Meta tags:
description= The behavior, billing, and limitations of the performance monitoring unit (PMU) for Compute Engine instances.;
Headings (most frequently used words):
pmu, and, pricing, events, supported, cpu, intel, rapids, overview, stay, organized, with, collections, save, categorize, content, based, on, your, preferences, understand, the, limitations, what, next, machine, series, platforms, by, platform, google, axion, emerald, granite, nvidia, grace, products, support, resources, engage,
Text of the page (most frequently used words):
standard (649), create (116), the (112), and (110), vms (103), for (80), enhanced (76), with (74), 0x01 (69), troubleshoot (62), 0x02 (61), about (59), instance (53), architectural (51), 0x04 (49), manage (46), performance (44), disk (43), 0x10 (42), 0x08 (42), using (41), pmu (39), 0x24 (39), server (39), 0x20 (38), view (38), compute (36), use (35), disks (35), configure (34), l2_rqsts (33), overview (33), windows (33), mig (33), google (30), 0x40 (30), machine (28), set (28), cloud (27), engine (27), reservation (27), br_misp_retired (24), 0xc5 (24), from (23), events (22), 0x80 (22), instances (21), cpu (21), hyperdisk (21), sql (21), 0x11 (20), fp_arith_inst_retired (20), 0xc7 (20), 0xc4 (20), monitor (20), images (20), type (19), network (19), resources (18), 0x12 (18), 0xe7 (18), mem_inst_retired (18), 0xd0 (18), br_inst_retired (18), access (18), gpu (18), that (17), 0x21 (17), event (17), image (17), linux (17), policies (16), enable (16), 0x13 (16), int_vec_retired (16), reservations (16), you (15), 0x03 (15), exe_activity (15), 0xa6 (15), cycle_activity (15), 0xa3 (15), snapshots (15), groups (15), ssh (15), recommendations (15), configuration (15), mem_load_retired (14), 0xd1 (14), fp_arith_inst_retired2 (14), 0xcf (14), rtm_retired (14), 0xc9 (14), uops_dispatched (14), 0xb2 (14), management (14), workload (14), creation (14), host (14), load (14), virtual (14), apply (14), custom (14), connect (14), code (13), any (13), core (13), bulk (13), microsoft (13), cluster (13), inst_retired (12), dtlb_store_misses (12), dtlb_load_misses (12), time (12), workloads (12), storage (12), sole (12), availability (12), persistent (12), best (12), regional (12), cpu_clk_unhalted (11), start (11), 0xc1 (11), processor (11), number (11), monitoring (11), based (11), delete (11), practices (11), stateful (11), following (10), types (10), can (10), 0x47 (10), topdown (10), 0xa4 (10), itlb_misses (10), your (10), high (10), optimize (10), hpc (10), metadata (10), migs (10), machines (10), attached (10), migrate (10), add (10), logs (10), zones (10), this (9), 0x48 (9), 0xc0 (9), assists (9), sw_prefetch_access (9), offcore_requests (9), instructions (9), application (9), licenses (9), managed (9), tenant (9), multiple (9), openshift (9), shutdown (9), all (8), license (8), 0x79 (8), nvidia (8), 0xec (8), mem_load_l3_miss_retired (8), 0xd3 (8), mem_load_l3_hit_retired (8), 0xd2 (8), 0x00 (8), int_misc (8), 0xad (8), memory_activity (8), service (8), import (8), resource (8), login (8), internal (8), choose (8), install (8), disaster (8), recovery (8), agent (8), existing (8), update (8), guest (8), boot (8), keys (8), h4d (8), flexibility (8), dns (8), optimized (8), tpu (8), understand (7), thumb (7), 0x41 (7), 0x3c (7), 0x26 (7), supported (7), reference (7), offcore_requests_outstanding (7), intel (7), when (7), data (7), usage (7), applications (7), networking (7), commitments (7), startup (7), balancing (7), benchmark (7), manually (7), idle (7), addresses (7), request (7), scale (7), deploy (7), back (7), mysql (7), zonal (7), remove (7), maintenance (7), templates (7), snapshot (7), backup (7), send (6), 0x83 (6), 0x82 (6), 0x51 (6), 0x1b (6), name (6), machine_clears (6), 0xc3 (6), 0xc2 (6), fp_arith_dispatched (6), 0xb3 (6), uops_executed (6), 0xb1 (6), arith (6), 0xb0 (6), idq (6), tx_mem (6), 0x54 (6), l1d_pend_miss (6), stlb_hit (6), walk_active (6), walk_completed (6), 0x0e (6), walk_completed_2m_4m (6), walk_completed_4k (6), ld_blocks (6), generation (6), rapids (6), cycles (6), series (6), measure (6), local (6), available (6), default (6), security (6), migration (6), updates (6), instant (6), files (6), nested (6), gpus (6), across (6), group (6), mode (6), demand (6), pools (6), run (6), failover (6), extension (6), change (6), new (6), transfer (6), customized (6), spot (6), samples (5), down (5), details (5), clusters (5), 0x87 (5), 0x60 (5), 