Personal computing discussed
Moderators: renee, Flying Fox, morphine
LostCat wrote:I'd guess they've delayed the six core parts just to have more stock of the stuff people are more likely to buy right away.
Welch wrote:It would then cement that 4 core parts are the new norm.
sophisticles wrote:Welch wrote:It would then cement that 4 core parts are the new norm.
Did you mean "It would then cement that 8 core parts are the new norm"?
zzz wrote:I hate to be be that guy but: Cores and threads are meaningless without knowing how they perform individually or grouped and we all know what apps actually use multicore cpu's effectively. Even then, we're not buying the CPU on raw numbers: what added parts like AVX does it support? VT-d? FBRAM? The linked article doesn't even specify how many lanes of PCIE or where they're coming from (chipset or cpu), just that (which is that they're 3.0 but in reality it's going to be a mix of 2 and 3.)
You're not just buying a cpu, you're buying the motherboard it sits in and both of those give you a feature-set and performance, buy accordingly.
whm1974 wrote:LostCat wrote:I'd guess they've delayed the six core parts just to have more stock of the stuff people are more likely to buy right away.
Personally I don't see the point of even having a 6 core CPU if the 8 is just two 4 core processors glued together. Just have a 4 and 8 core ones
DancinJack wrote:Those prices just make no sense. If their IPC was on par with what we saw in the demos (caveats included), why on Earth would they be selling an 8C/16T part for 400 bucks? It just doesn't compute.
Also, has anyone heard ANYTHING about die size for an 8C/16T part? I can't imagine that's a small die, and Intel is still selling its cheapest 6C/12T part for over 400. AMD needs to stop playing the budget game with their top parts unless of course they suck.
synthtel2 wrote:Bleh. I'd rather buy 6C12T than 8C8T (not that it matters since I'll probably go 8C16T anyway), and even if it were still a stupid market segmentation move, at least with a 6C option they've got plausible deniability that it's just binning. (I presume that the absence of a 6C option means yields are good.)
Vhalidictes wrote:(Initial Release) RyZen doesn't have a iGPU. That means that the final die size could be a lot smaller than we've come to expect from Intel.
Lordhawkwind wrote:Reports are coming out that prices are more likely to be from $580 to $720. IMHO there's no way AMD would be able to offer any Ryzen chips at the prices the OP is stating.
LostCat wrote:Lordhawkwind wrote:Reports are coming out that prices are more likely to be from $580 to $720. IMHO there's no way AMD would be able to offer any Ryzen chips at the prices the OP is stating.
I haven't seen any credible report that prices would be that high or what's in the OP. At this point it seems more like everyone wants to speculate without any actual data to go off of.
derFunkenstein wrote:Oh, I thought those prices came from the previous article by the author linked in the article... But I see they are different...There will still be dual-core, quad-thread parts. They'll just be Raven Ridge APUs where that configuration makes more sense.
Also, where did the prices in the OP come from? Wet dreams? They don't appear in the article. Just someone else in a hurry to run AMD out of money, it seems.
drfish wrote:AMD is gunning for Broadwell-E, right? We have low-ish clocks, purportedly similar IPC, and the same number of cores. IMO that means the best case scenario is that we get 6800K performance for $350, 6850K for $550 and 6900K for $999. I think that's going to leave a lot of people doing what I already have, picking up a high-clock quad from Intel for gaming. If you're not as concerned with gaming, or if you believe the Vulkan/DX12 will save the world, then Ryzen could be a good option for heavy-lifting and "good enough" for gaming. We'll see...
Mr Bill wrote:Either the Forbes article is confused, or
MOSFET wrote:Mr Bill wrote:Either the Forbes article is confused, or
I think we have a winner. I don't dislike the author, but he doesn't have any information that the readers here don't have. Just more conjecture and headlining.
Krogoth wrote:It is possible that "6-core" chips aren't feasible due to how the Ryzen silicon is built.
Krogoth wrote:It is possible that "6-core" chips aren't feasible due to how the Ryzen silicon is built.