#sapphirerapids — Public Fediverse posts

Gerade hat #Intel damit begonnen, die zwölfte Core-Generation #AlderLake einzustellen. Finale Auslieferungen der #Prozessoren erfolgen am 22. Januar 2027. Auch #Xeon #SapphireRapids Server-#CPUs gehen in Rente. https://winfuture.de/news,156358.html?utm_source=Mastodon&utm_medium=ManualStatus&utm_campaign=SocialMedia

Gerade hat #Intel damit begonnen, die zwölfte Core-Generation #AlderLake einzustellen. Finale Auslieferungen der #Prozessoren erfolgen am 22. Januar 2027. Auch #Xeon #SapphireRapids Server-#CPUs gehen in Rente. https://winfuture.de/news,156358.html?utm_source=Mastodon&utm_medium=ManualStatus&utm_campaign=SocialMedia

Update on In-Field-Scan adventures: Following my message, Intel has now published In-Field-Scan "scan tests" firmware images for Emerald Rapids (Xeon 5th gen), and updated the Linux kernel documentation accordingly.
Still missing are the images for already released CPU families Sierra Forrest and Granite Rapids (Xeon 6th gen). 4th gen Xeon (Sapphire Rapids) won't have a public release of the scan tests firmware images.

They can still do the Array Built-In Self Tests (BIST) which require no firmware at all to test L1/L2 caches, register files, and (SRAM?) data arrays.

#LinuxKernel #InFieldScan #Intel #Xeon #EmeraldRapids #SapphireRapids #GraniteRapids #SierraForrest

Update on In-Field-Scan adventures: Following my message, Intel has now published In-Field-Scan "scan tests" firmware images for Emerald Rapids (Xeon 5th gen), and updated the Linux kernel documentation accordingly.
Still missing are the images for already released CPU families Sierra Forrest and Granite Rapids (Xeon 6th gen). 4th gen Xeon (Sapphire Rapids) won't have a public release of the scan tests firmware images.

They can still do the Array Built-In Self Tests (BIST) which require no firmware at all to test L1/L2 caches, register files, and (SRAM?) data arrays.

#LinuxKernel #InFieldScan #Intel #Xeon #EmeraldRapids #SapphireRapids #GraniteRapids #SierraForrest

#Nvidia #GraceSuperchip loses to #Intel #SapphireRapids in #HPC performance #benchmarks, promises greater efficiency
Barcelona #Supercomputing and #SUNY published benchmarks showing Grace Superchip, couldn't quite match TWO 48-core Sapphire Rapids #CPU. Despite not having earth-shattering performance, Grace promises to be competitive datacenter and HPC processor thanks to its efficiency. Built on Arm architecture. A single Grace CPU comes with 72c and 480GB of LPDDR5X RAM
https://www.tomshardware.com/pc-components/cpus/nvidia-grace-superchip-loses-to-intel-sapphire-rapids-in-hpc-performance-benchmarks-but-promises-greater-efficiency

#Nvidia #GraceSuperchip loses to #Intel #SapphireRapids in #HPC performance #benchmarks, promises greater efficiency
Barcelona #Supercomputing and #SUNY published benchmarks showing Grace Superchip, couldn't quite match TWO 48-core Sapphire Rapids #CPU. Despite not having earth-shattering performance, Grace promises to be competitive datacenter and HPC processor thanks to its efficiency. Built on Arm architecture. A single Grace CPU comes with 72c and 480GB of LPDDR5X RAM
https://www.tomshardware.com/pc-components/cpus/nvidia-grace-superchip-loses-to-intel-sapphire-rapids-in-hpc-performance-benchmarks-but-promises-greater-efficiency

#Microsoft #Azure Eagle is a Paradigm Shifting Cloud #Supercomputer
While traditional #HPC clusters are built in places like national labs and access is limited to those who are approved by the institutions that run them, Microsoft’s is a cloud solution. With 14,400 #NVIDIA #H100 #GPU (14400/8 = 1800 nodes?), InfiniBand, and #Intel #Xeon #SapphireRapids #CPU, the new system turned in a #3 result only around 4% less than #2 on the November 2023 #Top500 list.
https://www.servethehome.com/microsoft-azure-eagle-is-a-paradigm-shifting-cloud-supercomputer-nvidia-intel/

#Microsoft #Azure Eagle is a Paradigm Shifting Cloud #Supercomputer
While traditional #HPC clusters are built in places like national labs and access is limited to those who are approved by the institutions that run them, Microsoft’s is a cloud solution. With 14,400 #NVIDIA #H100 #GPU (14400/8 = 1800 nodes?), InfiniBand, and #Intel #Xeon #SapphireRapids #CPU, the new system turned in a #3 result only around 4% less than #2 on the November 2023 #Top500 list.
https://www.servethehome.com/microsoft-azure-eagle-is-a-paradigm-shifting-cloud-supercomputer-nvidia-intel/

Seeing every review of #genoa #bergamo and #SapphireRapids coming out with #spr looking foolish is concerning but cathartic. #spr with #hbm being clowned on by #VCache is hard to not laugh about.

