I play on MAX, I play on AMD

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Feb 2, 2009
12,997
283
126
#76
Except that this thread is not about SFF.
Where does it says its only about Full Towers ??

The thread is about iGPUs but you had no problem talking about CPU + dGPUs.

double standards ??
 
Feb 2, 2009
12,997
283
126
#77
For idle power consumption, I do wonder how the APUs compare to CPU + dGPU?

Lets say Athlon x 4 860K + R7 250 vs. A10-7850K......idle power consumption difference?
If i remember correctly i believe the A10-7850K has lower.
 

Seba

Golden Member
Sep 17, 2000
1,357
20
91
#78
I build a Pentium G3220 (3.0GHz) + GTX 760 PC for someone in my family. The games (like Borderlands 2, Left 4 Dead 2) are running better on that PC than on my PC with Core i5-4440 (3.1GHz) + GTX 660 (that GTX 660 graphics card is now sold).

Before Core i5-4440, I had that GTX 660 paired with a Wolfdale Core 2 Duo E8400 3.0GHz (which is much slower than G3220). When I replaced the platform with i5-4440 one, the GTX 660 performed slightly better in the games I played, but not by a huge margin.

So I'll say from my experience that a Pentium Haswell CPU like G3220 is not a bottleneck in games for a card like GTX 750. With that in mind, any GTX 750 benchmark from various review sites can be considered representative for that card performance when paired with a Pentium Haswell, even if they tested the card paired with a better CPU (the GTX 750 is the bottleneck with the CPUs used by reviewers).
 

Seba

Golden Member
Sep 17, 2000
1,357
20
91
#79
Where does it says its only about Full Towers ??

The thread is about iGPUs but you had no problem talking about CPU + dGPUs.

double standards ??
This thread is not about full towers either. You are the one who tries to steer the direction of this thread towards the case size discussion.

This thread is about performance in games, not about case size. The performance in games does not depend on case size, but it depends on CPU and GPU (integrated or discrete) performance.
 
Mar 27, 2009
12,968
36
106
#80
If i remember correctly i believe the A10-7850K has lower.
Yes, I figured the APU would have lower idle power consumption....I just wonder how much lower?

Is the cost saving difference enough that maybe the APU is worth going to for a 24/7 server that can also play games, use Open CL apps?

Or maybe the lack of dGPU helps the user puts more hard drives in the enclosure.
 
Last edited:
Feb 2, 2009
12,997
283
126
#81
This thread is about performance in games, not about case size.
This thread was about AMDs iGPU performance being able to play games at higher Image Quality and or Higher Resolutions than the competition. Read the OP again.
 

Erenhardt

Diamond Member
Dec 1, 2012
3,251
0
101
#82
Because this thread is about the (only) selling point of APUs: gaming at a low budget.


I asked you to show games that do not run on a dual core CPU (because it is only a dual core), but are playable on a [more than dual core] CPU with only the integrated graphics.

You posted a GTA 5 benchmark with various CPUs, all paired with a GTX 970. "Jeez..", indeed.
FC4 and DA:I from top of my head.

Both can be played on APU on medium settings.

I'm not sure about the GTA5 2 threads stutter at lower settings, but I guess it doesn't depend on graphics settings as it is CPU bottleneck. So yea, playing GTA5 on APU is better experience than on the 2 core pentium:

How it is to play GTA5 on pentium:
https://www.youtube.com/watch?v=RGpkNPbSa2Q
 
Feb 2, 2009
12,997
283
126
#83
Yes, I figured the APU would have lower idle power consumption....I just wonder how much lower?

Is the cost saving difference enough that maybe the APU is worth going to for a 24/7 server that can also play games, use Open CL apps?

Or maybe the lack of dGPU helps the user puts more hard drives in the enclosure.
I will see my numbers tomorrow but i dont believe the difference is that much higher.
 

