Sunday, October 17, 2021

RIG stock: My weekend research | RIG financial information | Anna Sokolidou | Seekingalpha

 

Transocean Earnings - The Offshore Recovery Is Starting To Take Shape

May 06, 2021 3:08 PM ETTransocean Ltd. (RIG)63 Comments17 Likes

Summary

  • Transocean reported its first-quarter earnings that were broadly in line with the analysts' estimates.
  • GAAP EPS totaled -$0.16, which was a $0.05 estimate beat, whilst the revenue figure was $653M, a miss of $0.71M.
  • The management is highly optimistic about the company's performance, whereas the heavy debt load Transocean used to suffer from is much lower now.
  • The vaccination campaigns in many countries are running at full speed and there is an economic recovery.
  • Other offshore drillers have gone bankrupt and many of their rigs have been scrapped. The Green agenda will most probably be a limit to rising oil production. The outlook for both the oil prices and Transocean is much brighter now.
Transocean (RIG) reported its first-quarter earnings that were quite in line with what many analysts and investors had expected. The management sounded highly optimistic during the company's conference call. The market for offshore drilling rigs is getting tighter, whilst the day rates have already started rising. The macroeconomic environment, meanwhile, is improving. I am personally inclined to believe the management since they do whatever they can to deliver good value to the stockholders.

The earnings release

The sales revenues were slightly below the previous quarter but the operating and maintenance expenses were lower as well.

  • time range: 4Q2020
  • Total contract drilling revenues: $690 million
  • Revenue Efficiency: 97.2%
  • Operating and maintenance expenses: $465 million

  • time range: 1Q2021
  • Total contract drilling revenues: $653 million
  • Revenue Efficiency: 97.4%
  • Operating and maintenance expenses: $435 million
At the same time, the company's adjusted contract drilling revenues of $709 million were above the management's guidance, mostly due to stronger than expected revenue efficiency as well as higher than forecasted reimbursable expenses. Operating and maintenance expenses in Q12021 totaled $135 million. This was slightly below the management's expectations, mostly due to the timing of some shipyard projects. The total net loss was $99 million or $0.16 per diluted share. However, if we adjust the results for the retirement of debt, disposal assets, and discrete tax items, the reported net tax loss was $117 million or $0.19 per diluted share.

In my view, one of the brightest spots of the first quarter's earnings is the adjusted EBITDA of $245 million figure, a $35 million improvement compared to 4Q2020. I agree with the management that this shows sound cost control. Fleetwide revenue efficiency totaled 97.4%, which was quite in line with the company's usually excellent backlog conversion. Even more important, in my opinion, was the fact there was positive cash flow from operating activities, which totaled $96 million.

As concerns the cash cushion, Transocean ended the fourth quarter with total liquidity of about $2.7 billion, including unrestricted cash and cash equivalents of approximately $1.1 billion, roughly $300 million of restricted cash for debt service, and $1.3 billion from the company's undrawn revolving credit facility. It is quite a sound cash cushion, whereas the management has shown creativity, whilst retiring some debt. So, I truly believe Transocean will explore all the possible options to reduce the debt load and raise the cash levels.

The backlog totaled $7.4 billion, a fall from $7.8 billion. The falling backlog problem has been often pointed out here on Seeking Alpha. However, if we dig a bit deeper, we will see that new contracts were awarded in the past quarter and none were canceled.

For example, in the Gulf of Mexico, the Deepwater Asgard was awarded a three-well fixture with Beacon Offshore Energy. Two of these wells were priced at $240,000, whilst the third well was priced at $280,000 a day. This led to a $30 million addition in the company's backlog. Transocean's CEO Jeremy Thigpen also pointed out tightening market conditions in the Gulf of Mexico.

In Trinidad the DD3 is going to start her next contract with BHP in June and, including options, the DD3 could remain on contract with BHP in Trinidad through September.

In Brazil, the Petrobras 10000 is scheduled to end her contract with Petrobras in September. However, Transocean's management was said to be "in the middle of discussions" with Petrobras about a long-term extension.

As concerns Norway, the Transocean Norge was awarded another one-well extension by Equinor at a day rate of $297,000 plus bonus. The rig is expected to be on contract through June. But it is actively being offered multiple opportunities in the Norwegian market. The Transocean Barents is starting her work with MOL Norge soon, which is expected to run to the fourth quarter and probably even beyond.

