If you commented after the link, instead of just linking it, that would have been great, and maybe I would not have commented. Just putting in my 2 cents.
If you commented after the link, instead of just linking it, that would have been great, and maybe I would not have commented. Just putting in my 2 cents.I'm only a simple messenger buddy, nothing more, nothing less. I just recall others caring about Linux. Since this is a review thread, I posted a simple review link. But still, they did not review as thorough as Phoronix seems to have did.
In order to try and minimize the number of test variables, all RAM was kept at 2133 MHz for the testing
Well, first off, they gimp the Ryzen, and put faster memory in the 6700, not even the PC3200 that they put in the 6700 ! Let alone on the newer bios, you might get 3600 to run, like other benchmarks,. And they couldn't even get 4 ghz like many of our members get ? This review bites. But they still gave a few kudos to Ryzen
Why i've timestamped it is for what the reviewer says, and it either means that Intel processors have either been suddenly optimised for this game, or either the Nvidia driver or the Mafia 3 driver has nerfed Ryzen performance.
Has anyone got this game and noticed a drop in performance?
I just throw this in here. I have collected the numbers of many recent gaming benchmarks (1080p, only reviews published 04/11 or later) and put them together in this database:
https://drive.google.com/open?id=15s-e_FxSbvhpR8m5JtHLBOTDm-pQHVe0gAG6lWYnfuM
Is really the most telling. Nearly across the board all the Intel 4c see improvements in benches when switching to DX12 and on nearly all the Ryzen benches, except 2 and 1 of them had a counter point, they lose performance when switching to DX12.
I'd have liked to see more 6900 tests. Only one of them acts like a Ryzen (though It would ignore the one with a 40% increase as that site seems a crazy outlier for that benchmark). Computerbase did have some 720p tests with the 6900 that had the same performance characteristics as Ryzen almost across the board.
You forgot to mention that all those tests are using an Nvidia GPU.Is really the most telling. Nearly across the board all the Intel 4c see improvements in benches when switching to DX12 and on nearly all the Ryzen benches, except 2 and 1 of them had a counter point, they lose performance when switching to DX12.
Well I didn't think I needed to point that out because it was pretty much exactly what Oddzz was collecting the information for. But yes all of them appear to be using a 1080 or higher.You forgot to mention that all those tests are using an Nvidia GPU.
Well, so far the best I have gotten to via Aida 5.90.4200 is 81.9ns on the memory latency bench. 2933mhz @14-15-15 timings. 1600 non-x @ 3500mhz. I'm running G-Skill 3000+ Hynix ram.
The averages are different overall but I cracked into the very low 80's for once.
Anyone able to get in the 70ns area?
German retailer Mindfactory is showing how many processors they have sold and it basically confirms what the article is saying - 1700 is the most popular (with the 1700X) atm but the 1600 is catching up fast.Good to see Ryzen selling well .
https://mybroadband.co.za/news/hardware/207708-amd-catching-up-to-intel-in-south-africa.html
I hope AMD can get to 20-25% desktop / notebook market share over the next few years. The key is going to be AMD's execution with Raven Ridge, Pinnacle Ridge and a timely launch of 7nm products in 2019.
Looks like at least one OEM/boutique builders is preparing to sell Ryzen rigs:
For once, AMD marketing did a good job with that chart.
https://twitter.com/iBUYPOWER/status/857741376770850818
The low ram speed alone almost invalidates the testing, but he still did say, even handicapped like that it did better.
someone attempted to compare "real world" usage of 1600 vs 6600K
only compares 2 games , uses stock clocks, and uses low speed RAM though