Ryzen 2500u benchmark by The Tech Report (img heavy)

[DocSwag]

I'm willing to bet the thing is heavily memory bandwidth and latency constrained. I thought HBCC could help with that but it looks like it's not enough.

[Likwid]

18 hours ago, Prysin said:

 

Lukewarm reception. Seems like AMD cannot deliver. Add this ontop of leaks showing terrible battery life and well, Ryzen with "Vega" graphics (worst marketing name EVER) seems to be a dud. Sure it has a strong GPU, and decent CPU, but its in mobile, and its battery life is meh.

 

Why didn't you include power consumption graph?

Spoiler

	HP Envy x360 15m-bq1xx 2500U, Vega 8, HGST Travelstar 7K1000 HTS721010A9E630, , 1920x1080, 15.6	Asus FX550IU-WSFX FX-9830P, Radeon RX 460 (Laptop), SK Hynix HFS128G32TND, TN LED, 1920x1080, 15.6	HP Spectre x360 15t-bl100 8550U, GeForce MX150, Samsung PM961 NVMe MZVLW512HMJP, IPS, 3840x2160, 15.6	HP Spectre x360 15-bl002xx 7500U, GeForce 940MX, Toshiba XG4 NVMe (THNSN5512GPUK), IPS, 3840×2160 , 15.6	HP Spectre x360 15-ap012dx 6200U, HD Graphics 520, Sandisk X300 SD7SN6S-256G-1006, IPS, 3840x2160, 15.6
Power Consumption
Idle Minimum *	5.9	5.9	11.6	8.14	6.5
Idle Average *	8.7	8.8	13.8	10.91	13.3
Idle Maximum *	10.5	9.7	14.2	12.26	15.1
Load Average *	45.7	78.1	67.9	45.67	30.8
Load Maximum *	49.4	117	76.9	74.94	38.5
Witcher 3 ultra *	50.2	112.4	59.04

As far I think It's not worth to pay 26% more to get max 10% performance in this price segment when you can't get really good gpu.

P.S Your post seems like BS for now after checking review

[WMGroomAK]

22 hours ago, AluminiumTech said:

The testing methodology used in that review for battery life is deeply flawed

Funny thing is, if we go with the flawed testing methodology, then the R5 2500U with a ~56 Wh battery is getting the same battery life as a i5-8250U w/ a MX150 using a ~51 Wh battery or within 4% of the i5-8250U with integrated graphics.  While this would suggest that it may not be quite as power efficient as the Intel counter part, it certainly isn't as horrible as people would make it out to be...  I would personally like to see how much of this relates to how the Intel parts divides up it's work between the iGPU and the dGPU vs AMD having to rely on the Vega GPU.  One test might be to see how long each system would last on a single charge while playing something like DOTA or Rocket League, which would probably require the Intel systems to always use the MX150.

 

https://hothardware.com/reviews/intel-8th-gen-kaby-lake-refresh-mobile-review?page=5

 

[Whispre]

OP seems to have cherry picked graphs to make the CPU look worse than it is. If you read the full review it's not all that bad.

 

Doom is a good example:

 

 

 

 

 

And then theres this in the summary:

 

Quote

The Vega 8 IGP is the brightest spot for the Ryzen 5 2500U. Where Intel's UHD Graphics 620 IGP might run out of gas even at the lowest possible resolutions and graphics settings in some titles, Vega 8 usually provides playable experiences even with more demanding games and sometimes even higher resolutions. It's no replacement for a GeForce MX150, but Vega 8 certainly sets a high new bar for integrated graphics performance in this power envelope.

 

 

Reading for the win.

 

 

 

[Trixanity]

12 minutes ago, WMGroomAK said:

Funny thing is, if we go with the flawed testing methodology, then the R5 2500U with a ~56 Wh battery is getting the same battery life as a i5-8250U w/ a MX150 using a ~51 Wh battery or within 4% of the i5-8250U with integrated graphics.  While this would suggest that it may not be quite as power efficient as the Intel counter part, it certainly isn't as horrible as people would make it out to be...  I would personally like to see how much of this relates to how the Intel parts divides up it's work between the iGPU and the dGPU vs AMD having to rely on the Vega GPU.  One test might be to see how long each system would last on a single charge while playing something like DOTA or Rocket League, which would probably require the Intel systems to always use the MX150.