0x44 (5), 0x42 (5), stall (5), 0x23 (5), tracking (5), 0x09 (5), miss (5), l2_lines_out (5), scalable (5), how (5), branch (5), databases (5), accounts (5), errors (5), future (5), drivers (5), virtualization (5), operations (5), placement (5), node (5), highly (5), build (5), multi (5), modify (5), capacity (5), schedules (5), audit (5), symphony (5), options (5), email (5), scripts (5), manager (5), between (5), operating (5), systems (5), capabilities (5), stop (5), tenancy (5), restore (5), regions (5), control (5), workstation (5), faq (5), flex (5), português (4), español (4), support (4), see (4), are (4), one (4), sve_inst_spec (4), 0x8006 (4), l1i_cache_lmiss (4), 0x4006 (4), 0x88 (4), 0x81 (4), 0x76 (4), 0x75 (4), 0x70 (4), 0x61 (4), 0x4c (4), 0x43 (4), l1d_cache_rd (4), 0x3f (4), l1d_cache_lmiss_rd (4), 0x39 (4), 0x2d (4), 0x25 (4), stall_backend (4), stall_frontend (4), 0x22 (4), chain (4), 0x1e (4), 0x1c (4), inst_spec (4), br_pred (4), cpu_cycles (4), br_mis_pred (4), 0x8 (4), l1d_cache (4), 0x4 (4), l1d_cache_refill (4), 0x3 (4), sw_incr (4), 0x0 (4), table (4), lists (4), processors (4), cores (4), 0x0c (4), l2_hit (4), indirect (4), near_taken (4), cond (4), cond_ntaken (4), cond_taken (4), all_branches (4), uops_retired (4), slots (4), port_1 (4), port_0 (4), stalls_l1d_miss (4), cycles_l1d_miss (4), stalls_l3_miss (4), stalls_l2_miss (4), 0x05 (4), resource_stalls (4), 0xa2 (4), stalls (4), walk_completed_1g (4), list (4), platform (4), level (4), within (4), retired (4), pmcs (4), specific (4), software (4), counters (4), tools (4), distributed (4), authentication (4), issues (4), ubuntu (4), console (4), bandwidth (4), tcp (4), ssd (4), get (4), share (4), projects (4), upgrade (4), live (4), autoscaling (4), cross (4), region (4), requests (4), calendar (4), single (4), metrics (4), health (4), active (4), containers (4), container (4), database (4), uses (4), postgresql (4), extensions (4), work (4), zone (4), suspend (4), location (4), resize (4), graceful (4), locations (4), plans (4), replication (4), review (4), bound (4), terms (3), pricing (3), other (3), information (3), more (3), content (3), apache (3), kubernetes (3), ultramem (3), costs (3), 0x50 (3), 0x4f (3), 0x18 (3), grace (3), arm (3), vector (3), scalar (3), count (3), hardware (3), 0xa5 (3), decode (3), l1d (3), longest_lat_cache (3), 0x2e (3), granite (3), emerald (3), axion (3), a4x (3), c4a (3), platforms (3), cache (3), halted (3), runs (3), changes (3), identify (3), bottlenecks (3), throughput (3), such (3), guides (3), infrastructure (3), observability (3), analytics (3), automatic (3), troubleshooting (3), encryption (3), pro (3), serial (3), configurations (3), patterns (3), higher (3), discounts (3), cuds (3), manual (3), resilient (3), autoscale (3), balancer (3), prevent (3), project (3), policy (3), testing (3), ibm (3), spectrum (3), perform (3), deploying (3), plan (3), during (3), prepare (3), deployment (3), rhel (3), environment (3), protect (3), certificates (3), query (3), state (3), disable (3), properties (3), account (3), increase (3), schedule (3), lifecycle (3), design (3), size (3), non (3), pool (3), move (3), creating (3), workstations (3), tpus (3), product (3), 한국어 (2), 日本語 (2), עברית (2), brasil (2), italiano (2), indonesia (2), français (2), américa (2), latina (2), deutsch (2), english (2), sign (2), site (2), youtube (2), center (2), started (2), system (2), products (2), need (2), last (2), updated (2), 2026 (2), utc (2), licensed (2), under (2), its (2), feedback (2), next (2), 224 (2), 112 (2), megamem (2), only (2), limitations (2), ase_sve_int64_spec (2), 0x80ef (2), ase_sve_int32_spec (2), 0x80eb (2), ase_sve_int16_spec (2), 0x80e7 (2), ase_sve_int8_spec (2), 0x80e3 (2), fp_fixed_ops_spec (2), 0x80c1 (2), fp_scale_ops_spec (2), 0x80c0 (2), sve_ldff_fault_spec (2), 0x80bd (2), sve_ldff_spec (2), 0x80bc (2), sve_pred_not_full_spec (2), 0x8079 (2), sve_pred_partial_spec (2), 0x8077 (2), sve_pred_full_spec (2), 0x8076 (2), sve_pred_empty_spec (2), 0x8075 (2), sve_pred_spec (2), 0x8074 (2), fp_dp_spec (2), 0x801c (2), fp_sp_spec (2), 0x8018 (2), fp_hp_spec (2), 0x8014 (2), ase_inst_spec (2), 0x8005 (2), st_align_lat (2), 0x4022 (2), ld_align_lat (2), 0x4021 (2), ldst_align_lat (2), 0x4020 (2), l2d_cache_lmiss_rd (2), 0x4009 (2), stall_backend_mem (2), 0x4005 (2), cnt_cycles (2), 0x4004 (2), rc_st_spec (2), 0x91 (2), rc_ld_spec (2), 0x90 (2), exc_trap_fiq (2), 0x8f (2), exc_trap_irq (2), 0x8e (2), exc_trap_other (2), 0x8d (2), exc_trap_dabort (2), 0x8c (2), exc_trap_pabort (2), 0x8b (2), exc_hvc (2), 0x8a (2), exc_smc (2), exc_fiq (2), exc_irq (2), 0x86 (2), exc_dabort (2), 0x84 (2), exc_pabort (2), exc_svc (2), exc_undef (2), dmb_spec (2), 0x7e (2), dsb_spec (2), 0x7d (2), isb_spec (2), 0x7c (2), br_indirect_spec (2), 0x7a (2), br_return_spec (2), br_immed_spec (2), 0x78 (2), crypto_spec (2), 0x77 (2), pc_write_spec (2), vfp_spec (2), ase_spec (2), 0x74 (2), dp_spec (2), 0x73 (2), st_spec (2), 0x71 (2), ld_spec (2), strex_spec (2), 0x6f (2), strex_fail_spec (2), 0x6e (2), strex_pass_spec (2), 0x6d (2), ldrex_spec (2), 0x6c (2), unaligned_ldst_spec (2), 0x6a (2), unaligned_st_spec (2), 0x69 (2), unaligned_ld_spec (2), 0x68 (2), mem_access_wr (2), 0x67 (2), mem_access_rd (2), 0x66 (2), bus_access_wr (2), bus_access_rd (2), l2d_tlb_wr (2), 0x5f (2), l2d_tlb_rd (2), 0x5e (2), l2d_tlb_refill_wr (2), 0x5d (2), l2d_tlb_refill_rd (2), 0x5c (2), l2d_cache_inval (2), 0x58 (2), l2d_cache_wb_clean (2), 0x57 (2), l2d_cache_wb_victim (2), 0x56 (2), l2d_cache_refill_wr (2), 0x53 (2), l2d_cache_refill_rd (2), 0x52 (2), l2d_cache_wr (2), l2d_cache_rd (2), l1d_tlb_wr (2), l1d_tlb_rd (2), 0x4e (2), l1d_tlb_refill_wr (2), 0x4d (2), l1d_tlb_refill_rd (2), l1d_cache_inval (2), l1d_cache_wb_clean (2), l1d_cache_wb_victim (2), 0x46 (2), l1d_cache_refill_outer (2), 0x45 (2), l1d_cache_refill_inner (2), l1d_cache_refill_wr (2), l1d_cache_refill_rd (2), l1d_cache_wr (2), stall_slot (2), stall_slot_frontend (2), 0x3e (2), stall_slot_backend (2), 0x3d (2), op_spec (2), 0x3b (2), op_retired (2), 0x3a (2), itlb_walk (2), 0x35 (2), dtlb_walk (2), 0x34 (2), remote_access (2), 0x31 (2), l2d_tlb (2), 0x2f (2), l2d_tlb_refill (2), l1i_tlb (2), l1d_tlb (2), br_mis_pred_retired (2), br_retired (2), l2d_cache_allocate (2), bus_cycles (2), 0x1d (2), ttbr_write_retired (2), bus_access (2), 0x19 (2), l2d_cache_wb (2), l2d_cache_refill (2), 0x17 (2), l2d_cache (2), 0x16 (2), l1d_cache_wb (2), 0x15 (2), l1i_cache (2), 0x14 (2), mem_access (2), cid_write_retired (2), 0xb (2), exc_return (2), 0xa (2), exc_taken (2), 0x9 (2), l1d_tlb_refill (2), 0x5 (2), l1i_tlb_refill (2), 0x2 (2), l1i_cache_refill (2), 0x1 (2), neoverse (2), c0_wait (2), pause (2), c02 (2), c01 (2), 256bit (2), 0xac (2), mul_256 (2), shuffles (2), vnni_256 (2), 128bit (2), vnni_128 (2), add_256 (2), add_128 (2), mem_uop_retired (2), 0xe5 (2), misc2_retired (2), lfence (2), 0xe0 (2), mem_load_misc_retired (2), 0xd4 (2), remote_fwd (2), remote_hitm (2), remote_dram (2), local_dram (2), xsnp_none (2), xsnp_fwd (2), xsnp_no_fwd (2), xsnp_miss (2), fb_hit (2), l3_miss (2), l2_miss (2), l1_miss (2), l3_hit (2), l1_hit (2), all_stores (2), all_loads (2), split_stores (2), split_loads (2), lock_loads (2), stlb_miss_stores (2), stlb_miss_loads (2), 512b_packed_half (2), 256b_packed_half (2), 128b_packed_half (2), complex_scalar_half (2), scalar_half (2), aborted_events (2), aborted_memtype (2), aborted_unfriendly (2), aborted_mem (2), aborted (2), commit (2), 512b_packed_single (2), 512b_packed_double (2), 256b_packed_single (2), 256b_packed_double (2), 128b_packed_single (2), 128b_packed_double (2), scalar_single (2), scalar_double (2), ret (2), indirect_call (2), far_branch (2), near_return (2), near_call (2), smc (2), memory_ordering (2), heavy (2), sse_avx_mix (2), page_fault (2), macro_fused (2), rep_iteration (2), nop (2), any_p (2), exe (2), amx_busy (2), 0xb7 (2), port_5 (2), port_7_8 (2), port_6 (2), port_5_11 (2), port_4_9 (2), port_2_3_10 (2), x87 (2), cycles_ge_1 (2), idiv_active (2), uops_issued (2), 0xae (2), clear_resteer_cycles (2), mba_stalls (2), uop_dropping (2), recovery_cycles (2), lsd (2), cycles_active (2), 0xa8 (2), exe_bound_0_ports (2), bound_on_stores (2), bound_on_loads (2), 4_ports_util (2), 3_ports_util (2), 2_ports_util (2), 1_ports_util (2), 0x07 (2), memory_bound_slots (2), br_mispredict_slots (2), bad_spec_slots (2), backend_bound_slots (2), slots_p (2), cycles_mem_any (2), 0x06 (2), stalls_total (2), cycles_l2_miss (2), scoreboard (2), idq_uops_not_delivered (2), 0x9c (2), lcp (2), icache_tag (2), icache_data (2), ms_cycles_any (2), dsb_cycles_any (2), mite_cycles_any (2), uops_decoded (2), dec0_uops (2), inst_decoded (2), decoders (2), dsb2mite_switches (2), penalty_cycles (2), abort_capacity_read (2), abort_capacity_write (2), abort_conflict (2), hwpf_miss (2), replacement (2), load_hit_prefetch (2), swpf (2), fb_full (2), pending (2), mem_store_retired (2), mem_load_completed (2), l1_miss_any (2), 0xfd (2), prefetchw (2), t1_t2 (2), nta (2), thread_p (2), full_cycles (2), non_silent (2), silent (2), l2_lines_in (2), 0x1f (2), references (2), 0xff (2), all_demand_references (2), all_code_rd (2), 0xe4 (2), all_rfo (2), 0xe2 (2), all_demand_data_rd (2), 0xe1 (2), swpf_hit (2), 0xc8 (2), code_rd_hit (2), rfo_hit (2), demand_data_rd_hit (2), swpf_miss (2), 0x28 (2), all_demand_miss (2), 0x27 (2), code_rd_miss (2), rfo_miss (2), demand_data_rd_miss (2), all_requests (2), data_rd (2), demand_data_rd (2), cycles_with_demand_rfo (2), no_sr (2), store_forward (2), address_alias (2), umask (2), 6th (2), amx_ops_retired (2), 0xce (2), 5th (2), xeon (2), including (2), isn (2), example (2), unhalted (2), evaluate (2), were (2), then (2), specify (2), low (2), after (2), learning (2), address (2), documentation (2), sdk (2), languages (2), frameworks (2), industry (2), solutions (2), hybrid (2), multicloud (2), pipelines (2), hosting (2), development (2), quota (2), commitment (2), consumption (2), full (2), registration (2), sles (2), export (2), port (2), collecting (2), nodes (2), report (2), api (2), rdp (2), general (2), app (2), latency (2), analyze (2), visible (2), automatically (2), committed (2), underutilized (2), insights (2), utilization (2), globally (2), web (2), autohealing (2), iis (2), external (2), shared (2), reserve (2), autoscaler (2), autoscalers (2), activity (2), provider (2), place (2), running (2), directory (2), setting (2), services (2), redis (2), cloning (2), writer (2), always (2), alwayson (2), file (2), volumes (2), authenticate (2), method (2), synchronization (2), global (2), switch (2), byol (2), payg (2), byos (2), families (2), guide (2), securing (2), simulate (2), topology (2), together (2), repairs (2), repair (2), check (2), failures (2), template (2), suspended (2), stopped (2), target (2), distribution (2), physical (2), cancel (2), once (2), limit (2), interfaces (2), ipv6 (2), static (2), decrease (2), basic (2), consistent (2), scoped (2), copy (2), appliances (2), asynchronous (2), encrypt (2), names (2), format (2), mount (2), memory (2), extreme (2), balanced (2), credentials (2), organization (2), user (2), through (2), advanced (2), rdma (2), created (2), preemptible (2), provisioning (2), models (2), rtx (2), vws (2), constraints (2), accelerator (2), apis (2), technology (2), areas (2), close (2), subscribe, newsletter, our, third, decade, climate, action, join, cookies, privacy, tech, twitter, blog, engage, training, certification, architecture, getting, github, status, release, notes, community, forums, contact, sales, marketplace, easy, easytounderstand, solved, problem, solvedmyproblem, otherup, hard, hardtounderstand, incorrect, sample, incorrectinformationorsamplecode, missing, missingtheinformationsamplesineed, otherdown, tell, except, otherwise, noted, page, java, registered, trademark, oracle, affiliates, developers, creative, commons, attribution, what, 144, 288, vcpus, there, associated, stlb_hit_stores, 0x0a, stlb_hit_loads, 0xfc, 8_flops, 4_flops, indirect_cost, near_taken_cost, cond_cost, cond_ntaken_cost, ret_cost, all_branches_cost, indirect_call_cost, cond_taken_cost, div_active, fpdiv_active, 2_3_ports_util, 0xc, empty, empty_resource, cycles_l3_miss, ms_busy, baclears, l2_stalls, 0xf, ref_tsc_p, useless_hwpf, all_hwpf, 0xf0, hit, 0xdf, 0x30, l2_trans, l2_wb, l3_miss_demand_data_rd, demand_rfo, demand_code_rd, cycles_with_l3_miss_demand_data_rd, cycles_with_data_rd, cycles_with_demand_code_rd, cycles_with_demand_data_rd, bf16, int8, divider_active, fp_divider_active, rs_empty, l2_stall, all_data_rd, emr, armv9, tables, outside, inside, fetching, operate, fixed, frequency, providing, stable, even, speed, preserve, energy, spent, task, mwait, hlt, thread, due, power, interrupts, overall, pipeline, used, simultaneously, execute, efficiently, top, successfully, processes, instruction, mispredicted, causing, discard, fetched, likely, misses, execution, potential, disabled, track, enabling, background, continuously, optionally, thresholds, preferred, pmc, exceeds, designated, threshold, notifies, composed, called, these, each, occurs, misprediction, read, vtune, profiler, model, registers, document, explains, behavior, billing, learn, unit, component, monitors, approach, lets, sensitive, computing, helping, save, categorize, preferences, stay, organized, collections