#AVX10_2
LightNews LightNews
instlatx64.bsky.social
The current VNNI landscape
#AMD #Intel #Zen6 #AVX10_2 #AVX_VNNI #AVX_VNNI_INT8 #AVX_VNNI_INT16 #DiamondRapids #NovaLake
November 9, 2025 at 8:44 PM
instlatx64.bsky.social
Unfortunately, #Intel does not answer openly Game.Keeps.Loading 's question regarding #NovaLake ISA:
community.intel.com/t5/Mobile-and-Desktop-Processors/Intel-APX-ACX-10-2/m-p/1723564#M86387

#APX #AVX10_1 #AVX10_2 #AVX512
October 27, 2025 at 9:21 AM
instlatx64.bsky.social
The Netwide Assembler 3.01 is ready and supports the full #DiamondRapids and #NovaLake instruction sets:
nasm.us
#AVX10_2 #AVX10_VNNI_INT #APX #AMX_FP8 #AMX_TF32 #AMX_COMPLEX #AMX_AVX512 #AMX_MOVRS
October 11, 2025 at 10:07 AM
instlatx64.bsky.social
#Intel projects 2025+ v53
#PantherLake #NovaLakeU #NovaLakeS #RazerLake #TitanLake #BartlettLake #WildcatLake #GraniteRapids #GraniteRapidsD #DiamondRapids #DiamondRapidsD #CoralRapids #ClearwaterForest #RogueRiverForest #APX #AVX10_1 #AVX10_2 #AVX512
Thx for the feedback @mintsuki.com, G_melo_ding!
July 28, 2025 at 5:58 PM
instlatx64.bsky.social
An explicit mention of 512b #AVX10_2 for future #Intel desktop CPUs ( #NovaLake, what else?):
github.com/uxlfoundatio...
July 28, 2025 at 10:03 AM
instlatx64.bsky.social
#Intel projects 2025+ v50
#PantherLake #NovaLakeU #NovaLakeS #RazerLake #TitanLake #BartlettLake #WildcatLake #GraniteRapids #GraniteRapidsD #DiamondRapids #DiamondRapidsD #CoralRapids #ClearwaterForest #RogueRiverForest #APX #AVX10_1 #AVX10_2 #AVX512
July 17, 2025 at 9:08 AM
instlatx64.bsky.social
#Intel refreshed the #AVX10_2 specification to 5.0:
cdrdv2-public.intel.com/856721/36105...
VDPPHPS is not #AVX10_VNNI_INT, it's much more logical
June 9, 2025 at 8:50 AM
instlatx64.bsky.social
#Intel #DiamondRapids new vector instructions:
#AVX10_2 #AVX10_VNNI_INT
#MOVRS #SM4 #AMX_TRANSPOSE
#AMX_AVX512 #AMX_FP8 #AMX_MOVRS #AMX_TF32
June 7, 2025 at 10:43 AM
instlatx64.bsky.social
#Intel projects 2025+ v45
#PantherLake #NovaLakeU #NovaLakeS #RazerLake #TitanLake #BartlettLake #WildcatLake #GraniteRapids #GraniteRapidsD #DiamondRapids #DiamondRapidsD #CoralRapids #ClearwaterForest #RogueRiverForest #APX #AVX10_1 #AVX10_2 #AVX512
June 2, 2025 at 2:48 PM
instlatx64.bsky.social
#Intel refreshed the #AVX10_2 specification to 4.0:
cdrdv2-public.intel.com/855340/36105...
CPUID.24h.ECX=1.ECX[2] = AVX10_VNNI_INT
#MOVRS #SM4 #AMX_AVX512 #AVX10_VNNI_INT are decoupled from AVX10.2
May 22, 2025 at 8:43 AM
instlatx64.bsky.social
#VisualStudio2022 17.14.0 Preview 3.0 has almost complete #AVX10_2 / #DiamondRapids support, only a few VNNI and media processing instructions are missing.

#MOVRS #AMX_TRANSPOSE #AMX_AVX512 #AMX_FP8 #AMX_MOVRS #AMX_TF32
April 22, 2025 at 12:26 PM
instlatx64.bsky.social
#VisualStudio2022 17.14.0 Preview 2.0 got #MOVRS #AMX_TRANSPOSE #AMX_AVX512 #AMX_FP8 #AMX_MOVRS #AMX_TF32 support
#DiamondRapids #AVX10_2
March 26, 2025 at 9:24 AM
instlatx64.bsky.social
#Intel projects 2025+ v41
#PantherLake #NovaLake #RazerLake #TitanLake #BartlettLake #WildcatLake #GraniteRapids #GraniteRapidsD #DiamondRapids #CoralRapids #ClearwaterForest #RogueRiverForest #APX #AVX10_1 #AVX10_2 #AVX512
March 21, 2025 at 9:18 PM
instlatx64.bsky.social
Another round of #Intel ISA defragmentation:
Finally, #AVX10_2 is a common ISA for all CPU segments!
March 19, 2025 at 3:30 PM
instlatx64.bsky.social
#Intel refreshed the "The Converged Vector ISA: Intel® Advanced Vector Extensions 10" Technical Paper to 3.0:
cdrdv2-public.intel.com/849709/35636...
Great! This opens the possibility of the unified 512b #AVX10_2 ISA for hybrids too!
March 19, 2025 at 1:38 PM
instlatx64.bsky.social
@fclc.bsky.social: #AVX10/128 description is in sync with #AVX10_2
February 28, 2025 at 12:35 PM
instlatx64.bsky.social
#Intel projects 2025+ v37
#PantherLake #NovaLake #RazerLake #TitanLake #BartlettLake #WildcatLake #GraniteRapids #DiamondRapids #CoralRapids #ClearwaterForest #RogueRiverForest #APX #AVX10_2
February 1, 2025 at 5:04 PM
instlatx64.bsky.social
#VisualStudio2022 17.13.0 Preview 4 got the #EVEX-encoded, zmm/512b #SM4 cipher support: #DiamondRapids #AVX10_2
January 29, 2025 at 3:35 PM
instlatx64.bsky.social
#Intel refreshed the #AVX10_2 specification to 3.0:
cdrdv2-public.intel.com/844829/36105...
AVX10.2 is unified now with levels of #GraniteRidge + #AMX_AVX512 #SM4 #MOVRS,
#AMX_TRANSPOSE #AMX_FP8 #AMX_TF32 still missing
January 17, 2025 at 9:11 AM
instlatx64.bsky.social
#Intel ISA defragmentation:
SDE 9.48 internally shows, that under the #AVX10_2 label there are 15 instruction groups in #DiamondRapids
December 22, 2024 at 10:24 PM
instlatx64.bsky.social
If there was #Intel #SSE4_2, #AVX2 and now there is #AVX10_2, then why won't there be #AMX2?
December 22, 2024 at 10:21 PM
instlatx64.bsky.social
#VisualStudio2022 17.13.0 Preview 2.1 #AVX10_2 256-bit embedded rounding intrinsics support:
#DiamondRapids
December 19, 2024 at 8:38 PM