 

https://hothardware.com/reviews/intel-8th-gen-kaby-lake-refresh-mobile-review?page=5

 

Well, for starters it's flawed because in all likelihood it's probably using hardware decoding (fixed-function block) which does not correspond to either CPU or GPU power efficiency. Of course the hardware on one chip can be better than the next and so can the software but the power usage should be pretty close. It's more likely that differences in display, network, disk drive and the like causing the small difference between them.

[RagnarokDel]

23 hours ago, Prysin said:

Lukewarm reception. Seems like AMD cannot deliver. Add this ontop of leaks showing terrible battery life and well, Ryzen with "Vega" graphics (worst marketing name EVER) seems to be a dud. Sure it has a strong GPU, and decent CPU, but its in mobile, and its battery life is meh.

 

Source: https://techreport.com/review/32877/amd-ryzen-5-2500u-apu-reviewed

 

Is your superpower hating on AMD??

23 hours ago, ScratchCat said:

From this source it seems that it is closer to the i5 8250U , more in line with the expectations. Note: Although the MX150 is notably ahead of Vega the power consumption is greater by ~25W.

https://www.notebookcheck.net/Our-first-Ryzen-5-2500U-benchmarks-are-in-and-Intel-has-every-reason-to-worry.266618.0.html

 

The MX150 is a discrete GPU, which puts the total power enveloppe at 40 watts, Raven Ridge is more efficient.

 

[Prysin]

5 hours ago, Likwid said:

Why didn't you include power consumption graph?

  Reveal hidden contents

	HP Envy x360 15m-bq1xx 2500U, Vega 8, HGST Travelstar 7K1000 HTS721010A9E630, , 1920x1080, 15.6	Asus FX550IU-WSFX FX-9830P, Radeon RX 460 (Laptop), SK Hynix HFS128G32TND, TN LED, 1920x1080, 15.6	HP Spectre x360 15t-bl100 8550U, GeForce MX150, Samsung PM961 NVMe MZVLW512HMJP, IPS, 3840x2160, 15.6	HP Spectre x360 15-bl002xx 7500U, GeForce 940MX, Toshiba XG4 NVMe (THNSN5512GPUK), IPS, 3840×2160 , 15.6	HP Spectre x360 15-ap012dx 6200U, HD Graphics 520, Sandisk X300 SD7SN6S-256G-1006, IPS, 3840x2160, 15.6
Power Consumption
Idle Minimum *	5.9	5.9	11.6	8.14	6.5
Idle Average *	8.7	8.8	13.8	10.91	13.3
Idle Maximum *	10.5	9.7	14.2	12.26	15.1
Load Average *	45.7	78.1	67.9	45.67	30.8
Load Maximum *	49.4	117	76.9	74.94	38.5
Witcher 3 ultra *	50.2	112.4	59.04

As far I think It's not worth to pay 26% more to get max 10% performance in this price segment when you can't get really good gpu.

P.S Your post seems like BS for now after checking review

9 minutes ago, RagnarokDel said:

Is your superpower hating on AMD??

 

The MX150 is a discrete GPU, which puts the total power enveloppe at 40 watts, Raven Ridge is more efficient.

 

fanboys will always hate those who speak against their ideals

[Prysin]

25 minutes ago, Whispre said:

OP seems to have cherry picked graphs to make the CPU look worse than it is. If you read the full review it's not all that bad.

 

Doom is a good example:

 

 

 

 

 

And then theres this in the summary:

 

 

 

Reading for the win.