, home, concurrent, operation, renewal, nvme, resizing, common, pay, viewing, output, diagnostic, sudoers, screenshots, generate, bug, blackwell, soft, lockups, bus, locks, rescue, inaccessible, dumps, kernel, panic, fstab, unresponsive, suspension, reboots, shutdowns, connectivity, tips, nics, dpdk, improve, resiliency, communication, reduce, compact, idpf, interface, irdma, driver, nic, customize, threads, per, sustained, merge, split, extend, renew, without, purchase, cost, savings, dynamic, overcommit, cpus, floating, reliable, udp, backends, routing, https, scaling, consuming, consume, combine, cud, allow, sharing, decisions, signals, serving, predictions, organize, labels, replica, states, observe, reports, tensorflow, inference, tensorrt5, monte, carlo, methods, spark, forwarding, others, integrate, clustering, exchange, 2016, sharepoint, strategies, passive, inactive, setups, built, integrations, validation, deployed, transition, test, hammerdb, bucket, aws, ec2, pacemaker, s2d, block, netapp, hot, standby, drbd, architectures, client, private, mailjet, mailgun, sendgrid, sending, blue, green, deployments, lamp, joomla, asp, net, interactive, mongodb, flask, terraform, servers, over, mtls, random, generator, virtio, rng, accurate, protocol, ntp, append, els, restrictions, packages, deprecate, versions, family, trusted, red, hat, knowledgebase, functionality, risks, command, vpc, controls, confidential, kek, secure, expiration, shielded, attributes, predefined, handle, notices, process, faulty, unmanaged, affect, preserved, turn, off, alternate, fails, repairing, maintain, click, upgrades, override, selectively, applying, accelerate, out, outage, rebalance, reenable, proactive, redistribution, shape, distribute, define, info, accidental, deletion, spread, edit, rename, reset, resume, restart, referrers, source, uuid, recover, corrupted, duplicate, clones, alerts, backups, vss, regionally, different, protection, considerations, synchronous, consistency, failback, replicate, make, provisioned, iops, limits, settings, help, kms, customer, supplied, symbolic, links, practice, device, ram, attach, additional, exapools, verify, identity, ptr, record, failure, rates, tags, restrict, automate, password, require, powershell, sac, securely, apps, root, vpn, bastion, iap, connection, browser, connections, networks, iso, imported, prerequisites, importing, exporting, bring, own, path, detach, reattach, constraint, accounting, provision, examples, slurm, reserved, team, slice, slices, adds, scenarios, deterministic, base, requirements, ready, aci, configured, preempted, historical, alternative, similar, displays, ops, logging, subnet, public, has, gpudirect, enabled, ultra, minimum, hostname, quickstarts, iam, roles, permissions, office, licensing, premium, strategy, gemini, bare, metal, accelerators, scores, purpose, discover, free, skip, main,
Text of the page (random words):
latency with load balancing use dpdk to improve network performance network performance and gpu vms networking and gpu machines use higher network bandwidth patterns for using multiple host nics troubleshoot general tips troubleshoot connectivity troubleshoot rdp troubleshoot ssh troubleshoot os login troubleshoot vms troubleshoot vm operations troubleshoot vm creation troubleshoot resource availability errors troubleshoot bulk api vm creation troubleshoot vm reboots and shutdowns troubleshoot vm suspension troubleshoot vm updates troubleshoot unresponsive vms troubleshoot vm startup troubleshoot fstab errors troubleshoot kernel panic collecting core dumps rescue an inaccessible vm troubleshoot cpu bus locks troubleshoot cpu soft lockups troubleshoot vm configurations troubleshoot arm vms troubleshoot gpu vms troubleshoot nvidia gpu errors generate a nvidia bug report for blackwell gpus troubleshoot nested virtualization troubleshoot using vm screenshots troubleshoot sole tenant nodes troubleshoot vm performance issues troubleshoot sudoers files troubleshoot windows vms troubleshoot windows vms collecting diagnostic information troubleshoot using the serial console troubleshoot using the serial console viewing serial port output troubleshoot instance groups troubleshoot managed instance groups migs troubleshoot os management troubleshoot licenses troubleshoot image import and export troubleshooting sles pay as you go registration troubleshooting ubuntu pro registration troubleshoot metadata server troubleshoot metadata server