This is a split board - You can return to the Split List for other boards.

Predict the future - GT 1030 edition

  • Topic Archived
You're browsing the GameFAQs Message Boards as a guest. Sign Up for free (or Log In if you already have an account) to be able to post messages, change how messages are displayed, and view media in posts.
  1. Boards
  2. PC
  3. Predict the future - GT 1030 edition

User Info: zhenghan

zhenghan
5 months ago#31
godplaysSNES posted...
JKatarn posted...
zhenghan posted...
Setzera posted...
Pineapple5 posted...
TC, can you go on YouTube and see if you can watch 4k 60fps videos smoothly? That GT 1030 is supposed to do all the heavy lifting.

Unfortunately my monitor only goes up to 1080p, but when I put YouTube on 4k anyway, the playback runs flawlessly at that setting. Not sure if that answers your question.

I see , so 1050 is not even needed for 4k playback.


Why would you think it would be? This is video decoding we're talking about, not pushing polys. Even iGPU can handle 4K video.


Skylake IGPs and certain Geforce 900 GPUs and later support HEVC and VP9, which 4k video typically use.
Otherwise, it's running on the CPU.

What else could they use besides HEVC and VP9? I guess that would be an interesting test at as well.
"dude i am like a 10 th grader,i am not smart i am not smart okay.i know basic english not level 100 english lol" -stephanielish

User Info: Setzera

Setzera
5 months ago#32
Lee7 posted...
I have similar setup too OP

I know I'm surprised how these systems perform considering their age.

darkus_f posted...
I dunno why exactly but I love such builds. :3

I'm really starting to appreciate this build more and more, started off as a "what if" type of project, but it's coming along nicely. :)

Speaking of which, more results to post in a few minutes!

User Info: Setzera

Setzera
5 months ago#33
Today I received the 8GB of RAM and it's been successfully installed, but before that, some unposted results from yesterday.. I didn't get a chance to do a full game library test to compare 32 bit to 64 bit using only 3GB of RAM, but I did test some.

-

Results simply from upgrading 32 bit OS to 64 bit OS:
- TW: Shogun 2 saw a 20-30 fps increase under the same settings (low 1366x768) bringing it within 70-90 fps with occasional drops to 40 fps for a few seconds.

-Sniper Elite 3 was also about 20-30 fps increase under the same settings (low 1920x1080) bringing it within 60-80 fps with occasional drops to 30 fps for a few seconds.

-Once again I didn't check fps scores, but my eyes could tell a notable improvement in Starcraft 1 and 2.

I'm guessing the frame drops were likely caused by having only 3 GB of RAM since it was a fairly noticeable drop when it happened, not gradual, which to me indicated some sort of memory management issue. Also possible that it's a Windows 10 issue, as I've heard of that being around still.

-

Now on to the results comparison with 8GB of RAM freshly installed today!
(DDR2 PC-6400 800Mhz - 2GB each, 4 sticks in 4 slots)

Setzera posted...
Passmark/Performance Test 9.0 results:

CPU: AMD Phenom x3 8550 - 3 core @ 2.2Ghz (95W)
32 bit OS / 3GB RAM score: 1677
64 bit OS / 3GB RAM score: 2036 (+359 from previous)
64 bit OS / 8GB RAM score: 2082 (+46 from previous)
Passmark Average as of original post: 1930 (59 samples)

GPU (3D Mark): Gigabyte Nvidia Geforce GT 1030 - 2GB DDR5 (30W)
32 bit OS / 3GB RAM score: 1949
64 bit OS / 3GB RAM score: 1942 (-7 from previous)
64 bit OS / 8GB RAM score: 1938 (-4 from previous)
Passmark Average as of original post: 2295 (225 samples)

Passmark system overall
32 bit OS / 3GB RAM score: 1031
64 bit OS / 3GB RAM score: 1194 (+163 from previous)
64 bit OS / 8GB RAM score: 1227 (+33 from previous)

RAM
32 bit OS / 3GB RAM score: 499
64 bit OS / 3GB RAM score: 716 (+217 from previous)
64 bit OS / 8GB RAM score: 814 (+98 from previous)

2D Mark GPU (285-308)
Disk Mark HDD (636-666)
Both of these results have remained almost the same throughout each test.

