Ener­gy Trans­fer keeps its ERP behind its own fire­walls. Med­line moved a live SAP estate onto Azure with­out a minute of down­time. Both made the right call for their busi­ness, and togeth­er, their sto­ries show how much choice SAP cus­tomers now have as they modernize.

That choice runs through a five-part video series Intel pro­duced with ASUG at this year’s SAP Sap­phire & ASUG Annu­al Con­fer­ence. Mod­er­at­ed by ASUG, the inter­views bring togeth­er lead­ers from SAP, Intel, Leno­vo, HPE, Dell Tech­nolo­gies, Microsoft, Infos­ys, and Pro­tera, along­side cus­tomers includ­ing Ener­gy Trans­fer and Medline.

Across the series, the con­ver­sa­tions move beyond cloud as a des­ti­na­tion and into the prac­ti­cal deci­sions enter­prise lead­ers are mak­ing every day: where data should sit, how much con­trol the busi­ness needs, what resilience looks like, and how infra­struc­ture choic­es can pre­pare SAP land­scapes for AI.

The result is an edi­to­r­i­al look at real cus­tomer trans­for­ma­tion, from RISE with SAP and hybrid cloud ERP to sov­er­eign infra­struc­ture, sup­ply chain intel­li­gence, and AI-ready SAP on Azure. For lead­ers weigh­ing their own path for­ward, the series offers a ground­ed view of what SAP mod­ern­iza­tion looks like in prac­tice, and how the right infra­struc­ture choic­es can sup­port resilience, con­trol, and readi­ness for AI.

From RISE to Results: Lenovo’s SAP Transformation

Leno­vo runs its own SAP estate on the Cus­tomer Data Cen­ter (CDC) option for SAP Cloud ERP Pri­vate Edi­tion, and helps SAP deliv­er that option to every­one else. SAP built it with pub­lic sec­tor and gov­ern­ment-affil­i­at­ed orga­ni­za­tions in mind, though demand has come from well beyond that, notably from pri­vate com­pa­nies across Europe.

In this inter­view, Arthur Hu, SVP and Glob­al CIO at Leno­vo, and Lalit Patil, EVP and CTO of Glob­al Cloud Oper­a­tions at SAP, describe work­ing sce­nario by sce­nario through Lenovo’s most crit­i­cal process­es, bench­mark­ing data­bas­es to the mil­lisec­ond and tun­ing long-run­ning jobs, so that nobody would dis­cov­er the new sys­tem by way of a longer month­ly close.

The two also cov­er the joint ref­er­ence archi­tec­ture beneath the offer­ing, why reli­a­bil­i­ty counts as much as raw speed once agen­tic AI is run­ning in pro­duc­tion, the case Hu made to his board, and a part­ner­ship that start­ed with SAP want­i­ng a deliv­ery part­ner with glob­al scale and local exper­tise in China.

Watch to learn how Leno­vo mod­ern­ized SAP with­out compromise.

Hybrid Cloud ERP: Con­trol, Choice & Confidence

Boards have start­ed ask­ing who con­trols their company’s data and under whose juris­dic­tion it falls. They want to know whether com­pli­ance and inno­va­tion can pro­ceed togeth­er. Sov­er­eign­ty now splits into four kinds: data, oper­a­tional, tech­ni­cal, and legal.

This pan­el — Rakesh Roushan, VP and Glob­al Head of Cloud ERP Pri­vate, Cus­tomer Data Cen­ter at SAP; Jim Loia­cono, Glob­al SAP Pri­vate Cloud Sales Lead at HPE; Joakim Zetterblad, Glob­al Head of SAP Alliance at Dell Tech­nolo­gies; and Graeme Tre­vayne, Alliance Direc­tor for SAP at Leno­vo — argues that few com­pa­nies chose hybrid as a strat­e­gy, but the shape of their oper­a­tions pro­duced it any­way. Man­u­fac­tur­ing and dis­tri­b­u­tion gen­er­ate data on the fac­to­ry floor; SaaS appli­ca­tions like Con­cur and Ari­ba only run in the cloud.

Some cus­tomers have already asked about com­ing back on site after run­ning into cloud capac­i­ty lim­its. The pan­el takes up that pres­sure along­side SAP’s sup­port for split­ting a land­scape so pro­duc­tion runs in the data cen­ter while dis­as­ter recov­ery sits with a hyper­scaler, and the ques­tion of where the tech­ni­cal bound­ary actu­al­ly falls, giv­en that the soft­ware stack is iden­ti­cal and only the oper­a­tor changes. Tre­vayne expects as much as 70% of enter­prise data to be gen­er­at­ed out­side the core ERP — his argu­ment for putting data strat­e­gy ahead of cloud strategy.

Watch to learn how SAP cus­tomers are mak­ing hybrid work.

