@aws-sdk/client-ecs 3.1156.0 → 3.1158.0

package/dist-cjs/index.js

@@ -69,7 +69,7 @@ const commonParams = {
     UseDualStack: { type: "builtInParams", name: "useDualstackEndpoint" },
 };
 
-var version = "3.1155.0";
+var version = "3.1157.0";
 var packageInfo = {
 	version: version};
 
@@ -833,6 +833,8 @@ const _GCLa = "GatewayConnectorList";
 const _GCR = "GatewayConnectionResources";
 const _GCa = "GatewayConnector";
 const _GCat = "GatewayConfiguration";
+const _GD = "GpuDevice";
+const _GDp = "GpuDevices";
 const _GEALC = "GatewayElbAccessLogConfiguration";
 const _GEALD = "GatewayElbAccessLogDestination";
 const _GEALSD = "GatewayElbAccessLogS3Destination";
@@ -982,6 +984,7 @@ const _NC = "NetworkConfiguration";
 const _NI = "NetworkInterface";
 const _NICR = "NetworkInterfaceCountRequest";
 const _NIe = "NetworkInterfaces";
+const _NMA = "NvidiaMpsAllocation";
 const _NNFE = "NamespaceNotFoundException";
 const _NUAE = "NoUpdateAvailableException";
 const _PA = "PutAttributes";
@@ -1090,6 +1093,7 @@ const _SS = "SensitiveString";
 const _SSD = "StopServiceDeployment";
 const _SSDR = "StopServiceDeploymentRequest";
 const _SSDRt = "StopServiceDeploymentResponse";
+const _SSh = "SharingStrategy";
 const _ST = "StartTask";
 const _STR = "StartTaskRequest";
 const _STRt = "StartTaskResponse";
@@ -1266,7 +1270,8 @@ const _cB = "circuitBreaker";
 const _cBTIM = "canaryBakeTimeInMinutes";
 const _cBr = "createdBy";
 const _cC = "canaryConfiguration";
-const _cCo = "containerConfigurations";
+const _cCo = "containerCount";
+const _cCon = "containerConfigurations";
 const _cD = "containerDefinitions";
 const _cDC = "computedDesiredCount";
 const _cDu = "currentDeployment";
@@ -1424,13 +1429,16 @@ const _gC = "gatewayConnectors";
 const _gCA = "gatewayClusterAssociations";
 const _gCa = "gatewayConnections";
 const _gCr = "grpcCode";
+const _gD = "gpuDevices";
 const _gDE = "guardDutyEnabled";
 const _gI = "gpuIds";
+const _gIp = "gpuId";
 const _gR = "gatewayResources";
 const _gRA = "gatewayRouteArn";
 const _gRAa = "gatewayRouteArns";
 const _gRa = "gatewayRoute";
 const _gRat = "gatewayRoutes";
+const _gSS = "gpuSharingStrategy";
 const _ga = "gateways";
 const _gr = "group";
 const _h = "hostname";
@@ -1533,6 +1541,7 @@ const _mAC = "maxAlbCount";
 const _mAN = "managedAgentName";
 const _mAe = "metricAlarms";
 const _mC = "minCapacity";
+const _mCP = "maxComputePercent";
 const _mCa = "maxCapacity";
 const _mCe = "metricConfigurations";
 const _mD = "managedDraining";
@@ -1570,6 +1579,7 @@ const _nDI = "neuronDeviceIds";
 const _nI = "networkInterfaces";
 const _nIC = "networkInterfaceCount";
 const _nM = "networkMode";
+const _nMv = "nvidiaMps";
 const _nT = "nextToken";
 const _na = "namespace";
 const _nam = "names";
@@ -1632,6 +1642,7 @@ const _rI = "resourceIds";
 const _rIC = "runningInstanceCount";
 const _rIe = "referenceId";
 const _rIu = "runtimeId";
+const _rMIMB = "remainingMemoryInMiB";
 const _rMT = "resourceManagementType";
 const _rO = "readOnly";
 const _rOHT = "resetOnHealthyTask";
@@ -1692,6 +1703,7 @@ const _sIe = "sessionId";
 const _sIn = "snapshotId";
 const _sKP = "s3KeyPrefix";
 const _sL = "softLimit";
+const _sM = "sharingMode";
 const _sMPPOLP = "spotMaxPricePercentageOverLowestPrice";
 const _sMS = "sharedMemorySize";
 const _sN = "serviceName";
@@ -1713,6 +1725,7 @@ const _sSGB = "storageSizeGiB";
 const _sSR = "sourceServiceRevisions";
 const _sSRC = "sourceServiceRevisionCleanup";
 const _sSV = "stringSetValue";
+const _sSh = "sharingStrategy";
 const _sSt = "stabilityStatus";
 const _sT = "startTimeout";
 const _sTc = "scalingTarget";
@@ -1770,6 +1783,7 @@ const _tLR = "testListenerRule";
 const _tLRA = "testListenerRuleArn";
 const _tLRAe = "testListenerRuleArns";
 const _tLSGB = "totalLocalStorageGB";
+const _tMIMB = "totalMemoryInMiB";
 const _tP = "terminationPolicy";
 const _tR = "totalResources";
 const _tRA = "taskRoleArn";
@@ -2130,8 +2144,8 @@ var ClusterSetting$ = [3, n0, _CS,
 ];
 var Container$ = [3, n0, _Co,
     0,
-    [_cAo, _tA, _n, _im, _iD, _rIu, _lS, _eC, _r, _nB, _nI, _hS, _mA, _cp, _me, _mR, _gI, _nDI],
-    [0, 0, 0, 0, 0, 0, 0, 1, 0, () => NetworkBindings, () => NetworkInterfaces, 0, () => ManagedAgents, 0, 0, 0, 64 | 0, 64 | 0]
+    [_cAo, _tA, _n, _im, _iD, _rIu, _lS, _eC, _r, _nB, _nI, _hS, _mA, _cp, _me, _mR, _gI, _nDI, _gSS],
+    [0, 0, 0, 0, 0, 0, 0, 1, 0, () => NetworkBindings, () => NetworkInterfaces, 0, () => ManagedAgents, 0, 0, 0, 64 | 0, 64 | 0, () => SharingStrategy$]
 ];
 var ContainerDefinition$ = [3, n0, _CD,
     0,
@@ -2150,8 +2164,8 @@ var ContainerImage$ = [3, n0, _CI,
 ];
 var ContainerInstance$ = [3, n0, _CIo,
     0,
-    [_cIA, _eII, _cPN, _ve, _vI, _rRe, _rReg, _st, _sR, _aC, _rTC, _pTC, _aUS, _at, _rAe, _a, _ta, _hS],
-    [0, 0, 0, 1, () => VersionInfo$, () => Resources, () => Resources, 0, 0, 2, 1, 1, 0, () => Attributes, 4, () => Attachments, () => Tags, () => ContainerInstanceHealthStatus$]
+    [_cIA, _eII, _cPN, _ve, _vI, _rRe, _rReg, _st, _sR, _aC, _rTC, _pTC, _aUS, _at, _rAe, _a, _ta, _hS, _gD],
+    [0, 0, 0, 1, () => VersionInfo$, () => Resources, () => Resources, 0, 0, 2, 1, 1, 0, () => Attributes, 4, () => Attachments, () => Tags, () => ContainerInstanceHealthStatus$, () => GpuDevices]
 ];
 var ContainerInstanceHealthStatus$ = [3, n0, _CIHS,
     0,
@@ -2978,6 +2992,11 @@ var GetTaskProtectionResponse$ = [3, n0, _GTPRe,
     [_pTro, _fa],
     [() => ProtectedTasks, () => Failures]
 ];
+var GpuDevice$ = [3, n0, _GD,
+    0,
+    [_gIp, _sM, _tMIMB, _rMIMB, _cCo],
+    [0, 0, 1, 1, 1]
+];
 var HealthCheck$ = [3, n0, _HC,
     0,
     [_com, _int, _ti, _ret, _sP],
@@ -3388,6 +3407,11 @@ var NetworkInterfaceCountRequest$ = [3, n0, _NICR,
     [_mi, _ma],
     [1, 1]
 ];
+var NvidiaMpsAllocation$ = [3, n0, _NMA,
+    0,
+    [_me, _mCP],
+    [1, 1], 1
+];
 var PlacementConstraint$ = [3, n0, _PC,
     0,
     [_t, _ex],
@@ -3505,8 +3529,8 @@ var Resource$ = [3, n0, _R,
 ];
 var ResourceRequirement$ = [3, n0, _RR,
     0,
-    [_v, _t],
-    [0, 0], 2
+    [_t, _v, _sSh],
+    [0, 0, () => SharingStrategy$], 1
 ];
 var Rollback$ = [3, n0, _Ro,
     0,
@@ -3565,7 +3589,7 @@ var ServiceConnectClientAlias$ = [3, n0, _SCCA,
 ];
 var ServiceConnectConfiguration$ = [3, n0, _SCC,
     0,
-    [_ena, _na, _serv, _lC, _aLCc, _cCo],
+    [_ena, _na, _serv, _lC, _aLCc, _cCon],
     [2, 0, () => ServiceConnectServiceList, () => LogConfiguration$, () => ServiceConnectAccessLogConfiguration$, () => ServiceConnectContainerConfigurations], 1
 ];
 var ServiceConnectService$ = [3, n0, _SCS,
@@ -4183,6 +4207,9 @@ var GatewayRouteSummaryList = [1, n0, _GRSL,
 var GatewaySummaryList = [1, n0, _GSL,
     0, () => GatewaySummary$
 ];
+var GpuDevices = [1, n0, _GDp,
+    0, () => GpuDevice$
+];
 var HostEntryList = [1, n0, _HEL,
     0, () => HostEntry$
 ];
@@ -4399,6 +4426,11 @@ var ServiceConnectContainerConfiguration$ = [4, n0, _SCCCe,
     [_lL],
     [0]
 ];
+var SharingStrategy$ = [4, n0, _SSh,
+    0,
+    [_nMv],
+    [() => NvidiaMpsAllocation$]
+];
 var UpdateGatewayConfiguration$ = [4, n0, _UGC,
     0,
     [_hSA],
@@ -5867,6 +5899,11 @@ const ClusterSettingName = {
 const TargetType = {
     CONTAINER_INSTANCE: "container-instance",
 };
+const GpuSharingMode = {
+    DEDICATED: "DEDICATED",
+    IDLE: "IDLE",
+    NVIDIA_MPS: "NVIDIA_MPS",
+};
 const InstanceHealthCheckState = {
     IMPAIRED: "IMPAIRED",
     INITIALIZING: "INITIALIZING",
@@ -6728,6 +6765,8 @@ exports.GetTaskProtection$ = GetTaskProtection$;
 exports.GetTaskProtectionCommand = GetTaskProtectionCommand;
 exports.GetTaskProtectionRequest$ = GetTaskProtectionRequest$;
 exports.GetTaskProtectionResponse$ = GetTaskProtectionResponse$;
+exports.GpuDevice$ = GpuDevice$;
+exports.GpuSharingMode = GpuSharingMode;
 exports.HealthCheck$ = HealthCheck$;
 exports.HealthStatus = HealthStatus;
 exports.HostEntry$ = HostEntry$;
@@ -6870,6 +6909,7 @@ exports.NetworkInterfaceCountRequest$ = NetworkInterfaceCountRequest$;
 exports.NetworkMode = NetworkMode;
 exports.NoUpdateAvailableException = NoUpdateAvailableException;
 exports.NoUpdateAvailableException$ = NoUpdateAvailableException$;
+exports.NvidiaMpsAllocation$ = NvidiaMpsAllocation$;
 exports.OSFamily = OSFamily;
 exports.PidMode = PidMode;
 exports.PlacementConstraint$ = PlacementConstraint$;
@@ -6986,6 +7026,7 @@ exports.Session$ = Session$;
 exports.Setting$ = Setting$;
 exports.SettingName = SettingName;
 exports.SettingType = SettingType;
+exports.SharingStrategy$ = SharingStrategy$;
 exports.SortOrder = SortOrder;
 exports.StabilityStatus = StabilityStatus;
 exports.StartTask$ = StartTask$;

