Coalesced TLB to Exploit Diverse Contiguity of Memory Mapping

Jiangfang Yi; Xu Cheng; Yikun Ban; Yuchen Zhou

REVIEW 4 major objections 6 minor 42 references

Coalesced TLB to Exploit Diverse Contiguity of Memory Mapping

T0 review · 4 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that K-bit Aligned TLB, which stores multiple sizes of contiguous page chunks in aligned page-table entries, reduces TLB misses by at least 27% over the prior state of the art on mixed-contiguity workloads.

desk verdict A solid, incremental extension of Anchor that deserves peer review, but the permission-homogeneity assumption is unmeasured and the single-trace evaluation overstates numeric confidence. read the letter →

arxiv 1908.08774 v2 pith:CD6WXQUE submitted 2019-08-22 cs.DC

Yikun Ban , Yuchen Zhou , Xu Cheng , Jiangfang Yi This is my paper · ORCID

classification cs.DC

keywords virtualmemoryaddresstranslationTLBcoalescingmixedcontiguityK-bitalignedentriespagetablecoverage

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Real applications often map memory in contiguous chunks of many different sizes at once, a situation the paper calls mixed contiguity. The paper argues that prior TLB coalescing schemes—huge pages for large chunks, hardware coalescing for small chunks, and the hybrid Anchor design for a single best chunk size—each leave coverage on the table when chunk sizes vary. To fix this, it proposes the K-bit Aligned TLB: page-table entries placed at several power-of-two VPN alignments record how many following pages are contiguous, and the OS fills the L2 TLB with the largest matching aligned entry. The paper reports that this design reduces TLB misses by at least 27% on average over Anchor, and by 69.2% over an unmodified TLB, across 16 benchmarks, with a small predictor keeping about 93% of aligned hits to a single L2 lookup.

What carries the argument

The load-bearing object is the k-bit aligned page-table entry: an entry whose VPN has its low k bits zero, which stores the contiguity of the following 2^k pages. The Rightward Compatible Rule makes an entry that satisfies several alignments act as the largest one, so a faulting VPN can always be checked against alignments in descending k and the first match gives maximal coverage. Algorithm 3 selects K by weighting each possible alignment by the number of contiguous pages in chunks of its size range, stopping when the chosen types cover 90% of contiguous pages or |K| reaches 4. These pieces do the work: the aligned entries carry the coverage, the compatible rule makes lookup ordering safe, the greedy selection keeps the TLB modifications small, and the 4-bit predictor hides the lookup cost.

What would settle it

Count, on the paper's 16 real-machine traces, how often a page inside a recorded contiguity chunk has r/w/x permission bits different from the chunk's first page; if that fraction is nontrivial, the contiguity values stored in aligned PTEs overstate coalescible pages and the reported ≥27% miss reduction over Anchor would not reproduce. A cheaper proxy: run the same simulator with permission heterogeneity injected at increasing rates and find the rate at which K Aligned loses its advantage over Anchor.

Watch

Extended reading notes

Core claim

The central claim is that a TLB can profitably exploit several contiguity granularities at the same time. K-bit Aligned TLB places, for each k in a set K, aligned page-table entries at VPNs whose low k bits are zero; each such entry records the number of contiguous pages within the next 2^k pages. On a page-table walk the OS inserts the aligned entry with the largest contiguity that covers the faulting VPN, so one L2 TLB entry translates a whole chunk. Lookup tries alignment types in descending k, with a 4-bit predictor that guesses the alignment and completes about 94% (|K|=2) to 93% (|K|=4) of aligned hits in one lookup. The K itself is chosen by a greedy algorithm from the OS's contiguity histogram, keeping the types that cover 90% of contiguous pages with at most 4 alignments. Evaluated on 10 billion instructions from standard workloads, the paper reports that |K|=2 reduces TLB misses by 27% over Anchor-Static on the real demand mapping, and larger K reduce misses further; on synthetic mixed-contiguity mappings the reduction over Anchor reaches 58%.

Load-bearing premise

The stated gains assume that pages inside a contiguity chunk almost always share the same read/write/execute permissions; the paper adopts this from a prior study instead of measuring it on its own traces, so if permission differences are frequent the real coalescing coverage shrinks.

