nlnet_2022_opf_isa_wg.mdwn

   1 # NL.net proposal
   2
   3 * 2022-08-051
   4 * [[nlnet_20222_opf_isa_wg/discussion]]
   5
   6 ## Project name
   7
   8 Libre-SOC OpenPOWER ISA RFCs
   9
  10 ## Website / wiki
  11
  12 <https://libre-soc.org/nlnet_2022_opf_isa_wg>
  13
  14 Please be short and to the point in your answers; focus primarily on
  15 the what and how, not so much on the why. Add longer descriptions as
  16 attachments (see below). If English isn't your first language, don't
  17 worry - our reviewers don't care about spelling errors, only about
  18 great ideas. We apologise for the inconvenience of having to submit in
  19 English. On the up side, you can be as technical as you need to be (but
  20 you don't have to). Do stay concrete. Use plain text in your reply only,
  21 if you need any HTML to make your point please include this as attachment.
  22
  23 ## Abstract: Can you explain the whole project and its expected outcome(s).
  24
  25 The current NLnet funding to date has allowed Libre-SOC to develop
  26 one of the most powerful Scalable Vector ISAs in the world.
  27 The 25-year-old Power ISA, developed and curated by IBM, was
  28 transferred to the OpenPOWER Foundation, and is the basis on
  29 which, with NLnet EU funding, we have based
  30 Simple-V, the Draft Scalable Vector Extension.
  31
  32 Simple-V *needs* to be submitted to the OPF ISA Working Group,
  33 for formal discussion and inclusion. Given that it is 380
  34 pages we expect this to be done carefully and incrementally.
  35 https://ftp.libre-soc.org/simple_v_spec.pdf
  36
  37 However the
  38 process of submitting RFCs (Requests For Change), at the time of writing,
  39 still has not been publicly announced and opened up. We expect it
  40 to be very soon, but obviously could not begin any RFC Submission
  41 as part of earlier NLnet funding.  The timing is now right.
  42
  43 We will become publicly informed very shortly of the procedures but anticipate
  44 it to include development and submission of Compliance Test Suites
  45 (already partly covered by Simple-V unit tests, kindly funded by NLnet)
  46 as well as ongoing work on the Simulator.
  47
  48 # Have you been involved with projects or organisations relevant to this project before? And if so, can you tell us a bit about your contributions?
  49
  50 A lot! a full list is maintained here <https://libre-soc.org/nlnet_proposals/>
  51 and includes
  52
  53 * the world's first FOSSHW IEEE754 Formal Correctness Proofs for fadd, fsub, and fma, with support for FP Formal Proofs added to symbiyosis;
  54 * the world's first in-place Discrete Cosine Transform algorithm;
  55 * Significant improvements to Europe's only silicon-proven FOSSHW VLSI toolchain (coriolis2, by LIP6 Labs of Sorbonne University)
  56  to do an 800,000 transistor fully automated RTL2GDSII
  57 tape-out;
  58 * development of a 180nm Power ISA 3.0 "Test ASIC", the largest fully FOSSHW
  59   ASIC ever taped-out in Europe (and funded by Horizon 2020)
  60 * development of an Interoperability "Test API" for Power ISA systems,
  61   with thousands of unit tests.
  62
  63 and much more. The side-benefits alone for EU citizens are enormous.
  64
  65 # Requested Amount
  66
  67 EUR 100,000.
  68
  69 # Explain what the requested budget will be used for?
  70
  71 Time and resource, primarily manpower, to prepare and submit the documentation
  72 to OPF. To give us legal compliance for the development
  73 work carried out over the past four years, as part of the
  74 transfer to the OpenPOWER Foundation.
  75
  76 * ongoing communication with the OpenPOWER Foundation ISA Working Group
  77 * preparation of a large number of RFCs (380 pages total so far) through
  78   the External RFC Process
  79 * for each RFC accepted, work needs to be done with IBM to submit Power ISA Spec
  80   changes
  81 * for each RFC accepted, a Compliance Test Suite must be written
  82 * for each Compliance Test Suite written the results must be
  83   confirmed correct by inspection (hence the Simulator) which has
  84   as we already discovered been quite a lot of work
  85 * Along the way we aim to continue developing the "Test API" which
  86   allows running thousands of unit tests on multiple systems and
  87   cross-checking the results.  Currently we have Simulator, some
  88   "Expected Results", and the Libre-SOC HDL as well as qemu.