#Linux Patches Updated For @Intel #SapphireRapids C0.x Idle States

https://www.phoronix.com/news/SPR-C0.1-C0.2-Idle-States-Linux

Original tweet : https://twitter.com/phoronix/status/1667844729004761088

The @OpenAtIntel @IntelSoftware One Line Of #Linux Code For Speeding-Up @Intel #SapphireRapids On #Ubuntu
https://www.phoronix.com/review/intel-spr-one-line

Original tweet : https://twitter.com/phoronix/status/1654124769388777472

Tomorrow's article: One Line Code Change To #Linux Yields Better @Intel #SapphireRapids Performance On @Ubuntu

Original tweet : https://twitter.com/phoronix/status/1653837699826982934

.@IntelSoftware @IntelDevTools #oneDNN 3.1 Further Optimizing For #SapphireRapids, Starts Tuning For #SierraForest

-- Lots of improvements in oneDNN 3.1!

https://www.phoronix.com/news/Intel-oneDNN-3.1

Original tweet : https://twitter.com/phoronix/status/1641898116151496731

The @Intel #SapphireRapids Performance With @GoogleCloud Compute Engine C3

-- 47~60% faster geo mean compared to prior Google Cloud N2/C2 instances at same vCPU count. Striking win with perf-per-dollar too.

https://www.phoronix.com/review/google-cloud-c3-sapphire-rapids

Original tweet : https://twitter.com/phoronix/status/1640386090803638274

.@intel Xeon #SapphireRapids Code Continues Landing For #Coreboot
https://www.phoronix.com/news/Intel-SPR-More-Coreboot

Original tweet : https://twitter.com/phoronix/status/1637769570864844803

.@Intel Preparing #IAA Crypto Compression Driver - Kernel Crypto API Use For #SapphireRapids Accelerators

-- Impressive results with now opening accelerator benefits to Zswap, zRAM, etc...

https://www.phoronix.com/news/Intel-IAA-Crypto-Linux-Driver

Original tweet : https://twitter.com/phoronix/status/1633096764239618048

.@Intel Preparing #SapphireRapids C0.2 Idle State Support For Better @OpenAtIntel #Linux Energy Efficiency

-- "SPR C0.2 idle state is 5-20% more energy-efficient than Linux POLL state, and its latency is close to POLL latency."

https://www.phoronix.com/news/Intel-SPR-C0.2-Idle-State-Linux

Original tweet : https://twitter.com/phoronix/status/1632849307236704256

Small I/O Performance Boost Coming For #Intel #IceLake & #SapphireRapids Servers

-- The Linux kernel Intel P-State driver mistakenly had a check that stopped getting updated after Skylake... This patch gets rid of that HWP I/O boost guard.

https://www.phoronix.com/news/Intel-ICL-SPR-IO-Perf-Boost

Original tweet : https://twitter.com/phoronix/status/1631617190850961408

.@IntelDevTools @Intel ISPC 1.19 Released With #SapphireRapids Support, Data Center GPU Max

https://www.phoronix.com/news/Intel-ISPC-1.19

Original tweet : https://twitter.com/phoronix/status/1630524928993685504

The @IntelSoftware @IntelDevTools Embree 4.0 Is Running Well On @Intel 4th Gen Xeon Scalable "#SapphireRapids"

-- Nice uplift over Embree 3.13 and a look at SPR SSE4.2 vs. AVX vs. AVX2 vs. AVX-512 usage with the new Embree 4.0.

https://www.phoronix.com/review/embree4-sapphire-rapids

Original tweet : https://twitter.com/phoronix/status/1624046450790342656

#Linux 6.3 Features, #Intel #SapphireRapids & More Made For An Exciting January

https://www.phoronix.com/news/January-2023-Highlights

Original tweet : https://twitter.com/phoronix/status/1620580247140995073

#Linux Support For #Intel #SapphireRapids' User Interrupts Still Awaiting Mainline

https://www.phoronix.com/news/Linux-SPR-Intel-User-Interrupts

Original tweet : https://twitter.com/phoronix/status/1617207890296639489

.@Intel Advanced Matrix Extensions [AMX] Performance With Xeon Scalable #SapphireRapids

-- The big #AI performance uplift and power efficiency benefits from #AMX w/ #oneAPI #oneDNN & #OpenVINO benchmarks

https://www.phoronix.com/review/intel-xeon-amx

Original tweet : https://twitter.com/phoronix/status/1615054211582164993

Exciting look at the @Intel AMX performance on/off with #SapphireRapids coming up in the next few hours.... Raw perf & power efficiency are great with supported software. Much better story than with the early days of AVX-512.

