1 of 49

La réglementation

et

les indicateurs

pour les datacenters

boavizta.org

2 of 49

Speakers

François Moriamez

Franck Pramotton

Catherine Guermont

Benoit Petit

2

2

3 of 49

Sommaire

3

  1. Intro

  • Les réglementations
    1. Décret tertiaire
    2. CSRD
    3. Energy Efficiency Directive (EED)

  • Les indicateurs et leur calcul
    • PUE
    • WUE
    • ERF
    • CER
    • CUE

  • Conclusion

4 of 49

Des tendances défavorables

4

4

  • Le dernier GPU Nvidia B200 : 208 milliards de transistors & 1000 W
  • Estimation du RTE UK : les DC vont faire X6 en 10 ans

5 of 49

Réglementations

6 of 49

Le cadre législatif est en construction

6

6

Autres actions Boavizta :

  • Chantier Référentiels d’indicateurs
  • bi-monthly 29 avril : la CSRD et son application au secteur du numérique

CSRD

Décret tertiaire

EED

Reporting

only

Reporting

&

Performance

Taxonomy

Code of Conduct

7 of 49

FRANCK

7

7

8 of 49

Le cadre législatif : “The Big Picture”

8

8

9 of 49

Décret Tertiaire (2019)

9

9

  • Approche “Relative”

ou

  • Approche “Absolue” (CABS)

10 of 49

Decret Tertiaire (2019+)

10

10

  • Oct 2019 : Decret éco-énergie tertiaire 1er octobre 2019, complété par :
    • Avril 2020 : Decret METHODE
      • Approche Relative ou Absolue
    • Nov 2020 : Decret Valeurs Absolues I
      • Bureaux, Enseignement, Logistique, Froid
    • Avr 2022 (CABS) : Val. Absolues II : Petite Enfance, Spectacles, Transport… Data Centers
    • Avr 2023 : Decret BACS �Dec 2023 : Val. Absolues III : Hotellerie, Restaurant, Data Centers
    • Fev 2024 : Val Absolues IV : Sport, Hopitaux…�

DECRET CABS : Les Coefficents absolus (Cabs) sont fixées par arrêté avant le début de chaque décennie. L’atteinte des objectifs est une obligation, les moyens à mettre en œuvre relève de la responsabilité des assujettis (propriétaires et, le cas échéant, les preneurs à bail, en référence aux MTD).

11 of 49

Performance exigée en 2030 pour les DC

11

11

Taille du DC

PUE cible pour Paris et Lille

PUE cible pour Marseille

Intensité max (kWh/m²/an)

< 20 m²

2,00

2,00

1156

entre 20 et 100 m²

1,80

1,80

1041

entre 100 et 500 m²

1,60

1,72

728

entre 500 et 1000 m²

1,60

1,71

833

entre 1000 et 5000 m²

1,40

1,51

1093

entre 5000 et 10000 m²

1,40

1,51

1214

plus de 10000 m²

1,20

1,29

2342

  • Soit une réduction de la consommation électrique de 40 % entre 2010 et 2030
  • Soit l’atteinte en 2030 d’un PUE cible et d’une densité énergétique maximale :

12 of 49

FIN/FRANCK

12

12

13 of 49

CSRD

13

13

Directive européenne CSRD sur le reporting extra-financier

Les “activités durables” de l’entreprise sont définies par la taxonomie européenne.

Le premier exercice (pour les entreprises concernées) est à effectuer en 2025 sur les activités de 2024.

Une activité IT “Data hosting” est considérée comme alignée avec la taxonomie si elle est hébergée dans un datacenter qui :

  • a mis en oeuvre le Code of Conduct on Datacenter Energy Efficiency (i.e la commission européenne lui a accordé le label “participant”)
  • a justifié les exigences du CoC non déployées
  • utilise des fluides réfrigérants dont le PRG (pouvoir de réchauffement global) est inférieur à 675

Les hyperscalers devraient publier un pourcentage d’alignement de leurs DC en 2024.