package/dist-es/models/enums.js

@@ -135,6 +135,11 @@ export const ClusterSettingName = {
 export const TargetType = {
     CONTAINER_INSTANCE: "container-instance",
 };
+export const GpuSharingMode = {
+    DEDICATED: "DEDICATED",
+    IDLE: "IDLE",
+    NVIDIA_MPS: "NVIDIA_MPS",
+};
 export const InstanceHealthCheckState = {
     IMPAIRED: "IMPAIRED",
     INITIALIZING: "INITIALIZING",

package/dist-es/schemas/schemas_0.js

@@ -286,6 +286,8 @@ const _GCLa = "GatewayConnectorList";
 const _GCR = "GatewayConnectionResources";
 const _GCa = "GatewayConnector";
 const _GCat = "GatewayConfiguration";
+const _GD = "GpuDevice";
+const _GDp = "GpuDevices";
 const _GEALC = "GatewayElbAccessLogConfiguration";
 const _GEALD = "GatewayElbAccessLogDestination";
 const _GEALSD = "GatewayElbAccessLogS3Destination";
@@ -435,6 +437,7 @@ const _NC = "NetworkConfiguration";
 const _NI = "NetworkInterface";
 const _NICR = "NetworkInterfaceCountRequest";
 const _NIe = "NetworkInterfaces";
+const _NMA = "NvidiaMpsAllocation";
 const _NNFE = "NamespaceNotFoundException";
 const _NUAE = "NoUpdateAvailableException";
 const _PA = "PutAttributes";
@@ -543,6 +546,7 @@ const _SS = "SensitiveString";
 const _SSD = "StopServiceDeployment";
 const _SSDR = "StopServiceDeploymentRequest";
 const _SSDRt = "StopServiceDeploymentResponse";
+const _SSh = "SharingStrategy";
 const _ST = "StartTask";
 const _STR = "StartTaskRequest";
 const _STRt = "StartTaskResponse";
@@ -719,7 +723,8 @@ const _cB = "circuitBreaker";
 const _cBTIM = "canaryBakeTimeInMinutes";
 const _cBr = "createdBy";
 const _cC = "canaryConfiguration";
-const _cCo = "containerConfigurations";
+const _cCo = "containerCount";
+const _cCon = "containerConfigurations";
 const _cD = "containerDefinitions";
 const _cDC = "computedDesiredCount";
 const _cDu = "currentDeployment";
@@ -877,13 +882,16 @@ const _gC = "gatewayConnectors";
 const _gCA = "gatewayClusterAssociations";
 const _gCa = "gatewayConnections";
 const _gCr = "grpcCode";
+const _gD = "gpuDevices";
 const _gDE = "guardDutyEnabled";
 const _gI = "gpuIds";
+const _gIp = "gpuId";
 const _gR = "gatewayResources";
 const _gRA = "gatewayRouteArn";
 const _gRAa = "gatewayRouteArns";
 const _gRa = "gatewayRoute";
 const _gRat = "gatewayRoutes";
+const _gSS = "gpuSharingStrategy";
 const _ga = "gateways";
 const _gr = "group";
 const _h = "hostname";
@@ -986,6 +994,7 @@ const _mAC = "maxAlbCount";
 const _mAN = "managedAgentName";
 const _mAe = "metricAlarms";
 const _mC = "minCapacity";
+const _mCP = "maxComputePercent";
 const _mCa = "maxCapacity";
 const _mCe = "metricConfigurations";
 const _mD = "managedDraining";
@@ -1023,6 +1032,7 @@ const _nDI = "neuronDeviceIds";
 const _nI = "networkInterfaces";
 const _nIC = "networkInterfaceCount";
 const _nM = "networkMode";
+const _nMv = "nvidiaMps";
 const _nT = "nextToken";
 const _na = "namespace";
 const _nam = "names";
@@ -1085,6 +1095,7 @@ const _rI = "resourceIds";
 const _rIC = "runningInstanceCount";
 const _rIe = "referenceId";
 const _rIu = "runtimeId";
+const _rMIMB = "remainingMemoryInMiB";
 const _rMT = "resourceManagementType";
 const _rO = "readOnly";
 const _rOHT = "resetOnHealthyTask";
@@ -1145,6 +1156,7 @@ const _sIe = "sessionId";
 const _sIn = "snapshotId";
 const _sKP = "s3KeyPrefix";
 const _sL = "softLimit";
+const _sM = "sharingMode";
 const _sMPPOLP = "spotMaxPricePercentageOverLowestPrice";
 const _sMS = "sharedMemorySize";
 const _sN = "serviceName";
@@ -1166,6 +1178,7 @@ const _sSGB = "storageSizeGiB";
 const _sSR = "sourceServiceRevisions";
 const _sSRC = "sourceServiceRevisionCleanup";
 const _sSV = "stringSetValue";
+const _sSh = "sharingStrategy";
 const _sSt = "stabilityStatus";
 const _sT = "startTimeout";
 const _sTc = "scalingTarget";
@@ -1223,6 +1236,7 @@ const _tLR = "testListenerRule";
 const _tLRA = "testListenerRuleArn";
 const _tLRAe = "testListenerRuleArns";
 const _tLSGB = "totalLocalStorageGB";
+const _tMIMB = "totalMemoryInMiB";
 const _tP = "terminationPolicy";
 const _tR = "totalResources";
 const _tRA = "taskRoleArn";
@@ -1586,8 +1600,8 @@ export var ClusterSetting$ = [3, n0, _CS,
 ];
 export var Container$ = [3, n0, _Co,
     0,
-    [_cAo, _tA, _n, _im, _iD, _rIu, _lS, _eC, _r, _nB, _nI, _hS, _mA, _cp, _me, _mR, _gI, _nDI],
-    [0, 0, 0, 0, 0, 0, 0, 1, 0, () => NetworkBindings, () => NetworkInterfaces, 0, () => ManagedAgents, 0, 0, 0, 64 | 0, 64 | 0]
+    [_cAo, _tA, _n, _im, _iD, _rIu, _lS, _eC, _r, _nB, _nI, _hS, _mA, _cp, _me, _mR, _gI, _nDI, _gSS],
+    [0, 0, 0, 0, 0, 0, 0, 1, 0, () => NetworkBindings, () => NetworkInterfaces, 0, () => ManagedAgents, 0, 0, 0, 64 | 0, 64 | 0, () => SharingStrategy$]
 ];
 export var ContainerDefinition$ = [3, n0, _CD,
     0,
@@ -1606,8 +1620,8 @@ export var ContainerImage$ = [3, n0, _CI,
 ];
 export var ContainerInstance$ = [3, n0, _CIo,
     0,
-    [_cIA, _eII, _cPN, _ve, _vI, _rRe, _rReg, _st, _sR, _aC, _rTC, _pTC, _aUS, _at, _rAe, _a, _ta, _hS],
-    [0, 0, 0, 1, () => VersionInfo$, () => Resources, () => Resources, 0, 0, 2, 1, 1, 0, () => Attributes, 4, () => Attachments, () => Tags, () => ContainerInstanceHealthStatus$]
+    [_cIA, _eII, _cPN, _ve, _vI, _rRe, _rReg, _st, _sR, _aC, _rTC, _pTC, _aUS, _at, _rAe, _a, _ta, _hS, _gD],
+    [0, 0, 0, 1, () => VersionInfo$, () => Resources, () => Resources, 0, 0, 2, 1, 1, 0, () => Attributes, 4, () => Attachments, () => Tags, () => ContainerInstanceHealthStatus$, () => GpuDevices]
 ];
 export var ContainerInstanceHealthStatus$ = [3, n0, _CIHS,
     0,
@@ -2434,6 +2448,11 @@ export var GetTaskProtectionResponse$ = [3, n0, _GTPRe,
     [_pTro, _fa],
     [() => ProtectedTasks, () => Failures]
 ];
+export var GpuDevice$ = [3, n0, _GD,
+    0,
+    [_gIp, _sM, _tMIMB, _rMIMB, _cCo],
+    [0, 0, 1, 1, 1]
+];
 export var HealthCheck$ = [3, n0, _HC,
     0,
     [_com, _int, _ti, _ret, _sP],
@@ -2844,6 +2863,11 @@ export var NetworkInterfaceCountRequest$ = [3, n0, _NICR,
     [_mi, _ma],
     [1, 1]
 ];
+export var NvidiaMpsAllocation$ = [3, n0, _NMA,
+    0,
+    [_me, _mCP],
+    [1, 1], 1
+];
 export var PlacementConstraint$ = [3, n0, _PC,
     0,
     [_t, _ex],
@@ -2961,8 +2985,8 @@ export var Resource$ = [3, n0, _R,
 ];
 export var ResourceRequirement$ = [3, n0, _RR,
     0,
-    [_v, _t],
-    [0, 0], 2
+    [_t, _v, _sSh],
+    [0, 0, () => SharingStrategy$], 1
 ];
 export var Rollback$ = [3, n0, _Ro,
     0,
@@ -3021,7 +3045,7 @@ export var ServiceConnectClientAlias$ = [3, n0, _SCCA,
 ];
 export var ServiceConnectConfiguration$ = [3, n0, _SCC,
     0,
-    [_ena, _na, _serv, _lC, _aLCc, _cCo],
+    [_ena, _na, _serv, _lC, _aLCc, _cCon],
     [2, 0, () => ServiceConnectServiceList, () => LogConfiguration$, () => ServiceConnectAccessLogConfiguration$, () => ServiceConnectContainerConfigurations], 1
 ];
 export var ServiceConnectService$ = [3, n0, _SCS,
@@ -3651,6 +3675,9 @@ var GatewayStatusList = 64 | 0;
 var GatewaySummaryList = [1, n0, _GSL,
     0, () => GatewaySummary$
 ];
+var GpuDevices = [1, n0, _GDp,
+    0, () => GpuDevice$
+];
 var GpuIds = 64 | 0;
 var HostEntryList = [1, n0, _HEL,
     0, () => HostEntry$
@@ -3883,6 +3910,11 @@ export var ServiceConnectContainerConfiguration$ = [4, n0, _SCCCe,
     [_lL],
     [0]
 ];
+export var SharingStrategy$ = [4, n0, _SSh,
+    0,
+    [_nMv],
+    [() => NvidiaMpsAllocation$]
+];
 export var UpdateGatewayConfiguration$ = [4, n0, _UGC,
     0,
     [_hSA],

package/dist-types/commands/DeleteTaskDefinitionsCommand.d.ts

@@ -205,8 +205,14 @@ declare const DeleteTaskDefinitionsCommand_base: {
  * //           ],
  * //           resourceRequirements: [ // ResourceRequirements
  * //             { // ResourceRequirement
- * //               value: "STRING_VALUE", // required
+ * //               value: "STRING_VALUE",
  * //               type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ * //               sharingStrategy: { // SharingStrategy Union: only one key present
+ * //                 nvidiaMps: { // NvidiaMpsAllocation
+ * //                   memory: Number("int"), // required
+ * //                   maxComputePercent: Number("int"),
+ * //                 },
+ * //               },
  * //             },
  * //           ],
  * //           firelensConfiguration: { // FirelensConfiguration

package/dist-types/commands/DeregisterContainerInstanceCommand.d.ts

@@ -121,6 +121,15 @@ declare const DeregisterContainerInstanceCommand_base: {
  * //         },
  * //       ],
  * //     },
+ * //     gpuDevices: [ // GpuDevices
+ * //       { // GpuDevice
+ * //         gpuId: "STRING_VALUE",
+ * //         sharingMode: "IDLE" || "NVIDIA_MPS" || "DEDICATED",
+ * //         totalMemoryInMiB: Number("int"),
+ * //         remainingMemoryInMiB: Number("int"),
+ * //         containerCount: Number("int"),
+ * //       },
+ * //     ],
  * //   },
  * // };
  *

package/dist-types/commands/DeregisterTaskDefinitionCommand.d.ts

@@ -202,8 +202,14 @@ declare const DeregisterTaskDefinitionCommand_base: {
  * //         ],
  * //         resourceRequirements: [ // ResourceRequirements
  * //           { // ResourceRequirement
- * //             value: "STRING_VALUE", // required
+ * //             value: "STRING_VALUE",
  * //             type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ * //             sharingStrategy: { // SharingStrategy Union: only one key present
+ * //               nvidiaMps: { // NvidiaMpsAllocation
+ * //                 memory: Number("int"), // required
+ * //                 maxComputePercent: Number("int"),
+ * //               },
+ * //             },
  * //           },
  * //         ],
  * //         firelensConfiguration: { // FirelensConfiguration

package/dist-types/commands/DescribeContainerInstancesCommand.d.ts

@@ -126,6 +126,15 @@ declare const DescribeContainerInstancesCommand_base: {
  * //           },
  * //         ],
  * //       },
+ * //       gpuDevices: [ // GpuDevices
+ * //         { // GpuDevice
+ * //           gpuId: "STRING_VALUE",
+ * //           sharingMode: "IDLE" || "NVIDIA_MPS" || "DEDICATED",
+ * //           totalMemoryInMiB: Number("int"),
+ * //           remainingMemoryInMiB: Number("int"),
+ * //           containerCount: Number("int"),
+ * //         },
+ * //       ],
  * //     },
  * //   ],
  * //   failures: [ // Failures

package/dist-types/commands/DescribeTaskDefinitionCommand.d.ts

@@ -205,8 +205,14 @@ declare const DescribeTaskDefinitionCommand_base: {
  * //         ],
  * //         resourceRequirements: [ // ResourceRequirements
  * //           { // ResourceRequirement
- * //             value: "STRING_VALUE", // required
+ * //             value: "STRING_VALUE",
  * //             type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ * //             sharingStrategy: { // SharingStrategy Union: only one key present
+ * //               nvidiaMps: { // NvidiaMpsAllocation
+ * //                 memory: Number("int"), // required
+ * //                 maxComputePercent: Number("int"),
+ * //               },
+ * //             },
  * //           },
  * //         ],
  * //         firelensConfiguration: { // FirelensConfiguration