Editorial extensions

If this is right

On the real demand mapping, |K|=2 Aligned reduces TLB misses by 27% relative to Anchor-Static and 69.2% relative to the base TLB; |K|=3 and |K|=4 reduce misses further, to 48% and 55% below Anchor respectively.
On the synthetic mixed-contiguity mapping, where chunk sizes span 0.4 small, 0.4 medium, and 0.2 large, |K|=2 Aligned reduces misses by 58% relative to Anchor (25% vs 60.5% of base) and |K|=4 brings it to 5.6% of base.
Aligned lookup stays cheap enough to be practical: 94.3% of aligned hits finish in one lookup at |K|=2 and 93.1% at |K|=4, so the extra coverage does not add a multi-lookup penalty for most translations.
The chosen K adapts to the contiguity histogram of each process, and updating K every five billion instructions is sufficient because contiguity distributions stabilize after initial memory allocation.
Coverage per TLB entry grows with |K| (e.g., mcf reaches 34.2x base coverage at |K|=2 vs 23.44x for Anchor), which is the mechanism behind the miss reduction.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The same 'match container size to chunk-size distribution' principle could be applied to nested page tables in virtualized systems or to GPU address translation, where contiguity is also fragmented; the paper does not evaluate those settings.
Because the predictor relies on spatial locality of consecutive requests within a coalesced range, workloads whose accesses alternate between widely separated chunks could see lower than the reported 93-94% accuracy; the per-benchmark range (83-98%) already points to omnetpp, sjeng, and xalancbmk as stress cases.
The contiguity histogram that Algorithm 3 reads could also guide the OS memory allocator: instead of just adapting the TLB to whatever contiguity exists, the OS could preferentially create chunks of the sizes K already covers, making coalescing and allocation cooperate.
If permission heterogeneity within contiguity chunks is higher than the prior observation [5] suggests, the effective coalescing coverage will be lower than the stored contiguity counts imply; measuring permission uniformity on the paper's own 16 traces would bound the true gain.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Best quick take: the paper introduces a real, incremental idea—multiple alignment granularities in a coalesced TLB—and it doesn't oversell it. The mixed-contiguity observation is well-supported by their pagemap data, and the K-bit aligned PTE design with the rightward-compatible rule is a clean way to let one entry cover multiple chunk sizes. Credit where due: the authors cite Anchor, COLT, Cluster, and RMM properly, explain why one anchor distance is limiting, and add a predictor that looks sensible for hiding aligned-lookup latency. The simulated relative misses are internally consistent across benchmarks, and the improvement grows monotonically with |K|, which is the sign of a mechanism that isn't just noise.

Now the soft spots, in order of severity. First, permission homogeneity: Section 3.4 admits that differing r/w/x bits impede coalescing and leans on reference [5] for the claim that permissions are usually homogeneous. The paper never measures this on its own traces. Because pagemap doesn't expose permission bits, the contiguity counts in Table 4 are upper bounds on coalescable coverage. If heterogeneous permissions are common in the 16 workloads, the 27% average miss reduction over Anchor shrinks. This isn't fatal—it's a testable assumption—but it deserves measurement, especially since it is load-bearing.

Second, the evaluation is one trace per benchmark, no error bars, and no released simulator or traces. That makes the precise numbers (30.8% vs 42%) look sharper than the methodology supports. The synthetic mixed mapping (0.4 small + 0.4 medium + 0.2 large) is also hand-picked to disadvantage Anchor, which biases the comparison in favor of the new scheme. Third, there's a numeric inconsistency: the abstract says 'at least 27%' average miss reduction over Anchor, while the introduction claims 36%; Table 4 gives 27% for |K|=2 and better for larger K. That should be reconciled. Minor: the theta and psi thresholds in Algorithm 3 are arbitrary, and the predictor accuracy is reported for a single design with no sensitivity analysis.

Bottom line: the mechanism is sound, the paper is honest about its own limitations, and the main weakness is evidentiary, not logical. Who is this for? Researchers working on TLB coalescing or OS memory mapping—they'll want to know this and may build on it. It deserves a serious referee: send it out, but require artifact release, permission-homogeneity analysis on the actual workloads, multi-trace runs, and a fix to the inconsistent headline numbers.

Referee Report

4 major / 6 minor

Summary. The paper observes that real process memory mappings often contain a mix of contiguity chunk sizes and argues that prior TLB-coalescing schemes, including Anchor, are tuned to a single size regime and therefore leave translation coverage on the table. It proposes K-bit Aligned TLB, in which the OS maintains aligned page-table entries for several alignment widths K; an aligned entry records how many following pages are contiguously mapped, and after a page-table walk the OS fills L2 TLB with the best-fitting aligned entry. A per-process heuristic (Algorithm 3) selects K from the observed contiguity histogram, and a small predictor is added to finish most aligned lookups in one TLB access. Using Pin traces and pagemap-derived mappings for 16 benchmarks, the paper reports that |K|=2 Aligned reduces average relative TLB misses from 42% (Anchor-Static) to 30.8%, a 27% reduction, with larger |K| giving further reductions, and that the predictor keeps about 93% of aligned hits within one lookup.

