(ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) TPS calculator

Open weights NTT Communication Science Laboratories,Tohoku University 185M parameters August 2018

Each card below is assessed against this model at the context length and minimum quality you choose. Speed is an estimate for a single request, calculated from the card's memory bandwidth and the size of the model once compressed.

Calculated for this model

818 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Tesla C1080

4 GB · Q8_0 · 199 tok/s

Fastest card

B200

18,315 tok/s · 180 GB

Which GPUs can run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2)?

Set the inputs, read the answer

Context length tokens

A longer conversation needs more memory, which can push this model off smaller cards.

Minimum quality

Hides cards that would only fit the model by compressing it below this point.

Search cards

Card memory

Also show cards that cannot run it

818 cards match

Calculating

					Needs	Quantisation	Fit
18,315 tok/s 10,989–29,304 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	0.9 GB	Q8_0	Comfortable
18,315 tok/s 10,989–29,304 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	0.9 GB	Q8_0	Comfortable
14,625 tok/s 8,775–23,400 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	0.9 GB	Q8_0	Comfortable
14,625 tok/s 8,775–23,400 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	0.9 GB	Q8_0	Comfortable
11,696 tok/s 7,018–18,714 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	0.9 GB	Q8_0	Comfortable
11,195 tok/s 6,717–17,912 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.9 GB	Q8_0	Comfortable
11,195 tok/s 6,717–17,912 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.9 GB	Q8_0	Comfortable
10,714 tok/s 6,428–17,143 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	0.9 GB	Q8_0	Comfortable
9,509 tok/s 5,705–15,214 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	0.9 GB	Q8_0	Comfortable
9,509 tok/s 5,705–15,214 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	0.9 GB	Q8_0	Comfortable
9,509 tok/s 5,705–15,214 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	0.9 GB	Q8_0	Comfortable
9,020 tok/s 5,412–14,432 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable
7,692 tok/s 4,615–12,308 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable
7,692 tok/s 4,615–12,308 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	0.9 GB	Q8_0	Comfortable
7,692 tok/s 4,615–12,308 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable
7,692 tok/s 4,615–12,308 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable
7,692 tok/s 4,615–12,308 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable
5,857 tok/s 3,514–9,371 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	0.9 GB	Q8_0	Comfortable
5,857 tok/s 3,514–9,371 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	0.9 GB	Q8_0	Comfortable
4,881 tok/s 2,929–7,809 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	0.9 GB	Q8_0	Comfortable
4,777 tok/s 2,866–7,643 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	0.9 GB	Q8_0	Comfortable
4,670 tok/s 2,802–7,472 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	0.9 GB	Q8_0	Comfortable
4,670 tok/s 2,802–7,472 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	0.9 GB	Q8_0	Comfortable
4,670 tok/s 2,802–7,472 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	0.9 GB	Q8_0	Comfortable
4,670 tok/s 2,802–7,472 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable

Speeds are estimates for a single request — one conversation at a time — calculated from memory bandwidth, model size and quantisation. Real throughput varies with the inference runtime and its version. Figures published by hardware vendors measure many simultaneous requests and are much higher.

On record

Full specification

Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.

Origin

Who built this model, where, and when it was published.

Organisation: NTT Communication Science Laboratories,Tohoku University
Organisation type: Industry,Academia
Country: Japan
Published: 30 August 2018
Authors: Sho Takase, Jun Suzuki, Masaaki Nagata

What it does

The problem areas the model was built for. A model can carry several of each.

Domain: Language
Task: Language modeling

Size

How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.

Parameters: 185M
Training data: 2,000,000 tokens
Epochs: 300

Training compute

The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.

Training compute: 6.7 × 10¹⁷ FLOP
How it was established: Operation counting

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights: Open — downloadable
Model access: Open weights (unrestricted)
Training code: Open source

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Why it is tracked: SOTA improvement
Record confidence: Confident
Citations: 37
Benchmark data: (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2)

Sources

Where this record came from and when it was last checked.

Reference: Direct Output Connection for a High-Rank Language Model
Last updated: 11 February 2026

The extremes

The ten fastest GPUs that run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2)

Ranked by estimated tokens per second, newest card first where speeds tie. Because generation is bound by memory bandwidth, this ordering follows bandwidth rather than any gaming benchmark.

The smallest GPUs that still run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2)

The cheapest route in, by memory capacity. A tight fit runs the model but leaves nothing spare for a longer conversation.

What the numbers mean

The hardware side

Minimum card

Tesla C1080

Memory needed

0.9 GB

Fastest

18,315 tok/s

(ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) is small enough at 185M parameters that hardware is rarely the obstacle — 818 of the cards we track can run it, including cards several years old.