Plus d’informations dans le bi-monthly du 29 avril 2024

14 of 49

Focus on the new requirements for DC introduced by the�Energy Efficiency Directive (EED)

15 of 49

15

Recast Energy Efficiency Directive (EED)

Obligations for high energy intensive enterprises (Article 11)

See directive 2023 here

The Directive 2012/27/EU establishes measures to help the EU reach its energy efficiency. It covers mandatory energy audit for large companies from tertiary, industry & transport sectors, smart meters roll out for electricity & gas, energy efficiency obligation scheme, requirements for District Heating & Cooling (DHC) infrastructures, protection of the rights of consumers to receive easy and free access to data on real-time and historical energy consumption information displaying.

The version adopted on September 2023 introduces

  1. The obligation for enterprises with annual av. annual consumption > 85 TJ (23,600 MWh) of energy over the previous three years to implement an energy management system (EMS) certified by an independent body by 11 October 2027.

  • The obligation for enterprises with an av. annual consumption >10 TJ of energy over the previous three years, which do not implement an energy management system are subject to an energy audit by 11 October 2026.

16 of 49

16

Article 32 obliges MS to lay down rules on penalties applicable to infringements.

These penalties shall be “effective, proportionate and dissuasive”. Each MS must notify the Commission of those rules by October 11, 2025

See directive 2023 here

  1. By 15 May 2024 and every year thereafter, owners and operators of data centres with power of the installed information technology (IT) ≥ 500kW must make the information set out in Annex VII publicly available, except for information subject to Union and national law protecting trade and business secrets and confidentiality, and DC used for defence and civil protection purpose.

Recast Energy Efficiency Directive (EED)

New obligations specific to data centers owners & operators (Article 12)

17 of 49

Objective: this first delegated regulation sets out the information and key performance indicators (KPI) needed from data centres as well as the first sustainability indicators that can be used for the assessment of the sustainability of data centres.

Scope: This delegated regulation includes six articles that cover subject matter and scope, definitions of terms, introduce the reporting mechanism for the sustainability of data centres, introduce the data centre sustainability indicators, and set out the specifications of the European database.

Content:

The regulation’s four Annexes define:

  • The information to be communicated to the European database on data centres (Annex I)
  • The key performance indicators to be monitored, gathered, and communicated and their measurement methodologies (Annex II)
  • The data centre sustainability indicators and calculation methodologies (Annex III)
  • The information that will be publicly available in an ‘aggregated form’ (Annex IV)

17

EED: 1st delegated act adopted on March 2024

Once the Commission has adopted the act, Parliament and Council generally have two months to formulate any objections so theoretically ... ‘some surprises’ may occur within May 13.

See the delegated act & its annexes adopted on March 14 2023 here

18 of 49

18

Information to be communicated to the EU database

Annex I of the delegated act

Description

1

a

Data centre name used to identify and describe the reporting data centre

b

Owner and operator of the data centre including the name and contact details

c

Location i.e. Local Administrative Unit Code (LAU code) of the location of the reporting data centre (building or site) in accordance with the most recent LAU tables of Eurostat

d

The type of data centre: ‘enterprise data centre’, ‘colocation data centre’ or ‘co-hosting data centre’, combined with one of the values ‘structure’ or ‘group of structures’.

If a colocation data centre also offers co-hosting services or if a co-hosting data

centre also offers colocation services, this shall be indicated.

e

Year and month of entry into operation

2

a

Electrical infrastructure redundancy level at high voltage level / at low voltage

level (line-up) / at rack level

b

Cooling infrastructure redundancy level at room level / at rack level.

For the redundancy levels, if "N" represents the baseline number of components or functions to satisfy the normal conditions, redundancy shall be expressed compared to that baseline "N", for example as "N+1," "N+2," "2N", etc. Facility redundancy can apply to an entire site (backup site), systems or components. Information technology redundancy can apply to hardware and software

See annex I to the delegated act adopted on March 14 2023 here

19 of 49

19

Key performance indicators to be communicated

Annex II of the delegated act

Description

1

Energy & sustaina-bility indicators

a

Installed information technology power demand (“PDIT” in kW)

NB: where the installed information technology power demand has changed during the reporting period,

a weighted average shall be used.

b

Data centre total floor area (“SDC” in m²).

c

Data centre computer room floor area (“SCR” in m²).

d

Total energy consumption (“EDC” in kWh)

by using the methodology in the CEN/CENELEC EN 50600-4-2 standard or equivalent.

