Arista Networks Inc (ANET) 2017 Q4 法說會逐字稿

Welcome to the Fourth Quarter 2017 Arista Networks Financial Results Earnings Conference Call. (Operator Instructions) As a reminder, this conference is being recorded and will be available for replay from the Investor Relations section at Arista website following this call.

歡迎參加 Arista Networks 2017 年第四季財務業績電話會議。 （操作員說明）謹此提醒，本次會議正在錄製中，並可在本次電話會議後透過 Arista 網站的投資者關係部分重播。

I will now turn the call over to Mr. Chuck Elliott, Director of Business and Investor Development. Sir, you may begin.

我現在將把電話轉給業務和投資者發展總監 Chuck Elliott 先生。先生，您可以開始了。

Thank you, operator. Good afternoon, everyone, and thank you for joining us. With me on today's call are Jayshree Ullal, Arista Networks' President and Chief Executive Officer; and Ita Brennan, Arista's Chief Financial Officer.

謝謝你，接線生。大家下午好，感謝您加入我們。參加今天電話會議的有 Arista Networks 總裁兼執行長 Jayshree Ullal；以及 Arista 財務長 Ita Brennan。

This afternoon Arista Networks issued a press release announcing the results for its fiscal fourth quarter and year ended December 31, 2017. If you would like a copy of the release, you can access it online at the company's website.

今天下午，Arista Networks 發布了一份新聞稿，公佈了截至 2017 年 12 月 31 日的第四財季和年度業績。如果您想要該新聞稿的副本，可以在該公司網站上在線訪問。

During the course of this conference call, Arista Networks' management will make forward-looking statements, including those relating to our financial outlook for the first quarter of the 2018 fiscal year. Industry innovation, our market opportunity and the impact of litigation, which are subject to the risks and uncertainties that we discuss in detail in our documents filed with the SEC, specifically in our most recent Form 10-Q and Form 10-K and which could cause actual results to differ materially from those anticipated by these statements. These forward-looking statements apply as of today, and you should not rely on them as representing our views in future. We undertake no obligation to update these statements after this call.

在本次電話會議期間，Arista Networks 管理層將做出前瞻性聲明，包括與我們 2018 財年第一季的財務前景相關的聲明。行業創新、我們的市場機會和訴訟的影響，這些都受到我們在向SEC 提交的文件中詳細討論的風險和不確定性，特別是在我們最新的10-Q 表格和10-K 表格中，並且可能導致實際結果與這些陳述預期的結果有重大差異。這些前瞻性陳述從今天起適用，您不應依賴它們來代表我們未來的觀點。我們不承擔在本次電話會議後更新這些聲明的義務。

Also, please note that certain financial measures we use on this call are expressed on a non-GAAP basis and have been adjusted to exclude certain charges. We have provided reconciliations of these non-GAAP financial measures to GAAP financial measures in our earnings press release.

另請注意，我們在本次電話會議中使用的某些財務指標是在非公認會計原則的基礎上表示的，並已進行調整以排除某些費用。我們在收益新聞稿中提供了這些非公認會計原則財務指標與公認會計原則財務指標的調節表。

With that, I will turn the call over to Jayshree.

這樣，我會將電話轉給 Jayshree。

Thank you, Chuck, and welcome back. Thank you, everyone, for joining us this afternoon on our fourth quarter 2017 earnings call.

謝謝你，查克，歡迎回來。感謝大家今天下午參加我們的 2017 年第四季財報電話會議。

As you can see, our profitability growth combination was once again demonstrated with our non-GAAP revenue of $467.9 million, while our non-GAAP EPS grew to a record $1.71. Service contributed approximately 13% of revenue.

正如您所看到的，我們的獲利能力成長組合再次得到證明，我們的非 GAAP 收入為 4.679 億美元，而我們的非 GAAP 每股收益成長至創紀錄的 1.71 美元。服務貢獻了約13%的收入。

We delivered non-GAAP gross margins of 65.9%, due to the favorable cost and lower cloud titan mix. Overall, in 2017, gross margin came in at 64.8%.

由於有利的成本和較低的雲端巨頭組合，我們的非 GAAP 毛利率為 65.9%。整體而言，2017年毛利率為64.8%。

In terms of customer trends, as of December 31, 2017, we have now acquired over 4,900 customers, with Microsoft becoming 16% of the total revenue for the year. The Q4 2017 international contribution was our strongest to date at 33%, with the Americas contributing 67%. For the full year, the Americas were 73% and the rest of international at 27%.

從客戶趨勢來看，截至2017年12月31日，我們已獲得超過4,900家客戶，其中微軟佔全年總收入的16%。 2017 年第四季的國際貢獻是迄今為止最強的，達到 33%，其中美洲貢獻了 67%。全年來看，美洲地區佔 73%，國際其他地區佔 27%。

In terms of verticals, in Q4 2017, enterprise was our #1 vertical for the very first time in our public company history. Cloud titans was the second largest vertical because of some 945 legal certifications, where some of the complex use cases testing spilled into Q1 2018.

就垂直產業而言，2017 年第四季度，企業在我們上市公司歷史上首次成為我們排名第一的垂直產業。雲端巨頭是第二大垂直產業，擁有約 945 項法律認證，其中一些複雜的用例測試延續到了 2018 年第一季。

In terms of new products, December 2017, Arista unveiled its next generation of cloud-grade routing. Arista EOS Version 4.20 with FlexRoute delivers transformative routing and automation. And with that, we can now address a myriad of peering use cases across cloud, content delivery and Internet exchanges. One such example in cable networks is operators migrating from traditional headend to service-rich virtualized access together with our technology partner, Harmonic.

新產品方面，2017年12月，Arista發表了新一代雲級路由。 Arista EOS 版本 4.20 與 FlexRoute 提供變革性的路由和自動化。有了這個，我們現在可以解決跨雲端、內容交付和互聯網交換的大量對等用例。有線網路中的一個這樣的例子是營運商與我們的技術合作夥伴 Harmonic 一起從傳統的前端遷移到服務豐富的虛擬化存取。

Undoubtedly, 2017 has been a significant year for Arista. We increased our employee strength to 1,800 with 45.8% annual sales growth and $1.646 billion in revenue. For the first time in nearly 2 decades of stagnant IP, Arista is offering that compelling alternative. Legacy vendors are on the wrong side of this trend with traditional places in the network. Instead what Arista is doing is converging PINS into an elastic places in the cloud or PICs with LAN, WAN, Layer 2 and Layer 3 switching integration.

毫無疑問，2017年對Arista來說是意義重大的一年。我們將員工人數增加到 1,800 人，年銷售額成長 45.8%，營收達到 16.46 億美元。在 IP 近 20 年來停滯不前的情況下，Arista 首次提供了令人信服的替代方案。傳統供應商在網路中的傳統位置處於這一趨勢的錯誤一邊。相反，Arista 正在做的是將 PINS 匯聚到雲端中的彈性位置或具有 LAN、WAN、第 2 層和第 3 層交換整合的 PIC。

2017 has also been an important inflection year for both our routing and 100 Gigabit Ethernet products. I expect robust growth to continue ahead in 2018 together with our entry into 400 Gigabit Ethernet next year.

2017 年對我們的路由和 100G 乙太網路產品來說也是重要的轉折年。我預計 2018 年將繼續強勁成長，明年我們將進入 400 Gigabit 乙太網路。

We also delivered many powerful software innovations in containerization, data analyzers and our hybrid Any Cloud offering with Amazon, Microsoft, Google, Oracle and Equinix to name a few. Coupled with meaningful traction of CloudVision, we see this as an emerging software category for us in a couple of years.

我們也在容器化、資料分析器以及與 Amazon、Microsoft、Google、Oracle 和 Equinix 等公司合作的混合 Any Cloud 產品方面提供了許多強大的軟體創新。再加上 CloudVision 的有意義的吸引力，我們認為這是幾年後我們的新興軟體類別。

As we exit 2017, we have also surpassed a cumulative of a total 15 million cloud networking ports and are now the #1 market share leader in 100 Gigabit Ethernet switching.

2017 年結束時，我們的雲端網路連接埠總數也已超過 1500 萬個，目前是 100 Gb 乙太網路交換市場份額第一的領導者。

Our focus on the top 5 verticals will continue this year. We foresee the balance across the verticals to be similar to 2017, where the cloud titans was #1, followed by the service providers, cloud specialty providers and enterprise tied in a 3-way for second place and financials at third place.

今年我們將繼續關注前 5 個垂直行業。我們預計垂直產業的平衡將與 2017 年類似，其中雲端巨頭排名第一，其次是服務供應商、雲端專業供應商和企業並列第二位，財務公司排名第三。

Scaling my leadership team is also an important goal to adapt to all this growth. Many of you may recall that we hired John McCool in 2017 to drive hardware platforms and manufacturing as our Chief Platform Officer. Today, it's my pleasure to warmly and officially welcome Manny Rivelo as our Chief Sales Officer who joined us this year in January. Manny brings seasoned general management and sales and SE leadership from F5 and Cisco. He comes at the right time for Arista. I have personally known Manny for 25 years, and he is working closely with and for our Chief Customer Officer, Anshul Sadana, to formulate international expansion and tailored go-to-market models for all our verticals.

