AMD Confirms Development of Consumer CPUs with Hybrid Architecture
Click here to post a comment for AMD Confirms Development of Consumer CPUs with Hybrid Architecture on our message forum
cucaulay malkin
Told you this had to happen, cant have x3d go against intels mainstream for the long run.
barbacot
Late to the party as with Nvidia?
Too bad - I hate big/small architecture.
cucaulay malkin
Performance per dollar and platform cost will be better.
moab600
Wonder how different it will be from intel's approach, but yeah a major change is needed.
wavetrex
CPU Market:
"Always two, there are. No more. No less. A Master and an apprentice."
H83
daffy101
I would like to see 6/8 big cores and the same amount of little cores. Wondering if GPUs will ultimately head this way as well.
barbacot
Undying
I also dislike it. I would rather see amd pushing more cores and more cache next generation.
Silva
This might be great for mobile devices, but what's the point for desktops when a 16c/32t chip isn't fully utilized by a game?
Single core speed is still king unfortunately, they should work in auto work parallelization so software utilizes cores more efficiently.
Laptops, tablets and phones all benefit from efficiency and every watt counts to keep them going. But for desktops, +/-10W make no difference (figuratively speaking).
Dribble
Single core speed is only king if you have a thread that will completely max out a core, for the rest then low power cores are a better use of silicon as they are more efficient (both in size and power usage). So in reality you need enough high power cores for whatever you are running (6-8) and the rest are better being lower power ones.
The only reason to not have low power cores is for software that can't work out how to use high power vs low power cores, but as Intel have made this standard for the last few years that should no longer be such an issue, although there will be the odd exception.
JamesSneed
barbacot
schmidtbag
E-cores in a desktop are perfectly fine so long as they're done right. If E-cores are substituting P-cores, that's a problem. So for example, if a 8600X were to have 4 P-cores and 4 E-cores, that's a reason to complain. If you don't lose any P-cores then great - we're basically getting free extra performance. If an 8600 non-X came with 4 P-cores and perhaps 8 E-cores, I think that would also be acceptable, since such a CPU isn't expected to do any heavy loads. Of course, this is assuming the scheduler is smart enough, but Intel and Qualcomm at this point have had a lot of time to get the Windows scheduler to catch up with asymmetric core configurations.
E-cores do serve a good purpose in desktop CPUs. If you really "need" that many P-cores, then you probably shouldn't be using a mainstream socket in the first place. In other words, if you're groaning about losing a couple P cores to a bunch of E-cores and think that's making a noteworthy dent in your workload, there's a good chance you should've been spending a lot more money on a lot more cores in the first place.
I'm also wondering this, because AMD's "P-cores" are (to my understanding) already a lot smaller than Intel's, and their chiplet design allows for relatively cheap production. I'm not sure how much smaller they really need to make things.
anticupidon
Krizby
Dribble
Alessio1989
NO THANK YOU. Intel already did a disaster with AVX-512 (well, actually intel did multiple disasters with AVX-512.. 1 is just the big.little bullshit). On desktop nobody cares spare 1w less in idle. Biggest consumptions are others.
On server side performance:watt are already winners. They should work instead on lowering MC latency and power.
barbacot
fantaskarsef
The issue with big / little architectures always was, is, and will be core parking and too aggressive power savings. THAT's why I don't want it, because I can already see that they will need multiple microcode / scheduler / bios updates until it works. And honestly, for the prices we pay, that's even worse, we're supposed to just use the hardware we buy.