package/dist-types/commands/DescribeTasksCommand.d.ts

@@ -120,6 +120,12 @@ declare const DescribeTasksCommand_base: {
  * //           neuronDeviceIds: [ // NeuronDeviceIds
  * //             "STRING_VALUE",
  * //           ],
+ * //           gpuSharingStrategy: { // SharingStrategy Union: only one key present
+ * //             nvidiaMps: { // NvidiaMpsAllocation
+ * //               memory: Number("int"), // required
+ * //               maxComputePercent: Number("int"),
+ * //             },
+ * //           },
  * //         },
  * //       ],
  * //       cpu: "STRING_VALUE",
@@ -162,8 +168,14 @@ declare const DescribeTasksCommand_base: {
  * //             memoryReservation: Number("int"),
  * //             resourceRequirements: [ // ResourceRequirements
  * //               { // ResourceRequirement
- * //                 value: "STRING_VALUE", // required
+ * //                 value: "STRING_VALUE",
  * //                 type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ * //                 sharingStrategy: {//  Union: only one key present
+ * //                   nvidiaMps: {
+ * //                     memory: Number("int"), // required
+ * //                     maxComputePercent: Number("int"),
+ * //                   },
+ * //                 },
  * //               },
  * //             ],
  * //           },

package/dist-types/commands/RegisterContainerInstanceCommand.d.ts

@@ -159,6 +159,15 @@ declare const RegisterContainerInstanceCommand_base: {
  * //         },
  * //       ],
  * //     },
+ * //     gpuDevices: [ // GpuDevices
+ * //       { // GpuDevice
+ * //         gpuId: "STRING_VALUE",
+ * //         sharingMode: "IDLE" || "NVIDIA_MPS" || "DEDICATED",
+ * //         totalMemoryInMiB: Number("int"),
+ * //         remainingMemoryInMiB: Number("int"),
+ * //         containerCount: Number("int"),
+ * //       },
+ * //     ],
  * //   },
  * // };
  *

package/dist-types/commands/RegisterTaskDefinitionCommand.d.ts

@@ -199,8 +199,14 @@ declare const RegisterTaskDefinitionCommand_base: {
  *       ],
  *       resourceRequirements: [ // ResourceRequirements
  *         { // ResourceRequirement
- *           value: "STRING_VALUE", // required
+ *           value: "STRING_VALUE",
  *           type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ *           sharingStrategy: { // SharingStrategy Union: only one key present
+ *             nvidiaMps: { // NvidiaMpsAllocation
+ *               memory: Number("int"), // required
+ *               maxComputePercent: Number("int"),
+ *             },
+ *           },
  *         },
  *       ],
  *       firelensConfiguration: { // FirelensConfiguration
@@ -466,8 +472,14 @@ declare const RegisterTaskDefinitionCommand_base: {
  * //         ],
  * //         resourceRequirements: [ // ResourceRequirements
  * //           { // ResourceRequirement
- * //             value: "STRING_VALUE", // required
+ * //             value: "STRING_VALUE",
  * //             type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ * //             sharingStrategy: { // SharingStrategy Union: only one key present
+ * //               nvidiaMps: { // NvidiaMpsAllocation
+ * //                 memory: Number("int"), // required
+ * //                 maxComputePercent: Number("int"),
+ * //               },
+ * //             },
  * //           },
  * //         ],
  * //         firelensConfiguration: { // FirelensConfiguration

package/dist-types/commands/RunTaskCommand.d.ts

@@ -82,8 +82,14 @@ declare const RunTaskCommand_base: {
  *         memoryReservation: Number("int"),
  *         resourceRequirements: [ // ResourceRequirements
  *           { // ResourceRequirement
- *             value: "STRING_VALUE", // required
+ *             value: "STRING_VALUE",
  *             type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ *             sharingStrategy: { // SharingStrategy Union: only one key present
+ *               nvidiaMps: { // NvidiaMpsAllocation
+ *                 memory: Number("int"), // required
+ *                 maxComputePercent: Number("int"),
+ *               },
+ *             },
  *           },
  *         ],
  *       },
@@ -237,6 +243,12 @@ declare const RunTaskCommand_base: {
  * //           neuronDeviceIds: [ // NeuronDeviceIds
  * //             "STRING_VALUE",
  * //           ],
+ * //           gpuSharingStrategy: { // SharingStrategy Union: only one key present
+ * //             nvidiaMps: { // NvidiaMpsAllocation
+ * //               memory: Number("int"), // required
+ * //               maxComputePercent: Number("int"),
+ * //             },
+ * //           },
  * //         },
  * //       ],
  * //       cpu: "STRING_VALUE",
@@ -279,8 +291,14 @@ declare const RunTaskCommand_base: {
  * //             memoryReservation: Number("int"),
  * //             resourceRequirements: [ // ResourceRequirements
  * //               { // ResourceRequirement
- * //                 value: "STRING_VALUE", // required
+ * //                 value: "STRING_VALUE",
  * //                 type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ * //                 sharingStrategy: {//  Union: only one key present
+ * //                   nvidiaMps: {
+ * //                     memory: Number("int"), // required
+ * //                     maxComputePercent: Number("int"),
+ * //                   },
+ * //                 },
  * //               },
  * //             ],
  * //           },

package/dist-types/commands/StartTaskCommand.d.ts

@@ -74,8 +74,14 @@ declare const StartTaskCommand_base: {
  *         memoryReservation: Number("int"),
  *         resourceRequirements: [ // ResourceRequirements
  *           { // ResourceRequirement
- *             value: "STRING_VALUE", // required
+ *             value: "STRING_VALUE",
  *             type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ *             sharingStrategy: { // SharingStrategy Union: only one key present
+ *               nvidiaMps: { // NvidiaMpsAllocation
+ *                 memory: Number("int"), // required
+ *                 maxComputePercent: Number("int"),
+ *               },
+ *             },
  *           },
  *         ],
  *       },
@@ -215,6 +221,12 @@ declare const StartTaskCommand_base: {
  * //           neuronDeviceIds: [ // NeuronDeviceIds
  * //             "STRING_VALUE",
  * //           ],
+ * //           gpuSharingStrategy: { // SharingStrategy Union: only one key present
+ * //             nvidiaMps: { // NvidiaMpsAllocation
+ * //               memory: Number("int"), // required
+ * //               maxComputePercent: Number("int"),
+ * //             },
+ * //           },
  * //         },
  * //       ],
  * //       cpu: "STRING_VALUE",
@@ -257,8 +269,14 @@ declare const StartTaskCommand_base: {
  * //             memoryReservation: Number("int"),
  * //             resourceRequirements: [ // ResourceRequirements
  * //               { // ResourceRequirement
- * //                 value: "STRING_VALUE", // required
+ * //                 value: "STRING_VALUE",
  * //                 type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ * //                 sharingStrategy: {//  Union: only one key present
+ * //                   nvidiaMps: {
+ * //                     memory: Number("int"), // required
+ * //                     maxComputePercent: Number("int"),
+ * //                   },
+ * //                 },
  * //               },
  * //             ],
  * //           },

package/dist-types/commands/StopTaskCommand.d.ts

@@ -115,6 +115,12 @@ declare const StopTaskCommand_base: {
  * //         neuronDeviceIds: [ // NeuronDeviceIds
  * //           "STRING_VALUE",
  * //         ],
+ * //         gpuSharingStrategy: { // SharingStrategy Union: only one key present
+ * //           nvidiaMps: { // NvidiaMpsAllocation
+ * //             memory: Number("int"), // required
+ * //             maxComputePercent: Number("int"),
+ * //           },
+ * //         },
  * //       },
  * //     ],
  * //     cpu: "STRING_VALUE",
@@ -157,8 +163,14 @@ declare const StopTaskCommand_base: {
  * //           memoryReservation: Number("int"),
  * //           resourceRequirements: [ // ResourceRequirements
  * //             { // ResourceRequirement
- * //               value: "STRING_VALUE", // required
+ * //               value: "STRING_VALUE",
  * //               type: "GPU" || "InferenceAccelerator" || "NeuronDevice", // required
+ * //               sharingStrategy: {//  Union: only one key present
+ * //                 nvidiaMps: {
+ * //                   memory: Number("int"), // required
+ * //                   maxComputePercent: Number("int"),
+ * //                 },
+ * //               },
  * //             },
  * //           ],
  * //         },

