• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

No Performance Difference Between My Old 5770 and New 6870 - Get My Money Back?

S

SkipCameron

New Member
Joined
Jun 22, 2012
Messages
14 (0.00/day)
Location
Boise, Idaho, USA
I finally built my first gaming PC since 1996 so that I could push my favorite games past their console limitations (I have a PS3). At first everything worked out great as I was able to run Fallout 3 and New Vegas at near maximum settings at 50-60 FPS. However, as I added more and more mods they really started to take a hit, dropping into the low 20s and worse in intensive areas. Because I love my mods I decided to upgrade my system so that it could handle all the high-res texture packs and the like I threw at it. I upgraded my CPU, Motherboard, and went from 4GB of RAM to 8GB about a month before I decided to do one 6870 rather than getting another 5770 to Crossfire.

Much to my disappointment, as I've been benchmarking my new XFX 6870 on Fallout New Vegas I've determined that there is almost no improvement over my Sapphire 5770. To be specific, running around the New Vegas Strip at night (a place I used to get a lot of lag) doing three 30 sec FRAPS benchmarks found that the 5770 averaged 53 FPS and the new 6870 averaged 54 FPS. The 5770 did better than usual in this area probably due to the CPU/Motherboard/RAM upgrades that I made since my last playthrough, but the 6870 didn't make any noticeable improvement over it (and it's A LOT noisier!).

My 5770 is able to run all of my non-Bethesda games just fine, but I'm hoping to get and mod the heck out of Skyrim after all the DLC is out and I can get the complete edition, but if my new 6870 isn't going to improve over my 5770, I might as well send it back. (BTW, my Fallout benchmarks were using only NMCs texture pack and the URWL lighting mod. I have been waiting to add all the others until my new GPU arrived.)

On another note, I ran some Furmark 1.10.1 and MSI Kombustor 2.3 tests, and the 6870 does show a significant improvement over the 5770 (Furmark 1440 x 900 w/ 0 AA: 5770 ran 22 FPS vs. 6870 at 36 FPS; at 8x AA the FPS were 4 and 7 respectively). However, I don't really care if it can test better if it doesn't improve my gaming experience in any meaningful way. It also didn't seem to improve The Witcher 2 much either, though I haven't benchmarked it yet to be sure.

I'm hoping someone can explain what I'm missing here. I mean, I almost bought a 6950 and wonder now if I would be in the same situation, or even if I did the 5770 Crossfire.

Old System Specs:
CPU: Intel Core2 Duo E7500 Wolfdale 2.93GHz
MoBo: Asus P5G41-M LE/CSM LGA 775 Intel G41 Micro ATX
GPU: Sapphire Vapor-X Radeon HD 5770 1GB
RAM: Corsair DDR2 2x 2GB
PSU: Rosewill RD500-2SB 500W
HD: Western Digital Caviar Blue 500GB 7200RPM
Monitor: 1440 x 900 Native

New System Specs:
CPU: Intel Core i3-2100 3.1GHz
MoBo: Gigabyte GA-Z68XP-UD3 ATX Intel
GPU: XFX Radeon HD 6870 1GB
RAM: G.Skill Ripjaw Series 8GB (2x 4GB) 240-Pin DDR3 1600
PSU: Apevia ATX-CB700W
HD: Same
Monitor: Same

I've heard stuff about "CPU bottlenecks" and "RAM bottlenecks" and the like and wonder if something like that might be going on, but nothing I found was very conclusive. I'm kind of a noob at modern hardware (like I said, I've been out of the game since about 1996) and am at a loss. Thanks!
 
Well both cards have 1gb and your new cpu is still a dual core, return the gpu and the cpu and for a 2gb gpu (6950,7850) and a quad core cpu like 2500k for example.
 
Last edited by a moderator:
Looks more like a CPU bottle neck to me.

Upgrade your CPU.
 
The monitor you're using probably isn't helping the situation. 1400x900?

Also, the 6870 is the equal line to the 5770 when ATi was finally absorbed by AMD. the X870 performance line was moved to an x970 numbering system.

5770 to 6870
5870 to 6970
5970 to 6990
 
I still have my Rosewill 500W PSU, which I think is capable of running one 6870 with my other components, so I can swap those if that's the problem. Also, my monitor is actually my wife's TV that we got for steal a few years ago that I only appropriated last year to be my monitor. I plan to upgrade eventually, but it works fine for now.

I also don't want this to turn into a PSU thread, but I got the Apevia PSU because after hours and hours of research on PSUs I concluded that most PSUs on the market have just about an equal amount of the following reviews:

"I bought a _______ PSU and it fried my system/was DOA/died after a week! Stay away!!!"

followed by

"Really? I bought a (same as above) PSU and it has been great! No complaints!!!"