Significance. If the results hold, the contribution is solid and useful. Mixed contiguity is a real obstacle for single-distance coalescing schemes, and the proposed mechanism of multiple aligned entry types is a natural extension of Anchor with modest OS and TLB changes. The paper earns credit for using real-machine mappings, comparing against six prior approaches, reporting coverage and CPI effects, and giving a concrete algorithm for choosing K. However, the claimed 27–69% reductions rest on an unmeasured permission-homogeneity assumption and on a single-trace, single-mapping evaluation without variance estimates, so the quantitative headline should be treated with caution until those gaps are addressed.

major comments (4)

[§3.4, Permission and Page Sharing] The load-bearing assumption that permissions are homogeneous inside contiguity chunks is asserted but not measured. The text explicitly says that pages with different r/w/x permissions "impede coalescing contiguous pages" and then relies entirely on reference [5] for the claim that permissions are commonly homogeneous. The paper's own traces are captured via pagemap (§4.1), which exposes physical frame numbers but not PTE permission bits. Therefore the contiguity values used in Tables 4 and 5 are upper bounds on coalescable coverage; if permission heterogeneity is non-negligible in the 16 workloads, aligned entries must be truncated at permission boundaries or inserted with incorrect permission attributes, and the claimed 27% average miss reduction over Anchor-Static would shrink. Please measure PTE permission homogeneity on the same traces or provide a sensitivity analysis that bounds the effect.
[§4.1–4.2, Table 4] The headline quantitative result is based on one Pin trace and one captured virtual-physical mapping per benchmark, with no error bars, no multiple runs, and no confidence intervals. The abstract's "at least 27%" is a single observed average, and Table 6 shows substantial per-benchmark variation (e.g., 83.2% for omnetpp versus 98.2% for bwaves with |K|=2). In addition, Algorithm 3 selects K from the same mapping that is then replayed in the miss simulation, so the reported K-Aligned misses incorporate knowledge of the final mapping; Anchor-Static is also given oracle status, but the comparison still does not show how K Aligned behaves on unseen mappings or under allocation/deallocation dynamics. Please add trace-level detail, multiple runs, or a sensitivity analysis before taking the 27–69% reductions at face value.
[§2.2 / §4.1, Table 3] The mixed synthetic mapping (0.4 small + 0.4 medium + 0.2 large) is constructed to disadvantage single-type schemes, not derived from the measured contiguity histograms in Figures 2 and 3. As a stress test this is legitimate, but the paper uses it to advertise a 58% reduction over Anchor for mixed contiguity in Table 4. That number is not a measured property of the real workloads, and it should be clearly labeled as a synthetic stress-test result, with the real-mapping results reported per benchmark rather than only as an aggregate.
[§3.2, Algorithm 2] Algorithm 2 is titled "L2 TLB Aligned Look-up," but line 3 reads "Entry←PageTable(VPNk)." If taken literally, an aligned hit always requires a page-table access, which would defeat the purpose of TLB coalescing and contradict the flow described in Figure 6 and Section 3.2. Please correct the pseudocode to specify a lookup in the L2 TLB (or its aligned-entry storage) and clarify whether aligned entries live in the same set-associative TLB as regular entries.

minor comments (6)

[Abstract and Introduction] There are typos in the abstract: "at lease 27%" should be "at least 27%," and "an primary approach" should be "a primary approach."
[Table 6] The three column headers in Table 6 all read "|K| = 2 Aligned"; they should presumably be |K| = 2, |K| = 3, and |K| = 4. The caption also says "Table 6: The accuracy" followed by a duplicated table caption line in the text.
[§2.2, Figures 2 and 3] Figures 2 and 3 show 15 benchmarks, while the abstract and Section 4 say 16 benchmarks; the paper should clarify how graph500 and gups are counted in the contiguity analysis.
[§4.1] The y-axis labels in Figures 2 and 3 and the accompanying text ("displayed by loдn+1;2") are garbled; the log scale and axis units should be stated precisely.
[§3.5] The "Future Work" section appears before the evaluation and discusses parallel page-table-walk speculation that is not evaluated; moving this to the conclusion or related-work discussion would improve the paper's organization.
[References] The reference list includes a duplicated entry for Navarro et al. ([25] and [26]) and has inconsistent formatting of conference names and volume numbers; the citations for [5] and [30] should be checked to ensure they support the specific claims about permission homogeneity and memory-mapping stability.

