#data-center — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #data-center, aggregated by home.social.
-
@bich
No town/city should approve the construction of a new #DataCenter unless it produces it's own (green) electricity and uses a "water-free" cooling system.(BTW: Building them in space resolves NEITHER of these problems.)
-
@bich
No town/city should approve the construction of a new #DataCenter unless it produces it's own (green) electricity and uses a "water-free" cooling system.(BTW: Building them in space resolves NEITHER of these problems.)
-
☁️ Apple expande Private Cloud Compute y colabora con Google y NVIDIA para ejecutar nuevas cargas de trabajo de Apple Intelligence en Google Cloud
#PrivateCloudCompute #GoogleCloud #DataCenter
https://mecambioamac.com/apple-expande-private-cloud-compute-mas-alla-de-sus-data-center/
-
A modern AI server runs five layers of memory, from nanosecond on-chip SRAM to petabyte-scale SSDs, and keeping the GPU fed is the whole engineering game. This piece walks the full hierarchy: why HBM became the bottleneck, why adding more isn't simple, and where HBF, CXL, and PIM fit into the next generation.
#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #ITAD #buysellram
-
Why can't a GPU just carry more HBM? Interposers max out in size, stacking 12 to 16 DRAM dies compounds yield losses, and every stack sits beside a kilowatt-class package that hates sharing heat. So capacity climbs in careful steps — 80 GB on the H100, 141 on the H200, 192 on the B200, 288 on Blackwell Ultra — while KV caches for long-context inference balloon past 40 GB per request.
That gap between what models demand and what packaging permits is reshaping server design. NVIDIA's Rubin platform treats CPU memory and HBM as one coherent pool. SanDisk and SK hynix are standardizing High Bandwidth Flash as a capacity tier under HBM, with first samples due this half. CXL 4.0 pooling hardware is landing in racks now.
This article walks the whole memory hierarchy, from on-chip SRAM to NVMe, and explains what each emerging technology actually solves — and what it doesn't.
#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #ITAD #technology
-
Why can't a GPU just carry more HBM? Interposers max out in size, stacking 12 to 16 DRAM dies compounds yield losses, and every stack sits beside a kilowatt-class package that hates sharing heat. So capacity climbs in careful steps — 80 GB on the H100, 141 on the H200, 192 on the B200, 288 on Blackwell Ultra — while KV caches for long-context inference balloon past 40 GB per request.
That gap between what models demand and what packaging permits is reshaping server design. NVIDIA's Rubin platform treats CPU memory and HBM as one coherent pool. SanDisk and SK hynix are standardizing High Bandwidth Flash as a capacity tier under HBM, with first samples due this half. CXL 4.0 pooling hardware is landing in racks now.
This article walks the whole memory hierarchy, from on-chip SRAM to NVMe, and explains what each emerging technology actually solves — and what it doesn't.
#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #ITAD #technology
-
A modern AI server runs five layers of memory, from nanosecond on-chip SRAM to petabyte-scale SSDs, and keeping the GPU fed is the whole engineering game. This piece walks the full hierarchy: why HBM became the bottleneck, why adding more isn't simple, and where HBF, CXL, and PIM fit into the next generation.
#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #tech
-
A modern AI server runs five layers of memory, from nanosecond on-chip SRAM to petabyte-scale SSDs, and keeping the GPU fed is the whole engineering game. This piece walks the full hierarchy: why HBM became the bottleneck, why adding more isn't simple, and where HBF, CXL, and PIM fit into the next generation.
#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #tech
-
Ach. War wohl nur eine Halluzination… #datacenter https://www.heise.de/news/Stargate-UK-20-Milliarden-Pfund-Investment-war-nicht-mehr-als-ein-PR-Stunt-11354198.html
-
Ach. War wohl nur eine Halluzination… #datacenter https://www.heise.de/news/Stargate-UK-20-Milliarden-Pfund-Investment-war-nicht-mehr-als-ein-PR-Stunt-11354198.html
-
Changed instance header once more to reflect the infrastructure changes. Also we are now in a really #energy efficient and #green #datacenter. See video for more info:
-
Changed instance header once more to reflect the infrastructure changes. Also we are now in a really #energy efficient and #green #datacenter. See video for more info:
-
RE: https://mastodon.green/@gerrymcgovern/116866846666195007
Water cooling systems are not so closed.
"Microsoft and Nvidia market sealed liquid loops as a near-zero-water alternative to evaporative cooling... Microsoft describes cooling systems that are filled once during construction and then recirculate the same water, while Nvidia's Rubin platform runs a coolant that is 75% water and 25% propylene glycol. That one-time fill, however, is the step that produces a discharge, and the flush water leaves the site before the loop is sealed"
-
RE: https://mastodon.green/@gerrymcgovern/116866846666195007
Water cooling systems are not so closed.
"Microsoft and Nvidia market sealed liquid loops as a near-zero-water alternative to evaporative cooling... Microsoft describes cooling systems that are filled once during construction and then recirculate the same water, while Nvidia's Rubin platform runs a coolant that is 75% water and 25% propylene glycol. That one-time fill, however, is the step that produces a discharge, and the flush water leaves the site before the loop is sealed"
-
Rad Web #Hosting Announces New Data Center Location in Tokyo, #Japan
New #Tokyo location expands Rad Web Hosting’s footprint, strengthens Asia-Pacific connectivity, and gives businesses options for low-latency hosting deployments.
Dallas, TX, United States — Rad Web Hosting, a provider of #cloud hosting, VPS hosting, dedicated servers, colocation, and managed infrastructure solutions, today announced the launch of ...
Continued 👉 https://blog.radwebhosting.com/new-data-center-location-in-tokyo/?utm_source=mastodon&utm_medium=social&utm_campaign=mastodon.raddemo.host #dedicatedserver #datacenter -
📣 Einen Tag nach der Räumung ihres Protestcamps gegen ein Rechenzentrum in Beringen (SH) schlagen Aktivisten ihr neues Camp in Tengen (D) auf. Die Gruppe «Aufstände der Allmende» will damit weiter gegen Datenzentren & Big Tech mobilmachen. 🌍⚡️ Klima-Kritik! #KI #Datacenter #Protest https://www.nau.ch/politik/bundeshaus/ki-rechenzentrum-beringen-sh-aktivisten-verlegen-camp-67145565
-
📣 Einen Tag nach der Räumung ihres Protestcamps gegen ein Rechenzentrum in Beringen (SH) schlagen Aktivisten ihr neues Camp in Tengen (D) auf. Die Gruppe «Aufstände der Allmende» will damit weiter gegen Datenzentren & Big Tech mobilmachen. 🌍⚡️ Klima-Kritik! #KI #Datacenter #Protest https://www.nau.ch/politik/bundeshaus/ki-rechenzentrum-beringen-sh-aktivisten-verlegen-camp-67145565
-
#CowboyStateDaily:
"
Cheyenne announced Thursday it's suspended taking wastewater associated with data center systems after a Meta contractor contaminated the city’s wastewater system, prompting months of cleanup. "It's a very, very unpleasant surprise,” said a city councilman.
"
https://cowboystatedaily.com/2026/07/02/cheyenne-wont-take-data-center-wastewater-after-meta-company-contaminated-system/2.7.2026
#Abwasser #BOPU #Cheyenne #CupriavidusGilardii #datacenter #GoatSystems #Meta #Pollution #ProjectCosmo #Rechenzentrum #Umweltverschmutzung #USA #Wastewater #Wasser #Water
-
#CowboyStateDaily:
"
Cheyenne announced Thursday it's suspended taking wastewater associated with data center systems after a Meta contractor contaminated the city’s wastewater system, prompting months of cleanup. "It's a very, very unpleasant surprise,” said a city councilman.
"
https://cowboystatedaily.com/2026/07/02/cheyenne-wont-take-data-center-wastewater-after-meta-company-contaminated-system/2.7.2026
#Abwasser #BOPU #Cheyenne #CupriavidusGilardii #datacenter #GoatSystems #Meta #Pollution #ProjectCosmo #Rechenzentrum #Umweltverschmutzung #USA #Wastewater #Wasser #Water
-
If anyone still has samples of the #ReflectingPool before Trump locked it down or the recovered carcasses they might want to test for heavy metal pollution + other common data center discharge pollutants. This is still theoretical but it wouldn't surprise me if Trump turned the Reflecting Pool & Potomac River into ecological AI disaster zones for his #ballroom bunker #dataCenter. PLEASE for your own health & safety stop touching the water at the pool without proper PPE!!!
-
The Cheyenne, Wyoming, Board of Public Utilities identified Goat Systems LLC, a contractor working on Meta ’s in-progress #DataCenter campus in the High Plains Business Park, as the source of the bacteria Cupriavidus gilardii in the city’s wastewater treatment system.
-
The Cheyenne, Wyoming, Board of Public Utilities identified Goat Systems LLC, a contractor working on Meta ’s in-progress #DataCenter campus in the High Plains Business Park, as the source of the bacteria Cupriavidus gilardii in the city’s wastewater treatment system.
-
The #PotomacRiver is regularly polluted with sewage overflow, farming fertilizer + pesticide runoff, industrial discharge, trash, etc, etc. To use that in a #dataCenter it needs to be cleaned up or it would clog up data center chillers almost immediately, which takes industrial water filtration, like ozone nanobubblers.
-
The #PotomacRiver is regularly polluted with sewage overflow, farming fertilizer + pesticide runoff, industrial discharge, trash, etc, etc. To use that in a #dataCenter it needs to be cleaned up or it would clog up data center chillers almost immediately, which takes industrial water filtration, like ozone nanobubblers.
-
TechChannel: preços de componentes para data centers estão subindo todo dia, toda hora
-
The Cheyenne, Wyoming, Board of Public Utilities identified Goat Systems LLC, a contractor working on #Meta ’s in-progress #DataCenter campus in the High Plains Business Park, as the source of the bacteria Cupriavidus gilardii in the city’s wastewater treatment system.
Cheyenne BOPU traces rare bact... -
The Cheyenne, Wyoming, Board of Public Utilities identified Goat Systems LLC, a contractor working on #Meta ’s in-progress #DataCenter campus in the High Plains Business Park, as the source of the bacteria Cupriavidus gilardii in the city’s wastewater treatment system.
Cheyenne BOPU traces rare bact... -
The Cheyenne, Wyoming, Board of Public Utilities identified Goat Systems LLC, a contractor working on #Meta ’s in-progress #DataCenter campus in the High Plains Business Park, as the source of the bacteria Cupriavidus gilardii in the city’s wastewater treatment system.
https://www.wyomingnews.com/news/local_news/cheyenne-bopu-traces-rare-bacteria-discharge-to-meta-data-center-contractor/article_1c538467-06ec-427d-9a6f-33797cd3c6ce.html -
The Cheyenne, Wyoming, Board of Public Utilities identified Goat Systems LLC, a contractor working on #Meta ’s in-progress #DataCenter campus in the High Plains Business Park, as the source of the bacteria Cupriavidus gilardii in the city’s wastewater treatment system.
https://www.wyomingnews.com/news/local_news/cheyenne-bopu-traces-rare-bacteria-discharge-to-meta-data-center-contractor/article_1c538467-06ec-427d-9a6f-33797cd3c6ce.html -
So is THIS the start of the #AI bubble burst?
Meh... I think there is a bubble, but it's a slow-rolling one. It'll be a fizzle (down) rather than a pop (burst) due to the incredible amount of investment being made.
That said, this does seem like an initial sign of taming expectations.
And I bet mass protestation of #DataCenter rollout (& holding it up in the courts) will slow down the AI-rollout process and help with the fizzle.
/1
#MyThoughts
--
The Al Trade Is Losing One of Its Key Signals -
So is THIS the start of the #AI bubble burst?
Meh... I think there is a bubble, but it's a slow-rolling one. It'll be a fizzle (down) rather than a pop (burst) due to the incredible amount of investment being made.
That said, this does seem like an initial sign of taming expectations.
And I bet mass protestation of #DataCenter rollout (& holding it up in the courts) will slow down the AI-rollout process and help with the fizzle.
/1
#MyThoughts
--
The Al Trade Is Losing One of Its Key Signals -
I'm thrilled to share that my new book #HYPERSCALE is hitting bookstores on October 20. You can preorder it now!
It's an account of how the data center boom is pushing our planet past its breaking point, and of the incredible efforts of people and movements pushing back against the profit-driven interests of Big Tech.
Get your copy and learn more: https://hyperscalebook.com/
#techindustry #techbook #newbook #reading #datacenter #datacenters #generativeai #politics #artificialintelligence
-
I'm thrilled to share that my new book #HYPERSCALE is hitting bookstores on October 20. You can preorder it now!
It's an account of how the data center boom is pushing our planet past its breaking point, and of the incredible efforts of people and movements pushing back against the profit-driven interests of Big Tech.
Get your copy and learn more: https://hyperscalebook.com/
#techindustry #techbook #newbook #reading #datacenter #datacenters #generativeai #politics #artificialintelligence
-
South Korea unveiled $550 billion in new memory fabs from Samsung & SK Hynix, part of a much larger $3.1 trillion corporate spending plan. Most of the new capacity targets server DRAM & HBM, not the commodity memory in ordinary PCs and servers. Industry forecasts point to real relief no earlier than late 2027, with some analysts already flagging oversupply risk...
https://www.buysellram.com/blog/will-samsung-and-sk-hynixs-550-billion-in-new-fabs-ease-the-memory-shortage/#DRAM #Semiconductors #MemoryShortage #SKHynix #Samsung #Micron #HBM #SupplyChain #DataCenter #buysellram
-
#DataCenter #AI #Microsoft #Espana #Zaragoza
Los abusadores de Micro$oft parece haber mojado las manos del gobierno de Aragón.
Además, en una zona tan árida los van a joder a la gente que queda ahí.Hablan los expropiados de Aragón ante el desembarco de Microsoft: "Fue un asalto, queremos un precio justo" | Empresas
https://www.elmundo.es/economia/empresas/2026/07/03/6a451434e9cf4add1d8b459f.html
> El 28 de febrero de 2025, Jorge Azcón se plantó en la Sala de Columnas del Edificio Pignatelli para anunciar que Microsoft iba a redoblar su apuesta en Aragón. El gigante...
-
#DataCenter #AI #Microsoft #Espana #Zaragoza
Los abusadores de Micro$oft parece haber mojado las manos del gobierno de Aragón.
Además, en una zona tan árida los van a joder a la gente que queda ahí.Hablan los expropiados de Aragón ante el desembarco de Microsoft: "Fue un asalto, queremos un precio justo" | Empresas
https://www.elmundo.es/economia/empresas/2026/07/03/6a451434e9cf4add1d8b459f.html
> El 28 de febrero de 2025, Jorge Azcón se plantó en la Sala de Columnas del Edificio Pignatelli para anunciar que Microsoft iba a redoblar su apuesta en Aragón. El gigante...
-
Samsung and SK Hynix just pledged $550 billion for new memory fabs — the concrete piece of a far larger $3.1 trillion long-term investment plan unveiled alongside South Korea's government on June 29. The number is easy to cheer and harder to interpret: most of the new capacity is aimed at server-grade DRAM and HBM, the highest-margin corners of the memory market, not the commodity memory in ordinary PCs and low-end servers.
Micron is running its own expansion on a similar timeline. China's CXMT is growing DRAM share faster than anyone but still years from qualifying for server and HBM business. Industry forecasters disagree on exactly when relief arrives — some point to late 2027, others flag oversupply risk once all this capacity actually lands.https://www.buysellram.com/blog/will-samsung-and-sk-hynixs-550-billion-in-new-fabs-ease-the-memory-shortage/
#DRAM #Semiconductors #MemoryShortage #SKHynix #Samsung #Micron #HBM #SupplyChain #DataCenter #CXMT #technology -
We don't need no #DataCenter in Pierce Co 🚫 Moratorium Now!
https://actionnetwork.org/letters/pass-a-data-center-moratorium-in-pierce-county
-
Meta will KI-Rechenleistung als Cloud-Dienst verkaufen
https://techupdate.io/dienste/meta-will-ki-rechenleistung-als-cloud-dienst-verkaufen/51272/
-
AI idzie do pracy. Autonomiczne boty przejmą stery w aplikacjach biznesowych
Do niedawna sztuczna inteligencja w biurze kojarzyła się z czatbotem, który potrafił napisać ładnego e-maila albo podsumować długi raport. Według analityków Gartnera ten etap mamy już za sobą.
Nadchodzi era tzw. agentów AI – autonomicznych systemów, które zamiast doradzać, same wykonają za nas pracę. Jak wynika z najnowszych prognoz firmy analitycznej Gartner, rynek oprogramowania czeka wstrząs, a korporacyjne działy IT już teraz w panice przebudowują swoje serwerownie.
Maszyna przestaje być asystentem. Zaczyna działać sama
Transformacja dokonuje się na naszych oczach w ekspresowym tempie. Według analityków z firmy Gartner, do końca 2026 roku aż 40 proc. oprogramowania klasy enterprise będzie zintegrowane z wyspecjalizowanymi agentami AI. Dla porównania – w 2025 roku wskaźnik ten wynosił zaledwie 5 proc. Oznacza to całkowitą zmianę technologicznego paradygmatu w biznesie.
Claude Science to nowa aplikacja na Maca. Anthropic zamienia komputer w potężne laboratorium
Sztuczna inteligencja przestaje być tylko wygodnym interfejsem. Modele wchodzą w warstwę operacyjną – potrafią samodzielnie analizować dane, podejmować decyzje i realizować procesy biznesowe od początku do końca, bez konieczności ciągłego zatwierdzania ich kroków przez człowieka. Zmiany mocno uderzą także w samych informatyków. Szacuje się, że do 2030 roku aż jedna czwarta procesów w działach IT będzie wykonywana całkowicie autonomicznie przez algorytmy. Eksperci przewidują, że do 2035 roku technologie te wygenerują ponad 450 mld dolarów przychodu.
Apetyt na prąd i wielki powrót prywatnych chmur
Przekazanie decyzyjności algorytmom rodzi jednak potężne wyzwania sprzętowe i infrastrukturalne. Autonomiczne boty potrzebują gigantycznej mocy obliczeniowej oraz dostępu do danych w czasie rzeczywistym. Z raportów McKinsey wynika, że zapotrzebowanie na centra danych gotowe do obsługi sztucznej inteligencji będzie rosło o 33 proc. rok do roku aż do 2030 roku.
Taka skala operacji na wrażliwych, firmowych danych sprawia, że korporacje zaczynają z rosnącą obawą patrzeć na globalnych dostawców chmury publicznej. Ze strachu przed uzależnieniem od jednego giganta technologicznego (tzw. vendor lock-in) oraz w trosce o cyfrową suwerenność, europejskie i polskie firmy masowo wracają do budowy chmur prywatnych oraz środowisk hybrydowych. Ścisła kontrola nad tym, gdzie fizycznie znajdują się firmowe sekrety przetwarzane przez algorytmy, staje się dziś dla biznesu ważniejsza niż wygoda płynąca z rozwiązań chmurowych.
#agenciAI #automatyzacjaIT #chmuraHybrydowa #dataCenter #gartner #oprogramowanieBiznesowe #sztucznaInteligencja -
Samsung and SK Hynix just pledged $550 billion for new memory fabs. Most industry forecasts still don't expect real DRAM/NAND relief before late 2027 — some analysts warn of a glut right after. The actual math behind the timeline, broken down.
https://www.buysellram.com/blog/will-samsung-and-sk-hynixs-550-billion-in-new-fabs-ease-the-memory-shortage/#DRAM #Semiconductors #MemoryShortage #SKHynix #Samsung #Micron #HBM #SupplyChain #DataCenter #ITAD #CXMT #AIInfrastructure #technology
