By using this site, you agree to our Privacy Policy and our Terms of Use. Close
fatslob-:O said:
Pemalite said:
Captain_Tom said:

I wish IBM would just jump in and stomp Intel.  They currently have an 8-core cpu and through hyperthreading each core can read 12 lines of code each for a total of 96 threads!  Oh and it runs above 4 GHz...


Intel's cores are easily better than IBM's, remember, clockspeed became irrellevant years ago for comparing CPU performance.
If IBM truly did have a powerhouse of a CPU, then they would be controlling a much much much larger share of the lucrative server pie, which values absolute performance over power in many scenario's.

Also, Intel has 8 and 16 core Xeon's.

It's not all about having higher IPC's either ... Once you need ridiculously large vectors that demand AVX-1024 the IPC argument becomes totally worthless as most applications are simply don't have this structure to take advantange of the high amount of instruction level parallelism. 

http://abinstein.blogspot.ca/2010/09/ipc-myths.html

Despite the fact that abinstein has a clear bias for AMD he's absolutely right about this issue. (This abinstein is probably the same one from AMDZone.) 

The only way we could probably realistically increase single threaded performance in the future is that we would basically need to increase the clocks or get instruction set extensions dedicated to increasing single threaded performance. 

I'm not sure if you would consider this to be a powerhouse but here it goes ... Each Power8 core has 2 VMX units and likely 4 single precision floating point units which basically means that the processor at hand is capable of 12 Fused Multiply-Add instructions so each core can pull off 24 flops/cycle. With 12 cores all clocked at 5 Ghz would mean that it can do 1.44 Teraflops tops! No that is not small and I think this may have being the first commercial CPU to do that. (Aside from the xeon phi ofcourse.)

I would agree with you about IBM controlling alot more of the server market than it has now because of more power but the tools and everything else is so much better on the x86 side of things. 


I wasn't actually talking about IPC specifically. :P
Intel and AMD as well as to a lesser extent, via... Actually spend more transisters on trying to hide latency and bandwidth deficits with various technologies than actuall compute hardware.
And for a good reason!

When IBM had it's PowerPC based processors inside the Mac, there was allot of flamboyant chit-chat with enthusiasts of how superior IBM PowerPC based processors were when compared to x86, in the end, when Apple switched processor architectures, there was stupidly massive gains in almost every area.
Granted, that was many many years ago and architectures of both camps have changed drastically since then.

Unfortunatly, we can't run 3D mark to find out how they compare. :P

Captain_Tom said:

You call this whipping the floor:

http://www.bf4blog.com/battlefield-4-retail-gpu-cpu-benchmarks/

http://www.tomshardware.com/reviews/core-i7-4960x-ivy-bridge-e-benchmark,3557-5.html

Like I said it is inbetween the i5's and i7's.  I don't know what benchmarks you are running.  I never claimed it would win in lazily threaded apps.


Anyone that has played a frostbite powered game, will know that even though it can utilise a ton of threads, doesn't mean it will utilise them all to it's greatest extent.
When CPU benchmarks like that are compressed, that usually implies a GPU bottleneck.

However, if you *really* wan't to see an Intel Core i7 hex core stretch it's legs whilst gaming, then run a heavily threaded game like Battlefield, throw an encoding job into the mix on a couple of cores, run Xsplit and a few other programs to bog the CPU down a bit.

An AMD system, regardless of it's clocks, will fold, it's the MAIN reason I shifted to a Core i7 3930K in the first place on my primary machine and in most instances, when all 12 threads are utilised I can see performance gains of in-excess of twice my Phenom 2 x6 and old FX 8120 systems.

Shall I run some benchmarks? I do have all the rigs in front of me.


Captain_Tom said:
So you just don't think it can compete at all?  I mean I'll just agree to disagree on that one lol.  Oh and IBM has 12 core CPU's at 5 GHz that run 12 threads each (On 22nm).  If you don't think that would stomp a measily 30 thread  (It has 15 cores buddy) at 3.8 GHz you really are an Intel Fanboy.


I was positive that calling someone a fanboy, even implied was a wrongfull offense on this forum?
Besides, how am I fanboy when I own more AMD systems than Intel? Or more ARM powered devices than x86? A fanboy implies that I would buy into a certain product line regardless of the competion, which simply isn't the case.


And you're putting words into my mouth, at this point there is no point continuing this discussion as it's moved beyond intelligent discussion and started to drop down into name calling.



--::{PC Gaming Master Race}::--