Note: Version R10a (aka 31.78) has replaced R10 (aka 31.76-77).
We are pleased to release R10a (version 31.78) update for Sandra 20/21 with the following changes:
We are releasing more benchmark updates for ARM/Arm64 after additional testing with support for additional hardware as tested in the field.
We are also announcing the development of the Apple Mac (OS X) version of Sandra – to benchmark those M1-class Arm64 processors! Now that we converted the code to Arm64, we can just use it on other platforms, e.g. Android, Mac, etc. We hope to be going public by the end of the year (2022 that is 😉
Is x86 dead? Well, that is for you, the users/customers to decide. Vote now! 😉
Please don’t forget to submit benchmark results to the Official SiSoftware Ranker! Many thanks for your continued support.
And please, don’t forget small ISVs (independent software vendors) like ourselves in these very challenging times. Please buy a copy of Sandra if you find our software useful. Your custom means everything to us!
Hardware Support, Updates & Fixes
- CPU Multi-Media Benchmarks
- Integer (Int16 / Int32) ARM/Arm64 128-bit NEON Benchmarks
- ~+41% Int16 (16×16 > 32) improvement [Pi4: 59MPix/s vs. 36.7]
- ~+28% Int32 (32×32 > 64) improvement [Pi4: 35MPix/s vs. 24.8]
- ~+5% Int64 (64×64 > 128) improvement [Pi4: 3.4MPix/s vs. 3.25]
- ~+41% Int16 (16×16 > 32) improvement [Pi4: 59MPix/s vs. 36.7]
- Preliminary Arm64 SVE/SVE2 Benchmarks
- Integer (Int16 / Int32) ARM/Arm64 128-bit NEON Benchmarks
- CPU Cryptography Benchmarks
- Multi-Buffer ARM/Arm64 128-bit NEON Hashing Benchmarks
- ~+100% improvement (SHA2-256, SHA1)
- Preliminary ARM/Arm64 SHA HWA (hardware accelerated) support (SHA2-256, SHA1)
- Resolved x86/x64 SHA HWA failing to engage – resulting in low performance for AMD Ryzen, Intel ADL/RKL (SHA2-256, SHA1) [Sandra 31.78]
- Multi-Buffer ARM/Arm64 128-bit NEON Hashing Benchmarks
- Memory & Cache Benchmarks
- Buffered ARM/Arm64 128-bit NEON Integer/Floating-Point Benchmarks
- ~~+33% bandwidth improvement [Pi4: 8GB/s vs. 6GB/s]
- ~~+33% bandwidth improvement [Pi4: 8GB/s vs. 6GB/s]
- Preliminary Arm64 SVE/SVE2 Benchmarks
- Buffered ARM/Arm64 128-bit NEON Integer/Floating-Point Benchmarks
- CPU Inter-Thread/Core/Module/Node/Package Transfers
- ARM/Arm64 128-bit NEON Transfers
- ~+25% bandwidth improvement
- ARM/Arm64 128-bit NEON Transfers
- Hardware
- AMD EPYC “Milan” CPU module/core detection issues
- AMD EPYC “Milan” CPU additional support
Reviews using Sandra 20/21:
- SiSoftware
- Intel
- Intel 12th Gen Core AlderLake Mobile (i7-12700H) Review & Benchmarks – big/LITTLE Performance
- Intel ARC (“Alchemist”) A380 GP-GPU Graphics – OpenCL Performance
- Microsoft SoC / Surface Pro X
- Qualcom / Snapdragon
- Broadcom SoC / Raspberry Pi
- TPMs
- Intel
- SiSoftware Official Ranker
- Intel Arc “Alchemist” A380 Graphics
- Intel Arc “Alchemist” A370M Mobile Graphics + Intel Iris Xe Internal Graphics
- AMD EPYC 72F3 8-Core/8-Module Processor (8M, 8C, 8x 32MB L3)
- Microsoft SQ2 Arm64 (Microsoft Surface X)
- Qualcomm Snapdragon 7c Arm64 (ECS SF40QT)
- Raspberry Pi 4 Model B Arm64 (Broadcom BCM2711)
- Raspberry Pi 3 Model B+ Arm64 (Broadcom BCM2837)
- Apple Silicon on Parallels VM
Update & Download
Commercial version customers can download the free updates from their software distributor; Lite users please download from your favourite download site.
