Wtf Is ryzen this bad? My friend has gtx 1080+ryzen 2700x and I have gtx 1070+i5 3570k. Still I managed to get better combined score in firestrike 6150 vs 7450 which is a bit over 20% better.
My cpu is 5 years old and it still beats couple months old cpu. What is going on here?
Other urls found in this thread:
3dmark.com
3dmark.com
3dmark.com
openbenchmarking.org
3dmark.com
3dmark.com
3dmark.com
3dmark.com
overclock.net
twitter.com
sir pls delet this post
I genuenely surprised. I was going to change to 2700x but after this I cancelled this plan.
I would guess that your friend's CPU cooler is not doing its job, as it is an 8-core part with 16 threads at and very power consumption. Check to see if his temps are exceedingly high or if it is throttling. the processors are not even comparable in performance, its more than 2 to 1 and obviously so. Your friend fucked up something very badly.
Here is Passmark results.
sorry I meant they are 4-cores with 8 threads. From AMD, only threadrippers and up are 8 physical core parts
It has the stock ryzen 2700x cooler. Also It peaked pretty good I'd say: Maximum turbo core clock 4420 MHz
Here are the test results.
this is blatantly untrue, look up any real benchmarks rather than what your "friend" says he has
he probably has really bad thermal throttling
either you're reading this wrong or this is total bait or shopped - look at the memory bus clock on his 1080.
even with stock cooler a 2700x blows the shit off a 3570k. even a r5-1600 does lol. i'm thinking his heatsink isn't properly seated, or he has some other severe problem (horribly shit ram, or something very wrong with his GPU such as the memory clock thing I mentioned)
all the scores are higher on AMD system, yet combined showing higher on Intel? #shopped #fake #IntelPaidForIt
>look at the memory bus clock on his 1080.
What it looks like just normal stock 1080. That gtx 1070 looks like overclocked one.
I can't believe I hadn't seen that gif before, it's absolutely fucking hilarious.
I wish I could do that. Just start screaming and doing that in random places.
What could be the problem here? Could it be the stock cooler which comes with 2700x?
3dmark.com
3dmark.com
Well look at these, high physics+gpu but combined is shit.
Shit RAM making the CPU underperform ?
Never go below 2666 for a ryzen
>inb4 half of the cores were disabled and this is an Intel shill thread
It's 3200 MHz ram though: Here also with 3200 MHz ram still shit combined for some reason: 3dmark.com
Might as well post the benchmark of my middle of the road build. Ryzen 7 2700x + GTX 1070.
3dmark.com
Not sure if Win8.1 gimped it. It kept bitching about my CPU with constant pop-ups.
What's stopping you?
This.
Results you are reporting are literally impossible op unless something is seriously wrong with his rig.
2700x is 8 cores you dunce
>implying
3dmark.com
Nothing wrong ram is 3200MHz etc.
telling.. perhaps something on the mobo is presenting a bottleneck
>i-it's the cpu cooler.....
>pissmark
What's next, geekbench?
>Wtf Is ryzen this bad
No. It's a really fast CPU:
openbenchmarking.org
> What is going on here?
That specific benchmark and/or his specific hardware, OS and drivers.
This is an older 1080 with GDDR5, the newer 1080 cards have much faster memory.
Literally the only part of Firestrike that's relevant to CPU benchmarking is the physics test, and his chip is 139.6% faster in that test. It's somewhat interesting that his graphics score is also fine in the lighter GPU tests, but everything goes to shit in the combined test. My bet would be a faulty/shitty power supply that can't handle ramping up both the CPU and GPU at the same time for the combined test. Alternatively, something is overheating somewhere and causing throttling.
really peculiar, tell him to run it again
Mm no, standard Asus GeForce GTX 1080 STRIX.
All gtx 1080 has gddr5x. It was gtx 980 which had gddr5.
PSU is super flower focus+ 750w gold. And I know the other test are fine but the combined is 800-1000 lower than it should be.
Having run it myself, their shitty benchmark is just broken.
>2700X and a 1080 Ti
7086 combined score
3dmark.com
>2600K and a 1080 Ti
7591 combined score
3dmark.com
>8400 and a 1080 Ti
9407 combined score
3dmark.com
>7700K and a 1080 (non-Ti)
9813 combined score
3dmark.com
I have no idea what the fuck they're doing in the combined score section, but it makes no sense based on the previous section results nor my observed real world gaming performance. Certainly the 1080 Ti smashes the regular 1080 across the board with all three processors, yet the highest combined score of all is with the regular 1080.
Even the Intel-only results don't make sense in context to each other. Shit's fucked. Use Time Spy or something else instead.
>Even the Intel-only results don't make sense in context to each other. Shit's fucked. Use Time Spy or something else instead.
Yeah that I thought because I have no idea what is going on with that combined score.
>ignoring the difference between a 1070 and 1080 on a fucking graphics card benchmark
Well 1080 performs better even 1070 is overclocked, nothing surprising there.
overclock.net
here is something about that low combined score.
make sure you have the motherboards extra 4 or 6 pin power connector plugged in and fully seated.
Well of course it is plugged in upper left corner.
Run cinebench and post your score
I was going to buy 2600X but now that I have seen this devastating bad result for AMD I will buy Intel instead.
>something's obviously wrong here
>but ill buy my cpu and mobo based on it
Go right ahead, you deserve it.
You can tell that to Hardware Canucks too