KI-Update: KI im Gesundheitswesen, Apple Intelligence, TikTok, Papst Leo zu KI
Das "KI-Update" liefert werktäglich eine Zusammenfassung der wichtigsten KI-Entwicklungen.
I wondered why so many people were taking photographs here but then I spotted this Ordnance Survey #Benchmark and it all made sense.
Samsung Galaxy Z Flip FE potrebbe integrare Exynos 2400
#Android16 #Benchmark #Chip #Chipset #Exynos2400 #FlipPhone #Foldable #GalaxyZFlip7FE #GalaxyZFlipFE #Geekbench #Notizie #OneUI8 #Prestazioni #Processore #Samsung #Smartphone #SmartphonePieghevoli #SoC #Specifiche #TechNews #Tecnologia
https://www.ceotech.it/samsung-galaxy-z-flip-fe-potrebbe-integrare-exynos-2400/
https://www.europesays.com/uk/89350/ The US is reviewing Benchmark’s investment into Chinese AI startup Manus #AI #ArtificialIntelligence #Benchmark #China #InBrief #ManusAI #Technology #UK #UnitedKingdom
We benchmarked 2k+ cloud servers for LLM inference speed (prompt processing and text generation) using models ranging from 135M to 70B parameters! 
Read tech details, use our open-source tools, learn from the results at https://sparecores.com/article/llm-inference-speed
Did benchmark my two laptops to see how they compare in terms of performance. All tests done under Fedora 42 with the Kernel 6.14 with power-profiles-daemon set to performance.
- ThinkPad T14s Gen4 AMD (Ryzen 7 PRO 7840U)
- ThinkPad X1 Carbon Gen12 (Core Ultra 7 165U)
In CPU power, the results were rather similar with the AMD Ryzen just slightly faster in single- and multi-core performance.
In GPU performance, the AMD Radeon 780M did anihilate the Intel graphics. Three(!) times the score and way faster in every test.
T14s-AMD:
CPU: https://browser.geekbench.com/v6/cpu/11781584
GPU: https://browser.geekbench.com/v6/compute/4098248
X1 Carbon-Intel:
CPU: https://browser.geekbench.com/v6/cpu/11781701
GPU: https://browser.geekbench.com/v6/compute/4098254
heise+ | Duell in der Mittelklasse-CPUs: Intel Core Ultra 200S gegen Ryzen 9000 im Test
Bei den Arrow-Lake-CPUs hat Intel nicht nur ein neues Namensschema eingeführt. Wir testen, wie gut sich die günstigen 65-Watt-Modelle gegen Ryzen 9000 schlagen.
OCCT Diagnosetool jetzt für Linux verfügbar
https://linuxnews.de/occt-diagnosetool-jetzt-fuer-linux-verfuegbar/ #monitoring #stresstest #benchmark
2/4 
Deep dive into #Sonar API architecture including how it integrates live #websearch and cited responses into your products
Examine recent #benchmark results covering latency, throughput, and accuracy metrics
Linux providing a better gaming performance than Microsoft Windows is no longer of any kind of anomaly 
AMD Radeon RX 9070 XT / Linux kernel 6.14 / Mesa 25 benchmarked on Arch Linux (Steam OS bases on BTW) vs. Windows 11.
Die #GeForce #RTX5060 Ti soll in Kürze erscheinen. Ein #Benchmark-#Leak zeigt jetzt erste #Leistungsdaten zur #Grafikkarte von #Nvidia. Kann die #GPU an den Erfolg ihrer Vorgängerinnen anknüpfen? https://winfuture.de/news,150301.html?utm_source=Mastodon&utm_medium=ManualStatus&utm_campaign=SocialMedia
LLM Benchmark for 'Longform Creative Writing'
#Mozilla stellt #Benchmark für lokale #KI vor.
Mozilla stellt mit #LocalScore einen einfach einsetzbaren #Benchmark für lokale #KI vor. Der Benchmark ist im Rahmen des Projekts Mozilla Builders entstanden, über das Mozilla-Blogger Sören Hentzschel im vergangenen Jahr berichtet hat. Im Rahmen der Initiative Mozilla Builders Accelerator wurden Förderungen für Open-Source-Projekte zum Thema Lokale KI ausgelobt.
https://linuxnews.de/mozilla-stellt-benchmark-fuer-lokale-ki-vor/
https://www.theverge.com/meta/645012/meta-llama-4-maverick-benchmarks-gaming
Meta gets caught cheating at AI benchmarks.