 

 

 

doom is also Vulkan, an API where AMD has a clear edge on a driver basis. No need to include benchmarks which is in AMDs favor by default. If AMD had performed poorly in DOOM, it would have been a travesty, given how the rest of their product stack performs

[RagnarokDel]

1 minute ago, Prysin said:

fanboys will always hate those who speak against their ideals

Yes, I am a fanboy 

[Prysin]

3 minutes ago, RagnarokDel said:

Yes, I am a fanboy 

and thus you invalidated your argument. And your counter argument of me being a hater is well.... provably false, as my posting history will show, at least 4000-5000 posts, or more, is in favor of whatever AMD is doing. But i dont sugar coat their actions and products anymore. Their products need to speak for themselves, and when they aint, they need to be called out on it.

 

Mindlessly supporting and cheering on a product from a company is as anti consumer as you get

[rawrdaysgoby]

First of all we all know it most likely have to due with its single channel memory and them not having dual channel memory 

 

Second of all why are there even i7-7700HQ numbers in there if we take that out and compare the CPU to their U versions it is exactly what we expected

 

Thirdly it is the same with The graphics it is exactly what we expected of the GPU better then intel worse then Nvda But it isn't the R7 2700u version it is the 2500u version

 

Fourth of all idk why but so far all of Ryzen 600$ offerings have touch screens even though intel barely has it with their higher price counter part with dgpu

 

Power consumption??? how is it even possible for a max 25 watt part APU would use more power then a intel+nvda solution that takes up far more space

 

We have already all speculated this before 2700u is the way to go. If we take away all the gimmicks that the intel + nvda option never really had on lower end laptops that AMD does have Their laptops could solidly have a place even at the 500$-600$ space for the R5 2500u and the R7 2700u would strike it at 700$-800$ range  

[Daniel644]

16 hours ago, Whispre said:

OP seems to have cherry picked graphs to make the CPU look worse than it is. If you read the full review it's not all that bad.

 

Doom is a good example:

 

 

 

 

 

And then theres this in the summary:

 

 

 

Reading for the win.

 

 

 

"best integrated graphics", thats like saying best towing capacity for a motorcycle, if the 4 cylinder car that is the MX150 can smoke the thing even as weak as it is, it's still MEANINGLESS to gamers looking for a budget laptop and only serves to boost performance in the few applications that can be accelerated by it. so you might see reduced render times compared to pure CPU rendering but beyond that it's useless.

[Trixanity]

2 hours ago, Daniel644 said:

"best integrated graphics", thats like saying best towing capacity for a motorcycle, if the 4 cylinder car that is the MX150 can smoke the thing even as weak as it is, it's still MEANINGLESS to gamers looking for a budget laptop and only serves to boost performance in the few applications that can be accelerated by it. so you might see reduced render times compared to pure CPU rendering but beyond that it's useless.

Integrated vs discrete. It's extremely difficult to integrate powerful graphics without making the die very big and power consuming. Also, the MX150 mostly wins through GDDR5 memory, TDP and clock speed.

[Daniel644]

1 minute ago, Trixanity said:

Integrated vs discrete. It's extremely difficult to integrate powerful graphics without making the die very big and power consuming. Also, the MX150 mostly wins through GDDR5 memory, TDP and clock speed.

winning is still winning, either way both are pretty useless outside of certain productivity apps that can actually use the GPU.

[Trixanity]

Just now, Daniel644 said:

winning is still winning, either way both are pretty useless outside of certain productivity apps that can actually use the GPU.

Yes and a 6950X also wins against a Celeron processor.

[Daniel644]

1 minute ago, Trixanity said:

Yes and a 6950X also wins against a Celeron processor.

but then you are talking about orders of magnitude higher costs and orders of magnitude higher power, not small percentage point differences.

[Trixanity]

6 hours ago, Daniel644 said:

but then you are talking about orders of magnitude higher costs and orders of magnitude higher power, not small percentage point differences.

Winning is still winning. 

And you're looking at higher power and higher cost for the winning hardware in the original scenario. 

[Daniel644]

4 minutes ago, Trixanity said:

Winning is still winning. 