Circularity Check

1 steps flagged · score 4.0 of 10

K is selected from the same pagemap-derived mappings used to report TLB misses, and the coverage metric that explains the win is the same quantity Algorithm 3 maximizes; the headline 27% reduction is therefore partly an in-sample fit. Permission homogeneity is asserted via an external citation, not measured on the evaluated traces.

fitted input called prediction [Sec. 3.3 (Algorithm 3) and Sec. 4.2 (Coverage Improvement; Effectiveness of K)]
"Once the initial memory allocation phase is stabilized, the status of memory mapping is used to ensure K. ... the TLB coverage is the number of inserted entries plus the sum of contiguity values of every coalesced entries. ... the coverage of TLB must be larger in face of the same mapping with the increasing of |K|, on account of Algorithm 1 that chooses the optimal aligned entries for coalescing."

Algorithm 3 chooses K by maximizing the number of contiguous pages covered by aligned entries, using the contiguity histogram of the same pagemap-derived memory mapping on which the simulator later reports TLB misses. The Coverage Improvement section then defines TLB coverage as the number of inserted entries plus the sum of the very contiguity values that Algorithm 3 maximizes. With more alignment types, Algorithm 1 always fills the aligned entry with maximal contiguity, so coverage mechanically increases with |K|. The paper uses that mechanically increasing coverage to explain the monotonic TLB-miss reduction and the headline improvement over Anchor.

full rationale

The core mechanism is internally coherent and not definitionally equivalent to the reported TLB-miss rates: misses come from a trace-driven simulator using Pin-generated access streams, which is an input distinct from the contiguity histogram. The K-bit aligned PTE design, fill rule, and aligned-lookup arithmetic are self-consistent. However, the evaluation is partly self-referential: Algorithm 3 selects K from the same mapping used for the reported misses, and the paper's coverage metric is the sum of contiguity values that Algorithm 3 is designed to maximize. Larger |K| therefore mechanically raises the coverage metric, and the paper openly says coverage 'must be larger' with growing |K| and uses that coverage to explain the miss reduction. This makes the central performance comparison in-sample rather than an independent confirmation. The permission-homogeneity premise in Sec. 3.4 is load-bearing but rests on reference [5] and is not measured on the evaluated traces; that is a missing-support concern, not a circular derivation. There is no load-bearing self-citation: the cited prior works, including Anchor [30] and direct segments [5], are by different authors and provide independent context. Overall, the design has independent technical content, but the headline miss-reduction claim is partly built into the parameter-fit and coverage-accounting loop, so a moderate circularity score is warranted.

Assumptions & free parameters 6 free parameters · 4 assumptions · 2 invented entities

The central claim rests on the existence of exploitable contiguity in OS-managed memory mappings and on the assumption that permission and sharing effects do not undermine it. Several quantitative choices (theta, psi, Table 1 mapping, synthetic mapping weights) are hand-selected and directly influence the reported coverage and miss rates. No invented entity has independent evidence outside the paper.

free parameters (6)

K, the set of alignment types = Varies per benchmark; experiments use |K| = 2, 3, 4
Chosen by Algorithm 3 from the contiguity histogram of each workload; the value is tuned to the evaluated mappings rather than fixed by an external standard.
theta = 0.9
Stops adding alignment types once 90% of contiguous pages are covered by the chosen K. Hand-set in evaluation.
psi = 4
Upper bound on the number of alignment types K; hand-set in evaluation.
Table 1 size-to-alignment mapping = e.g., size 2-16 to 4-bit; 17-64 to 6-bit; >1024 to 11-bit
Heuristic mapping that defines which alignment is 'matching' for a contiguity chunk before K is selected. It directly shapes the computed coverage.
Mixed synthetic mapping weights = 0.4 small + 0.4 medium + 0.2 large
The synthetic mixed mapping is constructed with these weights to match the paper's description of mixed contiguity, and it is used to show the largest advantage over Anchor.
Predictor size = 4 bits
Records the most recent alignment type; chosen under the assumption that |K| is at most 8. Affects the latency overhead model.

assumptions (4)

domain assumption The buddy allocator produces contiguity chunks that are visible in the page table and stable after initial memory allocation.
Sections 2 and 3.3 rely on this to justify computing contiguity once and then updating K only every five billion instructions.
domain assumption Permissions and page sharing do not significantly reduce coalescing opportunities.
Section 3.4 acknowledges that permission differences impede coalescing but cites reference [5] for homogeneity instead of measuring it on the evaluated workloads.
domain assumption Pin traces with fixed TLB latencies adequately represent real CPU behavior.
Section 4.1 uses a trace-based simulator with a fixed 50-cycle page-table walk and fixed L1/L2 latencies; no validation against real hardware is reported.
domain assumption Anchor-Static, which exhaustively searches anchor distances, is a fair state-of-the-art baseline.
Section 4.1 compares against this optimistic Anchor variant; if Anchor-Static is stronger than any practical online Anchor, the reported improvement over the true state of the art may be smaller.