Erenhardt

Diamond Member
Dec 1, 2012
3,251
0
101
#84

Seba

Golden Member
Sep 17, 2000
1,357
20
91
#85
Also, I enjoyed BattleFront beta immensely, as well as console peasants did.
How well did it run on pentium + gtx750?

http://www.computerbase.de/2015-10/star-wars-battlefront-erste-benchmarks-der-open-beta/

There are pictures if you can't read it.

Even console peasants could try it for free.
900p on apu:
https://www.youtube.com/watch?v=u0ckfy4n6z8
In your first link I see a GTX 750 Ti that performs at 35fps in a 40 player multiplayer game, at 1920x1080 with unknown game settings (since they test cards like R9 Fury X and GTX 980 Ti, it could be at maximum settings). Probably they've done the tests with that i7-4770k. Not seen a Pentium mentioned there.

In your second link (Youtube link), when that guy tests A10-7870k in a 16 player multiplayer game, he gets 17-27fps at 1920x1080 with Medium settings. That would translate to even poorer results at higher settings (similar with those used in the GTX 750 Ti benchmark maybe).

So what are you trying to prove with this comparison?
 
Last edited:

Erenhardt

Diamond Member
Dec 1, 2012
3,251
0
101
#86
In your first link I see a GTX 750 Ti that performs at 35fps in a 40 player multiplayer game, at 1920x1080 with unknown game settings (since they test cards like R9 Fury X and GTX 980 Ti, it could be at maximum settings). Probably they've done the tests with that i7-4770k. Not seen a Pentium mentioned there.
And for a good reason!

gtx750 paired with pentium scores 1 fps shy from the even 1 frame per second. APU is infinite times faster...
Basic German:
Die Anzahl der physischen Kerne ist dagegen offensichtlich zweitrangig, so lange es vier logische sind.
 
Last edited:

Seba

Golden Member
Sep 17, 2000
1,357
20
91
#87
There is no Pentium in that benchmark. So they disabled two cores and HT from that i7 and the game crashed. Sounds like a bug in that game (which is still in beta), not an indication of the performance of a dual core CPU. If it is not a bug, then it is a game that will not run on a dual core CPU. So you have how many? 3 games that do not run on a dual core? On the other hand, you have thousands of games that do run on a dual core.
 
Last edited:
Mar 13, 2006
10,138
132
126
#90
Except that this thread is not about SFF.
His ability to move the goalposts in unrivaled. He'll move them at least three times in every thread.

It's happened twice already, expect a third time in the next 24 hours :)
 
Last edited:
Apr 27, 2000
12,759
1,563
126
#91
i3-6100's MSRP is $117, and you can, in fact, get them around MSRP from 3rd party Amazon sellers.
Good catch. Though you've exposed another point: at least right now, availability of the Skylake i3s in NA is not top-notch. I can't even find one on NewEgg. The fact that you can get them at a decent price from Amazon resellers is positive, and it's reasonable to conclude that some intrepid buyers would eventually find one there.

In contrast, the 7870k is widely-available, and has been for some time. And, if you really want to, you can overclock it . . . something you can't do with a Skylake i3. You can color me extremely disappointed. Boo on Intel for getting my hopes up.

So you need AMD's top of the line APU for $140 to get less performance than you would from a dual-core Pentium Haswell and a GTX 750 as Seba pointed out.

Where's the draw? Even if you're building a SFF system, you can get ITX versions of a lot of NVIDIA cards.
The draw is that instead of getting a 2c/2t processor, you either get 2c/4t from the i3 or 2m/4t from the 7870k. I would not buy a Pentium for my own use, even if low-end gaming were the primary function of the machine. For end-users that do not want to spend a lot of time thinking about what to buy or otherwise configuring their machine, being able to pick out one product to take care of all computing and graphics functions is also highly-desirable.