package/dist-types/commands/UpdateContainerAgentCommand.d.ts

@@ -120,6 +120,15 @@ declare const UpdateContainerAgentCommand_base: {
  * //         },
  * //       ],
  * //     },
+ * //     gpuDevices: [ // GpuDevices
+ * //       { // GpuDevice
+ * //         gpuId: "STRING_VALUE",
+ * //         sharingMode: "IDLE" || "NVIDIA_MPS" || "DEDICATED",
+ * //         totalMemoryInMiB: Number("int"),
+ * //         remainingMemoryInMiB: Number("int"),
+ * //         containerCount: Number("int"),
+ * //       },
+ * //     ],
  * //   },
  * // };
  *

package/dist-types/commands/UpdateContainerInstancesStateCommand.d.ts

@@ -124,6 +124,15 @@ declare const UpdateContainerInstancesStateCommand_base: {
  * //           },
  * //         ],
  * //       },
+ * //       gpuDevices: [ // GpuDevices
+ * //         { // GpuDevice
+ * //           gpuId: "STRING_VALUE",
+ * //           sharingMode: "IDLE" || "NVIDIA_MPS" || "DEDICATED",
+ * //           totalMemoryInMiB: Number("int"),
+ * //           remainingMemoryInMiB: Number("int"),
+ * //           containerCount: Number("int"),
+ * //         },
+ * //       ],
  * //     },
  * //   ],
  * //   failures: [ // Failures

package/dist-types/models/enums.d.ts

@@ -349,6 +349,28 @@ export declare const TargetType: {
  * @public
  */
 export type TargetType = (typeof TargetType)[keyof typeof TargetType];
+/**
+ * @public
+ * @enum
+ */
+export declare const GpuSharingMode: {
+    /**
+     * <p>One container has exclusive use of the GPU. Amazon ECS assigns GPUs this way to containers that specify <code>value</code> instead of <code>sharingStrategy</code> for the <code>GPU</code> resource type.</p>
+     */
+    readonly DEDICATED: "DEDICATED";
+    /**
+     * <p>No containers are assigned to the GPU, and all of its memory is available.</p>
+     */
+    readonly IDLE: "IDLE";
+    /**
+     * <p>One or more containers share the GPU by using NVIDIA Multi-Process Service (MPS). While the GPU is in this mode, Amazon ECS assigns it only to containers that use the <code>nvidiaMps</code> sharing strategy.</p>
+     */
+    readonly NVIDIA_MPS: "NVIDIA_MPS";
+};
+/**
+ * @public
+ */
+export type GpuSharingMode = (typeof GpuSharingMode)[keyof typeof GpuSharingMode];
 /**
  * @public
  * @enum

