Excluding the current cpu generation, was there any point in time where a brute force approach took the performance crown over a more efficient architecture?
>>61199376
Are we talking something like a lower IPC 2ghz CPU beating a higher IPC 1ghz one, or something like Bulldozer shoving twice as many cores as Intel into the same price range and winning in multithreaded workloads but failing horribly in singlethreaded?
>what is netburst
>>61199472
Well the moar coarz race has only seen a couple gens so far (if we consider the side with the least recent update cycles), so I mean mostly about focus on GHz vs. focus on IPC (a very simple way to put it, but you know what I mean). At least in recent times I remember, when we had a competing Speed Demon architecture and a Brainiac architecture from either side, Speed Demon lost on both accounts (P4 vs. Athlon XP/64, and Bulldozer vs. Nehalem and forward).
>>61199376
>He doesn't remember the GHz wars
dark times those were.Both wonderous and terrible.
>>61199564
I remember it. I mean when each side took a different approach.
>>61199376
Not for CPUs per se, but the PowerVR Kyro GPU series for PC in 2000/2001 were utterly annihilated by Nvidia and ATi's offerings despite the Kyro being way way more efficient.
Like, a Kyro could match Nvidia/ATi GPUs that had 3 times higher clock/memory speeds.
>>61199564
>delicious ceramic package
Now Intel uses shitty thin PCB's that bend themselves. Disgusting.
>>61199606
They actually had a great product,way ahead,but you dont comme with empty pockets in war where land is already divided and forces entrenched.
Honestly, i wouldnt mind more tthan 2 big guys(3 if you count intel) , if voodoo,maxton?!,powerVR. were still here, we would have way faster,cheaper gpus, and DirectX D3D wouldnt have so much say in gayman world(opengl and linuxwould actuallyy matter in gaming world).
At least powervr is making great mobile gpus.