The amount of EDC coming from backup generators (EDC-BG in kWh) shall be measured separately.

NB: the measurement points shall be set at the primary & secondary supply of energy & at every additional supply

e

Total energy consumption of information technology equipment (“EIT” in kWh) shall

be measured in accordance with the category 1 methodology for the calculation of

the PUE set out in the CEN/CENELEC EN 50600-4-2 standard or eq.

f

Electrical grid functions is the information on whether any functions that support the

stability, reliability, and resilience of the electrical grid are provided by the data

centre, such as peak demand shifting or firm frequency response (FFR)

g

Average battery capacity (“CBtG” in kW) is the average capacity of the DC

batteries that were offered to the grid via a “relevant market” or “contracts” for elec. grid functions

See annex II to the delegated act adopted on March 14 2023 here

20 of 49

20

KPI to be communicated

Annex II of the delegated act

Description

1

h

Total water input (“WIN” in m3) as defined by CEN/CENELEC EN 50600-4-9 standard WUE

Category 2, or if not possible, the methodology set out in Category 1 or eq std.

i

Total potable water input (“WIN-POT” in m3)

See annex II to the delegated act adopted on March 14 2023 here

21 of 49

21

KPI to be communicated

Annex II of the delegated act

Description

1

Energy & sustainability indicators

j

Waste heat reused (“EREUSE” in kWh) following CEN/CENELEC EN 50600-4-6 std or eq.

NB1: defining the boundaries of DC is key since only energy reused outside the boundaries of the DC is counted.

NB2: if part of the waste heat is reused for cooling the DC that part must be subtracted from the reused waste heat

k

Average waste heat temperature (“TWH” in C°) shall be measured as the

temperature of the fluid used to cool the ICT equipment in the DC computer room averaged over the year, and across every measurement points.

l

Average setpoint information technology equipment intake air temperature

(“TIN” in C°) measured as the average setpoint temperature in all DC rooms, set as a setpoint command to the cooling system used for the ICT equipment averaged over the year

m

Types of refrigerants used in the cooling and air conditioning equipment

n

Cooling degree days (“CDD” in degree-days) for the location of the reporting DC during the last calendar year, by using the methodology used by Eurostat and the JRC or equ. and with a base temperature of 21 C°

NB: open access sources shall be used to determine the CDD

See annex II to the delegated act adopted on March 14 2023 here

ERF*

*acronym used in CEN/CENELEC EN 50600-4-2

22 of 49

22

Source & credits: annex I to the delegated act adopted on March 14 2023 here

KPI to be communicated

Measurement of water input & waste heat

CRAH (Computer Room Air Handler)

CRAC (Computer Room Air Conditioner)

23 of 49

23

KPI to be communicated

Annex II of the delegated act

Description

1

Energy & sustainability indicators

o

Total renewable energy consumption (“ERES-TOT” in kWh)

determined according the methodology set out in the CEN/CENELEC EN 50600-4-3 std or eq.

p

Total renewable energy consumption from Guarantees of Origin

(“ERES-GOO” in kWh)

NB: GO cannot be counted for more than one DC or be created from PPA or on-site renewables

q

Total renewable energy consumption from Power Purchasing Agreements

(“ERES- PPA” in kWh)

NB: any GO created as a result of such PPA must be included in ERES-PPA.

Otherwise, the concerned amount of energy shall be subtracted from the measured ERES-PPA.

r

Total renewable energy consumption from on-site renewables (“ERES-OS”, in kW)

NB: any GO created as a result of on-site Ren sources must be owned and retired by the reporting DC.

Otherwise, the amount of energy in question shall be subtracted from the measured ERES-OS.

See annex II to the delegated act adopted on March 14 2023 here

ERES-TOT = ERES-GOO + ERES-PPA + ERES-OS

NB: the delegated act does not clarify the definition of GO & PPA and does not link it to RED nor REMIT’s guidance for reporting Transaction Reporting User Manual (TRUM)

REF*

*acronym used in CEN/CENELEC EN 50600-4-2

24 of 49

24

ICT capacity indicators to be communicated

Annex II of the delegated act

Description

2

ICT capacity indicators

a

ICT capacity for servers (“CSERV”) = the sum of the SERT active state performance or eq for all servers as declared in the manufacturer information in accordance with Regulation 2019/424

The active state performance value for the configured server or group of servers in a DC computer room shall be either interpolated from the declared active state performance value for a configuration declared under the Regulation (EU) 2019/424, or provided by a server manufacturer, or provided by a table of values for CPU part numbers created from a large SERT dataset, or estimated from a large dataset of measured values where a “recognized” calculation method exists.

b

ICT capacity for storage equipment (“CSTOR” in petabytes)

shall be the storage capacity, namely the sum of the raw (addressable) capacity of all SSD and HDD storage devices installed in all the storage equipment as declared by the storage device manufacturer.