Short version, they submitted a different/tweaked version of their new Llama 4 models to benchmarking sites, than what they actually make available to the public.
NVIDIA RTX 5060 Ti: uscita imminente? Dettagli leak
#Benchmark #Componenti #FurMark #Gaming #GDDR7 #GPU #Hardware #Leak #Notizie #Novità #NVIDIA #PCGaming #Prestazioni #RTX5060Ti #Rumors #SchedaVideo #SchedeGrafiche #TechNews #Tecnologia
https://www.ceotech.it/nvidia-rtx-5060-ti-uscita-imminente-dettagli-leak/
AnTuTu marzo 2025: vivo X200 Pro batte tutti - La top 10
#Android #AnTuTu #Benchmark #Dimensity9400 #Flagship #IQOO13 #iQOONeo10Pro #Marzo2025 #Notizie #Novità #OnePlus13 #OnePlusAce5Pro #Prestazioni #Smartphone #Snapdragon8Elite #TechNews #Tecnologia #vivoX200Pro
https://www.ceotech.it/antutu-marzo-2025-vivo-x200-pro-batte-tutti-la-top-10/
Indeed, the CPU-only performance is even worse. The LocalScore on the tiny 1B model is only 16, with a text generation speed of 7.7 tokens/second.
https://www.localscore.ai/result/235
Let's see if I can run this on a Raspberry Pi for comparison...
My hobby: running LocalScore.ai to benchmark how fast (ehm) my 2018 laptop runs a tiny 1B LLM. The laptop has a NVIDIA MX150 mobile GPU, 2GB VRAM. I guess it was intended for Photoshop filters or CAD stuff.
I got a LocalScore of 101 on the tiny model using the GPU (13.5 tokens/second for generation). A value of around 1000 is considered passable.
https://www.localscore.ai/accelerator/234
Still, I think it's marginally better than CPU-only on the same laptop.
OnePlus 13T mostra i muscoli su AnTuTu superando i 3 milioni
#120Hz #6200mAh #80W #Android15 #AnTuTu #Benchmark #DisplayOLED #Flagship #Indiscrezioni #Leak #Notizie #Novità #OnePlus #OnePlus13T #Prestazioni #Rumors #Smartphone #Snapdragon8Elite #Specifiche #TechNews #Tecnologia
https://www.ceotech.it/oneplus-13t-mostra-i-muscoli-su-antutu-superando-i-3-milioni/
![library(rbenchmark)
library(dplyr)
# Initial Values
n <- 1:50000
value <- NULL
for(i in seq_along(n)) value[[i]] = n[[i]][[1]]
lapply(n, function (x) x[1])
lapply(n, \(x) x[1])
lapply(n, `[`, 1)
lapply(n, `[[`, 1)
# Benchmarking
benchmark(
`lapply` = lapply(n, function (x) x[1]),
`lapply with lambda` = lapply(n, \(x) x[1]),
`lapply with [ ]` = lapply(n, `[`, 1),
`lapply with [[ ]]` = lapply(n, `[[`, 1),
replications = 100,
columns = c("test", "replications", "elapsed", "relative")
) |>
arrange(relative)
test replications elapsed relative
1 lapply with [ ] 100 2.72 1.000
2 lapply with [[ ]] 100 2.75 1.011
3 lapply with lambda 100 4.90 1.801
4 lapply 100 5.10 1.875](https://media.mstdn.social/media_attachments/files/114/280/332/699/357/109/original/660198e41a5d5f03.png "library(rbenchmark)
library(dplyr)
# Initial Values
n <- 1:50000
value <- NULL
for(i in seq_along(n)) value[[i]] = n[[i]][[1]]
lapply(n, function (x) x[1])
lapply(n, \(x) x[1])
lapply(n, `[`, 1)
lapply(n, `[[`, 1)
# Benchmarking
benchmark(
`lapply` = lapply(n, function (x) x[1]),
`lapply with lambda` = lapply(n, \(x) x[1]),
`lapply with [ ]` = lapply(n, `[`, 1),
`lapply with [[ ]]` = lapply(n, `[[`, 1),
replications = 100,
columns = c(\"test\", \"replications\", \"elapsed\", \"relative\")
) |>
arrange(relative)
test replications elapsed relative
1 lapply with [ ] 100 2.72 1.000
2 lapply with [[ ]] 100 2.75 1.011
3 lapply with lambda 100 4.90 1.801
4 lapply 100 5.10 1.875")