擴大我的領導團隊也是適應所有這些成長的重要目標。許多人可能還記得，我們​​在 2017 年聘請 John McCool 擔任我們的首席平台官，負責推動硬體平台和製造。今天，我很高興正式歡迎 Manny Rivelo 成為我們的首席銷售官，他於今年 1 月加入我們。 Manny 帶來了來自 F5 和 Cisco 的經驗豐富的綜合管理、銷售和 SE 領導力。對阿里斯塔來說，他來得正是時候。我個人認識 Manny 已經 25 年了，他正在與我們的首席客戶長 Anshul Sadana 密切合作，為我們所有的垂直行業制定國際擴張和客製化的上市模式。

What is becoming crystal clear to me is that Arista is in the forefront of the movement to software-driven networking, beyond even data center and cloud boundaries. Our ecosystem of customers and partners are united in supporting us in this migration. And I want to give a special kudos and thank you to all my dedicated Arista employees and my leadership team for the systematic focus on results in the typical Arista way.

我越來越清楚的是，Arista 處於軟體驅動網路發展的最前沿，甚至超越了資料中心和雲端的邊界。我們的客戶和合作夥伴生態系統團結一致，為我們的遷移提供支援。我想向所有敬業的 Arista 員工和我的領導團隊表示特別的讚揚和感謝，感謝你們以典型的 Arista 方式系統地關注結果。

Now I'll turn it over to Ita for more financial specifics. Ita?

現在我將把它交給 Ita 以了解更多財務細節。伊塔？

Thanks, Jayshree, and good afternoon. This analysis of our Q4 and full year 2017 results and our guidance for Q1 '18 is based on non-GAAP and excludes all noncash, stock-based compensation impacts, legal costs associated with the ongoing lawsuits and onetime tax charges associated with the adoption of the Tax Cuts and Jobs Act of 2017.

謝謝，Jayshree，下午好。我們對2017 年第四季和全年業績的分析以及我們對18 年第一季的指導是基於非公認會計準則，不包括所有非現金、基於股票的薪酬影響、與正在進行的訴訟相關的法律成本以及與採用相關的一次性稅收費用。2017 年 減稅和就業法案。

A full reconciliation of our selected GAAP to non-GAAP results is provided in our earnings release.

我們的收益報告中提供了我們選擇的 GAAP 與非 GAAP 績效的全面對帳。

Total revenues in Q4 were $467.9 million, up 42.7% year-over-year and above our guidance of $450 million to $464 million. Services revenues represented approximately 13% of revenue, consistent with last quarter. International revenues for the quarter came in at $153.8 million or 33% of total revenue, up from 27% in the prior period. While some of this shift in geographical mix was driven by the timing of 945-related qualifications with our U.S. customers, we also experienced healthy growth in our international businesses in the period.

第四季總營收為 4.679 億美元，年增 42.7%，高於我們 4.5 億至 4.64 億美元的指導值。服務收入約佔營收的 13%，與上季持平。該季度的國際營收為 1.538 億美元，佔總營收的 33%，高於上一季的 27%。雖然這種地理組合的轉變在一定程度上是由我們的美國客戶獲得 945 相關資格的時間所推動的，但在此期間我們的國際業務也經歷了健康成長。

Overall gross margin in Q4 was 65.9%, up from 64.4% last quarter and above the midpoint of our guidance of 63% to 65%. This outperformance in gross margin for the quarter reflected a more favorable customer mix with a higher contribution from our enterprise and other verticals offsetting lower cloud titan revenue.

第四季的整體毛利率為 65.9%，高於上季的 64.4%，高於我們 63% 至 65% 指引值的中位數。本季毛利率的優異表現反映出更有利的客戶組合，以及我們的企業和其他垂直產業的更高貢獻，抵消了雲端巨頭的營收下降。

Operating expenses for the quarter were $139.3 million, up significantly from $112.9 million last quarter. R&D spending came in at $96 million or 20.5% of revenue, up from $68.6 million in the prior period. This reflected increased NRE spending combined with continued headcount growth. Sales and marketing expense was $33.5 million or 7.2% of revenue, down from $35.5 million last quarter, with reduced demo and sales costs.

本季營運費用為 1.393 億美元，較上季的 1.129 億美元大幅增加。研發支出為 9,600 萬美元，佔營收的 20.5%，高於上一期的 6,860 萬美元。這反映了 NRE 支出的增加以及員工人數的持續成長。銷售和行銷費用為 3,350 萬美元，佔收入的 7.2%，低於上季度的 3,550 萬美元，演示和銷售成本有所減少。

Our operating income for the quarter was $168.9 million or 36.1% of revenue. Other income and expense for the quarter was a favorable $2.2 million. And our effective tax rate was 19.8%, reflecting the favorable impact of the increased international revenue mix discussed above.

我們本季的營業收入為 1.689 億美元，佔營收的 36.1%。該季度的其他收入和支出為 220 萬美元。我們的有效稅率為 19.8%，反映了上述國際收入組合增加的有利影響。

This resulted in net income for the quarter of $137.3 million or 29.4%.

這使得該季度的淨利潤達到 1.373 億美元，即 29.4%。

Our diluted share number for the quarter was 80.24 million shares, resulting in a diluted earnings per share number of $1.71, up 64.3% from the prior year.

本季稀釋後股票數量為 8,024 萬股，稀釋後每股收益為 1.71 美元，較上年增長 64.3%。

For those of you focusing on our GAAP results, our GAAP tax rate for the quarter came in at 26.7%. This reflected the favorable impact of significant excess tax benefits on share-based awards in the quarter, combined with the favorable geographical mix mentioned above and offset by the inclusion of $51.8 million of additional tax charges associated with the recently enacted Tax Act. These share-related excess tax benefits and the onetime tax charges have been excluded from the non-GAAP tax rate discussed above.

對於那些關注我們 GAAP 業績的人來說，我們本季的 GAAP 稅率為 26.7%。這反映了本季度基於股票的獎勵的顯著超額稅收優惠的有利影響，加上上述有利的地域組合，並被與最近頒布的稅法相關的 5180 萬美元額外稅收費用所抵消。這些與股票相關的超額稅收優惠和一次性稅費已不包括在上述非公認會計原則稅率中。

Legal expenses associated with the ongoing lawsuits came in at $9.1 million for the quarter and are also excluded from our non-GAAP results.

本季與正在進行的訴訟相關的法律費用為 910 萬美元，也不包含在我們的非 GAAP 業績中。

Now turning to the balance sheet. Cash, cash equivalents and investments ended the quarter at approximately $1.5 billion. We generated $183.7 million from -- in cash from operations in the December quarter. This reflects strong net income performance combined with a decrease in supply chain-related working capital and deferred revenue amounts.

現在轉向資產負債表。本季末現金、現金等價物及投資約 15 億美元。我們在 12 月季度的營運中產生了 1.837 億美元的現金。這反映出強勁的淨利表現以及供應鏈相關營運資金和遞延收入金額的減少。

DSOs came in at 49 days, up from 45 days in Q3, reflecting the timing of billings in the quarter. Inventory turns were 1.8x, up from 1.7x in Q3. Inventory decreased to $306.2 million in the quarter, down from $333.2 million in the prior period. This primarily reflects reductions in raw materials as we continue to optimize our supply chain. In addition, consistent with last quarter, we maintained a further $34 million of inventory deposits recorded in other assets at the end of the quarter.

DSO 的周期為 49 天，高於第三季的 45 天，反映了本季的計費時間。庫存週轉率為 1.8 倍，高於第三季的 1.7 倍。本季庫存減少至 3.062 億美元，低於上一季的 3.332 億美元。這主要反映了我們不斷優化供應鏈時原料的減少。此外，與上季一致，我們在季度末還保留了其他資產中記錄的 3,400 萬美元的庫存存款。

Our total deferred revenue balance was $515.3 million, down from $565.1 million in Q3. Our product deferred revenue balance declined in the quarter due to the expiration of prior period new product and new customer acceptance clauses. The 2017 year-end product deferred revenue balance was essentially flat to 2016 levels. Accounts payable were 30 days, up from 19 days in Q3, reflecting the timing of inventory receipts and payments. Capital expenditures for the quarter were $2.9 million.

我們的遞延收入餘額總額為 5.153 億美元，低於第三季的 5.651 億美元。由於前期新產品和新客戶接受條款到期，我們的產品遞延收入餘額在本季有所下降。 2017年末產品遞延收入餘額與2016年基本持平。應付帳款期限為 30 天，高於第三季的 19 天，反映了庫存收付的時間。該季度的資本支出為 290 萬美元。

Now turning to our outlook for the first quarter and beyond. We're pleased with the strong execution underlying our 2017 financial performance with 46% revenue growth and 70% growth in earnings per share on a year-over-year basis.

現在轉向我們對第一季及以後的展望。我們對 2017 年財務表現的強勁執行力感到滿意，營收年增 46%，每股盈餘較去年同期成長 70%。

Looking forward to 2018, we're pleased -- we believe that the demand drivers for the business remain strong, and we are well positioned to benefit from the continuing growth in cloud networking across our customer base. That said, we will face some tough comparables for year-over-year revenue growth as we move through 2018. And with this in mind, I would reiterate Jayshree's comments from last quarter with respect to top line growth moderating to a more typical mid-20s rate for the year.