As the management noted on the fourth-quarter call, in West Africa the Deepwater Skyros was awarded Total's Rig of The Year. The company is discussing the possibility with the customer of a six-well option which is expected to last for more than a year for the rig.

If we look at the Asia-Pacific region, very recently the Deepwater Nautilus began her 90-day contract with POSCO. This means the Deepwater Nautilus will work at least through July but the management is actively looking for new opportunities for the Nautilus in the Asia-Pacific region.

Overall, the management expects the offshore recovery to take place. Jeremy Thigpen sees the market for offshore drilling rigs getting tight. It looks like the company is going to announce more contract awards pretty soon.

It is often mentioned by many analysts that RIG's competitors are emerging from bankruptcy proceedings. These companies, they argue, will be very strong rivals for Transocean since they will have few or no debts. So, they think the day rates will stay depressed since the market will not get tighter. But it is often forgotten that they emerge with little cash, whereas the offshore drilling business requires heavy capital expenditures. What is more, a majority of these newly emerged companies have very poor fleets with many of their rigs cold-stacked. As the management has mentioned, it is quite costly to activate cold-stacked rigs. That is why I would rather agree with the management who argue the market is getting tighter.

Transocean's activities and the macroeconomic outlook

Some analysts actually worry about the effects the Green agenda might have on Transocean's activities in the Gulf of Mexico. But, in my view, there is hardly any reason to worry about this.

First, Transocean does not rely too heavily on the Gulf of Mexico. What is more, there are a lot of growth opportunities not-to-be-missed in developing countries. Of particular interest are Guyana and Brazil where the environmental regulations are much more lenient, whereas the governments do whatever they can to support their offshore oil drillers.

Secondly, the Green agenda does not only combat offshore oil. It also targets other oil sources. If we are to talk about the US, much of the pre-pandemic oil production jump was due to shale, which is, arguably, more dangerous for the environment than offshore oil. This does not mean all American shale will disappear overnight but it does mean the conditions are getting less and less favorable.

For example, quite recently California decided to initiate a regulatory action to stop the issuance of new fracking permits by January 2024. It might not mean much in terms of oil production since not much of the state's oil is coming from fracking. However, it is quite a precedent and more states would probably follow California's example. This fits well with the Green agenda and the fact that the US joined the Paris Climate Accord. President Biden also pledged to make a 50% reduction in greenhouse emissions by 2030 compared to 2005. So, this suggests seriously limiting any new oil and gas production.

But the alternative green sources are not ready to fully replace the traditional fossil fuels. This winter, in my opinion, made it clear. We all know there were extreme weather conditions in Texas with extremely low temperatures and blackouts. Neither the solar panels nor the wind turbines were able to solve the electricity problems. However, not only did Texans suffer from the consequences of this unusually cold winter. Abnormally low temperatures were observed this winter in South Europe, Japan, and China. It is quite strange since these regions are associated with mild winters. If we continue to experience similar temperatures in the future, the high oil demand and especially gas will be here to stay until the alternative technologies get more advanced.

It is often assumed Transocean's backlog and drilling revenues only depend on the oil prices. This is not entirely so. According to the company description found on Reuters' website:

Transocean Ltd. is an international provider of offshore contract drilling services for oil and gas wells. The Company's primary business is to contract its drilling rigs, related equipment and work crews on a dayrate basis to drill oil and gas wells.

As you can see Transocean also relies on the demand for gas, which is a fossil fuel well-suited for heating homes. So, if the demand for gas stays high and its production goes down, Transocean will be here to gain.

Moreover, the oil prices are also stabilizing with the massive vaccination campaigns and the gradual post-pandemic reopening all over the world.

Conclusion

Transocean is a true leader in the offshore drilling industry. Transocean's management seems to be competent and willing to serve their stockholders. It was clearly evidenced by the debt exchange offer they did last year.

The market for drilling rigs is getting tighter, whilst the oil prices are getting higher. The macroeconomic outlook is improving due to the gradual post-pandemic reopenings and the favorable monetary conditions. The Green agenda will be quite a hurdle for raising oil production.



RIG stock: My weekend research | Crude oil price | Anna Sokolidou | Seeking Alpha

Oct. 17, 2021

Here is the article. 