At the low end, a Tesla C1080 handles it — 4 GB, at Q8_0, for about 199 tokens per second.

A B200 is the fastest we calculate for it: about 18,315 tokens per second, from 8,000 GB/s of memory bandwidth.

Where it came from

(ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) was published by NTT Communication Science Laboratories,Tohoku University, in Japan, in August 2018. industry,Academia is the category the publisher falls under.

It works in Language, and is recorded as doing language modeling.

The weights are published, so it can be downloaded and run on your own hardware indefinitely, offline, with no account attached.

Understanding the speeds

Across every card that can run it, the middle of the range is about 514.3 tokens per second, and 818 of them clear the ten tokens per second that roughly matches reading speed.

Being dense, it reads all of itself per token, which is why the ordering by speed below follows the ordering by memory bandwidth so closely.

Its internal architecture is not on file, so memory is approximated from the parameter count and marked accordingly. Expect the real figure to differ, more so at long context.

Training and provenance

The training run consumed about 6.7 × 10¹⁷ FLOP. That figure describes the cost of creating it and has no bearing on how quickly it generates text.

Around 2,000,000 tokens went into training it.

Its inclusion criterion is sOTA improvement.

Step by step

How to choose a GPU for (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2)

The table above has already assessed every card we hold specifications for against this model. Getting to your answer takes six steps.

01

Read the memory figure first

Every card here has been checked against (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) — around 0.9 GB at Q8_0. Capacity is the gate — a card either holds it or it does not.
02

Match the context to your actual use

Longer conversations cost memory on top of what the weights need. Move the slider to your real working length before trusting any row for (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2).
03

Choose how far you will compress it

Each card runs the least-compressed copy it can hold — Q8_0 on the smallest card that fits. Setting a floor drops the cards that only manage (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) by squeezing it further than you would want.
04

Rank by throughput rather than spec sheet

Sort by speed to see how cards rank for (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2). It will not match a gaming ordering — generation is bound by memory bandwidth, which is why the B200 tops it at 18,315 tok/s.
05

Check the fit verdict before buying

Tight means (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) loads and works, with no room to raise the context later. Comfortable means you can. The difference matters more than a few tokens per second.
06

See what else that card runs

Every card name links to its own page, which runs the same calculation across the whole model catalogue. Worth a look before buying for (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) alone — a card is usually bought for more than one model.

Answers

(ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) — common questions

Is (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) open source?

Its weights are published, so (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) can be downloaded and run on your own hardware. Note that open weights is not the same as open source in the full sense — it says nothing about the training data, the training code, or the commercial terms attached.

How many parameters does (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) have?

(ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) has 185M parameters. 185M (table 8). That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.

Who created (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2)?

(ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) was published by NTT Communication Science Laboratories,Tohoku University, based in Japan, categorised as industry,Academia.

When was (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) released?

(ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) was published in August 2018. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

What is (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) used for?

(ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) works in Language, and is recorded as handling language modeling. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2)?

The weights for (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

How much compute was used to train (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2)?

Around 6.7 × 10¹⁷ FLOP. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.

Can I run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) if it does not fit in my GPU?

Partly. Layers that do not fit sit in system memory and run at a fraction of the speed, so a mostly-offloaded (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) is rarely worth using. Every figure here assumes the whole model is on the card.

Would two GPUs run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) faster?

A second card roughly doubles the memory available but not the generation rate. With 818 cards already able to run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) alone, the case for pairing is weak.

Why does the quantisation differ between cards for (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2)?

Because capacity varies, so does how hard (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) has to be squeezed — 1 distinct levels appear in the table above. Set a minimum quality to compare at one.

How accurate are these (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) speed estimates?

Every figure is derived from memory bandwidth and model size, not benchmarked. That is why each is published as a range such as 10,989–29,304 tok/s on the B200 rather than a single number.

What GPU do I need to run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2)?

The smallest card in our catalogue that holds (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 0.9 GB, and produces roughly 199 tokens per second. 818 cards in total can run it.

How fast is (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 18,315 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 818 of the cards that can run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) clear that.

How much VRAM does (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) need?

About 0.9 GB at Q8_0 compression, which is what the smallest card that runs it uses. Less compression needs more: the figures in the memory column above are recalculated for each card, because each one holds the least-compressed version it can.

Can I run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 0.9 GB and generating roughly 3,411 tokens per second — a comfortable fit.

Can I run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 0.9 GB and generating roughly 2,089 tokens per second — a comfortable fit.

Can I run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 0.9 GB and generating roughly 2,587 tokens per second — a comfortable fit.

Can I run (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 0.9 GB and generating roughly 3,068 tokens per second — a comfortable fit.

Source

Original publication

Record last updated 11 February 2026

The other direction

Looking at it from the other side?

This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.

All AI models