展望 2018 年，我們很高興—我們相信業務的需求驅動因素仍然強勁，並且我們處於有利位置，可以從客戶群雲網路的持續成長中受益。也就是說，隨著 2018 年的到來，我們將面臨一些艱難的同比收入增長。考慮到這一點，我想重申 Jayshree 上個季度關於收入增長放緩至更典型的中期增長的評論。全年利率為20 美元。

On the gross margin front, we would reiterate our gross margin outlook of 63% to 65%, with customer mix being the key driver of where we operate within this range. This assumes a positive outcome on the remaining legal matters, with no disruption to our supply chain.

在毛利率方面，我們重申我們的毛利率前景為 63% 至 65%，客戶組合是我們在此範圍內營運的關鍵驅動因素。這假設剩餘的法律問題取得正面成果，而我們的供應鏈不會受到干擾。

While we will remain cautious in relation to our spending ramp, we also expect that over time, our operating expense investments will gravitate towards our long-term model of 20% R&D, 10% sales and marketing and 3% G&A. We will adopt ASC 606 in the first quarter of 2018. We expect this new guidance will have minimal impact on our operating model. We will reclass a small amount of sales commissions from retained earnings to other assets in the first quarter of 2018. And going forward, our CloudVision deferred revenue amounts will be reported as contract liabilities on the balance sheet.

雖然我們將對支出成長保持謹慎態度，但我們也預計，隨著時間的推移，我們的營運支出投資將傾向於 20% 研發、10% 銷售和行銷以及 3% G&A 的長期模式。我們將在 2018 年第一季採用 ASC 606。我們預計這項新指南將對我們的營運模式產生最小的影響。我們將在 2018 年第一季將少量銷售佣金從留存收益重新分類為其他資產。展望未來，我們的 CloudVision 遞延收入金額將在資產負債表上作為合約負債報告。

In addition, following the adoption of the Tax Act in December, we expect our go-forward non-GAAP tax rate to range from 19% to 21%, reflecting the lower tax rate on our U.S. business, partially offset by higher U.S. taxes on foreign earnings. This represents our current best estimates of the impacts, knowing that elements of the tax legislation are still being refined.

此外，在 12 月稅法通過後，我們預計未來的非公認會計準則稅率將在 19% 至 21% 之間，反映出我們美國業務的稅率較低，但部分被美國較高的稅率所抵銷。國外收入。這代表了我們目前對影響的最佳估計，因為我們知道稅收立法的要素仍在完善中。

With all of this as a backdrop, our guidance for the first quarter, which is based on non-GAAP results and excludes any noncash stock-based compensation impacts and any legal cost associated with the ongoing lawsuits, is as follows: revenues of approximately $450 million to $468 million, gross margin of approximately 63% to 65%, operating margin of approximately 32%. Our effective tax rate is expected to be approximately 20%, with diluted shares of approximately 81 million. Please note that based on our current outlook, we expect costs associated with the ongoing lawsuits to be approximately $8 million for the quarter.

以此為背景，我們對第一季的指導基於非 GAAP 業績，不包括任何非現金股票薪酬影響以及與正在進行的訴訟相關的任何法律成本，如下：收入約為 450 美元百萬至4.68億美元，毛利率約63%至65%，營業利益率約32%。我們的有效稅率預計約為20%，稀釋後股份約為8,100萬股。請注意，根據我們目前的展望，我們預計本季與正在進行的訴訟相關的成本約為 800 萬美元。

I'll now turn the call back to Chuck. Chuck?

我現在將電話轉回給查克。查克？

Thank you, Ita. We are now going to move to the Q&A portion of the Arista earnings call. (Operator Instructions)

謝謝你，伊塔。我們現在將進入 Arista 財報電話會議的問答部分。 （操作員說明）

(Operator Instructions) Your first question comes from the line of Mark Moskowitz from Barclays.

（操作員說明）您的第一個問題來自巴克萊銀行的馬克·莫斯科維茨（Mark Moskowitz）。

So my question really is around the glide path for 2018 full year revenue growth of the mid-20s. How should investors think about the composition in terms of enterprise? Will enterprise still continue to be a top 1 or 2 vertical? Or will there be other changes with respect to the main verticals? And then within that, how important is routing to the enterprise as far as adoption?

所以我的問題其實是圍繞著 20 世紀 20 年代中期的 2018 年全年營收成長的下滑路徑。投資人該如何看待企業的組成？企業仍將繼續維持業界第一或第二的地位嗎？或者主要垂直領域還會有其他變化嗎？那麼，就採用而言，路由到企業有多重要？

Thanks, Mark. I did try to say this in my script. We had an unusual bump in enterprise for Q4. But just to go back on 2017, our #1 vertical was cloud titans and our #2 vertical was tied between enterprise, service provider and cloud specialty providers. And #3 was financials. I pretty much expect the same mix in 2018. I can't exactly tell between 2, 3 and 4 how they'll go. They may be different in each quarter. But I fully expect cloud titans to be our #1 vertical in 2018. So the mix will remain quite similar. And in terms of routing, routing is really a horizontal across all our verticals. This is a transition, as you know, that has been going on for 2 years. I would say there are 2 forms of routing. There's the heavy-duty FlexRoute routing, which is largely for service providers for Internet peering use cases. And there's the more -- the Layer 2, Layer 3 switching, which frankly a majority of our customers use. So routing is quickly becoming an option across our entire customer base. But in terms of heavy-duty FlexRoute, what we've now seen is we've been steadily increasing our customer base. We're well over 200 customers in 2017 and I expect it to continue to be somewhere in the 15% to 20% of our new customer acquisitions.

謝謝，馬克。我確實嘗試在我的劇本中表達這一點。第四季度，我們的企業業務出現了不同尋常的成長。但回顧 2017 年，我們的第一大垂直產業是雲端巨頭，第二大垂直產業是企業、服務供應商和雲端專業供應商。第三是財務。我非常期待 2018 年會有同樣的組合。我無法準確判斷 2、3 和 4 之間它們會如何發展。每個季度的情況可能有所不同。但我完全預期雲端巨頭將成為我們 2018 年排名第一的垂直領域。因此，這種組合將保持非常相似。就路由而言，路由實際上是橫跨我們所有垂直領域的水平方向。如您所知，這種轉變已經持續了兩年。我想說有兩種形式的路由。還有重型 FlexRoute 路由，它主要適用於互聯網對等互連用例的服務提供者。還有更多——第 2 層、第 3 層交換，坦白說，我們的大多數客戶都在使用。因此，路由很快就成為我們整個客戶群的選擇。但就重型 FlexRoute 而言，我們現在看到的是我們的客戶群一直在穩定增加。 2017 年，我們的客戶數量遠遠超過 200 家，我預計該客戶數量將繼續占我們新客戶獲取量的 15% 到 20%。

Your next question comes from Rod Hall from Goldman Sachs.

你的下一個問題來自高盛的羅德·霍爾。

So I wanted to ask you, I guess, 1 question in 2 parts. One is related to 945. Can you just talk a little bit about, is there anything you can do to quantify that for us and also help us understand how 945 might have impacted the guidance? And then could you also talk about AI workloads and what you think your share there is? I know you've got 25% share in 100 gig, but how are you doing in particularly the cloud titan AI workloads? Are you -- do you feel like you have a higher share than what we see in some of the market data?

所以我想分兩部分問你 1 個問題。其中一個與 945 有關。您能否簡單談談，您可以做些什麼來為我們量化這一點，並幫助我們了解 945 可能如何影響指導？然後您能否談談人工智慧工作負載以及您認為您在其中所佔的份額？我知道您在 100 場工作中佔據了 25% 的份額，但您在雲端巨頭 AI 工作負載方面表現如何？您是否覺得自己的份額比我們在一些市場數據中看到的要高？

Welcome, Rod, to Goldman Sachs. I'll take the first question and then punt it to my General Counsel Marc Taxay here. In terms of 945, we've got some really good news on Valentine's Day yesterday that I'll let Marc elaborate on that pretty much affirms the PTAB's decision to invalidate the 668 Patent. But I think the patent you're talking about, the workaround, is really the ACO patent, the 577. And we expected most of the testing and certifications, the simple ones to take 2 to 4 weeks and the more complex ones to take 10 weeks. And because we ran into the holiday season, there were a couple of complex use cases that spilled into Q1 2018 that should have -- should really impact our Q2, Q3, Q4 2018 revenue more than 2018 Q1 itself. So the certifications did take longer. I'm very comfortable that there is a light at the end of the tunnel in Q1. And most of the cases have been completed, but the complex ones will be completed in Q1. Marc, do you want to add to the legal update?

羅德，歡迎來到高盛。我將回答第一個問題，然後將其轉給我的總法律顧問馬克·塔克賽 (Marc Taxay)。就 945 而言，昨天情人節我們得到了一些非常好的消息，我將讓 Marc 詳細說明這一消息，這幾乎肯定了 PTAB 宣布 668 專利無效的決定。但我認為你所說的專利，即解決方法，實際上是 ACO 專利，即 577。我們預計大部分測試和認證，簡單的需要 2 到 4 週，更複雜的需要 10 週幾週。由於我們正值假期，因此有一些複雜的用例滲透到了2018 年第一季度，這些用例對我們2018 年第二季度、第三季度、第四季度收入的影響應該比2018 年第一季本身還要大。所以認證確實需要更長的時間。我很高興第一季的隧道盡頭出現了曙光。大部分案例已經完成，但複雜的案例將在第一季完成。馬克，您想添加法律更新嗎？

Yes, sure. Yes, I think, first of all, we were very grateful that the Federal Circuit affirmed the PTAB's decision yesterday to invalidate the 668 patent. So this was a very important step for us in the 945 case. The other remaining patent that Jayshree was alluding to the TCAM [Apple] patent. That, as a reminder, is also subject to IPR appeal. We won the IPR on that last year. We do feel good about that as well. But regardless of outcome on that one, that patent expires on June 30. At this stage, despite the recent Federal Circuit decision, the PTAB decision, we still remain subject to the ITC order on the 945. So we continue to sell those redesigns. We will be seeking a suspension of the ITC orders based upon the recent 668 decision -- with respect to the 668. But in the interim, we continue to sell the redesigns and we continue to move forward with the modification proceeding.