invented entities (2)

K-bit aligned PTE
purpose: Store contiguity counts at multiple alignment granularities in unused page-table bits, enabling variable-size coalesced TLB entries.
This is the core proposed structure. Its only evidence is the in-paper simulation; no external implementation or independent measurement is provided.
Aligned-lookup predictor
purpose: Predict the alignment type of a requested VPN so that an aligned hit finishes in one TLB lookup.
A 4-bit hardware predictor added to L2 TLB. Its accuracy is measured only on the paper's own traces.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Coalesced TLB to Exploit Diverse Contiguity of Memory Mapping." pith.science (2026). https://pith.science/paper/CD6WXQUE

@misc{pith2026190808774,
  author       = {Pith},
  title        = {Pith review of: Coalesced TLB to Exploit Diverse Contiguity of Memory Mapping},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CD6WXQUE}},
  note         = {Machine review of arXiv:1908.08774}
}

read the original abstract

The miss rate of TLB is crucial to the performance of address translation for virtual memory. To reduce the TLB misses, improving translation coverage of TLB has been an primary approach. Many previous works focus on coalescing multiple contiguously mapped pages of the memory mapping into a modified entry, which function well if the assumed contiguity of memory mapping is given. Unfortunately, scenarios of applications are complicated and the produced contiguity diversify. To gain better performance of translation, in this paper, we first introduce a complex but prevalent type of contiguity, mixed contiguity. Then we propose a HW-SW hybrid coalesced TLB structure which works well on all observed types of contiguity including this type. In our evaluation, the proposed scheme, K-bit Aligned TLB, outperforms the state-of-the-art work by reducing at lease 27% TLB misses on average over it using 16 benchmarks.

Figures

Figures reproduced from arXiv: 1908.08774 by the authors.

**Figure 2.** THP[13] off. The distribution of contiguity chunks at first billionth instruction boundary for the used benchmarks. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** THP[13] on. The distribution of contiguity chunks at first billionth instruction boundary for the used benchmarks. [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Figures from the paper (7 more)

**Figure 4.** Figure 4: Page Table with K-bit Alignments table, every N (anchor distance) entries is placed an anchored entry which records the number of following contiguous pages ending to next anchored entry. For example, if memory pages are allocated in contiguity chunk of size 16, the op…

**Figure 6.** Figure 6: L2 TLB lookup: Regular lookup and Aligned [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]

**Figure 7.** Figure 7: Modified entry and index scheme. Based on this observation, we add a 4-bit predictor to the L2 TLB (assume the upper bound of |K| is 8), recording the latest used alignments. In the aligned look-up, the alignment kept by the predictor will be used to look up the align…

**Figure 8.** Figure 8: Relative misses of all compared approaches for [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]

**Figure 9.** Figure 9: Relative misses of varying |K| for demand mapping [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]

**Figure 10.** Figure 10: CPI breakdown of translation overhead for [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]

**Figure 11.** Figure 11: CPI breakdown of translation overhead for [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 42 canonical work pages

[5]

Efficient virtual memory for big memory servers

Arkaprava Basu, Jayneel Gandhi, Jichuan Chang, Mark D Hill, and Michael M Swift. Efficient virtual memory for big memory servers. In ACM SIGARCH Computer Architecture News, volume 41, pages 237–248. ACM, 2013

work page 2013
[1]

AMD Corporation. 2014. Compute cores. https://www.amd.com/Documents/ Compute_Cores_Whitepaper.pdf, 2014

work page 2014
[2]

Biobench: A benchmark suite of bioinfor- matics applications

Kursad Albayraktaroglu, Aamer Jaleel, Xue Wu, Manoj Franklin, Bruce Jacob, Chau-Wen Tseng, and Donald Yeung. Biobench: A benchmark suite of bioinfor- matics applications. In IEEE International Symposium on Performance Analysis of Systems and Software, 2005. ISPASS 2005. , pages 2–9. IEEE, 2005

work page 2005
[3]

Translation caching: skip, don’t walk (the page table)

Thomas W Barr, Alan L Cox, and Scott Rixner. Translation caching: skip, don’t walk (the page table). In ACM SIGARCH Computer Architecture News, volume 38, pages 48–59. ACM, 2010

work page 2010
[4]

Spectlb: a mechanism for spec- ulative address translation

Thomas W Barr, Alan L Cox, and Scott Rixner. Spectlb: a mechanism for spec- ulative address translation. In ACM SIGARCH Computer Architecture News , volume 39, pages 307–318. ACM, 2011

work page 2011
[6]

