Why the disparity?

[dpastuszczak]

A friend built a new, entry level PC with a Ryzen 3 2200G. My question is, why does my 5 year old Radeon R7 260X perform almost 0.75x faster than an RX Vega 8 integrated GPU?

 

[The Paladin]

1. because your video card is using 5 year old technology
2. the IGPU chipset are getting really good on the 8th gen Intel, and the new Ryzen motherboard

 

[dpastuszczak]

1. because your video card is using 5 year old technology
2. the IGPU chipset are getting really good on the 8th gen Intel, and the new Ryzen motherboard

You misunderstood, 0.75x faster means 75% faster, or 175% of the Vega 8's performance. The 5 year old technology still pumps more frames/sec than a brand new integrated GPU, as shown here

 

[Kalevra]

I'm pretty sure .75x = 3/4 the speed. Which means it's 25% slower.

 

[The Paladin]

why does my 5 year old Radeon R7 260X perform almost 0.75x faster than an RX Vega 8

this question with the affirmation you stated indicates the r7/260x is running at 3/4 of the power produced by the IGPU RX Vega 8.
75% faster than the vega 8 chipset does not equate 175%.. but I get what your trying to say.

based on : https://www.notebookcheck.net/AMD-Radeon-RX-Vega-8-GPU-Benchmarks-and-Specs.260180.0.html
based on that information I compared your R 7 to other benchmarks and its not 175% faster for sure.
https://www.videocardbenchmark.net/compare.php?cmp%5B%5D=2701&cmp%5B%5D=3763&cmp%5B%5D=3683

 

[dpastuszczak]

0.75x of 100 = 75
0.75x FASTER than 100 = 175.
1.75x FASTER than 100 = 275

You people obviously failed word problems in highschool.

 

[bystander]

Integrated GPU's have never come close to discrete video cards. That has always been the case. The very best iGPU's only compete with the low end of current discrete cards. You should not be surprised by the results you see.

 

[ElectrO_90]

0.75x of 100 = 75
0.75x FASTER than 100 = 175.
1.75x FASTER than 100 = 275

You people obviously failed word problems in highschool.

1x0.75 = 0.75
1x1.75 = 1.75

So therefore if something is multiplied by 0.75 it would be slower, if you say 1.75 it would be faster.
Your English is what is wrong when writing maths.

 

[dpastuszczak]

1x0.75 = 0.75
1x1.75 = 1.75

So therefore if something is multiplied by 0.75 it would be slower, if you say 1.75 it would be faster.
Your English is what is wrong when writing maths.

Wrong again, ask a math teacher.

■ Car A is travelling at 60MPH; car B is travelling 0.75x as fast as car A.
B = 60 x 0.75 = 45
■ Car A is travelling at 60MPH; car B is travelling 0.75x faster than car A.
B = 60 + (60 x 0.75) = 105
■ Car A is travelling at 60MPH; car B is travelling 1.75x as fast as car A.
B = 60 x 1.75 = 105

The equations for example 2 and 3 are interchangeable, what's important is how the problem is worded. I can't believe I'm teaching 7th grade math to a bunch of computer nerds...you're supposed to be smarter than this.

 

[bystander]

Any way you look at it, he said it was .75 faster. Or 75% faster. It didn't say it was 75% "as fast" as the original (this is what you insist was written) , he said "faster" than the original. That implies the base 100% + 75%. You also know damn well what he wants to say, so stop with the crap and stick to the question.

 

[bignastyid]

Let's stay on topic.

What are the rest of the systems specs? If the system was built with slow ram it can cripple the APUs graphics performance.

 

[dpastuszczak]

Any way you look at it, he said it was .75 faster. Or 75% faster. It didn't say it was 75% "as fast" as the original (this is what you insist was written) , he said "faster" than the original. That implies the base 100% + 75%. You also know damn well what he wants to say, so stop with the crap and stick to the question.

Thank you

Let's stay on topic.

What are the rest of the systems specs? If the system was built with slow ram it can cripple the APUs graphics performance.

The benchmark I'm referring to is here. It's an average consensus from thousands of benchmarks across any number of specs.

EDIT:

I never asked what benchmark you were using, I asked for the systems specs(your link does work btw).

I fixed the link above (oops).

 

[bignastyid]

I never asked what benchmark you were using, I asked for the systems specs(your link does work btw). When troubleshooting a performance problem knowing the systems specs is crucial. For all I know the system is running ddr4 2133 in single channel.

 

[dpastuszczak]

When troubleshooting a performance problem knowing the systems specs is crucial. For all I know the system is running ddr4 2133 in single channel.

Here's the highest scoring system w/ R7 260X graphics:
[UserBenchmarks: Game 33%, Desk 59%, Work 30%]

• ■ CPU: Intel Pentium G3258 - 60.4%
  ■ GPU: AMD R7 260X - 30.3%
  ■ SSD: SanDisk SSD 128GB - 26.1%
  ■ HDD: Seagate Desktop SSHD 1TB - 79%
  ■ RAM: G.SKILL Ares DDR3 1600 C9 2x4GB - 59.9%
  ■ MBD: Asrock Z97M Anniversary

Here's the highest scoring system w/ Vega 8 graphics:
[UserBenchmarks: Game 28%, Desk 63%, Work 41%]

• ■ CPU: AMD Ryzen 3 2200G - 77.8%
  ■ GPU: AMD RX Vega 8 (Ryzen iGPU) - 21%
  ■ SSD: Transcend SSD220 240GB - 51.7%
  ■ HDD: WD Blue 2.5" 320GB (2010) - 42.5%
  ■ HDD: Seagate ST3320613AS 320GB - 36.8%
  ■ RAM: Crucial CT8G4DFS824A.C8FHD1 2x8GB - 98.6%
  ■ MBD: MSI B350 PC MATE (MS-7A34)

 

[The Paladin]

showing us previously recorded scores from other people systems doesn't speak of anything on what your system is doing.

perhaps you can try using a benchmark such as the one at www.passmark.com called performance test, you cam run all the tests or just the video card one separately so we can see what's going on.

https://www.passmark.com/products/pt.htm

 

[DSzymborski]

When troubleshooting a performance problem knowing the systems specs is crucial. For all I know the system is running ddr4 2133 in single channel.