Original tweet : https://twitter.com/phoronix/status/1615024316034469888

.@Intel 4th Gen Xeon Scalable "#SapphireRapids" Support Coming To #Coreboot

-- Initial @OpenAtIntel SPR Coreboot patches published! ArcherCity reference board initially supported.

https://www.phoronix.com/news/Xeon-Sapphire-Rapids-Coreboot

Original tweet : https://twitter.com/phoronix/status/1614969728715431936

ServeTheHome's review of #SapphireRapids CPUs:

https://youtu.be/zIWlXjxyIuM

ServeTheHome's review of #SapphireRapids CPUs:

https://youtu.be/zIWlXjxyIuM

Setting Up @Intel 4th Gen Xeon Scalable "#SapphireRapids" For Accelerator Use

-- The steps needed to get these accelerators working under #opensource #Linux and the varying upstream software support for making use of the accelerators.

https://www.phoronix.com/review/intel-accelerators-linux

Original tweet : https://twitter.com/phoronix/status/1613972791455252485

There will be an absolute FLOOD of product announcements now that #Intel has announced #SapphireRapids #Xeon! Software, hardware, cloud, AI, everything. Everyone has been waiting for this!

There will be an absolute FLOOD of product announcements now that #Intel has announced #SapphireRapids #Xeon! Software, hardware, cloud, AI, everything. Everyone has been waiting for this!

Pat Gelsinger is back on stage for #Intel #Xeon #SapphireRapids launch. A little awkward mentioning God and then talking about the “soul of the machine”

Pat Gelsinger is back on stage for #Intel #Xeon #SapphireRapids launch. A little awkward mentioning God and then talking about the “soul of the machine”

Ok, props to Intel: with launching #sapphirerapids, they're also launching/releasing the specific *per port* throughput and latency per instruction.

If you want to wring out every last clock cycle, this is the way to do it.

#HPC #Xeon #avx512 #AVX512fp16 #sapphirerapids

Ok, props to Intel: with launching #sapphirerapids, they're also launching/releasing the specific *per port* throughput and latency per instruction.

If you want to wring out every last clock cycle, this is the way to do it.

#HPC #Xeon #avx512 #AVX512fp16 #sapphirerapids

Now promising 2x performance per watt and 30% better #5G using #Xeon #SapphireRapids at the #Edge

Side note: Telefonica guy takes our audience perspective…

Now promising 2x performance per watt and 30% better #5G using #Xeon #SapphireRapids at the #Edge

Side note: Telefonica guy takes our audience perspective…

Intel and Fujitsu suggest 4x improvement in #Xeon #SapphireRapids #AI accelerators vs. Skylake. Talking #OpenVINO, which is something we heard a lot about on the @UtilizingTech podcast last season.

Intel and Fujitsu suggest 4x improvement in #Xeon #SapphireRapids #AI accelerators vs. Skylake. Talking #OpenVINO, which is something we heard a lot about on the @UtilizingTech podcast last season.

.@Intel Xeon Platinum 8490H "#SapphireRapids" #Linux Performance Benchmarks

https://www.phoronix.com/review/intel-xeon-platinum-8490h

Original tweet : https://twitter.com/phoronix/status/1612876016913125378

Google Cloud talks about the specific accelerators in #Xeon #SapphireRapids while IBM Cloud claims to be one of the first to deploy it. Oracle Cloud also announces they’ll offer it everywhere too.

Now we have some initial “big number” #Xeon #SapphireRapids specs: 2.9x better per watt, 50%+ better performance in #ML inferencing, #database, and #HPC, and #QuickAssist encoding improvements (compared to Skylake?)

And Pat Gelsinger takes the stage and immediately confirms that this is the launch of the fourth-generation #Xeon processor known as #SapphireRapids. He’s also emphasizing the #XeonMax brand!

I'm going to be covering the Intel #Xeon #SapphireRapids announcement starting in a minute!

Intel's 2023 desktop & workstation processor lineup leak point sot Raptor Lake-S Refresh in the second half of the year in the 35W - 125W range and Sapphire Rapids & Emerald Rapids for the 250W+ workstation space.

https://www.tomshardware.com/news/intel-roadmap-leaks-raptor-lake-refresh-hedt-replacement-in-2023

#leaks #roadmap #intel #RaptorLake #SapphireRapids #EmeraldRapids