I think that Richland is in many ways a more appropriate choice than Kaveri *at this time* for a mainstream desktop intended for use as a "traditional" desktop PC (web/email/toob etc box). Kaveri is effectively "beta-hardware." Sort of like Zambezi was to Piledriver. It's been a rough launch with lots of hurdles (bios support on launch was especially bad). If it were priced more competitively I'd be more inclined to give it the benefit of the doubt but as such, it is priced alongside that mainstream i5 that is solidly planted as proven technology. I would buy Kaveri for myself
but I wouldn't put it in a machine I'm building for someone else at this time.
The Ripjaw Z series you have picked out is one of the finest memory kits available in it's price class right now. It would be the type of kit I would buy for myself, as an enthusiast performance tuner. It would not be the type of memory I would place in a machine that is intended to be configured to run at stock speeds. It's my opinion that non-overclocking/performance-tuning builds should be built with memory whose default XMP profile matches the supported speed of the memory controller. I repeat this mantra in pretty much all threads that are concerning a "stock clocked" rig. For Intel that means 1333-1866MT/s in most cases, and for AMD, that means 1600-2133 in most cases. The performance gains available by exceeding the memory controller's supported speeds on AMD are very narrow; But the chances of it introducing instability are high. Running at speeds above that which is supported can require voltage adjustments to the CPU-NB to stabilize. I've done extensive performance testing on 3 AMD platforms (Stars, Richland, and Vishera) and found in every case that performance scaling beyond rated speeds is minuscule, and sometimes even worse than just running tighter timings at the supported speeds, even on a Richland A10 with the "full" 384 shaders, performance scaling in gaming benchmarks tapered off above 2133MT/s.
Volumetrically speaking, the bitfenix case is enormous. It's almost as large as my mid-tower case. It will have a TON of totally wasted space on system with a single SSD. There are dozens of great cases that are a fraction of the size of the Prodigy that I believe would be a much better use of space. I'll share some ideas here a little later....
If you'll allow it, I'm going to continue my debate with Durq below, feel free to ignore the following...
Originally Posted by Durquavian
I am no shill being I have NEVER not once bashed Intel users for their choice nor have I lied in order to make the outcome favorable to my standpoint. Making such accusations is in poor choice and likely not one you would ever make present. Beyond that lets make some valid points:
Almost ALL of your posts have overtones of AMD favoritism bursting from between the words. I've never once heard you mutter a fair and balanced contrast of AMD vs Intel solutions.
Fact: This is your first post and notice no inquiring not even acknowledging the original stated use.
We had a fair idea of the intended use, I made a build that would be faster, quieter, and more efficient for the intended use with higher quality parts at a lower cost. If I had pointed out a way to do the build better with a different AMD CPU you'd have absolutely no qualms.
I pointed out where the problem started and this isn't the only thread you have done this. Ask any poster, when the original build starts in the AMD or Intel thread it is in VERY POOR taste to go in and without provocation (the OP asking for the information contradictory to the original post build) giving said contradictory information.
You're holding the CPU brand choice in a "special light" that only exists in your head. Quit trying to "lock in" and "protect" the AMD selection like your first born's life depends on it. If it's in bad taste to point out a more competitive CPU choice, then it would have to be in equally bad taste to point out ANY ALTERNATIVE hardware. You're trying to sell me a false social convention on computer hardware recommendations set aside for only the CPU selection. Such an idea could only be born of an absolutely entrenched loyalist.
I don't condone others going into rival threads using negative language in the ruse of JUST giving information.
In your mind, ANYTHING sounds negative if it doesn't paint the AMD solution is the better solution. There's no way in the world that I could state that an i5 is a better CPU than Kaveri for this build without you thinking I've slayed an innocent bunny.
As far as Kaveri I do not think Good enough covers it. It is AMAZING. With the 14.3beta that sucker screams. It isn't the step down you like to pretend it is. At 45W it performs very well at nearly 80% of the 95W performance. The APU just releasedin Feb and it has only been 2 months, almost. The performance from launch has increased dramatically.
If you think that Kaveri offers a compelling solution for this build, you should start explaining your reasoning to the OP and stop complaining about me. The ability to "Set" Kaveri to a lower supported TDP is a great feature that gives AMD a better entry to the mainstream in many ways.
ANY unlocked CPU can deliver ~80% of it's performance at about half of it's rated TDP with performance tuning. The only thing that differentiates Kaveri in this matter, is that they included it as a "setting" that doesn't require any sophisticated tuning. It's my opinion that the main advantage to budget AMD hardware is the ability to performance tune it, where Intel has things locked down. For many users, the "TDP settings" will just be redundant as they would have achieved improved compute efficiency or performance or both through performance tuning anyway. When the user isn't going to be performance tuning at all, then Intel offers better "stock" performance and compute efficiency for most non-specialized workloads. Kaveri's "supported" reduced TDP settings are like a short-cut to alternative "tunes" that mainstream builds can take advantage of. It is compelling and interesting, but as I pointed out, even in this "reduced" TDP mode, it's still out-performed by Haswell.
NOW I WILL GO ON RECORD TO SAY THAT I HAVE NO ISSUE WITH THE i5 BUILD. I was not even trying to change the OPs mind, they made that decision and it isn't my place to tell them what they need to do. It saddened me that they had, more so when I felt the attempt at changing their decision to the i5 was poorly placed given the criteria.
Honestly evaluate what you have said here. I believe that you are very self conflicted about the issue. If your feelings are being effected by someone else's CPU choice then I believe you have some favoritism that is deeply entrenched in your psyche. It's probably not particularly healthy. I encourage you to find an appreciation and a positive attitude about ALL computer hardware. When you come full circle (which will involve a session of "nothing matters") you can get back to honestly comparing and contrasting hardware and spend less time being hurt by the likes of someone like me.
But I was not trying and nor did I the moment you started your debate. I have been following the thread since minute one just like the 8350 build, both of which I refrained from saying a word. But based on your assessment of me I should have been all over your first post, but I wasn't. Don't you think you are wrong about me?
I'd like to believe that there's great opportunity for you to develop a broader appreciation for computer hardware that has "happiness" at the door to all sorts of brand choices.Edited by mdocod - 3/29/14 at 7:26am