Reducing memory refer- ence energy with opportunistic virtual caching

Arkaprava Basu, Mark D Hill, and Michael M Swift. Reducing memory refer- ence energy with opportunistic virtual caching. In ACM SIGARCH Computer Architecture News, volume 40, pages 297–308. IEEE Computer Society, 2012

work page 2012
[7]

Large-reach memory management unit caches

Abhishek Bhattacharjee. Large-reach memory management unit caches. In Proceedings of the 46th Annual IEEE/ACM International Symposium on Microarchi- tecture, pages 383–394. ACM, 2013

work page 2013
[8]

Shared last- level tlbs for chip multiprocessors

Abhishek Bhattacharjee, Daniel Lustig, and Margaret Martonosi. Shared last- level tlbs for chip multiprocessors. In 2011 IEEE 17th International Symposium on High Performance Computer Architecture, pages 62–63. IEEE, 2011

work page 2011

Show all 42 references

[9]

Benchmarking modern multiprocessors

Christian Bienia and Kai Li. Benchmarking modern multiprocessors. Princeton University Princeton, 2011

2011
[10]

IntelÂő 64 and ia-32 architectures optimization reference manual, 2016

Intel Corporation. IntelÂő 64 and ia-32 architectures optimization reference manual, 2016

2016
[11]

Efficient address translation for architectures with multiple page sizes

Guilherme Cox and Abhishek Bhattacharjee. Efficient address translation for architectures with multiple page sizes. architectural support for programming languages and operating systems , 51(2):435–448, 2017

2017
[12]

pagemap, from the userspace perspective

Linux Kernel Documentation. pagemap, from the userspace perspective. https: //www.kernel.org/doc/Documentation/vm/pagemap.txt, 2016

2016
[13]

Transparent hugepage support

Linux Kernel Documentation. Transparent hugepage support. https://www. kernel.org/doc/Documentation/vm/transhuge.txt, 2017

2017
[14]

Supporting superpages in non-contiguous physical memory

Yu Du, Miao Zhou, Bruce R Childers, Daniel Mossé, and Rami Melhem. Supporting superpages in non-contiguous physical memory. In 2015 IEEE 21st International Symposium on High Performance Computer Architecture (HPCA) , pages 223–234. IEEE, 2015

2015
[15]

Data tiering in heterogeneous memory systems

Subramanya R Dulloor, Amitabha Roy, Zheguang Zhao, Narayanan Sundaram, Nadathur Satish, Rajesh Sankaran, Jeff Jackson, and Karsten Schwan. Data tiering in heterogeneous memory systems. In Proceedings of the Eleventh European Conference on Computer Systems , page 15. ACM, 2016

2016
[16]

Devirtualizing memory in heterogeneous systems

Swapnil Haria, Mark D Hill, and Michael M Swift. Devirtualizing memory in heterogeneous systems. In In Proceedings of 2018 Architectural Support for Programming Languages and Oper- ating Systems (ASPLOSâĂŹ18) , volume 53, pages 637–650. ACM, 2018

2018
[17]

Spec cpu2006 benchmark descriptions.ACM SIGARCH Computer Architecture News, 34(4):1–17, 2006

John L Henning. Spec cpu2006 benchmark descriptions.ACM SIGARCH Computer Architecture News, 34(4):1–17, 2006

2006
[18]

Kandiraju and Anand Sivasubramaniam

Gokul B. Kandiraju and Anand Sivasubramaniam. Going the distance for tlb prefetching: An application-driven study. In International Symposium on Com- puter Architecture, 2002

2002
[19]

pvm: persistent virtual memory for efficient capacity scaling and object storage

Sudarsun Kannan, Ada Gavrilovska, and Karsten Schwan. pvm: persistent virtual memory for efficient capacity scaling and object storage. In Proceedings of the Eleventh European Conference on Computer Systems , page 13. ACM, 2016

2016
[20]

Redundant memory mappings for fast access to large memories

Vasileios Karakostas, Jayneel Gandhi, Furkan Ayar, Adrián Cristal, Mark D Hill, Kathryn S McKinley, Mario Nemirovsky, Michael M Swift, and Osman Ünsal. Redundant memory mappings for fast access to large memories. InACM SIGARCH Computer Architecture News, volume 43, pages 66–78...