Transocean Stock Is Trading Low, Given The Energy Crisis

Oct. 14, 2021 2:37 PM ETTransocean Ltd. (RIG)69 Comments9 Likes


























Anna Sokolidou

Summary

  • Right now, Transocean stock is trading far below the 52-week maximum of $5.13 reached in the beginning of July.
  • However, back then, the prices did not reach their all-time high of more than $84 per barrel of Brent oil, a situation we are facing now.
  • Given the global energy crisis and the coming heating season, it is highly likely the oil prices will be over and above the $80 mark.
  • The fossil fuels deficit is highly beneficial for Transocean's backlog, its revenue generation perspectives, and its general financial position.
  • Therefore, I firmly believe Transocean stock has plenty of growth potential.
It is true that offshore drillers and the oil industry, in general, had a tough time last year. For sure as my Seeking Alpha colleagues have noted, Transocean's (NYSE:RIG) backlog had a very challenging time in the past several years. It faced backlog and revenue declines. But at the same time, the backlog decline speed has recently fallen due to the industry recovery. If there is no financial crisis in the near future, Transocean has good chances to start growing since the oil industry is likely to see its golden days very soon. In this article, I will explain why.

The global energy deficit

Quite recently ahead of the heating season many countries discovered their natural gas reserves were empty. This was partly due to the fact they expected green energy sources to satisfy their energy needs. Lots of European countries, Germany in particular, rely heavily on wind energy. But, unfortunately, the weather was not windy enough. So, not enough energy was generated. Moreover, there was also a conflict around North Stream 2. So, not enough natural gas was imported to Europe as a result.

What is more, not all capacity can be quickly brought back online. The 2020 pandemic made many producers substantially reduce their capital spending to extract oil and gas. It is not always easy to bring everything back online. What is more, many American companies also want to improve their balance sheets and pay their stockholders juicy dividends.

As a result, we are facing a real energy supply crunch. You might think that it only applies to natural gas. On the surface, it does seem so. After all, as can be seen from the diagram below, the natural gas prices have almost doubled ahead of the heating season.

The oil price rally was much more moderate. However, not everyone realizes that oil and gas are substitute goods when it comes to heating.

As you can see from the diagram above, there was a dramatic spike in the spot price of natural gas during the 2020-2021 winter. The oil prices also rose. But it is not that obvious if we look at the graph. Instead of using natural gas which was in a very scarce supply, some people have switched to heating oil or diesel fuel to keep their houses warm during the last winter season.

It seems like it will be even more the case this winter season. That is why the demand for oil has further room to run.

The 2021 - 2022 winter

It is very hard to predict weather conditions. Many forecasts are highly inaccurate. However, it is highly likely this winter will be as freezingly cold as last winter. But the difference is that many countries seem to be even more unprepared for extremely cold weather conditions. This is not just Europe that is even willing to burn coal in spite of the very rigid environmental regulations. China is also struggling right now. This simply means that many countries are extremely vulnerable should this winter be colder than usual.

Many meteorologists argue that a La Nina will emerge. This could bring colder weather to the Northern USA and milder weather conditions in the south while making other parts of the world dry out. Meanwhile, the polar vortex that contains icy air above the North Pole is weaker than last year. All that could translate into a frigid cold in Asia, North America, and Europe, thus making this winter season very cold.

You could argue, of course, this fuel shortage is a temporary factor. However, that is not exactly the case. Whilst springtime will eventually come and the weather conditions will get better, the situation we observe now highlights a fundamental problem. First, the oil and gas industry has been highly underinvested. I am not even talking about the year 2020 when the oil and gas production and capital investments in the industry fell to astonishingly low levels. I am talking about the oil investment history from 2003 to 2019.

As you can see from the graph above, oil investments have been struggling since 2016. The year 2020 has made matters even worse.

Secondly, green energy sources cannot fully solve the problem. Many large banks and other financial giants have recently reduced their investments in fossil fuels. Instead, they have actively started investing in environmentally clean energy sources. Some of them are more effective than others, of course. But many European countries, Germany and the UK in particular have recently seen how unreliable wind energy is. The same is true of solar batteries. They are excellent when the weather is sunny but they cannot generate enough energy when it is not sunny. Unfortunately, modern technologies still do not allow to accumulate solar energy and use it later.