NB: colocation data centre operators may calculate CSTOR by extrapolating the value that corresponds to at least 90% of the installed information technology power demand of all new storage equipment installed in the reporting data centre.

See annex II to the delegated act adopted on March 14 2023 here

ICT capacity is measured for servers and data storage products as servers and data storage products are defined in the Commission Regulation (EU) 2019/4244 that laydown ecodesign requirements for servers and data storage products

NB: it does not refer to the Regulation 2021/341 that amends part of the 2019 Directive

25 of 49

25

Focus on SERT metrics related to energy efficiency

For each of load levels, energy efficiency is calculated as follows :

See SERT’s Run and Reporting Rules @https://www.spec.org/sert2/SERT-JVM_Options-2.0.html

  • The SERT 2 Efficiency Score is a weighted geometric mean of the different workloads as follows:
  • Eff load is the ratio between the ‘Normalized performance’ ie. consumption normalized throughput for most worklets (with exception for memory worklets). It represents the amount of (normalized) transactions executed per unit of energy (Joule).
  • Worklet efficiency score Eff worklet is then calculated using the geometric mean of the separate load level scores

NB: contrary to the arithmetic mean, the geometric mean treats relative changes in efficiency equally at each load level.

  • Workload efficiency Eff workload is calculated by aggregating the efficiency scores of all worklets:

NB: SERT worklets are not designed to explicitly exercise General Purpose Graphics Processing Units (GPGPUs)

26 of 49

26

Data traffic indicators to be communicated

Annex II of the delegated act

See annex II to the delegated act adopted on March 14 2023 here

Data centre operators may base the monitoring and measurement of these indicators on any adequately reliable sources or combination of sources of data available, including data measured directly by the operator, data reported by data centre customers, or data provided by telecommunication operators and service providers !

Description

3

Data traffic indicators

a

Incoming traffic bandwidth (“BIN” in gigabytes per second) shall be measured as the total provisioned bandwidth for incoming traffic to the data centre compute.

b

Outgoing traffic bandwidth (“BOUT” in Gbs) shall be measured as the total provisioned bandwidth for outgoing traffic from the data centre computer room, aggregated for all the connectivity capacity, and averaged over the year

c

Incoming data traffic (“TIN”, in exabytes) shall be measured as the total incoming data to the data centre computer room, aggregated over the course of the reporting year, irrespective of the number of the data centre’s connections.

d

Outgoing data traffic (“TOUT”, in exabytes) shall be measured as the total outgoing data from the data centre computer room, aggregated over the course of the reporting year, irrespective of the number of the data centre’s connections.

27 of 49

Publicly available information in the EU database �Focus on Annex IV of the delegated act

27

Article 12 of Directive 2023/1791 states that information subject to Union and national law protecting trade and business secrets and confidentiality must not be made publicly available and requires that the European database be publicly available on an aggregated level. KPI communicated pursuant to Article 3 shall also be considered as confidential information

=> Annex IV of the delegated act introduces 5 size categories for the aggregation.

28 of 49

Publicly available information in the EU database �Focus on Annex IV of the delegated act

28

29 of 49

Timeline & specific guidances

  • By 15 September 2024, then by 15 May 2025, and every year thereafter, reporting DC operators shall communicate to the European database the information and KPI set out in Annex I and Annex II.

The information and KPI shall cover the calendar year immediately preceding the reporting year

NB: This communication shall take place via a national reporting scheme if the Member State where the reporting data centre is located has established such a scheme. Otherwise, the DC operators shall communicate this information and KPI directly to the European database.

