Oh well, at least it's better than Sandy Bridge.

>this shit again
okay anon

>>59344533
>gayming benchmarks

>>>/v/

>>59344823
>>posts random game benchmarks
>gets told to fuck off back to /v/
>>posts random adobe single threaded benchmarks
>gets told to fuck off, r7 is for multithreaded
>>posts more adobe benchmarks.
>gets told to fuck off, again with more single threaded benches
>>posts price/performace bench that does not have the 6900k
>gets told to fuck off, not relevent

Cherry picked pottery.

STOP BUMPING

>>59344867
nah, everytime I see some gaym benchmarks, it's GTA5
I'm fine with shitting on Zen, it's not the literal best CPU line-up to ever happen, and prices are dropping on some older processors, but fuck off with this shilling

File: r7-1700-totalwar-warhammer.png (112KB, 806x580px) Image search: [Google]

112KB, 806x580px

>>59344894
At least it's better than an i3.
Oh wait...

>>59344533

1700 is 3.7ghz turbo and is only ~5% faster than a stock 2600k

even a mildly OC'ed 2600k can beat that in gaymen, and you can buy them for $100 now, maybe even lower since retards will upgrade to ryzen due to the hype.

it comes down to windows 10 scheduling issues. its also why there is an average 10% increase in performance on ryzen with 7 over 10.

ryzen 8 cores are actually two physical quad cores connected together with something similar to intel's quickconnect called infinite fabric. so essentially its the equivalent of having two socket processors on a single chip rather than two different sockets on the motherboard.

due to this you have to schedule taskes on it similar to dual socket setups. like dual xeons. you essentially do this with numa support.

so take a single thread. unless the load is high on that single thread, windows will schedule it and toss it around different threads. this is fine in most cases but when you have dual socket setups, this can cause a performance regression. to send information to the second processor it would have to travel across the interconnect. with intel, quickpath(20gb/s) and now with ryzen, infinite fabric(22gb/s). this is slower and increases latency.

so instead what you do is have the scheduler do its best to schedule the thread on the same processor that it original was spawned on. so for example, if you have two quad cores, and a single thread spawned on quad core #1, you do your absolute best to keep that thread scheduled on the four cores of quad core #1.

you don't see this regression in heavy multithreaded / workstation applications because "workstation" workloads because #1, the thread, even if single, will mostly likely be heavy and thus windows won't move it once it spawn on a core, and #2, a lot of workstation workloads are heavily multithreaded and will saturate all the available threads / cores to their max.

>tldr
ryzen has to be scheduled similar to how multi-socketed processors are.

>>59344910
literally negligible average gaym framerate, but much worse low?
also, why the fuck is the 5.1GHz OC 7700k 1 frame faster than the non-OC?
fuck, if a 2500k performs at almost 63 minimum at 1080p on high, 2600k/2700k must be pretty decent
fuck this APU A10 7890K shit

File: 1464035379201.jpg (167KB, 1699x292px) Image search: [Google]

167KB, 1699x292px

>>59344965
>also, why the fuck is the 5.1GHz OC 7700k 1 frame faster than the non-OC?
because GN sucks at testing

also here is what happens to 2600k at 4.5Ghz

File: 1489085225096.png (1MB, 2484x3052px) Image search: [Google]

1MB, 2484x3052px

>>59344910
Meanwhile in reality.

>>59345000
is 4.5GHz on a 2600k easy to get on a Cryorig H5 or something?

>>59345010
>look at first graph
>overclocked Ryzen has lower performance than stock
OpinionDiscarded.png

>>59345023
if you got good chip, sure

So does Intel subcontract though other companies for shills to shit post on this board like other companies?

File: 1468596943201.png (6KB, 620x369px) Image search: [Google]

6KB, 620x369px

GTA V

1800X
.1% 70 fps
1% 74 fps
10% 82 fps
25% 82 fps
50% 90 fps
75% 98 fps
90% 106 fps
99% 126 fps

7700K
.1% 82 fps
1% 90 fps
10% 102 fps
25% 106 fps
50% 110 fps
75% 118 fps
90% 126 fps
99% 138 fps

here is what happens on actual data set, basically it has hicups with latency, which are purely windows related

>>59344912
>2600k for $100
Where You Find That Processor? I Can't Find It On Ebay Or Amazon?

>>59345027
4.1 is not on all cores, stock performs better at multithreadeing

>>59345093
What a pathetic excuse for a weak processor?

>>59345103
still beats intel 8 core, doesn't it?

File: Ryzen.jpg (231KB, 1920x1020px) Image search: [Google]

231KB, 1920x1020px

>>59345174
In value, not performance.

And even then there's better value Intel chips for most people on the market.

>>59345052
https://www.pcper.com/reviews/Processors/AMD-Ryzen-and-Windows-10-Scheduler-No-Silver-Bullet

Windows is handling Ryzen cores correctly. Both SMT and CCX migration myths busted. AMD fags on suicide watch

>>59345189
Haha not this again. He always seems to pop up at AMD launches with special tools in hand to do some special testing.

>>59344533
sage & cage

>>59344533
man that's awful
poor amd fags

File: 1485576476450.jpg (17KB, 542x540px) Image search: [Google]

17KB, 542x540px

>>59345183
AYYMD BTFO

>>59345342
Bump & Hump

File: 07a899e5d0968ff4b118a018ad3dba2f.jpg (50KB, 1000x1000px) Image search: [Google]

50KB, 1000x1000px

>>59345189
>win10 is fine, AMD BTFO
>10-20% better performance with win7
okay

>>59345681
Let's see those 25% better performance windows 7 chart

>>59345784
even fucking gnoo/linucks performs better than win10

>>59345805
Let's see those gnoo/lincucks vs windows performance charts

>>59345010
where is gta 5?