That is why fossil fuels have really good growth potential. This means offshore drillers will also benefit from the recovery. In my view, developing countries will be even more likely to benefit since many developed countries have lots of environmental restrictions in place to limit oil production. I wrote about this in a lot more detail in my other article. Transocean clearly stands as a beneficiary here.

Transocean's undervaluation

Given the situation, it is not very clear why Transocean's stock is trading that low. After all, Transocean is an offshore oil driller. Its revenue and therefore the stock price relies heavily on the oil prices. Big Oil invests heavily in offshore drilling contracts if it is profitable for them to do so. It is therefore obvious that there is a correlation between the oil prices and the company's stock price.

If we see the recent years, there were times when a barrel of Brent used to cost much lower than it does now, whereas RIG stock traded much higher than it is trading now.

You could argue that it is the backlog problem that matters here. Transocean's backlog is now much lower than it used to be two or three years ago.

It might seem reasonable. However, I also composed a graph for the recent 6 months. After all, there were no significant changes in the company's backlog during the period. At the beginning of July, the company's stock price was above $5, whereas the Brent crude oil spot price was lower than it is now. However, as of the time of writing RIG stock is trading at a price below $4 per share. Even if we take the short-term perspective, the stock clearly deserves to trade higher.

Transocean's latest fleet status report stated the company's backlog declined from $7.4 billion in April to $7.3 billion in July. This was at the slowest pace in recent years. This was a really good result given the company had reported a rapidly declining backlog before. But how about this winter season? And what will happen in the future, given the oil prices will keep rising? In my view, Transocean's stock will skyrocket.

Conclusion

It is highly challenging to predict the future. However, it seems to me that the energy crisis will not last for one season only. It looks like we will face the fossil fuels deficit for a while. That is why the oil and natural gas industry will benefit from capital investments in the near future. Transocean will also clearly be here to gain.

RIG's stock price is unbelievably low, even given the current oil prices. But if we face permanent deficits in the future, Transocean's stockholders will face huge gains.

Of course, everything is possible. In my view, the greatest stumbling block for this prediction is the global economy's health. Many macroeconomists say the Fed and other central banks will most likely begin to tighten the monetary conditions in the near future.

But at this point, I would say Transocean has a bright future.


Sunday and Learn: Google cloud | Data Modeling for BigQuery (Google Cloud Next '17)

Oct. 17, 2021

Here is the link. 



Saturday, October 16, 2021

Crude oil 80+ | Winter is coming | My short research | What to work on?

 Oct. 16, 2021

Introduction

I spent over 30 minutes to filter out oil stocks from original 56 stocks, and three steps are applied. 

Stocks to work on | Start to work on purchase | Expectation 10% return | Hard work

It is better to start with a plan, what to purchase, and then how to work on research on those companies. As an investor, it is better for me to choose those stocks first, and then read more about Yahoo -> Finance -> Conversation. 

First step, choose maximum ava vol (10days)


Next step | Div/ Share | More volative 

Last step | 52-Wk Low Chg% | Top 6 stocks 




Friday, October 15, 2021

Breakfast and learn: From blobs to relational tables: Where do I store my Data? (Google Cloud Next '17)

Oct. 15, 2021

Here is the link. 

Google Cloud Platform (GCP) offers many options for storing your data. From Cloud Storage (Multi-regional, Nearline, Coldline) to the various database services (Cloud Datastore, Cloud SQL, Cloud Bigtable) and data warehousing (BigQuery), learn about the most common workloads and common patterns for data storage. Dominic Preuss and Dave Nettleton discuss decision criteria and the various tradeoffs to consider when choosing your method of data storage. Missed the conference? Watch all the talks here: https://goo.gl/c1Vs3h Watch more talks about Infrastructure & Operations here: https://goo.gl/k2LOYG

Wednesday, October 13, 2021

System | 分布式 | GFS | 腾讯 后端工程师 | Julia is a best learner | Excellent article in depth

Here is the article. 

gfs被称为谷歌的三驾马车之一,主要面向谷歌的大流量流式读取和append写,通过控制流与数据流解耦提升并发能力。

GFS架构

GFS核心在于,master只告诉你地址,不给你数据,要取数据?自己去问chunkserver。至于chunksever给不给,master说了算。


GFS的本质就是将数据流和控制流解耦,master只负责控制流,提供metadata,chunkserver只负责数据流,提供data。

Performance can be improved by scheduling expensive data flow based on the network topology。在数据推送过程中,可以选择最近的chunkserver,而不需要在意推送顺序。推送顺序由同步保证