64 bit OS / 8GB RAM: Game fps tests will be posted in a few days.

So it would seem that for Passmark, an extra 5GB of RAM is a minor improvement compared to switching from 32 bit OS to 64 bit OS. Something tells me that real world gaming performance will actually greatly benefit from the extra RAM, since Windows 10 was using 1.5 / 3.0 GB of it at idle after a reboot, before the upgrade to 8GB today.

I'm not really sure why my GPU isn't seeing any benefits from all of these upgrades yet. I have a few more things to try to get it caught up to average scores.

1) I have an AMD Phenom 9600 X4 quad core @ 2.3Ghz (95W) on it's way from an ebay refurbisher. That's the max this motherboard is confirmed to handle. That'll be happening within a week or two.

2) If that doesn't increase performance for the GPU, I'm going to try moving this PC to a different wall outlet, as from this position it's using a 15 foot extension cord (with surge protection).

3) Last resort may be to upgrade the power supply from 300W to 400W, even though it should be enough as it is.

-

I'm hoping the CPU upgrade will be the key, as I really have no plans of running this PC in another room, nor do I think I'm pushing the power supply limit here. I'm definitely not going to overclock anything.

Well, that's it for my massive report. I'm open to suggestions on increasing GPU performance though!

User Info: DarkZV2Beta

DarkZV2Beta
5 months ago#34
I mean, is there a reason you think the 1030 should be doing better than it is?
It's a very low-end card, and passmark scoring it below-average shouldn't be a shock when you consider most people benching the card are probably overclocking it for e-peen, or doing some other muddling to increase benchmark scores, like lowering image quality settings, switching Windows to the classic skin, turning off features, ect.
a quad core i7 was just a rebranded celeron -Pengu1n
Anything that has 3p fps or better is fine with me -mucloud

User Info: Setzera

Setzera
5 months ago#35
DarkZV2Beta posted...
I mean, is there a reason you think the 1030 should be doing better than it is?
It's a very low-end card, and passmark scoring it below-average shouldn't be a shock when you consider most people benching the card are probably overclocking it for e-peen, or doing some other muddling to increase benchmark scores, like lowering image quality settings, switching Windows to the classic skin, turning off features, ect.

Well, I'm satisfied with the gaming performance it's giving for the price, but because it's below average on the score sheet I'm more so wondering if I have a bottle neck that's keeping it from being as good as it could be.

I can see some people are overclocking it because those scores are about 1000 points higher than mine, but almost all the others are consistently 200-300 more than mine. Seems like something must be causing that.

Edit: mine is the 1942, yellow bar.
https://i.imgur.com/a7HrV9O.jpg

User Info: zhenghan

zhenghan
5 months ago#36
Setzera posted...
DarkZV2Beta posted...
I mean, is there a reason you think the 1030 should be doing better than it is?
It's a very low-end card, and passmark scoring it below-average shouldn't be a shock when you consider most people benching the card are probably overclocking it for e-peen, or doing some other muddling to increase benchmark scores, like lowering image quality settings, switching Windows to the classic skin, turning off features, ect.

Well, I'm satisfied with the gaming performance it's giving for the price, but because it's below average on the score sheet I'm more so wondering if I have a bottle neck that's keeping it from being as good as it could be.

I can see some people are overclocking it because those scores are about 1000 points higher than mine, but almost all the others are consistently 200-300 more than mine. Seems like something must be causing that.

Edit: mine is the 1942, yellow bar.
https://i.imgur.com/a7HrV9O.jpg

cpu bottleneck most likely
"dude i am like a 10 th grader,i am not smart i am not smart okay.i know basic english not level 100 english lol" -stephanielish

User Info: Setzera

Setzera
5 months ago#37
zhenghan posted...

cpu bottleneck most likely

That's what I'm also thinking/hoping, since I already have that quad core shipping to me. lol

User Info: Setzera

Setzera
5 months ago#38
I ran some more tests today using MSI Afterburner and RivaTuner so I could see where the bottleneck was during game play, ended up with some unexpected results.

