Do we have a willy waving benchmark thread?

[WelshBloke]

Link? Or should we start one?

What benchmark? Vellamo 2 is pretty, no idea if it's any good though.

 

[Oyeve]

Why? Until there is a unified benchmark program it's all moot.

 

[zerogear]

Browser mark maybe? LOL
http://browsermark.rightware.com/

 

[WelshBloke]

Why? Until there is a unified benchmark program it's all moot.

Why?

I thought I specified why in the title.

Anyway I shall start the willy waving then.

SGS3

 

[TuxDave]

Well if you're going to start willy waving, you better start posting clock speeds too if we're going to start competing for biggest virtual willy.

 

[BenSkywalker]

Vellamo is a browser benchmark, it has over 100% margin of error depending on what browser you use on some devices, pretty worthless honestly.

3DMark will be out soon, I'd say GeekBench is the best out until it hits.

 

[WelshBloke]

Vellamo is a browser benchmark, it has over 100% margin of error depending on what browser you use on some devices, pretty worthless honestly.

3DMark will be out soon, I'd say GeekBench is the best out until it hits.

The metal part is CPU.

 

[thecapsaicinkid]

Did the nexus 4 benchmarks improve, weren't they appalling on early reviews?

 

[Eug]

OK, here is some willy waving. Motorola Razr HD with dual-core 1.5 GHz Snapdragon S4 Plus in Chrome with Android ICS 4.0.4:

Note that for the browser bench, that would put it roughly at about half the speed of the iPhone 5 I'm guessing.

In BrowserMark 2.0, the iPhone 5 gets over 190000.

I haven't tested the RAZR HD, but I suspect it'd be in the 110000 range.

 

[shortylickens]

I'm sure my Note 2 is just as fast as all the other Note 2's.

 

[Eug]

For Google Octane v1, I get 1138. The iPhone 5 gets 1672. (The HTC One X and RAZR HD have the same CPU, so they should be in the same ballpark.)

Apple has really hit it out of the park with their SoC design, OS optimization and integration, and overall browser implementation.

 

[grkM3]

are these benchmarks even taken advantage of 4 cores?

 

[Eug]

Some yes, some no, but then again a lot of apps don't maximize 4 cores either. It makes me wonder if it makes more sense to push dual-core over quad-core for the time being for most stuff (but not all stuff).

 

[grkM3]

I can tell you first hand that the note 2 is a lot smoother than the gs3 with the s4 dual even though its using older a9 cores but with 4 of them.

 

[NikolaeVarius]

I got over 1580 on Google octane on my Galaxy S3.

You know why thats a terrible benchmark?

Because javascript performance is based on the damn browser.

 

[zerocool84]

are these benchmarks even taken advantage of 4 cores?

That Vellamo only uses two cores. I wish it used four.

 

[Eug]

I got over 1580 on Google octane on my Galaxy S3.

You know why thats a terrible benchmark?

Because javascript performance is based on the damn browser.

That's precisely why it's a reasonable benchmark, at least for that aspect. Much of mobile performance these days is dependent upon browsing speed.

I really don't care that CPU A is as fast as CPU B in synthetic benchmarks that don't reflect real world usage in any way, shape, or form. So what if Linpack is faster or slower? I have no use for Linpack, much less Linpack on my phone. I'd much rather know how well my browsing experience will be.

Android is pushing Chrome. Well, they really need to optimize it, to get better performance out of it (and the OS) than what most benchmarks suggest it gets.

BTW, which S3 are you running? If the European quad-core version, then that just illustrates what I'm talking about. If the North American dual-core version, then that's more like it.

 

[grkM3]

Lol this is as synthetic as it gets.in real world usage you can't tell a difference between a gs3 and say an iPhone 5 or a nexus 4 with its quad s4 pro.

 

[Eug]

Sure you can. The iPhone 5 with Safari feels faster than browsing in Chrome on the Razr HD, and the Razr HD has the same 1.5 GHz dual-core Snapdragon S4 Plus as the North American Galaxy S III.

On WiFi, Chrome on the Razr HD feels like the iPhone 4S with Safari IMO. Very decent, and way, way faster than the iPhone 4, but not as fast as the iPhone 5.

I don't have my Nexus 4 yet though, since I can't comment about the quad-core Snapdragon S4 Pro's performance. Mind you I probably won't bother testing it and will either just return it or sell it. LTE vs. HSPA+ makes a big difference in congested areas.

 

[WelshBloke]

For Google Octane v1, I get 1138. The iPhone 5 gets 1672. (The HTC One X and RAZR HD have the same CPU, so they should be in the same ballpark.)

Apple has really hit it out of the park with their SoC design, OS optimization and integration, and overall browser implementation.

I just got 2421 on octane on my SGS3.

 

[Eug]

I just got 2421 on octane on my SGS3.

Lemme guess, UK model.

 

[WelshBloke]

Lemme guess, UK model.

Yeah. 4.1.2. Stock rom. Rooted with perseus kernel.

That scores with the cpu overclocked as well.

 

[NikolaeVarius]

Sure you can. The iPhone 5 with Safari feels faster than browsing in Chrome on the Razr HD, and the Razr HD has the same 1.5 GHz dual-core Snapdragon S4 Plus as the North American Galaxy S III.

On WiFi, Chrome on the Razr HD feels like the iPhone 4S with Safari IMO. Very decent, and way, way faster than the iPhone 4, but not as fast as the iPhone 5.

I don't have my Nexus 4 yet though, since I can't comment about the quad-core Snapdragon S4 Pro's performance. Mind you I probably won't bother testing it and will either just return it or sell it. LTE vs. HSPA+ makes a big difference in congested areas.

Bullshit. I've tried it out myself with Dolphin. They feel exactly the same.

 

[Eug]

Bullshit. I've tried it out myself with Dolphin. They feel exactly the same.

Last I checked, Dolphin is not Chrome. Note that I said Chrome several times in my post.

Chrome is Google's promoted and default application, and hence my point stands. Google needs to put lots of effort into optimizing it.

Yeah. 4.1.2. Stock rom. Rooted with perseus kernel.

That scores with the cpu overclocked as well.

Yeah, that makes sense, as the UK model is quad, and it's overclocked. The North American model is dual-core.

 

[Raduque]

Yeah. 4.1.2. Stock rom. Rooted with perseus kernel.

That scores with the cpu overclocked as well.

Run it at stock clocks.

bone stock Note II: