Categories
News

Ordering servers with the latest 5th generation AMD EPYC processors available

AMD’s latest EPYC “Turin” processors, part of the 5th-gen EPYC 9005 series, bring significant upgrades with the new Zen 5 and Zen 5c architectures. These processors are designed for various high-performance workloads, with models tailored for either high single-thread performance or increased multi-core efficiency.

More information about Supermicro available server configurations with 5th generation AMD EPYC processors is available on Supermicro website.

For a presentation, offer, and more information, contact us by phone or email.

Categories
News

Supermicro JumpStart Program Enhanced with NVIDIA GPU Servers

Supermicro has expanded its JumpStart Program to include NVIDIA GPU servers, broadening the scope for businesses looking to accelerate their applications with cutting-edge graphics processing power. This inclusion allows enterprises to leverage AI, machine learning, and deep learning capabilities more efficiently.

This program provides businesses with a seamless way to adopt and integrate high-performance computing solutions tailored to their unique needs. It offers pre-configured systems, rapid deployment, and ongoing support to maximize productivity and efficiency.

It is particularly beneficial for IT professionals who can test and optimize their systems in real-world conditions before the hardware investment.

As a trusted partner, Xenya is well-equipped to guide customers through every stage of the JumpStart Program. With expertise in high-performance computing and server solutions, Xenya assists in selecting, configuring, and deploying the right Supermicro systems to meet business goals.

Furthermore Xenya, as an elite NVIDIA partner, specializes in integrating these advanced NVIDIA GPU servers within the Supermicro framework, offering customers a streamlined pathway to harnessing the power of AI and graphics-intensive applications. Xenya’s team provides expert advice on selecting and configuring the optimal NVIDIA GPU servers to match specific business needs, ensuring a smooth integration and maximizing system performance.

For more information on how Xenya can help you utilize Supermicro’s JumpStart program to enhance your systems, please contact us via phone or e-mail.

Categories
News

Ordering NVIDIA Blackwell Systems Now Available

We are pleased to announce that it is now possible to place an order and receive a delivery date for NVIDIA Blackwell Systems. Contact us at prodaja@xenya.si.

Systems like the NVIDIA Blackwell GB200 NVL72 and Supermicro NVIDIA HGX B200 8-GPU are bringing a new dimension to the world of supercomputers and artificial intelligence.

Additional resources:

Categories
News

Order Supermicro HGX servers with NVIDIA H100 or H200 GPUs

We are pleased to announce that Xenya can deliver Supermicro NVIDIA HGX™ H100 and Supermicro NVIDIA HGX™ H200 8-GPU servers within 20-40 days (typically 30 days)*. Hurry up with your order and get GPU super servers for discounted price.

For more information about Supermicro NVIDIA HGX H100/H200 8-GPU Systems, please visit the Supermicro website.

Note*: the exact delivery date may vary depending on many factors, such as system configuration and market availability, and will be confirmed at the time of order confirmation!

Categories
Uncategorized

Supermicro – tehnološke novosti na srečanju partnerjev 2013

Konec avgusta je v Milanu potekalo tradicionalno poletno srečanje partnerjev proizvajalca Supermicro, ki se ga je udeležila tudi štiričlanska delegacija iz našega podjetja. Kot vedno je Supermicro predstavil kar precej novosti, ki mu stalno zagotavljajo prednost pred ‘večjimi’ znamkami.

Trenutno že poteka tranzicija plošč iz podnožja 1155 (Bromolow) v podnožje 1150 (Denlow) s podporo za Xeone E3 tretje generacije. Prihajajo tudi nove plošče za delovne postaje, namizne računalnike in tudi novost – plošče in ohišja za igralne računalnike, kjer dobite s čimmanjšo porabo energije in prostora bodo na voljo miniaturne mini-ITX plošče s procesorjema BayTrail Celeron (10W) ali 8-jedrnim Atomom po imenu Avoton (20W). Novost pa so komunikacijske plošče, s katerimi se bo dalo sestaviti požarni zid ali usmerjevalnik po meri uporabnika. Prilagojena ohišja so zelo plitva (globina od 35cm pa celo do 25cm), da se jih da zlahka vgraditi v komunikacijske omarice.

Na voljo bodo tudi že krmilniki nove generacije (SAS3), s prenosom podatkov do 12Gb/s. Poleg izvedbe v obliki dodatne PCI-E kartice ali vgrajene na plošči bo dobavljiv še v obliki ‘mezzanin’ modula, ki se ga namesti na ploščo na poseben priključek in s tem ne zaseda PCI-E rež na plošči, sploh če je strežnik oblike 1U in s tem omejen v številu dodatnih kartic. V primeru, če je krmilnik vgrajen na plošči, pa je vključen tudi v sistem nadzora in upravljanja na daljavo (IPMI).

Za zmanjševanje stroškov za hlajenje se dobi tudi serija plošč in strežnikov, namenjenih za podatkovne centre, ki zanesljivo delujejo tudi pri temperaturi do 47°C.

Pri ohišjih tudi prihajajo novosti – ena glavnih je možnost hkratne vgradnje redundantnih napajalnikov in baterijskih UPS modulov, ki zagotavljajo delovanje v primeru okvare napajalnika ali kratkotrajno delovanje v primeru izpada elektrike.

Za sisteme, ki potrebujejo visoko gostoto podatkovnega prostora, so na voljo ohišja z globino 35” in višino 4U, kamor lahko vgradite neverjetnih 72 hot-swap diskov velikosti 3,5”. Če pa je ohišje brez plošče in deluje le kot zunanje diskovno polje za primarni sistem, se število diskov povzpne že na 90. S trenutno največjimi 4TB diski si zagotovite gostoto 360TB na višino 4U! Obstaja tudi posebno 1U ohišje, ki res ne podpira hot-swap priklopa diskov velikosti 3,5”, jih pa zato lahko vgradite kar dvanajst, za nameček pa še dva velikosti 2,5” za sistem! Vsa ohišja so tudi optimizirana in preizkušena glede zagotavljanja hlajenja in pretoka zraka.

Nazadnje naj omenimo še sisteme za masivne računske operacije, ki računajo s pomočjo dodanih posebnih računskih grafičnih kartic (GPU), kot so Nvidia Tesla, Nvidia Kepler, Xeon Phi in podobne. Sistemi se dobijo v obliki samostojnih vgradnih ali samostoječih strežnikov, pa tudi v rezinah (blade), kjer si lahko sestavimo svojo gručo od 10 do 20 rezin v enem ohišju z medsebojno povezljivostjo po Infiniband protokolu do 56Gb/s (standard FDR).

css.php