And you're looking at higher power and higher cost for the winning hardware in the original scenario. 

not in performance to cost EFFICIENCY.

[Trixanity]

1 minute ago, Daniel644 said:

not in performance to cost EFFICIENCY.

That too.

[Doobeedoo]

Hmm, interesting. Would like to see more tests and 2700u with dual channel too. Though yeah we'll see more in time. 

[NumLock21]

On 11/28/2017 at 5:18 AM, Prysin said:

i have toyed around a lot with APUs, got several of them. It REALLY depends, but generally you see a 8-10 FPS uplift.

 

Now, after looking over HP's site, it appears that they only provide some form of dual channel (2x4GB SODIMM, 1x4GB soldered + 1x8GB SODIMM or 2x 8GB SODIMM)

The solder ram setup is only single channel not dual channel. Don't know about those with 2 ram slots.

[Derangel]

No, actually its pretty decent. This is an ULTRABOOK class processor. Ofc it doesn't hang with higher-end models. We're also looking at only a single example from a manufacturer known to skimp on internal design in favor of making things look pretty. We really need to see more models tested. That's the thing with laptops, a lot depends on how the laptop itself is designed and that can have an effect on performance, especially things like battery life. Tech Report's review was very positive about the APU itself, saying that it is competitive. The 2500u's compeition is Intel's ultrabook chips with integrated graphics. MX150 is, of course, going to win massively. Any modern dedicated GPU is going to be stronger than something on board. The real test for Ryzen mobile is going to be when we start seeing more models come out so we can see where they fall in terms of price and performance. The R7 chip should be interesting as well, mostly due to it's stronger Vega component.

[Dabombinable]

2 hours ago, Derangel said:

No, actually its pretty decent. This is an ULTRABOOK class processor. Ofc it doesn't hang with higher-end models. We're also looking at only a single example from a manufacturer known to skimp on internal design in favor of making things look pretty. We really need to see more models tested. That's the thing with laptops, a lot depends on how the laptop itself is designed and that can have an effect on performance, especially things like battery life. Tech Report's review was very positive about the APU itself, saying that it is competitive. The 2500u's compeition is Intel's ultrabook chips with integrated graphics. MX150 is, of course, going to win massively. Any modern dedicated GPU is going to be stronger than something on board. The real test for Ryzen mobile is going to be when we start seeing more models come out so we can see where they fall in terms of price and performance. The R7 chip should be interesting as well, mostly due to it's stronger Vega component.

The standard mobile versions, not the ULV model, should be awesome as long as the manufacturer doesn't go full retard.

[Derangel]

2 hours ago, Dabombinable said:

The standard mobile versions, not the ULV model, should be awesome as long as the manufacturer doesn't go full retard.

Yeah. I'm really excited to see the full Ryzen mobile lineup. It has been a very, very, long time since there has been any competition in the laptop market.

[RagnarokDel]

 

21 hours ago, Daniel644 said:

"best integrated graphics", thats like saying best towing capacity for a motorcycle, if the 4 cylinder car that is the MX150 can smoke the thing even as weak as it is, it's still MEANINGLESS to gamers looking for a budget laptop and only serves to boost performance in the few applications that can be accelerated by it. so you might see reduced render times compared to pure CPU rendering but beyond that it's useless.

You realize that a CPU + discrete GPU is going to be considerably more expensive, right? Even if Nvidia was selling the MX150 for $1(they're not) it would add way more then that. You need to connect that discrete gpu to the CPU somehow and it's not like you're getting rid of the cost of the IGPU on Intel, it's still there.

On 28/11/2017 at 3:33 PM, Prysin said:

and thus you invalidated your argument.

I thought the emote was captain obvious AF but apparently not.

19 hours ago, Daniel644 said:

but then you are talking about orders of magnitude higher costs and orders of magnitude higher power, not small percentage point differences.

It's at the very least 50% more expensive but probably more in the range of 75% (I'm not talking about the whole machine, but CPU+IGPU vs CPU+IGPU+DGPU)