited in the sort of fake automated single player runs that reviewers use, and then you have to only look at average fps. Play on a large multiplayer server and look at min fps which is what you get when the action is hottest and the differences are much more pronounced. Equally there are still a number of huge single threaded games out there (e.g world of tanks basically needs a modern intel cpu to break 45fps).

Op just get an i5K and be done with it, then you know your cpu will always be fast.

http://i.imgur.com/1s6B44z.jpg

Keep in mind the OP's FX 6300 is at 4.6GHz. Also, the OP is using a three monitor setup. The argument isn't that the FX is better than an i5. The argument is that he already has the AM3+ setup and it probably would be a pretty minor upgrade for the money (and time) spent. To me this is what I post about fairly often in these types of threads... a difference in benches, on paper. Probably not so much a difference in real life gaming.

 

[Vesku]

I don't understand why the AMD defense squad don't just drop it sometimes. The new i5's are clearly superior gaming cpu's. It's only in another universe where you think otherwise, and as for all games being gpu limited, that's rubbish. They tend to be most gpu limited in the sort of fake automated single player runs that reviewers use, and then you have to only look at average fps. Play on a large multiplayer server and look at min fps which is what you get when the action is hottest and the differences are much more pronounced. Equally there are still a number of huge single threaded games out there (e.g world of tanks basically needs a modern intel cpu to break 45fps).

Op just get an i5K and be done with it, then you know your cpu will always be fast.

Or, you know, GPU limitations come up when someone is running on 3 monitors. But sure go with that whole "AMD defense squad" thing.

I initially suggested he get a i7 Devil's Canyon or similar during the Thanksgiving sales. After rereading and noting his triple monitors I then suggested he consider his GPU situation. Being GPU limited isn't some sort of fairytale. However, if OP intends to stick with his dual 7970s then an i7 would still be a good upgrade.

 

[janeuner]

I'd like to see an i5 priced at the same level as an FX-6300. Until that happens, AMD still has its niche.

 

[bunnyfubbles]

the answer is very simple, different testing scenarios, PClabs is pushing more CPU bound scenes,
if you are doing a CPU test with watch dogs at 60+ for every CPU you are doing it wrong.

but take both in consideration and the answer is obvious, the FX is good in many cases sure, but the haswell i5s are great basically everywhere when it comes to gaming,

from 4.6GHz FX 6300 the only logical upgrade is Haswell when it comes to gaming,

this

an upgrade to an 8300+ from an overclocked 6300 would only net tangibly better performance for things that can consistently use the extra thread(s), games really don't meet that stipulation

whether or not the upgrade is worth the $285 is another matter