 就是这么设计,master 只提供 url,但是 client 允许直接访问数据而没有加入校验的控制流,所以也不是很完善,而且返回url本身开销也很大 ( once per request ),增加控制流开销,也算是 trade-off 了。

相比之下,gfs 增加了 master 对数据的控制流,并且数据流远远大于控制流,在这种 workload下控制流和数据流分离就是绝佳的设计了。此时虽然 client 上控制流和数据流仍然耦合,但是因为他只是边缘端的,根本没有多少流量,所以也不会有什么性能瓶颈。

Single Master

存储文件系统的metadata于内存中

  • Namespace
  • Access Control Information
  • Mapping from files to chunks
  • Current locations of chunks

namespace 和 mapping 同时通过 operation log 持久化,location 则通过 chunkserver 的心跳来获得,决定系统的访问权限。

这里的 namespace 并非 inode 文件系统,没有 symlink 也没有 hardlink,路径名并非真正的目录,通过前缀压缩性能优化。(猜测数据结构为Trie,单词查找树,是哈希树的变种)每个节点都有一把读写锁。

master虽然有全局了解,简化设计,但不是bottleneck(后来的改进证明谷歌还是乐观了)

  • client 从不读写 master,只请求通信的 chunkserver 地址
  • master 可以提供后续 chunk 的额外信息,从而减少延迟
  • 这些后续的 chunk 的读并不需要访问 master

master 的状态通过 log 和 checkpoint 备份,宕机时启动备份并且修改 DNS 从而得到primary。

这些备份的 "shadow"master 提供只读权限,但不要求强一致性从而避免性能开销,允许延后根据日志来进行同步。因此master在恢复的时候也允许进行读取,提供 fault tolerance。

Multiple Chunkservers

存储文件系统的data于内存中,每个chunk大小固定64MB。

数据冗余采用 3-way mirror,分散在不同机器、不同 rack,防止同时崩溃。

通过心跳信息与 master 通信,由 master 决定访问权限,但是 chunk 是否存在则由chunkserver 自己决定而非 master。

存在 primary/backup,由 primary 决定写入顺序保证同步并减轻 master 负担,primary 通过抢锁获得,抢到锁的成为 primary,但是需要定期续约,否则会自动释放(primary崩溃)。

读操作

  • 应用向 client 发出读请求 read(filename, byte range)
  • client 翻译为 read(filename, chunk index) 并请求 master
  • master 响应 chunk handle 并告诉所有 replicas 的地址
  • client 选择一个 replicas 并请求 read (chunk handle, byte range)
  • chunk 根据访问权限,决定是否返回请求的数据
  • client 将数据返回给用户

如果选择 replicas 检测到 checksum 不正确,则会返回错误并要求 client 重试其他 replicas,并向 master 请求同步其他 replicas。

checksum具体实现复制 cnblogs.com/lushilin/p/

  • 读操作的Checksum:只取Chunk小部分额外相关数据进行校验,客户端每次把读取操作对齐在Chunk块的边界上

写操作

  1. 应用向client发出写请求write(filename,data)
  2. client翻译为write(filename, chunk index) 并请求master
  3. master响应chunk handle并告诉所有primary/backup replicas的地址
  4. client对所有replicas请求write(chunk handle, byte range, data)或单纯append(chunk handle , data)
  5. client请求primary 进行同步,进行持久化
  6. primary决定buffer数据写的顺序,并写入chunk
  7. primary把顺序告诉所有的backup要求他们以同样顺序执行
  8. primary等待所有backup响应
  9. primary返回响应给client,如果任意backup失败,client重试同步。

checksum具体实现复制 cnblogs.com/lushilin/p/

  • 记录追加操作的Checksum(高度优化):只增量更新最后一个不完整Chunk块的Checksum
  • 写操作的Checksum:先读取和校验被写操作覆盖的第一个和最后一个Chunk块,写操作完成后再重新计算和写入Chunksum

云时代的改进

Here is the link. 

