Where is VPERMB in AVX2?

by

I’m 99% sure the main factor is transistor cost of implementation. It would clearly be very useful, and the only reason it doesn’t exist is that the implementation cost must outweigh the significant benefit.

Coding space issues are unlikely; the VEX coding space provides a LOT of room. Like, really a lot, since the field that represents combinations of prefixes isn’t a bit-field, it’s an integer with most of the values unused.

They decided to implement it for AVX512VBMI, though, with larger element sizes available in AVX512BW and AVX512F. Maybe they realized how much it sucked to not have this, and decided to do it anyway. AVX512F takes a lot of die area / transistors to implement, so much that Intel decided not to even implement it in retail desktop CPUs for a couple generations.

(Part of that is that I think these days a lot of code that can take advantage of brand new instruction sets is written to run on known servers, instead of runtime dispatching for use on client machines).

According to Wikipedia, AVX512VBMI isn’t coming until Cannonlake, but then we will have vpermi2b, which does 64 parallel table lookups from a 128B table (2 zmm vectors)). Skylake Xeon will only bring vpermi2w and larger element sizes (AVX512F + AVX512BW).

I’m pretty sure that thirty two 32:1 muxers are a lot more expensive than eight 8:1 muxers, even if the 8:1 muxers are 4x wider. They could implement it with multiple stages of shuffling (rather than a single 32:1 stage), since lane-crossing shuffles get a 3-cycle time budget to get their work done. But still a lot of transistors.

I’d love to see a less hand-wavy answer from someone with hardware design experience. I built a digital timer from TTL counter chips on a breadboard once (and IIRC, read out the counter from BASIC on a TI-99/4A which was very obsolete even ~20 years ago whe), but that’s about it.

It’s pretty clear that the SSE PSHUFB instruction is pretty much among the most useful instructions of all time.

Yup. It was the first variable-shuffle, with a control mask from a register instead of an immediate. Looking up a shuffle mask from a LUT of shuffle masks based on a pcmpeqb / pmovmskb result can do some crazy powerful things. @stgatilov’s IPv4 dotted-quad -> int converter is one of my favourite examples of awesome SIMD tricks.

More Related Contents:

  1. Micro fusion and addressing modes
  2. How exactly do partial registers on Haswell/Skylake perform? Writing AL seems to have a false dependency on RAX, and AH is inconsistent
  3. What are the best instruction sequences to generate vector constants on the fly?
  4. How to read the Intel Opcode notation
  5. What is the stack engine in the Sandybridge microarchitecture?
  6. What is a Partial Flag Stall?
  7. Slow jmp-instruction
  8. What was the original reason for the design of AT&T assembly syntax?
  9. What is the point of SSE2 instructions such as orpd?
  10. Is there a complete x86 assembly language reference that uses AT&T syntax? [closed]
  11. Why does Intel hide internal RISC core in their processors?
  12. Using ymm registers as a “memory-like” storage location
  13. Difference between MOVDQA and MOVAPS x86 instructions?
  14. In what situation would the AVX2 gather instructions be faster than individually loading the data?
  15. Can PTEST be used to test if two registers are both zero or some other condition?
  16. Does Skylake need vzeroupper for turbo clocks to recover after a 512-bit instruction that only reads a ZMM register, writing a k mask?
  17. Referencing the contents of a memory location. (x86 addressing modes)
  18. What is the meaning of “non temporal” memory accesses in x86
  19. How do you use gcc to generate assembly code in Intel syntax?
  20. What is stack frame in assembly?
  21. Size of store buffers on Intel hardware? What exactly is a store buffer?
  22. How to generate assembly code with clang in Intel syntax?
  23. gas: too many memory reference
  24. How to multiply a register by 37 using only 2 consecutive leal instructions in x86?
  25. What is the latency and throughput of the RDRAND instruction on Ivy Bridge?
  26. Convert string to int. x86 32 bit Assembler using Nasm
  27. Segment size in x86 real mode
  28. Addressing Modes in Assembly Language (IA-32 NASM)
  29. What’s difference between number with $ or without $ symbol in at&t assembly syntax?
  30. Is processor can do memory and arithmetic operation at the same time?

Leave a Comment