是的，當然。是的，我認為，首先，我們非常感謝聯邦巡迴法院昨天確認了 PTAB 宣布 668 專利無效的決定。因此，這對我們在 945 案件中邁出了非常重要的一步。 Jayshree 提到的另一項剩餘專利是 TCAM [Apple] 專利。提醒一下，這也受到智慧財產權上訴的影響。我們去年贏得了該項目的知識產權。我們對此也感覺良好。但無論該專利的結果如何，專利將於6 月30 日到期。在現階段，儘管聯邦巡迴法院最近做出了PTAB 的決定，但我們仍然受到ITC 對945 的命令的約束。因此，我們繼續銷售這些重新設計的產品。我們將根據最近的 668 決定尋求暫停 ITC 訂單——就 668 而言。但在此期間，我們將繼續出售重新設計的產品，並繼續推進修改程序。

So switching gears to AI workloads, as you know, Andy Bechtolsheim spoke about this at the technology conference. And we see this as a huge long-term driver. It's still too early to talk about use cases in our cloud titans. Most of them are in experimental stage. But we are working very closely with storage vendors and NVIDIA. And the network is clearly going to require more and more bandwidth. So 100 Gigabit Ethernet workloads, and in the future, 400 will become an important requirement for the vast amount of data crunching and analytics required on AI workloads.

因此，如您所知，轉向人工智慧工作負載，安迪·貝托爾斯海姆 (Andy Bechtolsheim) 在技術會議上談到了這一點。我們認為這是一個巨大的長期驅動力。現在談論我們的雲端巨頭的用例還為時過早。其中大多數處於實驗階段。但我們正在與儲存供應商和 NVIDIA 密切合作。網路顯然需要越來越多的頻寬。因此，100 Gb 乙太網路工作負載，以及未來的 400 Gb 乙太網路工作負載將成為人工智慧工作負載所需的大量資料處理和分析的重要要求。

Your next question comes from Patrick Newton from Stifel.

您的下一個問題來自 Stifel 的帕特里克·牛頓。

I guess, recent commentary from your competitors is that they're addressing disruption in the routing market with routers architected to deliver cloud-scale economics like the PTX or NCS. So do you view these solutions as cost and performance competitive with your R Series? And then on the switching side, where competitors state their Spine support routing functionality, can you comment on how FlexRoute is fundamentally differentiated?

我猜想，您的競爭對手最近的評論是，他們正在利用旨在提供雲端規模經濟效益（如 PTX 或 NCS）的路由器來解決路由市場的混亂問題。那麼您認為這些解決方案的成本和效能是否與您的 R 系列相比具有競爭力？然後在交換方面，競爭對手聲稱他們的 Spine 支援路由功能，您能評論一下 FlexRoute 的根本差異化嗎？

Okay. So first of all, I think my -- some of my competitors have noted there's a routing transition and an architectural shift going on. I just want to be clear, it's been going on for 3 years, not 1 quarter. So legacy and core routers have been migrating to the Spine for quite some time. And the Arista 7500R and 7280R has really become the gold standard for universal Leaf and Spine, both switching and routing. When you need that kind of cloud-grade routing and performance, a large number of customers just love to go with that, both not just for the performance, but the unmatched availability and programmability. In terms of products that are in the market, there's a lot of routers in the market. Arista is not a router company. What we are really doing is moving to more and more use cases for routing and really peering use cases, peering across the Internet, across content, across the cloud. And the reality is our Leaf/Spine architecture has given us a real natural entry into that use case, because people have tested and qualified us for that and they don't have to have a separate box or a separate function. So I think the 7500R and 7280R really stand out for being not just great routing, but great switching, and as I said before, they're occupying multiple positions and places in the cloud for that. So there's not a like-for-like comparison. We think it's truly unique and differentiated, and that's reflected in why we've been so well accepted this last year.

好的。首先，我認為我的一些競爭對手已經注意到正在發生路由轉換和架構轉變。我只是想澄清一下，這種情況已經持續了三年，而不是一個季度。因此，傳統路由器和核心路由器遷移到 Spine 已經有一段時間了。 Arista 7500R 和 7280R 確實成為通用 Leaf 和 Spine（交換和路由）的黃金標準。當您需要這種雲端路由和效能時，大量客戶就會喜歡這樣做，不僅是為了效能，而且是為了無與倫比的可用性和可程式性。從市面上的產品來看，市面上有許多路由器。 Arista 不是一家路由器公司。我們真正要做的是轉向越來越多的路由用例和真正的對等用例，跨互聯網、跨內容、跨雲端進行對等。事實上，我們的葉/脊椎架構讓我們能夠真正自然地進入該用例，因為人們已經對此進行了測試並證明我們合格，並且他們不必擁有單獨的盒子或單獨的功能。因此，我認為 7500R 和 7280R 確實脫穎而出，不僅因為它們具有出色的路由功能，而且還具有出色的交換功能，正如我之前所說，它們為此在雲端中佔據了多個位置和位置。所以不存在同類比較。我們認為它確實是獨特且與眾不同的，這也反映了我們去年如此受歡迎的原因。

Your next question comes from Erik Suppiger with JMP.

您的下一個問題來自 JMP 的 Erik Suppiger。

Yes, coming back to the 945 case, I thought you had introduced the workarounds in the September quarter. So I'm a little confused on the timing. Can you just walk us through the timing on the -- what's stalling some of the titans?

是的，回到 945 案例，我以為您已經在 9 月季度介紹了解決方法。所以我對時間安排有點困惑。您能否向我們介紹一下—是什麼阻礙了一些巨頭的發展？

Yes. So we -- the first time we introduced it was absolutely the month of September. But majority of the certifications -- we did a little bit of it in September, but majority of the certifications were really completed in Q4. And then some of them spilled over to Q1. So September and Q1 was a few, majority was Q4. Do you want to add anything to it, Marc?

是的。所以我們——我們第一次介紹它絕對是在九月。但大多數認證——我們在 9 月做了一些認證，但大多數認證實際上是在第四季度完成的。然後其中一些溢出到了第一季。所以 9 月和第一季是少數，大多是第四季。馬克，你想加什麼嗎？

And you said September, you had some business that fell out of the September quarter in light of that. Did that business close in the December quarter and these are additional projects on top of that?

你說九月，有鑑於此，你的一些業務在九月季度中下降了。該業務是否在 12 月季度結束，除此之外還有其他項目嗎？

Yes, that's exactly correct. Those that we were testing in September closed in Q4. The ones we couldn't finish were the more complex use cases that we were not testing in September.

是的，這是完全正確的。我們在 9 月測試的那些在第四季度關閉。我們無法完成的是我們在 9 月沒有測試的更複雜的用例。

Okay. And those are going to probably fall into the Q1, is that correct?

好的。這些可能會落入第一季度，對嗎？

The completion of the testing would absolutely be in Q1. The business may be in Q1 or Q2 or Q3 or Q4. We can't say exactly.

測試絕對會在第一季完成。該業務可能處於第一季、第二季、第三季或第四季。我們不能準確地說。

Your next question comes from Steve Milunovich with UBS.

您的下一個問題來自瑞銀集團的史蒂夫·米盧諾維奇。

You indicated that Microsoft was 16% of revenue, which I think was the same number the previous year. And yet coming into the year, you suggested it would be less. So it sounds like it did better than you expected. Could you talk about why that was the case and what you're looking for in 2018?

您表示微軟佔收入的 16%，我認為與前一年的數字相同。然而進入今年，你建議它會更少。所以聽起來它的表現比你預期的還要好。您能談談為什麼會發生這種情況以及您在 2018 年的目標是什麼嗎？

Yes. No, I want to give a big shout-out to Anshul Sadana, my entire Microsoft engineering team and [REs] in Seattle. We didn't expect Microsoft spending to be so strong. And although we are still in the early to mid-innings with many of our cloud titans, we really attribute this to 2 things. One is the success of Microsoft itself and the expansion in Azure. And the second is Arista's expansion into a number of use cases. So we find ourselves going into a number of tiers ranging all the way from the SONiC tier, which is their Layer 0 tier to tiers 1, 2, 3, 4, 5. So I think the use cases grew for Arista and grew for Microsoft as well. So it was a win-win.

是的。不，我想向 Anshul Sadana、我在西雅圖的整個 Microsoft 工程團隊和 [RE] 表示大力讚揚。我們沒想到微軟的支出如此強勁。儘管我們的許多雲端巨頭仍處於早期到中期階段，但我們確實將其歸因於兩件事。一是微軟本身的成功以及Azure的擴張。第二個是 Arista 擴展到許多用例。因此，我們發現自己進入了許多層，從 SONiC 層（即第 0 層）到第 1、2、3、4、5 層。因此，我認為 Arista 和 Microsoft 的用例都在增長以及。所以這是雙贏的。

Any expectation for this year?