2015
[21]

Performance analysis of the memory management unit under scale-out workloads

Vasileios Karakostas, Osman S Unsal, Mario Nemirovsky, Adrian Cristal, and Michael Swift. Performance analysis of the memory management unit under scale-out workloads. In 2014 IEEE International Symposium on Workload Charac- terization (IISWC), pages 1–12. IEEE, 2014

2014
[22]

Leveraging sharing in second level translation-lookaside buffers for chip multiprocessors

Yang Li, Rami Melhem, and Alex K Jones. Leveraging sharing in second level translation-lookaside buffers for chip multiprocessors. IEEE Computer Architec- ture Letters, 11(2):49–52, 2011

2011
[23]

3d-stacked memory architectures for multi-core processors

Gabriel H Loh. 3d-stacked memory architectures for multi-core processors. In ACM SIGARCH computer architecture news , volume 36, pages 453–464. IEEE Computer Society, 2008

2008
[24]

Pin: building customized program analysis tools with dynamic instrumentation

Chi-Keung Luk, Robert Cohn, Robert Muth, Harish Patil, Artur Klauser, Geoff Lowney, Steven Wallace, Vijay Janapa Reddi, and Kim Hazelwood. Pin: building customized program analysis tools with dynamic instrumentation. InProceedings of the 2005 ACM SIGPLAN Conference on Programm...

2005
[25]

Juan Navarro, Sitararn Iyer, Peter Druschel, and Alan Cox. Practical, transparent operating system support for superpages.ACM SIGOPS Operating Systems Review, Conference’17, July 2017, Washington, DC, USA Yikun Ban, Yuchen Zhou, Xu Cheng, and Jiangfang Yi 36(SI):89–104, 2002

2017
[26]

Practical, transparent operating system support for superpages.ACM SIGOPS Operating Systems Review, 36(SI):89–104, 2002

Juan Navarro, Sitararn Iyer, Peter Druschel, and Alan Cox. Practical, transparent operating system support for superpages.ACM SIGOPS Operating Systems Review, 36(SI):89–104, 2002

2002
[27]

Border control: Sandboxing accelerators

Lena E Olson, Jason Power, Mark D Hill, and David A Wood. Border control: Sandboxing accelerators. In 2015 48th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO), pages 470–481. IEEE, 2015

2015
[28]

Prediction-based superpage-friendly tlb designs

Misel-Myrto Papadopoulou, Xin Tong, André Seznec, and Andreas Moshovos. Prediction-based superpage-friendly tlb designs. In 2015 IEEE 21st International Symposium on High Performance Computer Architecture (HPCA) , pages 210–222. IEEE, 2015

2015
[29]

Efficient synonym fil- tering and scalable delayed translation for hybrid virtual caching

Caching Hyun Park, Taekyung Heo, and Jaehyuk Huh. Efficient synonym fil- tering and scalable delayed translation for hybrid virtual caching. international symposium on computer architecture , 44(3):217–229, 2016

2016
[30]

Hybrid tlb coalescing: Improving tlb translation coverage under diverse fragmented memory allocations

Chang Hyun Park, Taekyung Heo, Jungi Jeong, and Jaehyuk Huh. Hybrid tlb coalescing: Improving tlb translation coverage under diverse fragmented memory allocations. In 2017 ACM/IEEE 44th Annual International Symposium on Computer Architecture (ISCA), pages 444–456. IEEE, 2017

2017
[31]

Hybrid memory cube (hmc)

J Thomas Pawlowski. Hybrid memory cube (hmc). In 2011 IEEE Hot Chips 23 Symposium (HCS), pages 1–24. IEEE, 2011

2011
[32]

Increas- ing tlb reach by exploiting clustering in page translations

Binh Pham, Abhishek Bhattacharjee, Yasuko Eckert, and Gabriel H Loh. Increas- ing tlb reach by exploiting clustering in page translations. In 2014 IEEE 20th International Symposium on High Performance Computer Architecture (HPCA) , pages 558–567. IEEE, 2014

2014
[33]

Colt: Coalesced large-reach tlbs

Binh Pham, Viswanathan Vaidyanathan, Aamer Jaleel, and Abhishek Bhattachar- jee. Colt: Coalesced large-reach tlbs. In Proceedings of the 2012 45th Annual IEEE/ACM International Symposium on Microarchitecture , pages 258–269. IEEE Computer Society, 2012

2012
[34]

Binh Pham, Ján Vesel `y, Gabriel H Loh, and Abhishek Bhattacharjee. Large pages and lightweight memory management in virtualized environments: Can you have it both ways? In Proceedings of the 48th International Symposium on Microarchitecture, pages 1–12. ACM, 2015

2015
[35]

Scalable high performance main memory system using phase-change memory technology

Moinuddin K Qureshi, Vijayalakshmi Srinivasan, and Jude A Rivers. Scalable high performance main memory system using phase-change memory technology. In ACM SIGARCH Computer Architecture News, volume 37, pages 24–33. ACM, 2009