1. CPU throttling on the Kaveri APU during iGPU load. (Athlon x4 860K doesn't have this problem)
You are a broken record. Most end-users won't even notice. Just out of curiosity, have you bothered benching a 7870k running TF2 or LoL or DotA2 etc. with and without amdmsrtweaker being used to defeat the iGPU throttling? Or have you seen anyone else do it? I'd like to know if all that throttling actually makes any difference in framerates.

2. Price of dual channel 2 x 4GB DDR3 2133/DDR4 2400 RAM vs. dual channel kit 2 x 4GB 1600 kit (or maybe even one 8GB DDR3 1600 stick)
Are we still talking about Skylake i3 here?

3. Resolution being tested. In the low resolution case shown below a Athlon x4 860K plus R7 240 DDR3 beats a A10-7850K with dual channel DDR4 2400:
Yeah, I figured that's where you were going with that. It's a nice thought, but you haven't got the 7870k in that benchmark, and some folks just don't want the hassle of a dGPU anyway. Someday all those deprecated low-end dGPUs will leave the channel, and Nvidia/AMD will do nothing to replace them. What'll you do then?

To other posters (AtenRa etc) who brought up games like Witcher3, GTAV, etc . . . what happened to games like LoL where you can run settings on MAX (as in the ad campaign) without difficulty? The campaign pertains to products like that, not AAA titles.
 

vissarix

Senior member
Jun 12, 2015
275
40
101
#93
Also, I enjoyed BattleFront beta immensely, as well as console peasants did.
How well did it run on pentium + gtx750?

http://www.computerbase.de/2015-10/star-wars-battlefront-erste-benchmarks-der-open-beta/

There are pictures if you can't read it.

Even console peasants could try it for free.
900p on apu:
https://www.youtube.com/watch?v=u0ckfy4n6z8
Hahahahah such a great experience..28 fps is gorgeus...apu won big time.

I would rather not play the game at all instead of playing at 28fps which gives you an headcache after 5 minutes.
 
Feb 2, 2009
12,997
283
126
#94
Hahahahah such a great experience..28 fps is gorgeus...apu won big time.

I would rather not play the game at all instead of playing at 28fps which gives you an headcache after 5 minutes.
I could play the game with the sub $100 A8-7600 at nice 40fps and above. No need to play at 28fps.

 
Feb 2, 2009
12,997
283
126
#95
what happened to games like LoL where you can run settings on MAX (as in the ad campaign) without difficulty? The campaign pertains to products like that, not AAA titles.
LoL is nothing special in graphics, there are many many more games that you can almost max out or play at way higher IQ/Resolutions with AMD APUs but not with Intel.

And many of them are free to play.

edit:
The funny thing is that AMD would only need to put HBM on Kaveri and without even changing the architecture or increase Steam Processors it would gain 30-50% higher performance.
Next year Carrizo with DDR-4 and Color compression will be enough for 1080p 30+fps at sub $150.00
 
Last edited:
Mar 10, 2004
28,523
238
126
#96
LoL is nothing special in graphics, there are many many more games that you can almost max out or play at way higher IQ/Resolutions with AMD APUs but not with Intel.

And many of them are free to play.

edit:
The funny thing is that AMD would only need to put HBM on Kaveri and without even changing the architecture or increase Steam Processors it would gain 30-50% higher performance.
Next year Carrizo with DDR-4 and Color compression will be enough for 1080p 30+fps at sub $150.00
You should call AMD when you get these ideas...
 
Mar 13, 2006
10,138
132
126
#98
I could play the game with the sub $100 A8-7600 at nice 40fps and above. No need to play at 28fps.

With vsync you will be playing at 30fps. That's unbearable.
 
Mar 13, 2006
10,138
132
126
nobody plays online fps with vsync
Nobody? I garantee you are wrong.

But thanks for pointing out the compromises you have to make to play with an APU.
 
Thread starter Similar threads Forum Replies Date
VirtualLarry CPUs and Overclocking 15

Similar threads



ASK THE COMMUNITY

TRENDING THREADS