troubleshoot networking issues troubleshoot common networking issues troubleshoot network drivers troubleshoot vm performance issues troubleshoot storage troubleshoot disk creation troubleshoot full disks and disk resizing troubleshoot disk encryption troubleshoot nvme disks troubleshoot instant snapshots troubleshoot standard snapshots troubleshoot reservations and commitments troubleshoot reservation creation troubleshoot reservation consumption troubleshooting reservation monitoring troubleshoot reservation updates troubleshoot future reservation creation and updates troubleshoot automatic commitment renewal troubleshoot quota errors troubleshoot concurrent operation quota errors troubleshoot workload authentication troubleshoot default service accounts troubleshoot workload to workload authentication ai and ml application development application hosting compute data analytics and pipelines databases distributed hybrid and multicloud industry solutions migration networking observability and monitoring security storage access and resources management costs and usage management infrastructure as code sdk languages frameworks and tools home documentation compute compute engine guides send feedback pmu overview stay organized with collections save and categorize content based on your preferences the performance monitoring unit pmu is a hardware component within the cpu core that monitors how the processor runs code when you enable the pmu in an a4x c4a c4 or m4 compute engine instance you can access the performance counters in the pmu by using performance monitoring software this approach lets you optimize performance sensitive workloads such as high performance computing hpc or machine learning ml by helping you identify and address performance bottlenecks in your applications this document explains the behavior billing and limitations of the pmu to learn how to enable the pmu in an a4x c4a c4 or m4 instance see enable the pmu in compute engine instances understand the pmu the pmu is composed of a set of hardware counters called performance monitoring counters pmcs these counters are model specific registers that count each time a low level processor event occurs within the cpu such as a branch misprediction or cache miss you can read and configure pmcs in the pmu by using performance monitoring software such as intel vtune profiler after you enable the pmu in a compute instance the pmu runs in the background continuously monitoring performance events using pmcs you can optionally configure thresholds for specific pmcs using your preferred performance monitoring software if a pmc exceeds its designated threshold then the pmu notifies the software pmu events by default the pmu is disabled in compute instances to enable it specify the types of low level cpu events to track by enabling one of the following pmu types architectural architectural you can measure the following architectural performance events branch instructions retired the number of branch instructions retired use this event to measure your code s execution and identify potential performance bottlenecks branch misses retired the number of branch instructions that were mispredicted causing the processor to stall and discard fetched instructions if you see a high number for this event then you can likely optimize the cpu performance instructions retired the number of instructions the cpu successfully processes use this event to measure the cpu s instruction throughput top down slots the number of available slots within a processor s pipeline that are used to simultaneously execute instructions use this event to understand how efficiently your code is using the processor s resources unhalted core cycles the number of core cycles when the thread isn t halted for example due to power management or interrupts use this event to evaluate the overall usage of the processor unhalted reference cycles the number of reference cycles when the core isn t halted for example when fetching data or instructions the core is halted when it runs the hlt or mwait instructions reference cycles operate at a fixed frequency providing a stable time reference even when the speed of the processor changes to preserve energy use this event to measure the time spent on a task and identify performance bottlenecks in your code standard standard you can measure all events from the architectural pmu type and any local events inside the cpu core including level 2 l2 cache events