  89   We aim to add cavatools, gem5, Microwatt and stand-alone binary
  90   auto-generation for running on IBM POWER9 as well as Libre-SOC
  91   and Microwatt FPGAs.
  92
  93 # Compare your own project with existing or historical efforts.
  94
  95 We are developing a Cray-style Scalable Vector ISA Extension for
  96 the Supercomputing-class Power ISA.  Similar historic ISAs include
  97 Cray Y/MP, ETA-10, Cyber CDC 205. More recent is the NEC SX Aurora.
  98 They are all proprietary systems: Libre-SOC's efforts are entirely
  99 FOSSHW.
 100
 101 Open Scalable Vector ISAs include MRISC32/64 (in early development) and
 102 RISC-V RVV. Advocates of RISC-V have been discovering to their dismay
 103 that RVV and RISC-V ISA has fundamental design issues that cannot be fixed.
 104 Additionally, submission of RISCV ISA modifications requires RISCV Foundation
 105 Membership which puts us under impossible conflict of interest with
 106 Full Transparency Conditions not only with NLnet but also with
 107 EU Auditing Requirements.  By direct contrast OPF External RFC Submission
 108 does not require Secrecy.
 109
 110 ## What are significant technical challenges you expect to solve during the project, if any?
 111
 112 The main challenge is one of communication. The majority of the technical
 113 development has been done thanks to NLnet
 114 but it was so complex and comprehensive that it risks overwhelming the ISA
 115 WG Members, whose primary driver has of course been IBM for the past 25
 116 years.
 117
 118 Libre-SOC proposes taking the Power ISA into mainstream computing,
 119 including Video Decode, 3D, GPU workloads, cryptography, and Desktop
 120 and Portable devices, all of which are far different from IBM's traditional
 121 Mainframe-style multi-billion-dollar Supercomputing business.
 122 We therefore have to be both deeply respectful of their achievements, and
 123 non-disruptive to their customer base, but
 124 also appropriately assertive now that the ISA is managed by the OpenPOWER
 125 Foundation.
 126
 127 ## Describe the ecosystem of the project, and how you will engage with relevant actors and promote the outcomes?
 128
 129 Partly covered above, Libre-SOC is exclusively FOSSHW and full transparency
 130 is paramount. That said we recognise that no FOSSHW team is going to
 131 manufacture FOSS ASICs in 7nm (unless several billion dollars is available
 132 to buy a Foundry and open up its PDK). To that end RED Semiconductor Ltd
 133 has been formed by us as an Independent Entity,
 134 which will commercialise Libre-SOC's designs and handle
 135 any Commercially-confidential matters that a Transparency-committed
 136 FOSSHW team simply
 137 cannot. Thus, RS will join the OpenPOWER Foundation and help ensure,
 138 from the "other side of the fence", that matters progress smoothly
 139 for IBM and other OPF Members.
 140
 141 RED Semiconductor Ltd will the commercial point of contact for Simple-V
 142 where Organisations are unable to deal with FOSS Entities. This maximises
 143 the broad market benefit of the technology, in line with European Objectives.
 144
 145 We are already set to submit presentations through multiple Conferences
 146 as has been ongoing since 2019 as can be seen at <https://libre-soc.org/conferences> and will continue to submit press releases to
 147 OPF <https://openpowerfoundation.org/blog/libre-soc-180nm-power-isa-asic-submitted-to-imec-for-fabrication/>. Our entire development is public
 148 so is accessible to all.
 149
 150 # Extra info to be submitted
 151
 152 the budget is high because we honestly do not know yet how much work
 153 IBM and the ISA WG expects us to do.  we do however know that there
 154 will be announcements very soon. If it turns out to be less work
 155 we are more than happy to go with a proportionately smaller budget.