package/dist-types/models/models_0.d.ts

@@ -1,5 +1,5 @@
 import type { DocumentType as __DocumentType } from "@smithy/types";
-import type { AcceleratorManufacturer, AcceleratorName, AcceleratorType, AccessType, AgentUpdateStatus, ApplicationProtocol, AssignPublicIp, AutoRepairActionsStatus, BareMetal, BurstablePerformance, CapacityOptionType, CapacityProviderField, CapacityProviderStatus, CapacityProviderType, CapacityProviderUpdateStatus, CapacityReservationPreference, ClusterField, ClusterSettingName, Compatibility, ContainerCondition, ContainerInstanceField, ContainerInstanceStatus, CPUArchitecture, CpuManufacturer, DaemonDeploymentRollbackMonitorsStatus, DaemonDeploymentStatus, DaemonIpcMode, DaemonPidMode, DaemonPropagateTags, DaemonStatus, DaemonTaskDefinitionRevisionFilter, DaemonTaskDefinitionStatus, DaemonTaskDefinitionStatusFilter, DeploymentControllerType, DeploymentLifecycleHookAction, DeploymentLifecycleHookStage, DeploymentLifecycleHookStatus, DeploymentLifecycleHookTargetType, DeploymentStrategy, DesiredStatus, DeviceCgroupPermission, EFSAuthorizationConfigIAM, EFSTransitEncryption, EnvironmentFileType, ExecuteCommandLogging, ExpressCpuArchitecture, ExpressGatewayServiceScalingMetric, ExpressGatewayServiceStatusCode, FirelensConfigurationType, GatewayAlbDesyncMitigationMode, GatewayAlbProtocol, GatewayClusterAssociationStatus, GatewayConnectorHealthCheckProtocol, GatewayElbIpAddressType, GatewayElbScheme, GatewayRouteStatus, GatewayStatus, InstanceGeneration, InstanceHealthCheckState, InstanceHealthCheckType, IpcMode, LaunchType, LocalStorage, LocalStorageType, LogDriver, ManagedAgentName, ManagedDraining, ManagedInstancesMonitoringOptions, ManagedScalingStatus, ManagedTerminationProtection, NetworkMode, OSFamily, PidMode, PlacementConstraintType, PlacementStrategyType, PlatformDeviceType, PropagateGatewayTags, PropagateMITags, ProxyConfigurationType, ResourceType, Scope, ServiceConnectAccessLoggingFormat, ServiceConnectContainerLogLevel, ServiceConnectIncludeQueryParameters, ServiceDeploymentLifecycleStage, ServiceDeploymentRollbackMonitorsStatus, ServiceDeploymentStatus, ServiceRevisionCleanup, SettingName, SettingType, SortOrder, TargetType, TaskDefinitionFamilyStatus, TaskDefinitionField, TaskDefinitionPlacementConstraintType, TaskDefinitionStatus, ThresholdType, TransportProtocol, UlimitName, VersionConsistency } from "./enums";
+import type { AcceleratorManufacturer, AcceleratorName, AcceleratorType, AccessType, AgentUpdateStatus, ApplicationProtocol, AssignPublicIp, AutoRepairActionsStatus, BareMetal, BurstablePerformance, CapacityOptionType, CapacityProviderField, CapacityProviderStatus, CapacityProviderType, CapacityProviderUpdateStatus, CapacityReservationPreference, ClusterField, ClusterSettingName, Compatibility, ContainerCondition, ContainerInstanceField, ContainerInstanceStatus, CPUArchitecture, CpuManufacturer, DaemonDeploymentRollbackMonitorsStatus, DaemonDeploymentStatus, DaemonIpcMode, DaemonPidMode, DaemonPropagateTags, DaemonStatus, DaemonTaskDefinitionRevisionFilter, DaemonTaskDefinitionStatus, DaemonTaskDefinitionStatusFilter, DeploymentControllerType, DeploymentLifecycleHookAction, DeploymentLifecycleHookStage, DeploymentLifecycleHookStatus, DeploymentLifecycleHookTargetType, DeploymentStrategy, DesiredStatus, DeviceCgroupPermission, EFSAuthorizationConfigIAM, EFSTransitEncryption, EnvironmentFileType, ExecuteCommandLogging, ExpressCpuArchitecture, ExpressGatewayServiceScalingMetric, ExpressGatewayServiceStatusCode, FirelensConfigurationType, GatewayAlbDesyncMitigationMode, GatewayAlbProtocol, GatewayClusterAssociationStatus, GatewayConnectorHealthCheckProtocol, GatewayElbIpAddressType, GatewayElbScheme, GatewayRouteStatus, GatewayStatus, GpuSharingMode, InstanceGeneration, InstanceHealthCheckState, InstanceHealthCheckType, IpcMode, LaunchType, LocalStorage, LocalStorageType, LogDriver, ManagedAgentName, ManagedDraining, ManagedInstancesMonitoringOptions, ManagedScalingStatus, ManagedTerminationProtection, NetworkMode, OSFamily, PidMode, PlacementConstraintType, PlacementStrategyType, PlatformDeviceType, PropagateGatewayTags, PropagateMITags, ProxyConfigurationType, ResourceType, Scope, ServiceConnectAccessLoggingFormat, ServiceConnectContainerLogLevel, ServiceConnectIncludeQueryParameters, ServiceDeploymentLifecycleStage, ServiceDeploymentRollbackMonitorsStatus, ServiceDeploymentStatus, ServiceRevisionCleanup, SettingName, SettingType, SortOrder, TargetType, TaskDefinitionFamilyStatus, TaskDefinitionField, TaskDefinitionPlacementConstraintType, TaskDefinitionStatus, ThresholdType, TransportProtocol, UlimitName, VersionConsistency } from "./enums";
 /**
  * <p>The minimum and maximum number of accelerators (such as GPUs) for instance type selection. This is used for workloads that require specific numbers of accelerators.</p>
  * @public
@@ -1280,6 +1280,37 @@ export interface Attribute {
      */
     targetId?: string | undefined;
 }
+/**
+ * <p>The state of a GPU on a container instance, including its sharing mode and the memory that's available for new containers.</p>
+ * @public
+ */
+export interface GpuDevice {
+    /**
+     * <p>The ID of the GPU. This ID matches the GPU IDs in the <code>gpuIds</code> of the containers that Amazon ECS assigned to the GPU.</p>
+     * @public
+     */
+    gpuId?: string | undefined;
+    /**
+     * <p>How the GPU is shared between containers.</p> <p>Valid values:</p> <ul> <li> <p> <code>IDLE</code> - No containers are assigned to the GPU, and all of its memory is available.</p> </li> <li> <p> <code>NVIDIA_MPS</code> - One or more containers share the GPU by using NVIDIA Multi-Process Service (MPS).</p> </li> <li> <p> <code>DEDICATED</code> - One container has exclusive use of the GPU.</p> </li> </ul>
+     * @public
+     */
+    sharingMode?: GpuSharingMode | undefined;
+    /**
+     * <p>The total amount of GPU memory, in MiB, that the Amazon ECS container agent reported for the GPU. This is the memory that's available to containers, and it can be less than the nominal memory of the GPU.</p>
+     * @public
+     */
+    totalMemoryInMiB?: number | undefined;
+    /**
+     * <p>The amount of GPU memory, in MiB, that's available to allocate to new containers that use the <code>nvidiaMps</code> sharing strategy. Amazon ECS calculates this value from the memory that it allocated to containers on the GPU, not from the memory that the containers use.</p>
+     * @public
+     */
+    remainingMemoryInMiB?: number | undefined;
+    /**
+     * <p>The number of containers that Amazon ECS assigned to the GPU. This value is <code>0</code> when the sharing mode is <code>IDLE</code> and <code>1</code> when the sharing mode is <code>DEDICATED</code>. Amazon ECS counts the containers that it assigned to the GPU, not the processes that are running on it.</p>
+     * @public
+     */
+    containerCount?: number | undefined;
+}
 /**
  * <p>An object representing the result of a container instance health status check.</p>
  * @public
@@ -1479,6 +1510,11 @@ export interface ContainerInstance {
      * @public
      */
     healthStatus?: ContainerInstanceHealthStatus | undefined;
+    /**
+     * <p>The state of each GPU on the container instance, including its sharing mode and available memory. This field is present only for container instances that support GPU sharing.</p>
+     * @public
+     */
+    gpuDevices?: GpuDevice[] | undefined;
 }
 /**
  *
@@ -4183,21 +4219,75 @@ export interface PortMapping {
      */
     containerPortRange?: string | undefined;
 }
+/**
+ * <p>The GPU memory and compute limits for a container that shares a GPU by using NVIDIA Multi-Process Service (MPS). Amazon ECS assigns the container a single GPU that has enough memory available. Other containers that use MPS can share the same GPU.</p>
+ * @public
+ */
+export interface NvidiaMpsAllocation {
+    /**
