AMD Readies FX-8370, FX-8370E Microprocessors.

[PPB]

I know Shintai from older days. He dont like AMD many years and criticizes "all" from AMD. I agree with some stuffs but not with all. I like AMD slightly more but I like Intel CPus too of course (5960x is really fan for benchamrking ).

Ywah the blue tinted glasses are pretty easy to spot on some dudes. But that guy is special, never seen someone use so many hyperboles to compensate for the lack of arguments.

 

[MiddleOfTheRoad]

bullshits...

The problem of FXs is FPU, cache latencies, some single thread (not all situations)... But you and me know, the FX series 15h is ending. No more 15h in future. 4 years old cyclus ending. FX is effective in multithread as multiplication of performance effective. Of course, not overall.
The idea is good, but results is in half hearted .

Im sure AMD team know it and working some months at new microarchitecture.

Yes and no -- under Windows definitely. But My i7 3770k and FX-8320 score within 100 points of each other in the floating point benchmark for BOINC/World Community Grid (on Ubuntu 64 bit). There really isn't anything wrong with the FX's floating point at the hardware level -- IMO it appears to be more how Windows software utilizes it. The cache latencies do suck in every situation I've seen, though.

 

[Abwx]

Seems that the review provided an answer about 3D particle movement, it is rather a memory bandwith and latency test actualy.
This can be seen by comparing the scores in CB R15 and in 3DPM, in the latter the 990X and his tri channel inflate the score hugely, so much for a a bench that is supposed to check the FP perfs...

Also quite weird is that the X6 does best Piledriver and this latter does better than the more recent Steamroller...

 

[Enigmoid]

3DPM is inflated by HT, its something like 80% gain.

 

[Abwx]

3DPM is inflated by HT, its something like 80% gain.

Then it means that a non threaded core utilisation rate is low, something like 60% if your number is accurate.

 

[Enigmoid]

Then it means that a non threaded core utilisation rate is low, something like 60% if your number is accurate.

Yeah.

3DPM is pretty poor. But it does illustrate that sometimes software really doesn't scale well (like coremark for example).