2009
[36]

Recency-based tlb preloading

Ashley Saulsbury, Fredrik Dahlgren, and Per Stenström. Recency-based tlb preloading. In the 27th Annual International Symposium on Computer Architecture (ISCA âĂŹ00), volume 28. ACM, 2000

2000
[37]

Synergistic tlbs for high perfor- mance address translation in chip multiprocessors

Shekhar Srikantaiah and Mahmut Kandemir. Synergistic tlbs for high perfor- mance address translation in chip multiprocessors. In Proceedings of the 2010 43rd Annual IEEE/ACM International Symposium on Microarchitecture , pages 313–324. IEEE Computer Society, 2010

2010
[38]

Recency-based tlb preload- ing

Per Stenstr, Fredrik Dahlgren, and Ashley Saulsbury. Recency-based tlb preload- ing. In International Symposium on Computer Architecture , pages 117–127, 2000

2000
[39]

Surpassing the TLB performance of superpages with less operating system support , volume 29

Madhusudhan Talluri and Mark D Hill. Surpassing the TLB performance of superpages with less operating system support , volume 29. ACM, 1994

1994
[40]

Translation ranger: operating system support for contiguity-aware tlbs

Zi Yan, Daniel Lustig, David Nellans, and Abhishek Bhattacharjee. Translation ranger: operating system support for contiguity-aware tlbs. In Proceedings of the 46th International Symposium on Computer Architecture , pages 698–710. ACM, 2019
[41]

Revisiting virtual l1 caches: A practical design using dynamic synonym remapping

Hongil Yoon and Gurindar S Sohi. Revisiting virtual l1 caches: A practical design using dynamic synonym remapping. In 2016 IEEE International Symposium on High Performance Computer Architecture (HPCA) , pages 212–224. IEEE, 2016

2016
[42]

Enigma: architectural and operating system support for reducing the impact of address translation

Lixin Zhang, Evan Speight, Ram Rajamony, and Jiang Lin. Enigma: architectural and operating system support for reducing the impact of address translation. In Proceedings of the 24th ACM International Conference on Supercomputing , pages 159–168. ACM, 2010

2010

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [5]

Efficient virtual memory for big memory servers

Arkaprava Basu, Jayneel Gandhi, Jichuan Chang, Mark D Hill, and Michael M Swift. Efficient virtual memory for big memory servers. In ACM SIGARCH Computer Architecture News, volume 41, pages 237–248. ACM, 2013

work page 2013

[2] [1]

AMD Corporation. 2014. Compute cores. https://www.amd.com/Documents/ Compute_Cores_Whitepaper.pdf, 2014

work page 2014

[3] [2]

Biobench: A benchmark suite of bioinfor- matics applications

Kursad Albayraktaroglu, Aamer Jaleel, Xue Wu, Manoj Franklin, Bruce Jacob, Chau-Wen Tseng, and Donald Yeung. Biobench: A benchmark suite of bioinfor- matics applications. In IEEE International Symposium on Performance Analysis of Systems and Software, 2005. ISPASS 2005. , pages 2–9. IEEE, 2005

work page 2005

[4] [3]

Translation caching: skip, don’t walk (the page table)

Thomas W Barr, Alan L Cox, and Scott Rixner. Translation caching: skip, don’t walk (the page table). In ACM SIGARCH Computer Architecture News, volume 38, pages 48–59. ACM, 2010

work page 2010

[5] [4]

Spectlb: a mechanism for spec- ulative address translation

Thomas W Barr, Alan L Cox, and Scott Rixner. Spectlb: a mechanism for spec- ulative address translation. In ACM SIGARCH Computer Architecture News , volume 39, pages 307–318. ACM, 2011

work page 2011

[6] [6]

Reducing memory refer- ence energy with opportunistic virtual caching

Arkaprava Basu, Mark D Hill, and Michael M Swift. Reducing memory refer- ence energy with opportunistic virtual caching. In ACM SIGARCH Computer Architecture News, volume 40, pages 297–308. IEEE Computer Society, 2012

work page 2012

[7] [7]

Large-reach memory management unit caches

Abhishek Bhattacharjee. Large-reach memory management unit caches. In Proceedings of the 46th Annual IEEE/ACM International Symposium on Microarchi- tecture, pages 383–394. ACM, 2013

work page 2013

[8] [8]

Shared last- level tlbs for chip multiprocessors

Abhishek Bhattacharjee, Daniel Lustig, and Margaret Martonosi. Shared last- level tlbs for chip multiprocessors. In 2011 IEEE 17th International Symposium on High Performance Computer Architecture, pages 62–63. IEEE, 2011

work page 2011