Accel­er­at­ing SAP RISE: Cus­tomer Suc­cess with Pri­vate Edi­tion Cus­tomer Data Center

Ener­gy Trans­fer, among the largest mid­stream ener­gy oper­a­tors in North Amer­i­ca, runs SAP Cloud ERP Pri­vate Edi­tion, CDC option, on HPE infra­struc­ture-as-a-ser­vice pow­ered by HPE Green­Lake, inside its own data cen­ters. Rough­ly 99% of the load on that sys­tem comes from com­mer­cial appli­ca­tions already housed in those facil­i­ties, which also serve as back­up for pipeline control.

Ran­dall Gro­gan, Sr. Direc­tor IT SAP & Finan­cial Appli­ca­tions at Ener­gy Trans­fer, revis­its the migra­tion five months after go-live with Jim Loia­cono of HPE and Bal­a­sub­ra­man­ian V, Vice Pres­i­dent at Infos­ys. Gro­gan describes a lift-and-shift that reached pro­duc­tion rough­ly five months after the hard­ware was ready, at a total cost he puts on par with a hyper­scaler, with Edge Inte­gra­tion Cell hold­ing API traf­fic inside the data center.

Gro­gan wants the indus­try to count dif­fer­ent­ly. Egress and ingress charges, inte­gra­tion work, and con­nec­tiv­i­ty to the rest of the estate all belong in the fig­ure. Bal­a­sub­ra­man­ian, whose own com­pa­ny goes live on the same option lat­er in 2026, is can­did about the lim­its, point­ing to future scal­ing needs and M&A among the con­sid­er­a­tions that should shape the decision.

Watch to learn how con­trol and cost shaped Ener­gy Transfer’s SAP path.

Medline’s SAP Jour­ney to Sup­ply Chain Intelligence

Med­line runs ²⁴⁄₇ and could not stop for a migra­tion. It moved its live SAP estate onto Microsoft Azure any­way — no down­time, on sched­ule, on bud­get. The plat­form it land­ed on is where the next project started.

Mpow­er, Medline’s AI dig­i­tal con­trol tow­er, changed when sup­ply chain teams and the clin­i­cians they serve find out about a prob­lem. Inven­to­ry risk used to sur­face in batch reports, after a back­o­rder had already occurred; Mpow­er reads demand sig­nals upstream of the order, ear­ly enough to mit­i­gate the risk or line up an alter­na­tive product.

Jason Kaley, SVP of IT Oper­a­tions and Archi­tec­ture at Med­line, is joined by Matt Clip­pinger, VP of Strate­gic Alliances at Pro­tera, and Bryan Hie­s­tand, an archi­tect at Intel. They dis­cuss build­ing the plan around 32 TB vir­tu­al machines that did not yet exist when the project start­ed, Kaley’s insis­tence on run­ning the work as one team rather than ven­dor and cus­tomer, and Intel’s secu­ri­ty addi­tions on Azure, includ­ing Total Mem­o­ry Encryp­tion and con­fi­den­tial com­put­ing with Intel TDX.

Watch to learn how Med­line moved SAP to Azure at pace.

From Cloud ERP to AI-Ready SAP on Azure

What SAP HANA asks of infra­struc­ture has changed. Cus­tomers used to size their sys­tems around mem­o­ry. As more busi­ness process­es moved onto the plat­form, CPU became the bind­ing con­straint, and stor­age band­width has since emerged as a third; an in-mem­o­ry data­base still has to write to disk.

Juer­gen Thomas, Part­ner Archi­tect at Microsoft, Bryan Hie­s­tand, Archi­tect at Intel, and Elyse Ge Hylan­der, Senior Direc­tor of Azure Com­pute at Microsoft, dis­cuss what that shift demands of the hard­ware. CXL-attached mem­o­ry address­es the mem­o­ry wall direct­ly, sup­port­ing up to 8 TB data­bas­es on a two-sock­et plat­form where four sock­ets were pre­vi­ous­ly required, and flat mem­o­ry mode on Xeon 6 lets cus­tomers adopt it at a mea­sured per­for­mance cost of rough­ly 3%.

Get­ting enough pow­er into data cen­ters has become a lim­it­ing fac­tor for every hyper­scaler. A 50-watt sav­ing per serv­er sounds triv­ial until mul­ti­plied across the mil­lions Azure deploys each year. The three also cov­er scale-out designs for deploy­ments approach­ing 96 TB and the AMX accelerator’s role in agen­tic work­flows, where CPUs car­ry a sub­stan­tial share of the load. Thomas, who first sat down with Intel rep­re­sen­ta­tives at SAP in 1996, reflects on a part­ner­ship that pre­dates Azure itself.

Watch to learn how SAP is scal­ing for AI on Azure.