enhanced enhanced you can measure all events from the standard pmu type and any local events outside the cpu core including level 3 l3 cache events supported machine series and cpu platforms you can enable the pmu on compute instances that use the following machine series and cpu platforms cpu platform machine series pmu tracking types google axion processor c4a architectural and standard intel xeon scalable processor emerald rapids 5th generation c4 and m4 architectural standard and enhanced intel xeon scalable processor granite rapids 6th generation c4 architectural standard and enhanced nvidia grace processor a4x architectural and standard supported pmu events by cpu platform the following tables list the supported pmu events by cpu platform google axion the following table lists the supported pmu events for google axion processors with neoverse v2 armv9 cores event code pmu tracking type event name 0x0 architectural sw_incr 0x3 architectural l1d_cache_refill 0x4 architectural l1d_cache 0x8 architectural inst_retired 0x10 architectural br_mis_pred 0x11 architectural cpu_cycles 0x12 architectural br_pred 0x1b architectural inst_spec 0x1e architectural chain 0x23 architectural stall_frontend 0x24 architectural stall_backend 0x39 architectural l1d_cache_lmiss_rd 0x3c architectural stall 0x40 architectural l1d_cache_rd 0x4006 architectural l1i_cache_lmiss 0x8006 architectural sve_inst_spec 0x0 standard sw_incr 0x1 standard l1i_cache_refill 0x2 standard l1i_tlb_refill 0x3 standard l1d_cache_refill 0x4 standard l1d_cache 0x5 standard l1d_tlb_refill 0x8 standard inst_retired 0x9 standard exc_taken 0xa standard exc_return 0xb standard cid_write_retired 0x10 standard br_mis_pred 0x11 standard cpu_cycles 0x12 standard br_pred 0x13 standard mem_access 0x14 standard l1i_cache 0x15 standard l1d_cache_wb 0x16 standard l2d_cache 0x17 standard l2d_cache_refill 0x18 standard l2d_cache_wb 0x19 standard bus_access 0x1b standard inst_spec 0x1c standard ttbr_write_retired 0x1d standard bus_cycles 0x1e standard chain 0x20 standard l2d_cache_allocate 0x21 standard br_retired 0x22 standard br_mis_pred_retired 0x23 standard stall_frontend 0x24 standard stall_backend 0x25 standard l1d_tlb 0x26 standard l1i_tlb 0x2d standard l2d_tlb_refill 0x2f standard l2d_tlb 0x31 standard remote_access 0x34 standard dtlb_walk 0x35 standard itlb_walk 0x39 standard l1d_cache_lmiss_rd 0x3a standard op_retired 0x3b standard op_spec 0x3c standard stall 0x3d standard stall_slot_backend 0x3e standard stall_slot_frontend 0x3f standard stall_slot 0x40 standard l1d_cache_rd 0x41 standard l1d_cache_wr 0x42 standard l1d_cache_refill_rd 0x43 standard l1d_cache_refill_wr 0x44 standard l1d_cache_refill_inner 0x45 standard l1d_cache_refill_outer 0x46 standard l1d_cache_wb_victim 0x47 standard l1d_cache_wb_clean 0x48 standard l1d_cache_inval 0x4c standard l1d_tlb_refill_rd 0x4d standard l1d_tlb_refill_wr 0x4e standard l1d_tlb_rd 0x4f standard l1d_tlb_wr 0x50 standard l2d_cache_rd 0x51 standard l2d_cache_wr 0x52 standard l2d_cache_refill_rd 0x53 standard l2d_cache_refill_wr 0x56 standard l2d_cache_wb_victim 0x57 standard l2d_cache_wb_clean 0x58 standard l2d_cache_inval 0x5c standard l2d_tlb_refill_rd 0x5d standard l2d_tlb_refill_wr 0x5e standard l2d_tlb_rd 0x5f standard l2d_tlb_wr 0x60 standard bus_access_rd 0x61 standard bus_access_wr 0x66 standard mem_access_rd 0x67 standard mem_access_wr 0x68 standard unaligned_ld_spec 0x69 standard unaligned_st_spec 0x6a standard unaligned_ldst_spec 0x6c standard ldrex_spec 0x6d standard strex_pass_spec 0x6e standard strex_fail_spec 0x6f standard strex_spec 0x70 standard ld_spec 0x71 standard st_spec 0x73 standard dp_spec 0x74 standard ase_spec 0x75 standard vfp_spec 0x76 standard pc_write_spec 0x77 standard crypto_spec 0x78 standard br_immed_spec 0x79 standard br_return_spec 0x7a standard br_indirect_spec 0x7c standard isb_spec 0x7d standard dsb_spec 0x7e standard dmb_spec 0x81 standard exc_undef 0x82 standard exc_svc 0x83 standard exc_pabort 0x84 standard exc_dabort 0x86 standard exc_irq 0x87 standard exc_fiq 0x88 standard exc_smc 0x8a standard exc_hvc 0x8b standard exc_trap_pabort 0x8c standard exc_trap_dabort 0x8d standard exc_trap_other 0x8e standard exc_trap_irq 0x8f standard exc_trap_fiq 0x90 standard rc_ld_spec 0x91 standard rc_st_spec 0x4004 standard cnt_cycles 0x4005 standard stall_backend_mem 0x4006 standard l1i_cache_lmiss 0x4009 standard l2d_cache_lmiss_rd 0x4020 standard ldst_align_lat 0x4021 standard