  • For the first reporting period, if a DC operator cannot monitor and gather one or more of the KPI, points 1(d), 1(e), 1(h)-(l), and 1(o)-(r), for technical reasons, the DC operator may omit this information explaining the reasons for this omission.

  • For the first two reporting periods, if a colocation DC operator cannot monitor and gather the necessary data to KPI referred to in points 2(a) & 2(b) of Annex II, it shall estimate and indicate the percentage of the data centre computer room floor area that the information communicated covers.

  • Colocation DC operators may gather the KPI set out in Annex II, from their colocation customers, if necessary, by setting up an anonymous internal reporting mechanism.

29

See annex II to the delegated act adopted on March 14 2023 here

30 of 49

Main key points & limits�

30

  • By requesting the reporting of energy consumption of DC & IT equipment (EDC & EIT) separately + new ICT capacity indicators, the delegated Act allows to get access to more detailed information. Nevertheress, the ICT capacity for servers indicator (based on SERT) does not take into account all the components and DC operators & owners may not have access to all data.

  • By requesting the reporting of cooling degree days (CDD) the delegated act allows to consider the influence of the weather on the cooling needs but does not give a full picture of weather influence. Nevertheless, as Local Administrative Unit Codes are requested to localize the DC, one could theoretically have access to the heating degree days (HDD) in addition.

  • Even if recast EED obliges MS to lay down rules on penalties applicable to infringements that are “effective, proportionate and dissuasive”, the Directive does not mention the case of false statement and one may wonder how MS will be able to control all the declared data.

  • The indicators introduced by the EED do not take into account the all the environmental impacts over the life cycle of DC (depletion of abiotic resources, GHG emissions…) but only energy & water consumption during operation. The Ecodesign Directive and new Ecodesign for Sustainable Products Regulation (ESPR) fill some gaps but not all …

31 of 49

PUE

Power Usage Effectiveness

32 of 49

Le PUE, un indicateur d’efficacité énergétique

Source : Uptime Institute, “Global PUEs: Are they going anywhere ?”

33 of 49

Le PUE est variable dans l’année

Source: norme CEN-CENELEC 50600 / The Green Grid - PUE: A comprehensive examination of the metric

⇒ Les normes

EN 50660-4-2 &

ISO/EIC 30134-2 imposent

l’affichage d’une moyenne annuelle

34 of 49

Comment calculer son PUE ?

Formule de base :

“IT Equipment Energy” comprend également :

  • les KVM [keyboard-video-mouse] switches
  • les écrans
  • les desktops/laptops used to monitor or otherwise control the data center

“Total Facility Energy” comprend :

  • Les équipements destinés à délivrer la puissance électrique : UPS, commutateurs électriques, générateurs, les PDUs (power distribution unit), batteries, and distribution losses external to the IT equipment
  • Les équipements de refroidissement: chillers, cooling towers, pumps, computer room air-handling units (CRAHs), computer room air-conditioning units (CRACs), and direct expansion air-handler (DX) units • Le reste, comme l'éclairage, la sécurité incendie

Source: norme CEN-CENELEC 50600 / The Green Grid - PUE: A comprehensive examination of the metric

35 of 49

Comment calculer et ne pas calculer son PUE ?

Comment calculer son PUE ?

3 catégories avec une mesure de la consommation des serveurs de plus en plus proche des serveurs

Reco. : catégorie 2 pour PUE < 1,5, catégorie 3 pour PUE < 1,2

36 of 49

Comment calculer son PUE ?

Source: norme CEN-CENELEC 50600 / The Green Grid - PUE: A comprehensive examination of the metric

⚠️ La CEN-CENELEC 50600 indique

  • pour annoncer un PUE < 1.5 il faut que la méthode de calcul soit de niveau 2
  • pour annoncer un PUE < 1.2 il faut que la méthode de calcul soit de niveau 3

37 of 49

Limites du PUE

Le PUE permet de déterminer:

  • les opportunités d’amélioration de la performance opérationnelle d’un centre de traitement de données;
  • l’amélioration des conceptions et des processus d’un centre de traitement de données au cours du temps;
  • une valeur cible ou un objectif fixé(e) au moment de la conception des nouveaux centres de traitement de données pour la plage de charge IT anticipée.

Le PUE n’est pas:

  • un indicateur de la productivité du centre de traitement de données,
  • un indicateur autonome et complet du rendement d’utilisation des ressources.