對今年有什麼期待嗎？

Well, since I guessed poorly last year, I'm going to refrain from guessing. But I think there'll still be 10% concentration customer.

好吧，由於去年我猜得不好，所以我不會再猜了。但我認為還是會有10%的集中客戶。

Your next question comes from Alex Henderson from Needham.

您的下一個問題來自尼達姆的 Alex Henderson。

My primary question is just on the enterprise side. Can you give us some granularity on some of the rate of growth within that and how that splits domestically/internationally? But I was hoping you would give us some guidance on the 2018 tax rate, which I don't think I heard earlier?

我的主要問題是在企業方面。您能否詳細介紹其中的一些成長率以及國內/國際的成長？但我希望您能給我們一些關於 2018 年稅率的指導，我想我之前沒有聽說過？

Yes, actually, why don't I let Ita answer the tax rate question, while I rummage through some stats for you, Alex?

是的，實際上，為什麼我不讓伊塔回答稅率問題，而我為你翻閱一些統計數據，亞歷克斯？

Yes, Alex, we said we think it'll range from 19% to 21%. Probably with 20% at the midpoint is a good place to start, right, given everything that we know now about the new rules.

是的，Alex，我們說過我們認為這個比例會在 19% 到 21% 之間。考慮到我們現在對新規則的了解，中點 20% 可能是一個不錯的起點，對吧。

Okay. So looking at the enterprise sector, we just had a very strong sector just like we did in Q3. Another -- lots of new customers -- it was almost neck-to-neck with Q3. Record number of million-dollar customers in both Q3 and Q4. CloudVision is becoming a very important piece of the enterprise customer purchase. So as you know, this has been something we launched in 2016, but we are now well over 200. In fact, our customer count exceeds 250 cumulatively. So I'm going to challenge the team to double that in 2018. And I think a large contribution to that is our enterprise customers. In terms of international and domestic mix, Mark Foss is helping me here. International is higher. So I would say it's like a 60-40 split in the enterprise approximately. And that speaks to also the way -- our improved stats in international overall from 2016 to '17. Enterprise is clearly contributing there.

好的。因此，看看企業部門，我們的部門非常強大，就像第三季一樣。另一個——大量新客戶——幾乎與第三季並駕齊驅。第三季和第四季的百萬美元客戶數量均創歷史新高。 CloudVision正成為企業客戶購買中非常重要的一環。如您所知，這是我們在 2016 年推出的產品，但我們現在已經超過 200 家。事實上，我們的客戶數量累計超過 250 家。因此，我將挑戰團隊在 2018 年將這一數字翻倍。我認為對此做出巨大貢獻的是我們的企業客戶。在國際和國內組合方面，馬克·福斯（Mark Foss）在這裡為我提供幫助。國際化程度更高。所以我想說這就像是企業中大約 60-40 的分裂。這也說明了這一點——從 2016 年到 17 年，我們在國際比賽中的整體統計數據有所提高。 Enterprise 顯然在這方面做出了貢獻。

Could you give us any sense of the rough magnitude of the growth rate on that, or is that asking too much?

您能否讓我們了解一下成長率的大致幅度，或是這個要求是否過高？

I think I don't have the answers, Alex, but probably asking too much as always. But no problem, thanks for trying.

我想我沒有答案，亞歷克斯，但可能一如既往地問得太多了。不過沒問題，謝謝你的嘗試。

Your next question comes from James Fish with Piper Jaffray.

您的下一個問題來自 James Fish 和 Piper Jaffray。

I guess, first, you guys had a very strong free cash flow this quarter. How should we think about free cash flow going forward and the sustainability of these free cash flow margins? And then any update to the 944 case at this point?

我想，首先，你們本季的自由現金流非常強勁。我們該如何看待未來的自由現金流以及這些自由現金流利潤率的可持續性？那麼現在 944 案件有什麼更新嗎？

Yes. I'll take the cash flow question. I mean, our free cash flow is going to pretty much follow our operating -- our net income performance, right, and the performance of the business. There's very little CapEx and other adjustments that happen to that. So I think as the business performs, you'll see the free cash flow follow that pretty closely. I mean, there's the services deferred revenue, that will continue to grow. Obviously, that generates cash off the balance sheet. But outside of that, it's pretty much well tied to the net income performance. And Marc, if you want to take the other question?

是的。我來回答現金流問題。我的意思是，我們的自由現金流將在很大程度上跟隨我們的營運——我們的淨利潤表現，對的，以及業務表現。與之相關的資本支出和其他調整很少。因此，我認為隨著業務的表現，你會看到自由現金流緊隨其後。我的意思是，服務遞延收入將持續成長。顯然，這會產生資產負債表外的現金。但除此之外，它與淨利潤表現密切相關。馬克，你想回答另一個問題嗎？

Sure. On the 944, there was a remand hearing in January. We believe that went well, and we continue to feel upbeat about where we stand in that litigation. The judge is supposed to issue his decision in June, with the final decision by the Commission in August -- early August.

當然。關於 944，一月舉行了還押聽證會。我們相信進展順利，並且我們繼續對我們在該訴訟中的立場感到樂觀。法官預計在 6 月做出決定，委員會將在 8 月初做出最終決定。

Your next question comes from Alex Kurtz with KeyBanc Capital Markets.

您的下一個問題來自 KeyBanc Capital Markets 的 Alex Kurtz。

This is Steve Enders on for Alex. I was hoping we could maybe get a better understanding of what you're anticipating from a revenue glide path maybe over the year? How do you expect growth rates to kind of trend there? And also, I just wanted to clarify, are you guys expecting a 400 gigabit adoption hit this year or is that more of a 2019 event?

我是史蒂夫·恩德斯（Steve Enders）為亞歷克斯發言。我希望我們能夠更了解您對今年營收下滑的預期？您預計那裡的成長率會如何？而且，我只是想澄清一下，你們是否預計今年 400 Gb 的採用率會達到 400 Gb，還是 2019 年才會發生？

Ita is going to take the first question.

伊塔將回答第一個問題。

Yes. I mean, I think, Steve, that you've seen the guidance for Q1. So obviously, I think the comparables for Q1 have us growing 40% year-over-year with the guide -- the upper end of the guidance that we just gave you. I think for the rest of the year, we said kind of mid-20s for the total year. I don't know if there's any particular linearity to that, that we can see at this stage. But that's kind of the 2 bookends, right?

是的。我的意思是，史蒂夫，我認為您已經看到了第一季的指導。顯然，我認為第一季的可比較數據使我們與去年同期相比增長了 40%，即我們剛剛向您提供的指導的上限。我認為在今年剩下的時間裡，我們說全年的時間大約是 20 多歲。我不知道我們現階段是否可以看到任何特定的線性關係。但這就是兩個書擋，對吧？

Yes. And on 400 gig, I'll just step back and give you a little bit more of a tutorial. We saw that 10 gig had a very long tail, for almost 10 years. And then many vendors jumped in with 40 gig as did Arista. But 100 gig, we believe, first of all has an extremely long tail unlike 40 gig. So we think the relevant bandwidth points is really going to be 100 gig for a very long time, with the option to do 10 gig, 25 gig or 50 gig on the service or storage IO connections. 400 gig is going to be very important in certain use cases. And you can expect Arista is working very hard at it, and we'll be as always first or early to market. The mainstream 400 gig market is going to take multiple years. I believe initial trials will be in 2019, but the mainstream market will be even later. And just because 400 gig comes, by the way, doesn't mean 100 gig goes away. They're really going to be in tandem. The more we do 400 gig, the more we also do of 100 gig. So they're really together. It's not either/or.

是的。對於 400 場演出，我將退後一步，為您提供更多一點的教學。我們看到 10 gig 的尾巴很長，幾乎持續了 10 年。然後許多供應商加入了 40 演出，Arista 也是如此。但我們認為，100 gig 首先有一個與 40 gig 不同的極長的尾巴。因此，我們認為相關頻寬點在很長一段時間內實際上將是 100 GB，並且可以選擇在服務或儲存 IO 連接上使用 10 GB、25 GB 或 50 GB。在某些用例中，400 GB 將非常重要。您可以期待 Arista 正在努力工作，我們將一如既往地率先或儘早進入市場。主流 400 場演出市場將需要數年時間。我相信初步試驗會在2019年，但主流市場會更晚。順便說一句，僅僅因為 400 演出的到來，並不意味著 100 演出的消失。他們真的會齊頭並進。我們做的 400 場演出越多，我們做的 100 場演出就越多。所以他們真的在一起了。這不是非此即彼。

Your next question comes from Paul Silverstein with Cowen and Company.

您的下一個問題來自 Cowen and Company 的 Paul Silverstein。

So going back to the delay that you referenced last quarter and this quarter with respect to some of your cloud customers in connection with the workaround, can you give us a sense for the magnitude of the revenue that got displaced? I mean, clearly from your comments, not all the revenue was displaced, but some of it in connection with the more complex use cases. Any boundaries you can put on that growth?