+     * <p>The amount of GPU memory, in MiB, to allocate to the container. Amazon ECS assigns the container a GPU that has at least the specified amount of memory available, and sets the <code>CUDA_MPS_PINNED_DEVICE_MEM_LIMIT</code> environment variable in the container to limit its memory to this amount. The memory must be available on a single GPU. If more than one container in a task uses <code>nvidiaMps</code>, Amazon ECS assigns them the same GPU, which must have enough memory available for all of them.</p>
+     * @public
+     */
+    memory: number | undefined;
+    /**
+     * <p>The maximum percentage of the GPU's compute capacity that the container can use. Amazon ECS sets the <code>CUDA_MPS_ACTIVE_THREAD_PERCENTAGE</code> environment variable in the container to this value. This value is a limit, not a reservation.</p> <p>If you don't specify this parameter, Amazon ECS doesn't set <code>CUDA_MPS_ACTIVE_THREAD_PERCENTAGE</code>, and the active thread percentage of the MPS server on the container instance applies. By default, this is 100 percent.</p>
+     * @public
+     */
+    maxComputePercent?: number | undefined;
+}
+/**
+ * <p>The strategy that a container uses to share a GPU with other containers. Specify exactly one member.</p>
+ * @public
+ */
+export type SharingStrategy = SharingStrategy.NvidiaMpsMember | SharingStrategy.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace SharingStrategy {
+    /**
+     * <p>Shares a GPU between containers by using NVIDIA Multi-Process Service (MPS). You specify the GPU memory to allocate to the container and, optionally, the maximum percentage of the GPU's compute capacity that the container can use.</p>
+     * @public
+     */
+    interface NvidiaMpsMember {
+        nvidiaMps: NvidiaMpsAllocation;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        nvidiaMps?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        nvidiaMps: (value: NvidiaMpsAllocation) => T;
+        _: (name: string, value: any) => T;
+    }
+}
 /**
  * <p>The type and amount of a resource to assign to a container. The supported resource types are GPUs, Neuron devices, and Elastic Inference accelerators. For more information, see <a href="https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ecs-gpu.html">Working with GPUs on Amazon ECS</a> or <a href="https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ecs-inference.html">Working with Amazon Elastic Inference on Amazon ECS</a> in the <i>Amazon Elastic Container Service Developer Guide</i> </p>
  * @public
  */
 export interface ResourceRequirement {
     /**
-     * <p>The value for the specified resource type.</p> <p>When the type is <code>GPU</code>, the value is the number of physical <code>GPUs</code> the Amazon ECS container agent reserves for the container. The number of GPUs that's reserved for all containers in a task can't exceed the number of available GPUs on the container instance that the task is launched on. You can also specify <code>ALL</code> to allocate all available GPUs on the instance to the container.</p> <p>When the type is <code>NeuronDevice</code>, the value must be <code>ALL</code>. This allocates all available Neuron devices on the instance to the container. Only one container in a task can specify <code>NeuronDevice</code> resources. This resource type is only supported on Managed Instances.</p> <p>When the type is <code>InferenceAccelerator</code>, the <code>value</code> matches the <code>deviceName</code> for an <a href="https://docs.aws.amazon.com/AmazonECS/latest/APIReference/API_InferenceAccelerator.html">InferenceAccelerator</a> specified in a task definition.</p>
+     * <p>The value for the specified resource type.</p> <p>You must specify this parameter unless the type is <code>GPU</code> and you specify <code>sharingStrategy</code>.</p> <p>When the type is <code>GPU</code>, the value is the number of physical <code>GPUs</code> the Amazon ECS container agent reserves for the container. The number of GPUs that's reserved for all containers in a task can't exceed the number of available GPUs on the container instance that the task is launched on. You can also specify <code>ALL</code> to allocate all available GPUs on the instance to the container.</p> <p>When the type is <code>NeuronDevice</code>, the value must be <code>ALL</code>. This allocates all available Neuron devices on the instance to the container. Only one container in a task can specify <code>NeuronDevice</code> resources. This resource type is only supported on Managed Instances.</p> <p>When the type is <code>InferenceAccelerator</code>, the <code>value</code> matches the <code>deviceName</code> for an <a href="https://docs.aws.amazon.com/AmazonECS/latest/APIReference/API_InferenceAccelerator.html">InferenceAccelerator</a> specified in a task definition.</p>
      * @public
      */
-    value: string | undefined;
+    value?: string | undefined;
     /**
      * <p>The type of resource to assign to a container. </p>
      * @public
      */
     type: ResourceType | undefined;
+    /**
+     * <p>The strategy that the container uses to share a GPU with other containers. Specify a sharing strategy only when the type is <code>GPU</code>. For a GPU resource requirement, specify either <code>value</code> or <code>sharingStrategy</code>, but not both. If you don't specify a sharing strategy, the container has exclusive use of each GPU that Amazon ECS assigns to it. The containers in a task that request GPUs must all use the same option.</p> <p>You can use GPU sharing with tasks that run on Amazon EC2 instances or Amazon ECS Managed Instances. Task definitions that require <code>EXTERNAL</code> compatibility don't support GPU sharing.</p>
+     * @public
+     */
+    sharingStrategy?: SharingStrategy | undefined;
 }
 /**
  * <p>Details on a data volume from another container in the same task definition.</p>
@@ -7462,51 +7552,3 @@ export interface ServiceConnectClientAlias {
      */
     testTrafficRules?: ServiceConnectTestTrafficRules | undefined;
 }
-/**
- * <p>An object that represents the timeout configurations for Service Connect.</p> <note> <p>If <code>idleTimeout</code> is set to a time that is less than <code>perRequestTimeout</code>, the connection will close when the <code>idleTimeout</code> is reached and not the <code>perRequestTimeout</code>.</p> </note>
- * @public
- */
-export interface TimeoutConfiguration {
-    /**
-     * <p>The amount of time in seconds a connection will stay active while idle. A value of <code>0</code> can be set to disable <code>idleTimeout</code>.</p> <p>The <code>idleTimeout</code> default for <code>HTTP</code>/<code>HTTP2</code>/<code>GRPC</code> is 5 minutes.</p> <p>The <code>idleTimeout</code> default for <code>TCP</code> is 1 hour.</p>
-     * @public
-     */
-    idleTimeoutSeconds?: number | undefined;
-    /**
-     * <p>The amount of time waiting for the upstream to respond with a complete response per request. A value of <code>0</code> can be set to disable <code>perRequestTimeout</code>. <code>perRequestTimeout</code> can only be set if Service Connect <code>appProtocol</code> isn't <code>TCP</code>. Only <code>idleTimeout</code> is allowed for <code>TCP</code> <code>appProtocol</code>.</p>
-     * @public
-     */
-    perRequestTimeoutSeconds?: number | undefined;
-}
-/**
- * <p>The certificate root authority that secures your service.</p>
- * @public
- */
-export interface ServiceConnectTlsCertificateAuthority {
-    /**
-     * <p>The ARN of the Amazon Web Services Private Certificate Authority certificate.</p>
-     * @public
-     */
-    awsPcaAuthorityArn?: string | undefined;
-}
-/**
- * <p>The key that encrypts and decrypts your resources for Service Connect TLS.</p>
- * @public
- */
-export interface ServiceConnectTlsConfiguration {
-    /**
-     * <p>The signer certificate authority.</p>
-     * @public
-     */
-    issuerCertificateAuthority: ServiceConnectTlsCertificateAuthority | undefined;
-    /**
-     * <p>The Amazon Web Services Key Management Service key.</p>
-     * @public
-     */
-    kmsKey?: string | undefined;
-    /**
-     * <p>The Amazon Resource Name (ARN) of the IAM role that's associated with the Service Connect TLS.</p>
-     * @public
-     */
-    roleArn?: string | undefined;
-}