Il convient de ne pas utiliser le PUE pour une comparaison avec les valeurs d’autres centres de traitement de données.

37

37

38 of 49

WUE

Water Usage Effectiveness

39 of 49

Définition (EN 50600-4-9)

WUE = WU / EIT

WU = utilisation annuelle de l’eau du DC en m3

EIT = consommation annuelle des équipements IT en MWh

WU = Win - Wout

Win = débit entrant annuel

Wout = débit sortant annuel

39

39

40 of 49

3 catégories qui ne mesurent pas la même chose

40

40

Catégorie

Catégorie 1 basique

Catégorie 2 intermédiaire

Catégorie 3 avancée

Débit entrant

Entrée du DC

Entrée du DC

Entrée du DC

- eau de pluie

+ eau nécessaire à la production d’énergie

Débit sortant

Aucune réutilisation

(toute l’eau qui rentre sort en tant qu’eau usée)

Réutilisation non industrielle

Réutilisations industrielle et non industrielle

Déclaration supplémentaire

Aucune

Aucune

Niveau de stress hydrique à l’échelle régionale et consommation des sols

41 of 49

ERF, CER, CUE…

42 of 49

ERF : Energy Reuse Factor

EReuse :

  • énergie provenant du Datacenter, à priori sous forme d’énergie thermique
  • à convertir en kWh
  • utilisée à l’extérieur du DC, annuellement
  • qui substitue tout ou partie de l’énergie nécessaire “en dehors des limites du DC”

EDC :

  • l’énergie totale consommée annuellement par le DC�(comprend la consommation IT et non-IT)
  • en kWh

42

42

⚠️ EReuse ne doit pas comprendre :

  • l’énergie ré-utilisée pour faire fonctionner un système de refroidissement nécessaire au DC
  • l’énergie ré-utilisée pour générer de l’électricité qui sert au DC
  • l’énergie ré-utilisée pour chauffer des bureaux ou espaces de support liés au DC

43 of 49

CER: Cooling Efficiency Ratio

  • la chaleur totale “retirée” et l’énergie électrique utilisée pour le refroidissement
  • the volume of the coolant and its heat capacity shall be measured
  • La consommation d’énergie électrique de tous les composants de refroidissement doit être prise en compte: pompes, valves, etc.
  • mesuré et inclus dans l’énergie utilisée
  • la 50600-4-7 ne précise pas la fréquence de mesure
  • sur 12 mois
  • en kWh

43

43

44 of 49

CUE: Carbon Usage Effectiveness

44

44

Phase d’usage uniquement !

  • Exprimé en kgCO2eq / kWh

45 of 49

La chaîne de normalisation

45

45

Source: APL Datacenter

46 of 49

Conclusion

47 of 49

Exigences de reporting

CSRD

Décret Tertiaire

EED�(si version actuelle confirmée,�si >500KW, dès mai 2024)

Loi REEN

Énergie consommée

Non�% d’alignement au CoC

Oui *1

Privé (ADEME operat)

Oui, CENELEC50600Public (UE platform)

Non

PUE

Non�% d’alignement au CoC

Oui *1

Privé (ADEME operat)

Oui, CENELEC50600Public (UE platform)

? : donne mandat à l’ARCEP ?

WUE

Non�% d’alignement au CoC

Non

Oui, CENELEC50600Public (UE platform)

Non

ERF

Non�% d’alignement au CoC

Non

Oui, CENELEC50600Public (UE platform)

Non

CUE

Non�% d’alignement au CoC

Non

Oui, CENELEC50600Public (UE platform)

? : donne mandat à l’ARCEP ?

CER

Non�% d’alignement au CoC

Non

Oui, CENELEC50600Public (UE platform)

Non

Indicateurs plus fins

Non

Non

Oui

Public en partie (UE platform)

Non

48 of 49

Réglementation à venir en mai 2024 : EED

49 of 49

49

Actors in our ecosystems

Pictogram taken from Flaticon credits to freepik, nikita-golubev, bearicons, pixel-perfect, photo3idea-studio

boavizta.org

Thank you for your attention

Discuter sur notre chat public :

Suivre nos actualités :

Accéder à notre site web :

Contribuer à nos outils :