Storage software: Colossus

  • Next-generation cluster-level file system
  • Automatically sharded metadata layer
  • Data typically written using Reed-Solomn (1.5x)
  • Client-driven replication, encoding and replication
  • Metadata space has enabled availability analyses
Why Reed-Solomon?
  • Cost. Especially w/ cross cluster replication.
  • Field data and simulations show improved MTTF
  • More flexible cost vs. availability choices

对于Master采取数据切分,进一步提高并发能力

对于Chunk使用1.5倍的冗余编码Reed-Solomon减少冗余备份,提供纠错机制

Problem: 提供大规模文件的分布式存储

Related work: 组件故障是常态, 规模不够大, 没有append专门优化,应用API没有共同设计

Observation: 控制流数据流分离 + 先推送数据再执行同步

Solution: Master只寻址chunk,Primary只需要顺序同步不需要数据同步。

Evaluation: 单Master掌握全局,高度中心化

Comments: 如果每个区域都需要本地化的GFS,那么Master放哪里呢?如果全球通用GFS,那么如何保证时延平衡呢?

编辑于 2020-07-11



Google search | How search works

 Here is the link. 

How Search organizes information

Before you search, web crawlers gather information from across hundreds of billions of webpages and organize it in the Search index.

The fundamentals of Search

The crawling process begins with a list of web addresses from past crawls and sitemaps provided by website owners. As our crawlers visit these websites, they use links on those sites to discover other pages. The software pays special attention to new sites, changes to existing sites and dead links. Computer programs determine which sites to crawl, how often and how many pages to fetch from each site.

We offer Search Console to give site owners granular choices about how Google crawls their site: they can provide detailed instructions about how to process pages on their sites, can request a recrawl or can opt out of crawling altogether using a file called “robots.txt”. Google never accepts payment to crawl a site more frequently — we provide the same tools to all websites to ensure the best possible results for our users.

Finding information by crawling

The web is like an ever-growing library with billions of books and no central filing system. We use software known as web crawlers to discover publicly available webpages. Crawlers look at webpages and follow links on those pages, much like you would if you were browsing content on the web. They go from link to link and bring data about those webpages back to Google’s servers.


Organizing information by indexing

When crawlers find a webpage, our systems render the content of the page, just as a browser does. We take note of key signals — from keywords to website freshness — and we keep track of it all in the Search index.

The Google Search index contains hundreds of billions of webpages and is well over 100,000,000 gigabytes in size. It’s like the index in the back of a book — with an entry for every word seen on every webpage we index. When we index a webpage, we add it to the entries for all of the words it contains.

With the Knowledge Graph, we’re continuing to go beyond keyword matching to better understand the people, places and things you care about. To do this, we not only organize information about webpages but other types of information too. Today, Google Search can help you search text from millions of books from major libraries, find travel times from your local public transit agency, or help you navigate data from public sources like the World Bank.



Kirk McKusick and Sean Quinlan | Two computer scientists

 August 7, 2009

  Download PDF version of this article PDF

Case Study

GFS: Evolution on Fast-forward

A discussion between Kirk McKusick and Sean Quinlan about the origin and evolution of the Google File System.

Actionable items

I did spend a day to read the paper. I just could not believe that it is so important for me to read the conversation instead of formula or data structure algorithm. I should have good time to learn more later about GFS. 

To explore the reasoning behind a few of the more crucial initial design decisions as well as some of the incremental adaptations that have been made since then, ACM asked Sean Quinlan to pull back the covers on the changing file-system requirements and the evolving thinking at Google. Since Quinlan served as the GFS tech leader for a couple of years and continues now as a principal engineer at Google, he’s in a good position to offer that perspective. As a grounding point beyond the Googleplex, ACM asked Kirk McKusick to lead the discussion. He is best known for his work on BSD (Berkeley Software Distribution) Unix, including the original design of the Berkeley FFS (Fast File System).
 

Garth Gibson | Professor of CMU | My GFS study

Summary Biograph - From the webpage http://www.cs.cmu.edu/~garth/

 I joined the faculty of CMU's Computer Science Department in 1991. Previously I received a Ph.D. and a M.Sc. in Computer Science in 1991 and 1987, respectively, from the University of California at Berkeley. Prior to Berkeley, I received a Bachelor of Mathematics in Computer Science and Applied Mathematics in 1983 from the University of Waterloo in Ontario, Canada.