package/dist-types/models/models_1.d.ts

@@ -1,5 +1,53 @@
 import type { AccessType, AvailabilityZoneRebalancing, Compatibility, Connectivity, DeploymentRolloutState, EBSResourceType, ExpressCpuArchitecture, ExpressGatewayServiceInclude, GatewayRouteStatus, HealthStatus, IpcMode, LaunchType, ManagedAgentName, ManagedResourceStatus, NetworkMode, PidMode, PropagateTags, ResourceManagementType, ScaleUnit, SchedulingStrategy, ServiceDeploymentStatus, ServiceField, SortOrder, StabilityStatus, StopServiceDeploymentStopType, TaskDefinitionStatus, TaskField, TaskFilesystemType, TaskSetField, TaskStopCode } from "./enums";
-import type { Attachment, Attribute, CapacityProviderStrategyItem, ContainerDefinition, CreatedAt, DeploymentConfiguration, DeploymentController, ECSExpressGatewayService, EnvironmentFile, EphemeralStorage, ExpressGatewayContainer, ExpressGatewayScalingTarget, ExpressGatewayServiceConfiguration, ExpressGatewayServiceNetworkConfiguration, ExpressGatewayServiceStatus, Failure, GatewayConnector, InferenceAccelerator, KeyValuePair, LoadBalancer, LogConfiguration, MonitoringConfiguration, NetworkBinding, NetworkConfiguration, PlacementConstraint, PlacementStrategy, ProxyConfiguration, ResourceRequirement, RuntimePlatform, ServiceConnectAccessLogConfiguration, ServiceConnectClientAlias, ServiceConnectContainerConfiguration, ServiceConnectTlsConfiguration, Tag, TaskDefinition, TaskDefinitionPlacementConstraint, TimeoutConfiguration, Volume } from "./models_0";
+import type { Attachment, Attribute, CapacityProviderStrategyItem, ContainerDefinition, CreatedAt, DeploymentConfiguration, DeploymentController, ECSExpressGatewayService, EnvironmentFile, EphemeralStorage, ExpressGatewayContainer, ExpressGatewayScalingTarget, ExpressGatewayServiceConfiguration, ExpressGatewayServiceNetworkConfiguration, ExpressGatewayServiceStatus, Failure, GatewayConnector, InferenceAccelerator, KeyValuePair, LoadBalancer, LogConfiguration, MonitoringConfiguration, NetworkBinding, NetworkConfiguration, PlacementConstraint, PlacementStrategy, ProxyConfiguration, ResourceRequirement, RuntimePlatform, ServiceConnectAccessLogConfiguration, ServiceConnectClientAlias, ServiceConnectContainerConfiguration, SharingStrategy, Tag, TaskDefinition, TaskDefinitionPlacementConstraint, Volume } from "./models_0";
+/**
+ * <p>An object that represents the timeout configurations for Service Connect.</p> <note> <p>If <code>idleTimeout</code> is set to a time that is less than <code>perRequestTimeout</code>, the connection will close when the <code>idleTimeout</code> is reached and not the <code>perRequestTimeout</code>.</p> </note>
+ * @public
+ */
+export interface TimeoutConfiguration {
+    /**
+     * <p>The amount of time in seconds a connection will stay active while idle. A value of <code>0</code> can be set to disable <code>idleTimeout</code>.</p> <p>The <code>idleTimeout</code> default for <code>HTTP</code>/<code>HTTP2</code>/<code>GRPC</code> is 5 minutes.</p> <p>The <code>idleTimeout</code> default for <code>TCP</code> is 1 hour.</p>
+     * @public
+     */
+    idleTimeoutSeconds?: number | undefined;
+    /**
+     * <p>The amount of time waiting for the upstream to respond with a complete response per request. A value of <code>0</code> can be set to disable <code>perRequestTimeout</code>. <code>perRequestTimeout</code> can only be set if Service Connect <code>appProtocol</code> isn't <code>TCP</code>. Only <code>idleTimeout</code> is allowed for <code>TCP</code> <code>appProtocol</code>.</p>
+     * @public
+     */
+    perRequestTimeoutSeconds?: number | undefined;
+}
+/**
+ * <p>The certificate root authority that secures your service.</p>
+ * @public
+ */
+export interface ServiceConnectTlsCertificateAuthority {
+    /**
+     * <p>The ARN of the Amazon Web Services Private Certificate Authority certificate.</p>
+     * @public
+     */
+    awsPcaAuthorityArn?: string | undefined;
+}
+/**
+ * <p>The key that encrypts and decrypts your resources for Service Connect TLS.</p>
+ * @public
+ */
+export interface ServiceConnectTlsConfiguration {
+    /**
+     * <p>The signer certificate authority.</p>
+     * @public
+     */
+    issuerCertificateAuthority: ServiceConnectTlsCertificateAuthority | undefined;
+    /**
+     * <p>The Amazon Web Services Key Management Service key.</p>
+     * @public
+     */
+    kmsKey?: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the IAM role that's associated with the Service Connect TLS.</p>
+     * @public
+     */
+    roleArn?: string | undefined;
+}
 /**
  * <p>The Service Connect service object configuration. For more information, see <a href="https://docs.aws.amazon.com/AmazonECS/latest/developerguide/service-connect.html">Service Connect</a> in the <i>Amazon Elastic Container Service Developer Guide</i>.</p>
  * @public
@@ -2632,6 +2680,11 @@ export interface Container {
      * @public
      */
     neuronDeviceIds?: string[] | undefined;
+    /**
+     * <p>The GPU sharing strategy that the task definition specifies for the container. This field is present only for containers whose resource requirements include <code>sharingStrategy</code>.</p>
+     * @public
+     */
+    gpuSharingStrategy?: SharingStrategy | undefined;
 }
 /**
  * <p>The amount of ephemeral storage to allocate for the task.</p>

package/dist-types/schemas/schemas_0.d.ts

@@ -230,6 +230,7 @@ export declare var GatewaySummary$: StaticStructureSchema;
 export declare var GatewayTestTrafficHeader$: StaticStructureSchema;
 export declare var GetTaskProtectionRequest$: StaticStructureSchema;
 export declare var GetTaskProtectionResponse$: StaticStructureSchema;
+export declare var GpuDevice$: StaticStructureSchema;
 export declare var HealthCheck$: StaticStructureSchema;
 export declare var HostEntry$: StaticStructureSchema;
 export declare var HostSharedAlbConfiguration$: StaticStructureSchema;
@@ -312,6 +313,7 @@ export declare var NetworkBinding$: StaticStructureSchema;
 export declare var NetworkConfiguration$: StaticStructureSchema;
 export declare var NetworkInterface$: StaticStructureSchema;
 export declare var NetworkInterfaceCountRequest$: StaticStructureSchema;
+export declare var NvidiaMpsAllocation$: StaticStructureSchema;
 export declare var PlacementConstraint$: StaticStructureSchema;
 export declare var PlacementStrategy$: StaticStructureSchema;
 export declare var PlatformDevice$: StaticStructureSchema;
@@ -444,6 +446,7 @@ export declare var GatewayElbAccessLogDestination$: StaticUnionSchema;
 export declare var GatewayInfrastructureResources$: StaticUnionSchema;
 export declare var GatewayRouteResources$: StaticUnionSchema;
 export declare var ServiceConnectContainerConfiguration$: StaticUnionSchema;
+export declare var SharingStrategy$: StaticUnionSchema;
 export declare var UpdateGatewayConfiguration$: StaticUnionSchema;
 export declare var AssociateGatewayWithCluster$: StaticOperationSchema;
 export declare var ContinueServiceDeployment$: StaticOperationSchema;