So basically just decided that unless > 90% of people agree a given PSU/Company is crap, it's pretty much luck on whether or not your PSU works as advertised, so I went for price. I've had Corsair, Rosewill, and now Apevia and they've all worked fine.
 
DarkOCean said:
Well both cards have 1gb and your new cpu is still a dual core, return the gpu and the cpu and for a 2gb gpu (6950,7850) and a quad core cpu like 2500k for example.

i doubt is the psu.
Click to expand...

I've been wondering if it's a CPU issue, but I just read so many times that "modern games are so much more GPU intensive that CPU bottlenecking is less and less of an issue" etc. that I thought a modest CPU upgrade (basically just one that would work with my new MoBo since my prior MoBo wasn't Crossfire capable) and a significant GPU upgrade would solve the problem.

So would an i5 quad core really give my 5770 a boost or would I need to keep the 6870 to see the boost? I know the 5770 and 6870 are related but all the test sites I looked at show the 6870 having about a 1.5-2x FPS improvement over the 5770.
 
DannibusX said:
The monitor you're using probably isn't helping the situation. 1400x900?

Also, the 6870 is the equal line to the 5770 when ATi was finally absorbed by AMD. the X870 performance line was moved to an x970 numbering system.

5770 to 6870
5870 to 6970
5970 to 6990
Click to expand...

so your trying to tell us 5770 its the same performance to a 6870, if so i say that a lie, i suggest you do some research the 5770 vs 6870 and then you will see a 6870 is at least 30% faster
 
SkipCameron said:
I still have my Rosewill 500W PSU, which I think is capable of running one 6870 with my other components, so I can swap those if that's the problem. Also, my monitor is actually my wife's TV that we got for steal a few years ago that I only appropriated last year to be my monitor. I plan to upgrade eventually, but it works fine for now.

I also don't want this to turn into a PSU thread, but I got the Apevia PSU because after hours and hours of research on PSUs I concluded that most PSUs on the market have just about an equal amount of the following reviews:

"I bought a _______ PSU and it fried my system/was DOA/died after a week! Stay away!!!"

followed by

"Really? I bought a (same as above) PSU and it has been great! No complaints!!!"

So basically just decided that unless > 90% of people agree a given PSU/Company is crap, it's pretty much luck on whether or not your PSU works as advertised, so I went for price. I've had Corsair, Rosewill, and now Apevia and they've all worked fine.
Click to expand...

Well yes and no... people live in different houses with different power systems, different wiring, and their components put different loads on PSUs. If you buy something crap with no Active PFC for instance, good luck (I don't think that apevia has PFC). At the end of the day it is a gamble, with your shiny new rig.

Also, if you do not see an improvement, you should definitely take it back and get your money lol. Alot of times a 30% improvement is not easily noticeable to the naked eye, especially if you were expecting more. Take it back, save your dough, and buy something with a bigger improvement. That apevia PSU really is shit tho, I would take that back while you're at it, even if you have to get something a bit lower wattage... a cooler master 550W or a Lepa 650W for a few bucks more would make sense - especially for that rig, which is really nice.
 
Last edited:
SkipCameron said:
I've been wondering if it's a CPU issue, but I just read so many times that "modern games are so much more GPU intensive that CPU bottlenecking is less and less of an issue" etc. that I thought a modest CPU upgrade (basically just one that would work with my new MoBo since my prior MoBo wasn't Crossfire capable) and a significant GPU upgrade would solve the problem.

So would an i5 quad core really give my 5770 a boost or would I need to keep the 6870 to see the boost? I know the 5770 and 6870 are related but all the test sites I looked at show the 6870 having about a 1.5-2x FPS improvement over the 5770.
Click to expand...

Use msi afterburner in games to see how much vram is loaded when you play your moded NV to see if is a vram issue.
 
I went from an HD5750 to a HD6870 and dident notice much diffrence on my core 2 quad Q6600 @3200mhz but now i have upgraded to a core I5 2550K and now it makes a big diffrence so there is a chance the I3 is holding it back on your system.
 
Corduroy_Jr said:
so your trying to tell us 5770 its the same performance to a 6870, if so i say that a lie, i suggest you do some research the 5770 vs 6870 and then you will see a 6870 is at least 30% faster
Click to expand...

So. What part of my post stated that a 5770 was equal to a 6870 in performance?

Equal "line" doesn't translate to equal performance. Rather, I was merely pointing out that AMD shifted their product line coding when they finally and officially absorbed ATi. Since the OP hasn't built a computer in quite some time, the OP says 1996, I just wanted to make sure they weren't confused and expected flagship GPU performance from a lower budget GPU.

Thanks for playing!
 
