Benchmarked a Radeon 460 vs. the 455 - 12% faster, 17% more power usage

[drvelocity]

I had a chance to benchmark an i7-6920HQ Pro 460 system against an i7-6820HJQ 455 system. At least in OSX, the Pro 460 was only 12% faster in Heaven (31.1 FPS vs. 27.8 FPS) than the 455 and power consumption was 48.8 watts vs. 41.8 watts, a more than 16% increase.

I find it interesting that despite the fact that the TDP is technically the same for both i7's and both Radeon's, actual power use under load is still very different. I haven't seen this addresses elsewhere.

In my personal opinion, the 3 extra FPS aren't as much as I expected and nearly insignificant. Considering the power savings under full load of the 455 and 6820HQ, I think I'll save my $300.

Screenshots of benchmark results:

 

[Ma2k5]

I'm certain the benchmarks don't translate well into real world performance. Isn't there also a difference of 2GB RAM?

 

[Xombie11]

$300? It's $100 to upgrade.

 

[drvelocity]

$300 in

$300? It's $100 to upgrade.

$300 includes the CPU upgrade from 2.7 to 2.9 ghz.

 

[Patrick-Photo]

$300 in


$300 includes the CPU upgrade from 2.7 to 2.9 ghz.

I'd skip the CPU upgrade anyway...


And with regards to this single benchmark, testing out some rendering/exporting of footage, windows performance, and some light gaming might show otherwise, at least with vRAM hungry work.

 

[dk808]

Its only a $100 upgrade not $300

 

[leman]

Thanks for the test, that was very interesting! I wonder though how much the factor is the CPU. Also, could you repeat this when battery is closer to 100% on both machines? I have a suspicion that the system power consumption might get slightly throttled when the battery is getting low, even on power adapter, to prevent battery damage.

Also, is there a reason you have chosen Medium quality and disabled the tessellation? Could you do the same with Ultra?

 

[KirekkusuPT]

CPU upgrade isn't worth, but the GPU upgrade is. And in real world performance the diference is actually way bigger.

Radeon Pro 455 is around a Xbox One level of performance, the 460 is around the PS4's GPU power too. You can see that many games run better on the PS4. The Radeon 455 is like a 960M and the 460 is like a 965M and everyone will tell you a 965M is way better.

 

[leman]

The Radeon 455 is like a 960M and the 460 is like a 965M and everyone will tell you a 965M is way better.

The 460 is actually closer to the 460M than to 465M in actual performance. It would perform much better if the VRAM would be clocked faster...

 

[drvelocity]

CPU upgrade isn't worth, but the GPU upgrade is. And in real world performance the diference is actually way bigger.

Radeon Pro 455 is around a Xbox One level of performance, the 460 is around the PS4's GPU power too. You can see that many games run better on the PS4. The Radeon 455 is like a 960M and the 460 is like a 965M and everyone will tell you a 965M is way better.

Right but for me it's like comparing a Civic to an Accord when I have my Ferrari PC ready for serious gaming. I'll never game with the Macbook other than maybe some basic stuff that would run on pretty much anything (2d games, etc.). But I can see how the $90 would be worth it to someone doing 3d modeling work. Then again if you're doing that you should have bought a laptop with a 1080.

 

[jackoatmon]

They're not gaming computers though. If you want to benchmark their power the place to do it is in renders, exports, transcodes, lookup tables, etc.

It goes to the meaning of the word benchmark – why use diagnostic criteria that don't apply to the intention of the engineering?

 

[Papakaliati]

They're not gaming computers though. If you want to benchmark their power the place to do it is in renders, exports, transcodes, lookup tables, etc.

It goes to the meaning of the word benchmark – why use diagnostic criteria that don't apply to the intention of the engineering?

The fact that they are not gaming computers doesn't mean I don't want to very occusionally play a game or two. For me a desktop computer is a complete waste of space and money, I just can't find the reason (or even the place! )for it. With MBP 2016 I can play overwatch and battlefield 1 with quite high settings, even though I game for only maybe 3 weekends a year or so.

 

[wmk]

I had a chance to benchmark an i7-6920HQ Pro 460 system against an i7-6820HJQ 455 system. /.../ In my personal opinion, the 3 extra FPS aren't as much as I expected and nearly insignificant. Considering the power savings under full load of the 455 and 6820HQ, I think I'll save my $300.

It's not about the frame rate only. You won't be able to run many games in higher settings with 2GB of video memory, and I'm not talking about the textures only.

 

[leman]

The fact that they are not gaming computers doesn't mean I don't want to very occusionally play a game or two. For me a desktop computer is a complete waste of space and money, I just can't find the reason (or even the place! )for it. With MBP 2016 I can play overwatch and battlefield 1 with quite high settings, even though I game for only maybe 3 weekends a year or so.

Yep, exactly. The main theme of the MBP is versatility. So its nice to know that it can also be used for occasional entertainment such as gaming quite well, if one wants.

 

[xmonkey]

1 benchmark test?