package/dist-types/ts3.4/models/enums.d.ts

@@ -170,6 +170,12 @@ export declare const TargetType: {
   readonly CONTAINER_INSTANCE: "container-instance";
 };
 export type TargetType = (typeof TargetType)[keyof typeof TargetType];
+export declare const GpuSharingMode: {
+  readonly DEDICATED: "DEDICATED";
+  readonly IDLE: "IDLE";
+  readonly NVIDIA_MPS: "NVIDIA_MPS";
+};
+export type GpuSharingMode = (typeof GpuSharingMode)[keyof typeof GpuSharingMode];
 export declare const InstanceHealthCheckState: {
   readonly IMPAIRED: "IMPAIRED";
   readonly INITIALIZING: "INITIALIZING";

package/dist-types/ts3.4/models/models_0.d.ts

@@ -57,6 +57,7 @@ import {
   GatewayElbScheme,
   GatewayRouteStatus,
   GatewayStatus,
+  GpuSharingMode,
   InstanceGeneration,
   InstanceHealthCheckState,
   InstanceHealthCheckType,
@@ -401,6 +402,13 @@ export interface Attribute {
   targetType?: TargetType | undefined;
   targetId?: string | undefined;
 }
+export interface GpuDevice {
+  gpuId?: string | undefined;
+  sharingMode?: GpuSharingMode | undefined;
+  totalMemoryInMiB?: number | undefined;
+  remainingMemoryInMiB?: number | undefined;
+  containerCount?: number | undefined;
+}
 export interface InstanceHealthCheckResult {
   type?: InstanceHealthCheckType | undefined;
   status?: InstanceHealthCheckState | undefined;
@@ -444,6 +452,7 @@ export interface ContainerInstance {
   attachments?: Attachment[] | undefined;
   tags?: Tag[] | undefined;
   healthStatus?: ContainerInstanceHealthStatus | undefined;
+  gpuDevices?: GpuDevice[] | undefined;
 }
 export interface DeregisterContainerInstanceResponse {
   containerInstance?: ContainerInstance | undefined;
@@ -1083,9 +1092,29 @@ export interface PortMapping {
   appProtocol?: ApplicationProtocol | undefined;
   containerPortRange?: string | undefined;
 }
+export interface NvidiaMpsAllocation {
+  memory: number | undefined;
+  maxComputePercent?: number | undefined;
+}
+export type SharingStrategy = SharingStrategy.NvidiaMpsMember | SharingStrategy.$UnknownMember;
+export declare namespace SharingStrategy {
+  interface NvidiaMpsMember {
+    nvidiaMps: NvidiaMpsAllocation;
+    $unknown?: never;
+  }
+  interface $UnknownMember {
+    nvidiaMps?: never;
+    $unknown: [string, any];
+  }
+  interface Visitor<T> {
+    nvidiaMps: (value: NvidiaMpsAllocation) => T;
+    _: (name: string, value: any) => T;
+  }
+}
 export interface ResourceRequirement {
-  value: string | undefined;
+  value?: string | undefined;
   type: ResourceType | undefined;
+  sharingStrategy?: SharingStrategy | undefined;
 }
 export interface VolumeFrom {
   sourceContainer?: string | undefined;
@@ -1914,15 +1943,3 @@ export interface ServiceConnectClientAlias {
   dnsName?: string | undefined;
   testTrafficRules?: ServiceConnectTestTrafficRules | undefined;
 }
-export interface TimeoutConfiguration {
-  idleTimeoutSeconds?: number | undefined;
-  perRequestTimeoutSeconds?: number | undefined;
-}
-export interface ServiceConnectTlsCertificateAuthority {
-  awsPcaAuthorityArn?: string | undefined;
-}
-export interface ServiceConnectTlsConfiguration {
-  issuerCertificateAuthority: ServiceConnectTlsCertificateAuthority | undefined;
-  kmsKey?: string | undefined;
-  roleArn?: string | undefined;
-}

package/dist-types/ts3.4/models/models_1.d.ts

@@ -63,13 +63,24 @@ import {
   ServiceConnectAccessLogConfiguration,
   ServiceConnectClientAlias,
   ServiceConnectContainerConfiguration,
-  ServiceConnectTlsConfiguration,
+  SharingStrategy,
   Tag,
   TaskDefinition,
   TaskDefinitionPlacementConstraint,
-  TimeoutConfiguration,
   Volume,
 } from "./models_0";
+export interface TimeoutConfiguration {
+  idleTimeoutSeconds?: number | undefined;
+  perRequestTimeoutSeconds?: number | undefined;
+}
+export interface ServiceConnectTlsCertificateAuthority {
+  awsPcaAuthorityArn?: string | undefined;
+}
+export interface ServiceConnectTlsConfiguration {
+  issuerCertificateAuthority: ServiceConnectTlsCertificateAuthority | undefined;
+  kmsKey?: string | undefined;
+  roleArn?: string | undefined;
+}
 export interface ServiceConnectService {
   portName: string | undefined;
   discoveryName?: string | undefined;
@@ -669,6 +680,7 @@ export interface Container {
   memoryReservation?: string | undefined;
   gpuIds?: string[] | undefined;
   neuronDeviceIds?: string[] | undefined;
+  gpuSharingStrategy?: SharingStrategy | undefined;
 }
 export interface TaskEphemeralStorage {
   sizeInGiB?: number | undefined;

package/dist-types/ts3.4/schemas/schemas_0.d.ts

@@ -230,6 +230,7 @@ export declare var GatewaySummary$: StaticStructureSchema;
 export declare var GatewayTestTrafficHeader$: StaticStructureSchema;
 export declare var GetTaskProtectionRequest$: StaticStructureSchema;
 export declare var GetTaskProtectionResponse$: StaticStructureSchema;
+export declare var GpuDevice$: StaticStructureSchema;
 export declare var HealthCheck$: StaticStructureSchema;
 export declare var HostEntry$: StaticStructureSchema;
 export declare var HostSharedAlbConfiguration$: StaticStructureSchema;
@@ -312,6 +313,7 @@ export declare var NetworkBinding$: StaticStructureSchema;
 export declare var NetworkConfiguration$: StaticStructureSchema;
 export declare var NetworkInterface$: StaticStructureSchema;
 export declare var NetworkInterfaceCountRequest$: StaticStructureSchema;
+export declare var NvidiaMpsAllocation$: StaticStructureSchema;
 export declare var PlacementConstraint$: StaticStructureSchema;
 export declare var PlacementStrategy$: StaticStructureSchema;
 export declare var PlatformDevice$: StaticStructureSchema;
@@ -444,6 +446,7 @@ export declare var GatewayElbAccessLogDestination$: StaticUnionSchema;
 export declare var GatewayInfrastructureResources$: StaticUnionSchema;
 export declare var GatewayRouteResources$: StaticUnionSchema;
 export declare var ServiceConnectContainerConfiguration$: StaticUnionSchema;
+export declare var SharingStrategy$: StaticUnionSchema;
 export declare var UpdateGatewayConfiguration$: StaticUnionSchema;
 export declare var AssociateGatewayWithCluster$: StaticOperationSchema;
 export declare var ContinueServiceDeployment$: StaticOperationSchema;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@aws-sdk/client-ecs",
-  "version": "3.1156.0",
+  "version": "3.1158.0",
   "description": "AWS SDK for JavaScript Ecs Client for Node.js, Browser and React Native",
   "homepage": "https://github.com/aws/aws-sdk-js-v3/tree/main/clients/client-ecs",
   "license": "Apache-2.0",
@@ -48,12 +48,12 @@
     "test:index": "tsc -p tsconfig.test.json && node ./test/index-objects.spec.mjs"
   },
   "dependencies": {
-    "@aws-sdk/core": "^3.978.5",
-    "@aws-sdk/credential-provider-node": "^3.972.88",
+    "@aws-sdk/core": "^3.978.6",
+    "@aws-sdk/credential-provider-node": "^3.972.89",
     "@aws-sdk/types": "^3.975.0",
-    "@smithy/core": "^3.36.1",
+    "@smithy/core": "^3.37.0",
     "@smithy/fetch-http-handler": "^5.8.0",
-    "@smithy/node-http-handler": "^4.12.1",
+    "@smithy/node-http-handler": "^4.13.0",
     "@smithy/types": "^4.19.0",
     "tslib": "^2.6.2"
   },