SkipCameron said:
able to run Fallout 3 and New Vegas at near maximum settings at 50-60 FPS. However, as I added more and more mods they really started to take a hit, dropping into the low 20s and worse in intensive areas. Because I love my mods I decided to upgrade my system so that it could handle all the high-res texture packs and the like

(snip)

hoping to get and mod the heck out of Skyrim after all the DLC is out and I can get the complete edition,
Click to expand...

You like to play my favorite games, and use texture mods like me.

Forget about 'upgrading' to another dual core and a graphics card with 1GB vram!

Currently I'm using a dual core at 4Ghz and a 5970 (which is a dual graphics card with only 1GB vram utilizable) and all the games you mention suck if I add texture packs!

Don't get any graphics card with less than 2GB vram.
 
OK guys, we need to stop blaming his PSU, his CPU, vram and other nonsense. If his performance hasnt improved it's most likely driver related. The 6870 is significantly faster thant he 5770 so much so that he should see an obvious enhancement in performance. TPU shows a 35% difference.


http://tpucdn.com/reviews/HIS/Radeon_HD_6870/images/perfrel.gif
 
DannibusX said:
The monitor you're using probably isn't helping the situation. 1400x900?

Also, the 6870 is the equal line to the 5770 when ATi was finally absorbed by AMD. the X870 performance line was moved to an x970 numbering system.

5770 to 6870
5870 to 6970
5970 to 6990
Click to expand...

Black Panther said:
You like to play my favorite games, and use texture mods like me.

Forget about 'upgrading' to another dual core and a graphics card with 1GB vram!

Currently I'm using a dual core at 4Ghz and a 5970 (which is a dual graphics card with only 1GB vram utilizable) and all the games you mention suck if I add texture packs!

Don't get any graphics card with less than 2GB vram.
Click to expand...
These guys got it
 
I am assuming those mods are eating all your VRAM, i think you should swap it for a 2GB card, you'll see a fair boost.

I understand this will sound stupid but did you format your HDD when you swapped out your parts?
 
You should read this thread.

Added to that, I still think your CPU is bottlenecking you (as is the same in my case presently). Even though it's got 4 threads, it's still a dual core...

What you can try in the meantime is to make sure you have the barest minimum of applications running in the background. Just stop whatever is not necessary.
 
First, the 6870 is in a different league than the 5770, it should be at least 35% faster in 95% of the games at a decent resolution (1080p).

Second, the i3 is in a different league than the E7500 and it should give considerable performance gains in 95% of the games at a decent resolution.

The 5% of the games where this combination is less effective are those poorly coded and those modded where loading textures need more than 1GB of VRAM. There is no doubt that your current machine is more powerful than your last one and your less than stellar PSU has nothing to do with it. I don't know if that particular game would run better on an overclocked Ivy quad and a 500$ GPU, maybe yes. But is it worth it?
 
Could you try a different game or benchmark please?

Fallout isn't all that hard to run and I believe it has a frame cap of 60 fps ( so you won't see a performance difference here)

my 6870 was a great card upgrade from my 5770, I was nearly laughing at some benchmarks being double the speed.
 
Fallout hardly taxes my Core i3 2100 and overclocked HD5850 1GB. It hardly taxes the CPU!!

I had a Q6600 at 3GHZ with the same card and performance is similar - I got the Core i3 2100 as a stop gap but it seems to do the job well.

OTH,with a game like Crysis2 I am getting more consistent and higher framerates and in RTS games the Core i3 2100 is faster.

The OP should also get a better PSU IMHO.
 
Last edited:
If you can catch a P55 setup cheap or a cheap H61/2100 then that would be your best bet for gaming.

I just switched one of my crunchers out from a Athlon X4 630 OCed to 3.5Ghz to a I3-530 OCed to 4ghz and it increased FPS by almost 15-20 in some games.

Were are you located?
 
your CPU is bottlenecking i would think.

Get a better CPU.
 
de.das.dude said:
your CPU is bottlenecking i would think.

Get a better CPU.
Click to expand...

Neither one of his CPUs would bottle neck a 6870.

Well his older one would but even that at 3.4 pretty much eliminates possible bottle necks.

Only crappy coded ( skyrim ) games will be bottle necked.
 
the perfect pairing IMO is a core i5 2400/3450 and a 6870. I came from a Q9450 and 2 x 4850s to my current setup and noticed the improvement immediately. Even Gfx wise 2 4850 512mbs were holding my CPU back.
 
OP, I think the main reason that there isn't much visible difference between your old card and your new card is because of the relatively low resolution of your monitor. Faster video cards aren't really that much faster at lower resolutions like you are using. You will likely not notice any significant differences between them unless you run it at higher resolutions. You could also try bumping up the graphic detail, if you have not already. You will likely begin to see the more powerful GPU at work at a lower resolution by doing that. The fact that your new card is benchmarking much faster on Furmark likely means that everything is working just fine.
 
You must log in or register to reply here.
Back
Top