ld_align_lat 0x4022 standard st_align_lat 0x8005 standard ase_inst_spec 0x8006 standard sve_inst_spec 0x8014 standard fp_hp_spec 0x8018 standard fp_sp_spec 0x801c standard fp_dp_spec 0x8074 standard sve_pred_spec 0x8075 standard sve_pred_empty_spec 0x8076 standard sve_pred_full_spec 0x8077 standard sve_pred_partial_spec 0x8079 standard sve_pred_not_full_spec 0x80bc standard sve_ldff_spec 0x80bd standard sve_ldff_fault_spec 0x80c0 standard fp_scale_ops_spec 0x80c1 standard fp_fixed_ops_spec 0x80e3 standard ase_sve_int8_spec 0x80e7 standard ase_sve_int16_spec 0x80eb standard ase_sve_int32_spec 0x80ef standard ase_sve_int64_spec intel emerald rapids the following table lists the supported pmu events for intel 5th generation emerald rapids emr processors event code umask pmu tracking type event name 0x03 0x04 standard ld_blocks address_alias 0x03 0x82 standard ld_blocks store_forward 0x03 0x88 standard ld_blocks no_sr 0x11 0x02 standard itlb_misses walk_completed_4k 0x11 0x04 standard itlb_misses walk_completed_2m_4m 0x11 0x0e standard itlb_misses walk_completed 0x11 0x10 standard itlb_misses walk_active 0x11 0x20 standard itlb_misses stlb_hit 0x12 0x02 standard dtlb_load_misses walk_completed_4k 0x12 0x04 standard dtlb_load_misses walk_completed_2m_4m 0x12 0x08 standard dtlb_load_misses walk_completed_1g 0x12 0x0e standard dtlb_load_misses walk_completed 0x12 0x10 standard dtlb_load_misses walk_active 0x12 0x20 standard dtlb_load_misses stlb_hit 0x13 0x02 standard dtlb_store_misses walk_completed_4k 0x13 0x04 standard dtlb_store_misses walk_completed_2m_4m 0x13 0x08 standard dtlb_store_misses walk_completed_1g 0x13 0x0e standard dtlb_store_misses walk_completed 0x13 0x10 standard dtlb_store_misses walk_active 0x13 0x20 standard dtlb_store_misses stlb_hit 0x20 0x04 enhanced offcore_requests_outstanding cycles_with_demand_rfo 0x20 0x08 enhanced offcore_requests_outstanding all_data_rd 0x21 0x01 enhanced offcore_requests demand_data_rd 0x21 0x08 enhanced offcore_requests data_rd 0x21 0x80 enhanced offcore_requests all_requests 0x24 0x21 standard l2_rqsts demand_data_rd_miss 0x24 0x22 standard l2_rqsts rfo_miss 0x24 0x24 standard l2_rqsts code_rd_miss 0x24 0x27 standard l2_rqsts all_demand_miss 0x24 0x28 standard l2_rqsts swpf_miss 0x24 0x3f standard l2_rqsts miss 0x24 0xc1 standard l2_rqsts demand_data_rd_hit 0x24 0xc2 standard l2_rqsts rfo_hit 0x24 0xc4 standard l2_rqsts code_rd_hit 0x24 0xc8 standard l2_rqsts swpf_hit 0x24 0xe1 standard l2_rqsts all_demand_data_rd 0x24 0xe2 standard l2_rqsts all_rfo 0x24 0xe4 standard l2_rqsts all_code_rd 0x24 0xe7 standard l2_rqsts all_demand_references 0x24 0xff standard l2_rqsts references 0x25 0x1f standard l2_lines_in all 0x26 0x01 standard l2_lines_out silent 0x26 0x02 standard l2_lines_out non_silent 0x2d 0x01 enhanced xq full_cycles 0x2e 0x41 enhanced longest_lat_cache miss 0x3c 0x00 architectural cpu_clk_unhalted thread_p 0x40 0x01 standard sw_prefetch_access nta 0x40 0x02 standard sw_prefetch_access t0 0x40 0x04 standard sw_prefetch_access t1_t2 0x40 0x08 standard sw_prefetch_access prefetchw 0x43 0xfd standard mem_load_completed l1_miss_any 0x44 0x01 standard mem_store_retired l2_hit 0x47 0x02 standard memory_activity cycles_l1d_miss 0x47 0x03 standard memory_activity stalls_l1d_miss 0x47 0x05 standard memory_activity stalls_l2_miss 0x47 0x09 enhanced memory_activity stalls_l3_miss 0x48 0x01 standard l1d_pend_miss pending 0x48 0x02 standard l1d_pend_miss fb_full 0x48 0x04 standard l1d_pend_miss l2_stall 0x4c 0x01 standard load_hit_prefetch swpf 0x51 0x01 standard l1d replacement 0x54 0x01 enhanced tx_mem abort_conflict 0x54 0x02 enhanced tx_mem abort_capacity_write 0x54 0x80 enhanced tx_mem abort_capacity_read 0x61 0x02 standard dsb2mite_switches penalty_cycles 0x75 0x01 standard inst_decoded decoders 0x76 0x01 standard uops_decoded dec0_uops 0x79 0x04 standard idq mite_cycles_any 0x79 0x08 standard idq dsb_cycles_any 0x79 0x20 standard idq ms_cycles_any 0x80 0x04 standard icache_data stalls 0x83 0x04 standard icache_tag stalls 0x87 0x01 standard decode lcp 0x9c 0x01 standard idq_uops_not_delivered core 0xa2 0x02 standard resource_stalls scoreboard 0xa2 0x08 standard resource_stalls sb 0xa3 0x01 standard cycle_activity cycles_l2_miss 0xa3 0x04 standard cycle_activity stalls_total 0xa3 0x05 standard cycle_activity stalls_l2_miss 0xa3 0x06 enhanced cycle_activity stalls_l3_miss 0xa3 0x08 standard cycle_activity cycles_l1d_miss 0xa3 0x0c standard cycle_activit...
|