那麼，回到您提到的上季和本季與解決方法有關的一些雲端客戶的延遲問題，您能否讓我們了解被取代的收入的規模？我的意思是，從您的評論中可以清楚地看出，並非所有收入都被取代，但其中一些收入與更複雜的用例有關。您可以為這種增長設定任何界線嗎？

Yes. So Paul, I want to reiterate to everybody that no competitor displaced our revenue.

是的。所以保羅，我想向大家重申，沒有任何競爭對手取代了我們的收入。

No, I didn't mean to suggest competitive displacement, Jayshree. I just meant...

不，我並不是想建議競爭性取代，Jayshree。我只是想說...

Okay, okay. So I want to be really clear on that. But it did take longer, and it was really important as they're going into some mission-critical production sites in these complex use cases that the cloud operators, and the cloud titans especially, had extra testing. And so I would say let's -- roughly in terms of percentage, 20% of the use cases were complex and moved out.

好吧好吧。所以我想澄清這一點。但這確實需要更長的時間，而且這非常重要，因為他們要在這些複雜的用例中進入一些關鍵任務生產站點，雲端運營商，尤其是雲端巨頭，需要進行額外的測試。所以我想說，粗略地講，20% 的用例很複雜並被移出。

Can you translate that for us in dollar terms?

您能為我們翻譯成美元嗎？

No. Nice try.

不，不錯的嘗試。

All right. If I may, you've been kind enough to update us periodically on the competitive landscape with respect to cloud in general. So my question to you is, has there been any changes either in terms of competitive displacement among the web-scale folks as well as your specialty, as you call them, your specialty providers, either by other competitors or by any meaningful movement to white-box solutions? And can you tell us -- I think at your Analyst Day and periodically since, you told us that your hyperscale folks -- I think the number was 25% or thereabouts. Can you give us an update on what they were collectively for calendar '17?

好的。如果可以的話，請您定期向我們通報雲端領域競爭格局的最新情況。所以我問你的問題是，在網路規模的人們以及你的專業（正如你所說的那樣，你的專業提供者）之間的競爭位移方面是否有任何變化，無論是其他競爭對手還是任何有意義的白人運動盒式解決方案？您能否告訴我們 - 我想在您的分析師日以及此後您定期告訴我們您的超大規模人員 - 我認為這個數字是 25% 左右。您能否向我們介紹 17 年曆的最新情況？

We have seen no appreciable change in competitive landscape in our cloud titan or cloud scale customers. So the answer to all your questions is really in one nutshell no, we haven't seen any change.

我們沒有看到雲端巨頭或雲端規模客戶的競爭格局有明顯變化。因此，您所有問題的答案實際上一言以蔽之：不，我們沒有看到任何變化。

And the contribution for '17 from the hyperscale?

超大規模對 17 年的貢獻是什麼？

It was #1 vertical. The cloud titans was the #1 vertical out of our 5.

它是垂直領域第一名。雲泰坦在我們的 5 個垂直領域中排名第一。

Your next question comes from Stanley Kovler with Citi Research.

您的下一個問題來自花旗研究部的 Stanley Kovler。

Just to try on the quantification front one more time. 20% of workloads in cloud titan being pushed out. That sounds like about 3% or so of revenue, if that makes sense. But the broader question on the cloud titans is that if you look out at spending of some of your cloud titan customers, some are doubling their CapEx spending. Do you expect that to drive the international part of your business? Or will that be in the U.S.? As we look at the business growing mid-20s, just curious how you think about the mix between U.S. and international in that context?

只是為了再嘗試一次量化方面的工作。雲端巨頭 20% 的工作負載被淘汰。如果這有意義的話，聽起來大約是收入的 3% 左右。但關於雲端巨頭的更廣泛的問題是，如果你留意一些雲端巨頭客戶的支出，你會發現有些客戶的資本支出增加了一倍。您認為這會推動您業務的國際化嗎？還是會在美國？當我們看到 20 多歲中期的業務成長時，只是好奇您如何看待美國和國際在這種背景下的混合？

I think that's a really good question. So clearly -- first of all, I've made this point before and I'd really like to make it again. I know it's easy to put a one-to-one correlation between the cloud CapEx and the Arista spend. I think it's important not to look at it so precisely, because the cloud CapEx has a lot of components to it and the networking is a very small piece. And the second point I want to make is our visibility with our cloud customers, although getting better, is still much stronger for 1 to 2 quarters and nothing beyond that. So even though the cloud CapEx spend for 2018 may be great, we have much more visibility to what might happen in 1 or 2 quarters, not the rest -- not the whole year. So I think it's really important to understand that. Now the one thing I will say is one of the things we noticed as a trend in 2017 that we think will continue in 2018 with all the cloud titans is since they began a lot of their expansion early in the U.S., a lot of their later expansion will be in international sites. So we expect at least a 50-50, maybe even a higher contribution from international with our cloud titans.

我認為這是一個非常好的問題。很明顯——首先，我以前已經說過這​​一點，而且我真的很想再說一遍。我知道很容易在雲端資本支出和 Arista 支出之間建立一對一的關聯。我認為重要的是不要如此精確地看待它，因為雲端資本支出有很多組件，而網路只是一個非常小的部分。我想說的第二點是，我們對雲端客戶的可見度雖然有所改善，但在 1 到 2 個季度內仍然會更強，僅此而已。因此，儘管 2018 年的雲端資本支出可能很大，但我們對 1 或 2 個季度可能發生的情況有更多的了解，而不是其餘時間——而不是全年。所以我認為理解這一點非常重要。現在我要說的一件事是，我們在2017 年註意到了一種趨勢，我們認為所有雲端巨頭都會在2018 年繼續這種趨勢，因為他們很早就開始在美國進行大量擴張，其中很多是後來的擴張將在國際站點進行。因此，我們預計國際雲端巨頭的貢獻至少為 50-50，甚至可能更高。

Your next question comes from Simon Leopold with Raymond James.

您的下一個問題來自西蒙·利奧波德和雷蒙德·詹姆斯。

Just check on something of whether or not your forecast for the first quarter reflects an assumption of some catch-up from the titans or whether that's an event you think is later in the year? And in terms of big picture question, I wanted to see if you could comment on your thoughts about the pros and cons of Arista participating in the campus switching market? I assume that's something you're not that interested in, but just wanted to understand strategically how you would frame that?

只需檢查一下您對第一季的預測是否反映了巨頭們追趕的假設，或者您認為這是否是今年晚些時候的事件？就大局問題而言，我想看看您是否可以評論一下您對 Arista 參與校園交換市場的利弊的看法？我認為您對此並不感興趣，只是想從戰略上了解您將如何建立它？

Thanks, Simon. So as you know, our first quarter is usually our toughest quarter. It's seasonally slow, and our guidance reflects that. It does incorporate cloud titans, like it always does, but nothing unusual in that guidance. It's normal. And we think any completion of certifications will not have an immediate impact on Q1, but will have a more gradual impact over the rest of the year. In terms of campus, I think one of the phenomenas I'm seeing is, and I tried to say this in my opening script, that the well-defined boxes and places in the network where you have a campus box, a data center box, a branch box, a core box is really changing. And what our customers are really pushing us to do is look at campus as an extension of the cloud or look at new drivers in the campus. And so while we have no intentions of participating at this time in any kind of traditional campus market, we have every intention of disrupting the traditional campus market as currently defined.

謝謝，西蒙。如您所知，我們的第一季通常是最艱難的季度。這是季節性的緩慢，我們的指導反映了這一點。它確實像往常一樣納入了雲端巨頭，但該指南中沒有什麼不尋常的。這是正常的。我們認為任何認證的完成都不會立即對第一季產生影響，但會在今年剩餘時間內產生更漸進的影響。就園區而言，我認為我看到的現象之一是，我試圖在我的開場腳本中說這一點，即網絡中定義明確的盒子和位置，其中有校園盒子、數據中心盒子，一個分支箱，一個核心箱，真的在改變。我們的客戶真正推動我們做的是將園區視為雲端的延伸，或是尋找園區中的新驅動力。因此，雖然我們目前無意參與任何類型的傳統校園市場，但我們完全有意擾亂目前定義的傳統校園市場。

Your next question comes from James Faucette with Morgan Stanley.

您的下一個問題來自摩根士丹利的詹姆斯·福賽特。

I just wanted to go back to the international markets. And maybe can you help us understand -- they're bouncing around a fair amount in terms of contribution right now. Can you help us understand how those -- where you think those might end up? And also give us any insight as to which geographies and types of customers you're seeing internationally and how we should think about the evolution of your opportunities outside the U.S.?

我只是想回到國際市場。也許你能幫助我們理解——他們現在的貢獻量在相當大的範圍內波動。你能幫助我們了解這些——你認為這些最終可能會怎麼樣嗎？也請讓我們了解您在國際上遇到的客戶的地理位置和類型，以及我們應該如何考慮您在美國以外的機會的演變？

Yes, I think, James, I'd definitely look at the numbers for the year as a better way of looking at it. But if you look at it, we've got -- we were at 27% international, which is up from our kind of typical 25% historically. I think that's a better way to think about it than some of the swings that we saw in the last couple of quarters, right? So I think we're seeing some good, steady progress internationally. That business is growing well, like what I would call the in-region business, even if we carve out some of the cloud titan effects that we just talked about, right? The in-region business is growing well and it's still relatively small. So you can see it be lumpy between APAC and Europe. But both are actually, if you look at it over time, growing at a faster rate than the U.S..

