Unofficially, it is the Basilisk APU. Nothing is known about it except that it is after the Carrizo APU.Also, whats coming after that? I don' recall any successor having been announced.
Unofficially, it is the Basilisk APU. Nothing is known about it except that it is after the Carrizo APU.
AMD hasn't shown any signs of a fifth generation dozer. Which then it is a very safe assumption to assume Basilisk is Excavator also.Bazilisk is Excavator TOO?
No, there should be a fifth generation dozer after Excavator. It just hasn't been disclosed publicly as of yet.Ok, so is there a new family coming after the dozer family?
AMD hasn't shown any signs of a fifth generation dozer. Which then it is a very safe assumption to assume Basilisk is Excavator also.
Dozer family ends with Excavator.
Cat family ends with Margay.
Islands ends with Pirate.
Lets believe in AMD. AMD Turned junk(R600) to gold(R700) one time. Lets believe they can fix Bulldozer.
Yes, there were rumors about it . Codename "Zen".Ok, so is there a new family coming after the dozer family?
Yep. I tried to guesstimate Steamroller's performance and failed miserably. I was off by a good 100%+ on IPC. The architecture is just so broken that what would normally be a huge improvement results in a headfirst slam into another bottleneck. At this point, the issue is either with the stupid high cache and memory access latencies, or the whole thing is unsalvageable.
AMD is severely lacking in performance. 20nm provides more performance. New nodes are just as big of a deal as architecture updates, if not more so. 20nm is a bit of an outlier in this regard because of the somewhat higher cost, but the performance benefits are still there.
However, 20nm at GloFo is not a high performance process. The only flavor of 20nm they offer is 20nm Low Power Mobility.
Lets begin the speculation!(Again)
Carrizo:
-same 28nm process, 10-20% bigger clocks.
-less(10 or 20 percent lesser) die size, less power consumption at the same performance level;
-65w TOP APU CPU part is 15-20% faster than A10-7850k CPU, 65w TOP GPU part is 20% faster than A10-7850k GPU;
-45w TOP APU CPU part is 5-10% slower than A10-7850k on both CPU and GPU sides.
Doubtful. High density libraries limit maximum clock speed, not increase it. Expect the CPU to perform equivalently to Kaveri, if not potentially a bit slower, but with a lower TDP. Unless they use the extra die space to beef up the iGPU, though on FM2+ the GPU is constrained by bandwidth.
Doubtful. High density libraries limit maximum clock speed, not increase it.
Yep. True. But IPC gains we still don't know.
HDL/HDC reduces the energy required to do an operation; it also shrinks the area and reduces heat leakage.High density libraries limit maximum clock speed, not increase it.
Kaveri:Lets begin the speculation!(Again)
Carrizo:
-same 28nm process, 10-20% bigger clocks.
Basilisk:
-16nm FinFet process node;
Lets begin the speculation!(Again)
Carrizo:
-same 28nm process, 10-20% bigger clocks.
-less(10 or 20 percent lesser) die size, less power consumption at the same performance level;
-65w TOP APU CPU part is 15-20% faster than A10-7850k CPU, 65w TOP GPU part is 20% faster than A10-7850k GPU;
-45w TOP APU CPU part is 5-10% slower than A10-7850k on both CPU and GPU sides.
Basilisk:
-16nm FinFet process node;
-close to Carrizo Die Size, bigger clocks, much less power consumption at the same performance level;
-65w TDP APU is 30-40% faster than Carrizo CPU side and 50-80% faster at Carrizo GPU side;
-45w TDP APU is 10-15% faster than TOP APU Carrizo on both sides.
Lets begin the speculation!(Again)
Carrizo:
-same 28nm process, 10-20% bigger clocks.
-less(10 or 20 percent lesser) die size, less power consumption at the same performance level;
-65w TOP APU CPU part is 15-20% faster than A10-7850k CPU, 65w TOP GPU part is 20% faster than A10-7850k GPU;
-45w TOP APU CPU part is 5-10% slower than A10-7850k on both CPU and GPU sides.
Maybe we'll see something much better at 20nm, but that will likely be going up against a mature 14nm Skylake CPU. Maybe by the time Win9 comes out, all the XP people will finally upgrade and give AMD a little bit of breathing room.
Why would people migrating from XP go for AMD? Given that a lot of people and companies are replacing desktops with notebooks, I'd say that people upgrading will mean even less market share for AMD.
And why do you think IPC gains will come for free? Making units bigger/more complex also hurts power consumption, which either demands higher TDP or lower clocks. Adding more units even more, so some IPC gains might actually hurt more AMD than it helps.
...in the process node and in power management, the former is impossible because they do not have any control over the process design flow
and the latter is also impossible, because AMD spends today about two thirds of what it used to spend when the Bulldozer pipeline was full steam ahead.
Carrizo:
28nm-A(HPP/SLP) (Standard Bulk)
2.592 Billion transistors
Well, if for real, 10-20% higher clocks would be a nice bump. I just doubt we'll see both a smaller die *and* higher clocks.
As has been pointed out, AMD's APU line is in definite need of more bandwidth. A decent size L3$ or small e-dram LLC would help, but would cost more in terms of effective die size. Multi-channel DRAM would be a big help, be seems to be a no go for the lower cost systems that APUs go into.
Maybe we'll see something much better at 20nm, but that will likely be going up against a mature 14nm Skylake CPU. Maybe by the time Win9 comes out, all the XP people will finally upgrade and give AMD a little bit of breathing room.
Doubtful. High density libraries limit maximum clock speed, not increase it. Expect the CPU to perform equivalently to Kaveri, if not potentially a bit slower, but with a lower TDP. Unless they use the extra die space to beef up the iGPU, though on FM2+ the GPU is constrained by bandwidth.