Trying out the 2700X, a Gigabyte or AsRock x470 board, 32GB 2x16 3200 kit, maybe a Vega card if I can find a deal.
I just pre-ordered 2700x, X470Taichi, and this 3600 cas 15 3600 memory https://www.newegg.com/Product/Product.aspx?Item=N82E16820232306
You beat me to making a Ryzen2 builders thread dang it.
The memory already shipped. The CPU is due to ship on 4/19, and the motherboard on 4/26. I hate waiting a week, but I wanted to let the chip have its full potential on the new series motherboard and the fastest memory that motherboard/CPU can handle. Its rated 3600 cas15 and I am going to try 3466 cas 14. 3466 is the best the motherboard is advertised to handle in the specs.
By the time I build another rig, I might end up with the 3000 or 4000 series Ryzen.
So can anyone list all of the advantages, of the 2000 series over the first gen chips?
So I just ordered a big heatsink. The CM Hyper 612 v2. It cools most of my E5-2683v3's (14 cores). So first I will test stock, then OC'ed on the stock HSF, then OC'ed on the 612. I had to spend $10 just to have the adapter shipped from CM, it used to be 7. They need to include it in the cooler !
I will be swapping my 1700 to 2700X in few weeks time. It will go to into Asus X370 Prime, so I hope for good BIOS support.
Just put the results here, you can even do pics.I might build a 2700x as well as an 8700k system just so I can review them. I'm finding that reviewers are doing very poor jobs these days. I have a laundry list of things I'd like to see reviewers do, one of the least is disclosing any scripts they used, GPU control panel settings, etc. Anandtech (they are an excellent site, not bashing them at all btw, but every site has it's own fault) for example has a rocket league script that they have not disclosed. They claim Threadripper only gets a 99th percentile of 163.12 fps (Note they also originally only used a 1060, but now the bench says 1080...they should be using a 1080ti to eliminate bottlenecks). I've played several games (hundreds of hours) of rocket league, it's up there with one of my favorite games, and I do not even come close to going this low. Different GPUs, memory, speeds, SSDs, etc. all influence framerate. NUMA vs UMA, and much more. Testing should be transparent. A good benchmark is user reproducible. While this can make it hard to come up with good benchmarks, it's not impossible. I've set my eyes on a hardware site for a while, and I have a budget to buy the products, I just would rather spend that money on other, more fun things, but perhaps it's time to jump into the game. Many reviewers think that having custom 'private' workloads give them an edge, but all it does is create the illusion of bias. Companies like Nvidia also can't optimize their drivers for the shear number of games that exist today, so being more extensive by testing more games can make it impossible for manufacturers to 'rig' tests.
Absolute IPC (normalized MHz/GHz) is important as well as chip performance at stock speeds (within the manufacturers recommended specs) as well as overclocking potential. All this stuff can be automated, with an 8 hour work day I could put out all kinds of reviews.
I think purch has to do with a lot of this, I suspect these folks aren't getting the dollars they need for proper reviews. I'm going to set up a budget for this pretty soon and will likely order a 2700x after I recover from my surgery on May 7th. After that, we'll see. I've already degraded my threadripper, but I still may do limited testing with it to compare. I also have a 2600k in my posession. The domain name has been registered and, being a developer and technology guru, it won't take me long to set up a site. If this interests you, Stay tuned. You'll see my post on reddit and elsewhere when I'm ready to go.