是的，我想，詹姆斯，我肯定會以更好的方式看待今年的數字。但如果你看一下，我們的國際化比例為 27%，高於歷史上典型的 25%。我認為這是比我們在過去幾季看到的一些波動更好的思考方式，對嗎？所以我認為我們在國際上看到了一些良好、穩定的進展。該業務成長良好，就像我所說的區域內業務一樣，即使我們消除了我們剛才談到的一些雲端巨頭效應，對嗎？區域內業務成長良好，但規模仍相對較小。所以你可以看到亞太地區和歐洲之間的情況是不穩定的。但如果你長期觀察的話，實際上兩者的成長速度都比美國快。

Your next question comes from Vijay Bhagavath with Deutsche Bank.

您的下一個問題來自德意志銀行的 Vijay Bhagavath。

I would like to get like a state of state from you in the sense of international and also telcos here in the U.S. At least, our view is that's kind of your next phase of growth? So give us a state of state on how business is proceeding, how is sales, any interesting projects you could share with us in the international markets and also at the telcos?

我希望從國際和美國電信公司的角度從您那裡得到一種國家狀態。至少，我們認為這是您下一階段的成長？那麼請給我們介紹一下業務進度、銷售情況以及您可以在國際市場和電信公司與我們分享的任何有趣的項目嗎？

Sure. I think in international, we're doing very well in several countries. And in general, I would tell you we're acquiring customers faster internationally than we are in the U.S. I think 60% of our customer acquisitions in 2017 was international. So the projects are many and the size of deal is smaller, just to put it bluntly, right? In terms of service providers, switching gears, those are not small deals by any means. And we view service providers in the midst of a transformation where the traditional traffic engineering is going to start looking more and more like cloud networks and resemble what we're doing with Ethernet and IP and Leaf and Spine. So these Leaf/Spine cloud architectures are now permeating the service providers. We're starting to see some big RFPs, and we've experienced some wins in 2017. I think 2018 will be an important year. Ken Kiser, Jonathan and the entire team is really working hard on this worldwide with Manny. And what we really see is that deterministic performance we're getting out of Leaf/Spine is going to transform the SP networks for many of the peering use cases. And this is a tremendous Layer 2 and Layer 3 opportunity for us. So we expect to see service provider contribute better in 2018 than 2017.

當然。我認為在國際上，我們在幾個國家做得很好。總的來說，我會告訴你，我們在國際上獲取客戶的速度比在美國更快。我認為 2017 年我們 60% 的客戶獲取來自國際。那麼專案多，規模小，說穿了是吧？就服務提供者而言，切換設備無論如何都是一筆不小的交易。我們認為服務供應商正處於轉型之中，傳統的流量工程將開始變得越來越像雲端網絡，並且類似於我們在乙太網路、IP 以及 Leaf 和 Spine 所做的事情。因此，這些 Leaf/Spine 雲端架構現在正在滲透到服務供應商。我們開始看到一些大型 RFP，我們在 2017 年取得了一些成果。我認為 2018 年將是重要的一年。 Ken Kiser、Jonathan 和整個團隊正在與 Manny 一起在全球範圍內努力工作。我們真正看到的是，我們從 Leaf/Spine 中獲得的確定性效能將改變 SP 網路的許多對等用例。這對我們來說是一個巨大的第 2 層和第 3 層機會。因此，我們預計 2018 年服務提供者的貢獻將優於 2017 年。

Your next question comes from Aaron Rakers with Wells Fargo.

您的下一個問題來自富國銀行的 Aaron Rakers。

A lot of questions have already been answered, but -- or asked and answered. But I want to go kind of back to the guidance and kind of how you're thinking about the flow-through effect of deferred revenue as we progress through 2018. I can appreciate that you've communicated that you would have expected deferred revenue to decline here going forward. It looks like short-term deferred was down 20-some-odd percent sequentially. So I'm just curious of how we should think about that relative to the revenue expectations over the next couple of quarters?

很多問題已經得到解答，但是——或者已經提出並得到解答。但我想回顧一下指導意見，以及您如何看待 2018 年遞延收入的流通效應。我很欣賞您已傳達的訊息，即您預計遞延收入將未來這裡會下降。短期遞延似乎比上一季下降了 20% 左右。所以我只是好奇我們該如何考慮未來幾季的營收預期？

Yes, I mean, I think now that kind of the deferred balance -- the product deferred balance is back in line with what we've seen historically and it's pretty much consistent with where we were at this time last year, we're not going to try to honestly guide the movements in the product deferred going forward, right? We've kind worked through the acceptances, et cetera, that we had with the -- associated with the R Series that had caused it to move around so much. But I think we're through that now and it's just back to a normal kind of deferred level. But I'm not sure that we're going to try to forecast that as we go forward.

是的，我的意思是，我認為現在這種遞延餘額 - 產品遞延餘額又回到了我們歷史上看到的水平，並且與去年這個時候的情況非常一致，我們沒有我們會嘗試誠實地指導推遲的產品的未來發展，對嗎？我們已經完成了與 R 系列相關的驗收等工作，這導致它發生瞭如此大的變化。但我認為我們現在已經度過了這個階段，並且剛剛恢復到正常的延遲水平。但我不確定我們是否會在前進過程中嘗試預測這一點。

Your next question comes from George Notter with Jefferies.

你的下一個問題來自傑弗里斯的喬治·諾特。

I guess, I just wanted to follow up on that last question. So was the step-down in deferred sequentially, was that consistent with what you guys were looking for coming into the quarter? And then also, I assume that all of that R Series product revenue is largely out of the deferred number now, is that correct?

我想，我只是想跟進最後一個問題。那麼，降級是否會依序推遲，這是否符合你們對本季的預期？另外，我假設所有 R 系列產品收入現在大部分都超出了遞延數字，對嗎？

Yes. I mean, I think pretty much in line with what we were expecting coming in, a lot of it -- a lot of what's left in the deferred revenue balance at this stage, a portion of that is going to be 945 qualification related and then some other stuff. But kind of the concentration of R Series, new product type stuff that was there during the year last year has definitely been processed at this point.

是的。我的意思是，我認為這與我們的預期非常一致，其中很多——現階段遞延收入餘額中剩下的很多，其中一部分將與 945 資格相關，然後一些其他的東西。但 R 系列的集中度，去年的新產品類型在這一點上肯定已經得到了處理。

Your next question comes from James Kisner with Loop Capital Markets.

您的下一個問題來自 Loop Capital Markets 的 James Kisner。

I understand your comment on tough comps in 2018. I'm just wondering if you could talk about whether you see any gating factors to growth in 2018, whether they be external or internal, availability of optics, availability of silicon, CapEx or construction constraints? And conversely, perhaps you could talk about any major swing factors that could lead to upside to your [25%] growth forecast in 2018 and whether you would have any constraints to address potential upside opportunities?

我理解您對2018 年艱難的競爭的評論。我只是想知道您是否能談談您是否認為2018 年的增長有任何限制因素，無論是外部的還是內部的、光學器件的可用性、矽的可用性、資本支出或建設限制？相反，也許您可以談談任何可能導致您 2018 年 [25%] 增長預測上行的主要波動因素，以及您是否會受到任何限制來應對潛在的上行機會？

James, welcome to your first Arista call. No, we don't see anything major. We've done a good job of planning and forecasting our inventory. And from a supply chain perspective, there's been a lot of work that's gone on. With our top customers, we have a great deal of intimacy on the forecast, and they let us know even before they book orders sometimes, what kind of demand they expect. Of course, it can change dramatically. But we feel good about the projects ahead and the opportunity ahead. And we are putting all the right plans in place for it.

James，歡迎您第一次接到 Arista 電話。不，我們沒有看到任何重大的事情。我們在庫存規劃和預測方面做得很好。從供應鏈的角度來看，已經做了很多工作。對於我們的頂級客戶，我們對預測有很大的了解，有時他們甚至在預訂訂單之前就讓我們知道他們期望什麼樣的需求。當然，它可能會發生巨大的變化。但我們對未來的項目和機會感到滿意。我們正在為此制定所有正確的計劃。

I guess, any clarification there. I mean, I'm looking at the growth in 100-gig ports and optics, for example, in 2018. And obviously, they're going to be up a lot more than 25. And just kind of wondering if it's perhaps because switches are deployed in advance of the optics, and obviously, not all are 100 gig. But just seems like quite a gap in the difference in volumes. Maybe you can clarify that a little bit.

我想，那裡有任何澄清。我的意思是，我正在關注 100G 端口和光學器件的增長，例如 2018 年。顯然，它們的增長將遠遠超過 25。只是想知道這是否可能是因為交換器是在光學器件之前部署的，顯然，並非全部都是100 gig。但在數量上的差異似乎存在相當大的差距。也許你可以稍微澄清一下。

So the Layer 1 optics decisions aren't always made in conjunction with Arista switching and routing. Sometimes they're together and sometimes they're entirely separate like Data Center Interconnect decisions. So you won't see a one-to-one correlation. Secondly, sometimes you won't even see a one-to-one correlation in our applications, because they may already have acquired the optics from another vendor or they may have (inaudible). So that's one area where you really can't correlate a lot between the routers and the accessories.

