Arm Helium: New vector extension for the M-Profile Architecture
community.arm.comThe blog post https://community.arm.com/arm-research/b/articles/posts/maki... has more useful content.
So, Helium is similar to Neon, but optimized for simpler CPUs with fewer gates.
Meantime, in the royalty-free side, RISCV's work on V extension continues: https://www.embecosm.com/2018/09/09/supporting-the-risc-v-ve...
Hard to tell, with ARM's incredibly vague language, whether MVE is a vector extension, or just a packed SIMD extension.
Added: I'm sufficiently insulted by how tedious they've made it to download related documents, that I'll leave it to somebody else to care enough. [0]
I wonder what level of concern RISCV raises internally at ARM. Are they panicked, or just adjusting strategy in a more routine way?
It seems the moves from Western Digital would be concerning to them.
Though I suppose on the other side, there's not yet any credible/volume RISCV general purpose CPUs yet (just MCUs), so maybe they aren't yet scrambling.
Given their public reactions, I imagine they see it as an existential threat.
ARM has done an incredible job legitimizing RISC-V.
I downloaded the spec and took a look.
TLDR: it's pretty much a traditional short-register SIMD, but with the addition of predication, including handling the tail of random-length loops using the vector processing body (as in RISC-V and Cray), not an extra scalar loop as previously needed.
- provides 8 "Q" vector registers, always exactly 128 bits each
- overlays the FP register file (32 "S" registers of 32 bits each / 16 "D" registers of 64 bits each)
- MVE-I (8/16/32 bit integer) and MVE-F (16 and 32 bit FP) subsets.
- architecturally defined to execute each vector instruction in 4 beats
- 1, 2, or 4 beats per "architecture tick", and can vary during execution. An "architecture tick" might or might not be 1 clock cycle.
- two forms of predication, each with its own mask: "loop tail predication", which is like RISC-V/Cray "vl" (but described as a mask), and "VPT predication" for data-dependent conditions. The two masks are ANDed together.
- A VPT block is defined as the n instructions following a VPT or VPST instruction, where n <= 4
- can be predicated with the condition or the inverse of the condition. Similar to the existing If/Then/Else predicated execution. e.g. VPT, VPTT, VPTE, VPTTE, VPTEE, VPTEEE variants.
- "VPT can be considered as the vectorized combination of CMP and IT"
- predication is per-byte regardless of the element size.
- loads set predicated-off bytes to 0, other instructions leave them untouched
- VLD2/VLD4 and VST2/VST4 are provided for interleaving/deinterleaving. Each instruction always loads/stores exactly 128 bits to/from 2 or 4 consecutive Q registers.
- there is also scatter/gather
- there are some fancy operations. e.g. VCADD: Vector Complex Add with Rotate. This instruction performs a complex addition of the first operand with the second operand rotated in the complex plane by the specified amount, either 90 or 270 degrees. Also VCMLA: Vector Complex Multiply Accumulate.
Fun times, can't wait to have these in the Cortex M line. Of course the chips start to look more and more like the Pentium line of old.
The current dsp instructions are pretty limited, interested in seeing these!