-Passmark 3D Mark: (score 1952)
Avg 30 fps.
GPU load 100%.
CPU load 10-60% simultaneously across cores during direct x tests.
GPU Bottleneck.

-Sniper Elite 3:
Low 1080p.
75-80 fps.
GPU 100%.
CPU all cores around 75%.
GPU Bottleneck.

-Rocket League:
Low 768p.
30-40 fps.
GPU 100%.
CPU all cores 50-70%.
GPU Bottleneck.

-TW: Shogun 2 (battle):
Low 768p, but more added another AI opponent compared to previous tests.
25-30 fps.
GPU 50%.
CPU Core 1 100%, CPU Cores 2 and 3 at 30%.
Single core CPU Bottleneck.

-Civilization 5:
Low 768p.
40-60 fps.
GPU 30%.
CPU all cores around 80%.
Unknown/No Bottleneck.

-Civilization 6:
50-80 fps.
GPU and CPU all were at about 50-60%.
Unknown/No Bottleneck.

.

So, essentially, the GPU appeared to be it's own bottleneck at first, keeping itself away from other GT 1030 scores. But that's when it hit me.. This motherboard uses PCI-E x16 v1.0 or v1.1, which has a far slower bandwidth than modern PCI-E x16 v3.0+ slots. The power wattage is the same but not data bandwidth.

I'm still going to run more tests once I upgrade my CPU, but I think the real bottleneck here is in fact the motherboard! That's a bit disappointing since these parts were all ordered for this specific board and I know many would say not to upgrade an aging machine, however, because of my exploratory reason for doing this I'm not worried about it, compared to if I actually wanted to make this my main rig, then things might be a little different. I'm just happy to know that if I did pop this GT 1030 in a more modern machine some day, it still has some more power to show me, and that's cool.

That said, the quad core tests will still be coming once the CPU arrives in the mail!

.
EDIT: To clarify, all TW Shogun 2 results in this thread have been during battle, not campaign map.

User Info: DarkZV2Beta

DarkZV2Beta
5 months ago#39
Setzera posted...
DarkZV2Beta posted...
I mean, is there a reason you think the 1030 should be doing better than it is?
It's a very low-end card, and passmark scoring it below-average shouldn't be a shock when you consider most people benching the card are probably overclocking it for e-peen, or doing some other muddling to increase benchmark scores, like lowering image quality settings, switching Windows to the classic skin, turning off features, ect.

Well, I'm satisfied with the gaming performance it's giving for the price, but because it's below average on the score sheet I'm more so wondering if I have a bottle neck that's keeping it from being as good as it could be.

I can see some people are overclocking it because those scores are about 1000 points higher than mine, but almost all the others are consistently 200-300 more than mine. Seems like something must be causing that.

Edit: mine is the 1942, yellow bar.
https://i.imgur.com/a7HrV9O.jpg

Passmark is a bit s***. I ran it a few times on the same system and got a disparity of around 2000 points between GPU tests. Do you have anything running that caps framerates? Vsync enabled? Any display driver hooks?
a quad core i7 was just a rebranded celeron -Pengu1n
Anything that has 3p fps or better is fine with me -mucloud

User Info: zhenghan

zhenghan
5 months ago#40
If gpu doesnt run at 99% it is a cpu bottleneck. If gpu is at 99% it is gpu bottleneck. If cpu and gpu are both at 99% at the same time then no bottleneck/both are bottlenecking.'

With a strong cpu (such as i7 8700k) u should always see gpu @ 99% unless on a low res such as 720P and/or all low settings or paired with ultra high end GPU like 1080 ti.
"dude i am like a 10 th grader,i am not smart i am not smart okay.i know basic english not level 100 english lol" -stephanielish
  1. Boards
  2. PC
  3. Predict the future - GT 1030 edition

Report Message

Terms of Use Violations:

Etiquette Issues:

Notes (optional; required for "Other"):
Add user to Ignore List after reporting

Topic Sticky

You are not allowed to request a sticky.

Update Topic Flair

You are not allowed to update this topic's flair.

  • Topic Archived