Here's the highest scoring system w/ R7 260X graphics:
[UserBenchmarks: Game 33%, Desk 59%, Work 30%]

• ■ CPU: Intel Pentium G3258 - 60.4%
  ■ GPU: AMD R7 260X - 30.3%
  ■ SSD: SanDisk SSD 128GB - 26.1%
  ■ HDD: Seagate Desktop SSHD 1TB - 79%
  ■ RAM: G.SKILL Ares DDR3 1600 C9 2x4GB - 59.9%
  ■ MBD: Asrock Z97M Anniversary

Here's the highest scoring system w/ Vega 8 graphics:
[UserBenchmarks: Game 28%, Desk 63%, Work 41%]

• ■ CPU: AMD Ryzen 3 2200G - 77.8%
  ■ GPU: AMD RX Vega 8 (Ryzen iGPU) - 21%
  ■ SSD: Transcend SSD220 240GB - 51.7%
  ■ HDD: WD Blue 2.5" 320GB (2010) - 42.5%
  ■ HDD: Seagate ST3320613AS 320GB - 36.8%
  ■ RAM: Crucial CT8G4DFS824A.C8FHD1 2x8GB - 98.6%
  ■ MBD: MSI B350 PC MATE (MS-7A34)

This is a silly way to compare PC speed. Those numbers don't mean what you think they mean; more RAM and a better hard drive do not suddenly make an inferior GPU better.

Vega 8 graphics are a lot slower than a R7 260x. That's pretty much the answer here.

 

[dpastuszczak]

The R7 isn't in my system anymore, and the Vega 8 is a little more impressive on paper (more shaders, more than double the memory bandwidth, etc).

Now, I've posted this twice already, but this is where I got my original info:
http://gpu.userbenchmark.com/Compare/AMD-R7-260X-vs-AMD-RX-Vega-8-Ryzen-iGPU/3151vsm441833

I'm just floored that a GPU on the CPU is still so...awful. I thought the idea behind the "APU" was that they could make it faster than a PCIe card. Otherwise, what's the advantage of an APU if it's no better than an iGPU on the motherboard?

EDIT:
As I was doing some more research into the difference in specs, I found multiple sites reporting hugely different comparisons. Game-debate shows a much faster clock and a lot more shader units than hwbench for the same Vega GPU. My initial impression of the RX Vega 8 may have been exaggerated by incorrect information.

 

[DSzymborski]

The R7 isn't in my system anymore, and the Vega 8 is a little more impressive on paper (more shaders, more than double the memory bandwidth, etc).

Now, I've posted this twice already, but this is where I got my original info:
http://gpu.userbenchmark.com/Compare/AMD-R7-260X-vs-AMD-RX-Vega-8-Ryzen-iGPU/3151vsm441833

I'm just floored that a GPU on the CPU is still so...awful. I thought the idea behind the "APU" was that they could make it faster than a PCIe card. Otherwise, what's the advantage of an APU if it's no better than an iGPU on the motherboard?

EDIT:
As I was doing some more research into the difference in specs, I found multiple sites reporting hugely different comparisons. Game-debate shows a much faster clock and a lot more shader units than hwbench for the same Vega GPU. My initial impression of the RX Vega 8 may have been exaggerated by incorrect information.

No, the idea behind an integrated GPU is a basic GPU for people who did not need the power of a dedicated graphics card. It's not supposed to be a special high-performance option; you don't have the real estate on the CPU itself.

 

[dpastuszczak]

No, the idea behind an integrated GPU is a basic GPU for people who did not need the power of a dedicated graphics card. It's not supposed to be a special high-performance option; you don't have the real estate on the CPU itself.

I understand that an integrated GPU was just intended to be a "working" graphics solution, and not a performance option. What I don't get is the move from the integrated GPU on the mainboard, to the integrated GPU on the CPU.

 

[bystander]

The R7 isn't in my system anymore, and the Vega 8 is a little more impressive on paper (more shaders, more than double the memory bandwidth, etc).

Now, I've posted this twice already, but this is where I got my original info:
http://gpu.userbenchmark.com/Compare/AMD-R7-260X-vs-AMD-RX-Vega-8-Ryzen-iGPU/3151vsm441833

I'm just floored that a GPU on the CPU is still so...awful. I thought the idea behind the "APU" was that they could make it faster than a PCIe card. Otherwise, what's the advantage of an APU if it's no better than an iGPU on the motherboard?

EDIT:
As I was doing some more research into the difference in specs, I found multiple sites reporting hugely different comparisons. Game-debate shows a much faster clock and a lot more shader units than hwbench for the same Vega GPU. My initial impression of the RX Vega 8 may have been exaggerated by incorrect information.

There are different IGPU options on the CPU. Most are simply a basic, get you started GPU, to use at the desktop. Very few APU/iGPU's are designed to game at all and even those pale in comparison to a dGPU.

There may have been some talk about how the APU/iGPU could do something better than a dGPU, but those advantages are completely irrelevant because the game would have to be designed around the APU/iGPU, which won't happen as long as dGPU's are the such a big part of the gaming scene, and full of a lot more performance.

The only APU's which have real performance in mind are the gaming consoles, and they still can't compete with a dGPU, unless you talk about performance per watt.