DPDK patches and discussions
 help / color / mirror / Atom feed
From: "Varghese, Vipin" <Vipin.Varghese@amd.com>
To: Thomas Monjalon <thomas@monjalon.net>, dpdk-dev <dev@dpdk.org>
Cc: "Song, Keesang" <Keesang.Song@amd.com>
Subject: RE: DPDK release candidate 25.07-rc3
Date: Wed, 16 Jul 2025 10:05:29 +0000	[thread overview]
Message-ID: <PH7PR12MB8596207D1DBFF1D7A4DAE98D8256A@PH7PR12MB8596.namprd12.prod.outlook.com> (raw)
In-Reply-To: <086c9e27-466a-43b1-86e7-8dc4095512d0@linux.ibm.com>

[AMD Official Use Only - AMD Internal Distribution Only]

Hi Thomas,

Please find the AMD EPYC testing results for both rc-1 and rc-2 shared below

RC-1:

- tuning guide: https://doc.dpdk.org/guides/linux_gsg/amd_platform.html
- OS: Ubuntu 24.04.2, 6.8.0
- Grub: default_hugepagesz=1GB isolcpus=7-63,71-127 processor.max_cstate=0 nohz_full=7-63,71-127 rcu_nocbs=7-63,71-127 amd_iommu=on hugepagesz=1GB hugepages=12 irqaffinity=0-6,64-70
- Processor: AMD EPYC 9334 32-Core Processor, 2.7GHz (boost off)
- Memory: 4800MT/s,  4 DIMM
- Compiler: gcc 14.2
- NIC: CQDA2 Intel E810, firimware: 3.20, DDP: comms 1.3.53
- Application: testpmd, ICE_16B_DESC, AVX512 SIMD Bit Width
- Scenarios: io, mac-swp, 5tswp
- TX-RX descriptor: 2048
- packet size 64B

Diff: MPPs 25.07-rc1 vs 25.03 for burst size 32, 64, 128, 256, 512
++++++++++++++++++++++++++++++++++++++++++++++++++
 - IO: -0.23, -1.02, -2.81, -3.44, -2.11
 - MAC-SWP: -5.08, -2.6, -3.28, -2.07, -2
 - 5tswap: -2.39, -1.73, -1.95, -1.9, -1.9

RC-2:

- tuning guide: https://doc.dpdk.org/guides/linux_gsg/amd_platform.html
- OS: Ubuntu 24.04.2, 6.8.0
- Grub: default_hugepagesz=1GB isolcpus=7-63,71-127 processor.max_cstate=0 nohz_full=7-63,71-127 rcu_nocbs=7-63,71-127 amd_iommu=on hugepagesz=1GB hugepages=12 irqaffinity=0-6,64-70

- Processor: boost off
Turin: AMD EPYC 9755 128-Core Processor, 2.7 GHz(Boost off)
Genoa: AMD EPYC 9654 96-Core Processor, 2.4 GHz(Boost off
Siena: AMD EPYC 8534P 64-Core Processor,2.3 GHz(Boost off)
Milan: AMD EPYC 7713 64-Core Processor,2.0 GHz(Boost off)

- Memory:
Turin: 6400 MT/s
Genoa: 4800 MT/s
Siena: 4800 MT/s
Milan: 3200 MT/s

- Compiler: gcc 14.2
- NIC: CQDA2 Intel E810, firimware: 3.20, DDP: comms 1.3.53
- Application: testpmd, ICE_16B_DESC, AVX512 SIMD Bit Width
- Scenarios: io, mac-swp, 5tswp
- TX-RX descriptor: 2048
- packet size 64B

Diff: MPPs 25.07-rc2 vs 25.03 for burst 64
++++++++++++++++++++++++++++++++
Category: Turin: Siena: Genoa: Milan
 - TestPMD IO: 0.5%, -3.6%, 3.7%, -4.9%
 - L3FWD: 0.0%, -7.4%, -4.8%, -4.3%


Crypto+HMAC Perf:
++++++++++++++++

i. AES-CBC-128/SHA1-HMAC (Gbps) 
ii.AES-CBC-128/SHA2-256-HMAC (Gbps) 
iii.AES-GCM-128
iv.AES-CTR-Ciphering-Authentication
v.AES-CTR-Authentication-Ciphering
vi.AES-CTR-Ciphering-Encryption
vii.AES-CTR-Ciphering-Decryption

Variation in Cycles/Buffer:
-----------------------------------

GenoaAlgorithm, Buffer Size , cycles/buf variation for 32 burst for 1C1T, 1C2T, 8C16T
AES-128-CBC-SHA1, 256B, less than -1.5%
AES-128-CBC-SHA1, 512B, less than -1.1%
AES-128-CBC-SHA1, 1024B, less than -1.5%

Turin Algorithm, Buffer Size , cycles/buf variation for 32 burst for 1C1T, 1C2T, 8C16T
AES-128-CBC-SHA1, 512B, less than -2%
AES-128-CBC-SHA1, 512B, less than -2%
AES-128-CBC-SHA1, 1024B, less than -2%

      reply	other threads:[~2025-07-16 10:05 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-07-09 21:26 Thomas Monjalon
2025-07-13 14:52 ` Ali Alnubani
2025-07-15 14:15 ` Thinh Tran
2025-07-16 10:05   ` Varghese, Vipin [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=PH7PR12MB8596207D1DBFF1D7A4DAE98D8256A@PH7PR12MB8596.namprd12.prod.outlook.com \
    --to=vipin.varghese@amd.com \
    --cc=Keesang.Song@amd.com \
    --cc=dev@dpdk.org \
    --cc=thomas@monjalon.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).