因此，第 1 層光學決策並不總是與 Arista 交換和路由一起制定。有時它們是在一起的，有時它們是完全獨立的，就像資料中心互連決策一樣。所以你不會看到一對一的相關性。其次，有時您甚至不會在我們的應用程式中看到一對一的關聯，因為他們可能已經從其他供應商購買了光學器件，或者他們可能已經（聽不清楚）。因此，在這一領域中，路由器和配件之間確實無法進行大量關聯。

Your next question comes from Jeff Kvaal of Nomura.

你的下一個問題來自野村證券的傑夫·克瓦爾。

I wanted to ask about the mid-20% outlook for 2018, starting from the 40%-ish number in the March quarter. That seems like an awfully sharp deceleration. I get it that we're not in the same 100-g migration and you don't have a buildup of 7500 to work with. But why might the market -- why might your rhythm slow down that much? I mean, CapEx is going up quite dramatically, as was previously noted.

我想問一下 2018 年中期 20% 的前景，從 3 月季度的 40% 左右的數字開始。這似乎是一個非常急劇的減速。我知道我們不在同一個 100-g 遷移中，而且您沒有 7500 的累積可供使用。但為什麼市場——為什麼你的節奏會慢那麼多呢？我的意思是，正如之前所指出的，資本支出正在大幅增長。

Yes. I mean, Jeff, I think if you look at the comparables for last year, I mean, we really saw the spike in growth, kind of the 50%-plus kind of growth rates happen in the middle of the year, right? So I think Q1 is starting off a slightly different base, right? And again, this is our outlook as we sit here today. We haven't kind of looked at all of the information that we have. We feel good about that. It's a good solid outlook for next year. And then we'll see where we go from there.

是的。我的意思是，傑夫，我想如果你看看去年的可比情況，我的意思是，我們確實看到了增長的飆升，50% 以上的增長率發生在年中，對嗎？所以我認為第一季的起點略有不同，對嗎？再說一次，這就是我們今天坐在這裡的展望。我們還沒有查看我們所掌握的所有資訊。我們對此感覺很好。這是明年的良好前景。然後我們會看看接下來的發展方向。

Yes.

是的。

Okay. So maybe it would be fair to say that you don't have the visibility beyond mid-year with some of the biggest accounts, so -- which is normal. So therefore take a pretty cautious view of how things might proceed beyond the window of visibility?

好的。因此，也許可以公平地說，在年中之後，您對一些最大的客戶就沒有可見性了，所以——這是正常的。因此，請非常謹慎地看待事情在可見窗口之外可能會如何發展？

Yes. I mean, it's really hard to plan out the linearity a year out, right? I think...

是的。我的意思是，規劃一年後的線性確實很難，對吧？我認為...

Let me go back to -- if you ask Ita and me whether we would have 50% growth increases in 2017, we certainly didn't forecast it. It came upon us suddenly, because the cloud titans suddenly had a use case and application. So when we look at the overall year and when we look at how to plan the year, just like we did in 2017, we think a growth rate in the 20s to 30s, with an average of 25% is a very good growth rate. It's not cautious, it's realistic. And if something really good happens beyond that, that's great. Or something really bad happens, we'll let you know. Hope is not a strategy. This is our best effort at predicting what that growth would be.

讓我回到——如果你問 Ita 和我 2017 年我們是否會實現 50% 的增長，我們當然沒有預測到。它突然出現在我們面前，因為雲端巨頭突然有了一個用例和應用程式。所以當我們看整個一年的時候，當我們看如何規劃這一年的時候，就像我們在2017年所做的那樣，我們認為20到30歲的增長率，平均25%是一個非常好的增長率。這不是謹慎，而是現實。如果除此之外還有真正好的事情發生，那就太好了。或發生非常糟糕的事情，我們會讓您知道。希望不是策略。這是我們在預測成長方面所做的最大努力。

Yes. I didn't catch the 50% growth either.

是的。我也沒有看到50%的成長。

Your next question comes from Hendi Susanto from Gabelli.

您的下一個問題來自 Gabelli 的 Hendi Susanto。

Arista's cash has grown to $1.5 billion, and it generated strong free cash flow of over $600 million in 2017. How should investors think about the use of cash in the future?

Arista的現金已成長至15億美元，2017年產生了超過6億美元的強勁自由現金流。未來投資者該如何看待現金的使用？

Yes. I mean, I think at this stage, Hendi, we're still in the mode of growing and investing in the business. So I think we look at that as something that will continue to generate and use for the business. Beyond that, we're not opposed to looking at some capital allocation stuff later on just philosophically. But it's too early, right? We're still a relatively early stage, high-growth company. So I think we think about that as cash that we'll hold for the business in the near term.

是的。我的意思是，我認為在現階段，Hendi，我們仍處於業務成長和投資的模式。所以我認為我們將其視為將繼續為業務產生和使用的東西。除此之外，我們並不反對稍後從哲學角度看待一些資本配置問題。但現在還太早，對吧？我們仍然是一家相對早期、高成長的公司。因此，我認為我們將其視為短期內為業務持有的現金。

And I think we're going to continue to invest where we've always been strong in execution, which is our R&D and our go-to-market and systems engineers and sales and marketing. So we'll certainly put it to use and you saw us spending a lot more than you probably forecasted in Q4.

我認為我們將繼續在我們一直執行力較強的領域進行投資，即我們的研發、市場推廣、系統工程師以及銷售和行銷。因此，我們肯定會使用它，並且您看到我們在第四季度的支出比您可能預測的要多得多。

Your next question comes from Tal Liani with Bank of America Merrill Lynch.

您的下一個問題來自美銀美林的塔爾·利亞尼 (Tal Liani)。

Just a clarification on deferred. What's driving deferred down? Is it timing of revenue recognition following announcements of features or maybe following approvals of the workaround? I'm trying to understand if there's the 2 trends here. On one hand, you get the orders in an orderly manner. And then on the other hand, there is like odd fluctuation because of timing of certain things. So if it's not it, if you can just explain what are the puts and takes in the deferred revenues?

只是對延期的澄清。是什麼推動了延後？收入確認的時間是在功能公告之後還是在解決方案獲得批准之後？我想了解這裡是否有兩種趨勢。一方面，你有秩序地拿到訂單。另一方面，由於某些事情的發生時間，會出現奇怪的波動。那麼，如果不是這樣，您能否解釋一下遞延收入中的看跌期權和索取金額是多少？

I mean, there's probably 2 factors, right? One is we're closing out on features, et cetera, that we needed to deliver and on acceptances that we had in contracts. The R Series is over a year kind of in the market now. So we're seeing a lot of those close out. So we're not generating as much deferred revenue because of that. I think secondly, we were still qualifying some of these design-arounds in the fourth quarter, and that did affect the linearity of invoicing and stuff. I think the underlying business is solid. It's more the timing of shipments, billing and how it's flowing to the deferred.

我的意思是，可能有兩個因素，對嗎？一是我們正在關閉我們需要交付的功能等以及我們在合約中接受的承諾。 R 系列現已上市一年多了。所以我們看到很多這樣的公司倒閉了。因此，我們沒有產生那麼多的遞延收入。我認為其次，我們仍在第四季度對其中一些設計進行資格認證，這確實影響了發票和其他內容的線性度。我認為基礎業務很紮實。更多的是發貨時間、計費以及如何流入延期。

So going forward, you expect the deferred -- I think you said normal level. But what I'm trying to understand is going forward, you would expect the deferred to go more in line with the bookings kind of, with the orders you are getting? If the orders go up and you can recognize it goes up and vice versa, or did you still have these backlog of things that need to be recognized because of timing of features and approvals?

因此，展望未來，你預計會出現延遲——我認為你說的是正常水平。但我想了解的是，未來您會期望延期的訂單與您收到的訂單更加一致嗎？如果訂單增加，您可以認識到它增加了，反之亦然，或者您是否仍然有這些由於功能和批准的時間安排而需要識別的積壓事項？

Yes -- no, I think it's more the -- the deferred -- the product deferred revenue was never a topic of conversation prior to 2017. It was very much a factor of the new product features and new customers that we were engaging with in '17. Actually under the new accounting rules under 606, even a similar transition to that would not drive the same amount of deferred revenue going forward in the future. That guidance is very much geared towards an earlier recognition of revenue. So I think as we go forward, deferred revenue will be much less of a topic from a product perspective and you'll just see the services deferred revenue continue to grow.

是的 - 不，我認為這更多的是 - 遞延 - 產品遞延收入在 2017 年之前從來不是一個主題。這在很大程度上是我們正在接觸的新產品功能和新客戶的一個因素。 '17。實際上，根據 606 下的新會計規則，即使是類似的過渡也不會在未來帶來相同數量的遞延收入。該指引很大程度上是為了儘早確認收入。因此，我認為，隨著我們的前進，從產品角度來看，遞延收入將不再是一個話題，您只會看到服務遞延收入繼續成長。

This concludes the Arista Q4 2017 earnings call. I also want to mention that we have posted a presentation, which provides additional information on our fiscal results, which you can access on the Investors section of our website. Thank you to everyone for joining us today.

Arista 2017 年第四季財報電話會議到此結束。我還想提一下，我們已經發布了一份演示文稿，其中提供了有關我們財務業績的更多信息，您可以在我們網站的投資者部分訪問這些信息。感謝大家今天加入我們。

Thank you, everyone.

謝謝大家。

Thank you for joining, ladies and gentlemen. This concludes today's call. You may now disconnect.

女士們、先生們，感謝您的加入。今天的電話會議到此結束。您現在可以斷開連線。