Intel Core i7 4770K "Haswell" Benchmarks On Ubuntu Linux

[GreatEmerald]

Intel HD 4600 graphics

No, it's Intel HD Graphics 4600. Check Intel ARK if you want to be sure. Hence why it's appropriate to call it HDG 4600 and it won't be confused with AMD Radeon HD 4600 line of graphics cards.

The Xonotic temperature measurements are definitely not right. The A10 average temperature is 5 degrees? That's physically impossible...

[mendieta]

Sounds like the issue might just be a feature that defaults to off in radeon but on in the intel driver. Not sure what current state is, but floating point textures and texture compression/s3tc used to be the main areas.

Thanks, Bridgman, I'll take a look again. There is some debug info available on the command line, it's just a matter of having the time to chase it ... cheers!

[bartsch]

At least one propagated feature is missing: SMAP

Phoronix: Intel Core i7 4770K "Haswell" Benchmarks On Ubuntu Linux

This past weekend I shared the first experiences of running Intel's new Haswell CPU on Linux. While Intel Haswell is a beast and brings many new features and innovations to the new Core CPUs ...

http://www.phoronix.com/vr.php?view=18754

What I'm missing in the /proc/cpuinfo output of the i7 4770K and the engineering sample I've at hand is SMAP. Though it's uninteresting for graphics stuff and benchmarking, but it has been propagated in every article I can find about Haswell.

So, I'm wondering: where is it? Does anyone has serious informations why it's not there?

SMAP - Supervisor mode access prevention: http://lwn.net/Articles/517475/

[mendieta]

If you are going to test the IGPs I hope you will test

The common Intel HD 4600
The mobile/BGA Intel Iris Pro 5200
AMD HD 7660D with Catalyst drivers
AMD HD 7660D with Open Source drivers

There is also

AMD HD 8660D with Catalyst drivers (just released, Richland)
AMD HD 8660D with Open Source drivers

At least in Windows, for what I've seen, Richland is 25% or so faster than Trinity. It also has a higher overclocking ceiling (http://www.extremetech.com/gaming/15...or-steamroller)

[Michael]

There is also

AMD HD 8660D with Catalyst drivers (just released, Richland)
AMD HD 8660D with Open Source drivers

Except that I haven't gotten any hardware out of AMD in a long time now...

[mendieta]

Ok, then let's do a price comparison here, what the best bang for your computing buck? I'm using the same very popular shop for all parts.

$179.98 basic A10-5800K system

$264.98 basic FX-8350 system

$414.98 basic i7-4770K system

I'm assuming that ram, psu, drives and case are identical across all systems and thus left out of the comparison.

Sure, I actually said the same thing in my first post in this thread. AMD has better performance/price ratio for sure, for CPU. And as you noted, when you compound this with least expensive mobo's, the difference is really substantial, in terms of value. On the integrated graphics, AMD was much better, and now that Intel is catching up, the perforance is similar but AMD again provides better value.

Your comparison is lacking a bit, though, in hat the FX-8350 needs a ~$70 discreet card to compete with the i7-4770K. This brings it to $335, which is 20% cheaper than the Intel solution. Not a huge difference, but sure, a better value.

The A10-5800K is a better value as compared to Intel if you put apples to apples. But it's at a different ballpark altogether.

Thanks for the figures!

[curaga]

At least Richland has the same gpu as Trinity (Cayman), not GCN as was misreported by some sites - it uses the mature r600g driver, not the experimental radeonsi.

[duby229]

How do you propose getting a repeatable, offline test profile for a MMO?

it doesnt need to be repeatable. it needs to be actual gameplay. That is all that matters when benching games. You play games so gameplay on the newest release is what matters. MMOs should be benchmarked as such as often as releases are made.

It's not hard to understand that games are meant to be played so gameplay is what should be benchmarked. Obviously gameplay is not exactly repeatable.

[duby229]

Except that I haven't gotten any hardware out of AMD in a long time now...

It's pretty easy to see why after reading quite a few of your articles, though the last few has left out negative innuendo so they have improved.

[curaga]

it doesnt need to be repeatable. it needs to be actual gameplay. That is all that matters when benching games. You play games so gameplay on the newest release is what matters. MMOs should be benchmarked as such as often as releases are made.

It's not hard to understand that games are meant to be played so gameplay is what should be benchmarked. Obviously gameplay is not exactly repeatable.

If it isn't repeatable, it has questionable value. You can't tell whether it was an anomaly, a busy day on the server, or any other hiccup. Heck, you can't even tell whether anything was tested at all: if the test isn't repeatable, the writer could be pulling numbers out of his ass and nobody would be the wiser.