    In 1993 I founded CMU's Parallel Data Laboratory (PDL) and led it until April 1999. Today the PDL is led by Greg Ganger. The PDL is a community that typically comprises between 6 to 9 faculty, 2 to 3 dozen students and 4 to 10 staff. It receives support and guidance from a consortium of 15 to 25 companies with interests in parallel data systems, the Parallel Data Consortium. This community holds biannual retreats and workshops to exchange technology ideas, analysis and future directions. The publications of the PDL are available for your inspection.

    The principal contributions of my first twenty years of research: Redundant Arrays of Inexpensive Disks (RAID), Informed Prefetching and Caching (TIP) and Network-Attached Secure Disks (NASD), whose architectural basis shapes the Google File System and its descendents such as the Hadoop Distributed File System (HDFS) and the Parallel Network File System, pNFS, features in NFS v4.1 (video discussion), have all stimulated derivative research and development in academia and industry. RAID, in particular, is now the organizing concept of a 10+ billion-dollar marketplace (more on RAID in my 1995 RAID tutorial).

    In 1999 I started Panasas Inc., a scalable storage cluster company using an object storage architecture and providing 100s of TB of high-performance storage in a single management domain for national laboratory, energy sector, auto/aero-design, life sciences, financial modeling, digital animation, and engineering design markets (USENIX FAST08, PDSW07, SC04).

    In 2006 I founded a Petascale Data Storage Institute (PDSI) for the Department of Energy's Scientific Discovery through Advanced Computing (SciDAC). Led by CMU, with partners at Los Alamos, Sandia, Oak Ridge, Pacific Northwest and Lawrence Berkeley National Labs, and University of California, Santa Cruz and University of Michigan, Ann Arbor, this Institute gathers together leading experts in leadership class supercomputing storage systems to address the challenges involved in moving from today's terascale computers to the petascale computers of the next decade. PDSI has run its course, leaving ongoing collaboration among the community at the annual Parallel Data Storage Workshop (PDSW), between Los Alamos National Laboratory and CMU (IRHPIT), and an open source parallel checkpoint middleware file system (PLFS).

    In 2008 I turned to Data Intensive Scalable Computing, Clouds, and Scalable Analytics, participating in the design and installation of 2 TF, 2TB, 1/2PB of computing in an OpenCirrus and an OpenCloud cluster. We installed and operate a Hadoop cluster for any and all researchers at CMU and have published observations on their use of this facility and benchmarking tools for it. Astrophysics was a strong early user, computational biology and geophysics filling out a natural science slate, but the heaviest users were doing variants of machine learning and big data and the major collaboration has been the Intel Science and Technology Center for Cloud Computing (ISTC-CC).

    In 2011 I helped the New Mexico Consortium recycle retired Los Alamos National Laboratory supercomputing clusters into an NSF funded open platform for scalable systems development and testing (PRObE). PRObE offers multiple clusters with 1000s of cores in either low-core-count high-node-count clusters or high-core-count low-node-count clusters. Researchers at universities and labs from all around the country are using PRObE to demonstrate the scalability of their systems research ideas.

    In 2012 I rallied a team of Machine Learning and Distributed Systems researchers to form a Big Learning research group. Our premise is that Machine Learning on Big Data presents both theoretical (exploitation of the inherent search-iness of machine learning and ensuring convergence given concurrency induced error) and a practical (distributed systems latency hiding and load balancing given unusually flexible tolerance for bounded error) challenges.

    Also in 2012 I created the curriculum for and welcomed the first class of Big Data Systems masters students, now known as the Systems Major in the Master of Computational Data Sciences (MCDS). MCDS graduates are typically employed in the U.S. tech industry, earning an average salary of over $115,000 in their first post-MCDS job.

     

Howard Gobioff | Computer scientist | Ph.D. study

 In 1993, Howard Gobioff graduated Magna Cum Laude from the University of Maryland with double major in Computer Science and Mathematics. While he was an excellent undergraduate student, Howard distinguished himself as a gifted computer scientist before he even matriculated to UMD.

In high school at Montgomery Blair Magnet School, Howard was a part of a team of students who competed in a contest sponsored by ETA Systems in Minnesota in 1988. Howard and his teammates finished in the top four of 1480 teams of students from high schools all over the country. In addition to other prizes, Howard’s high school was awarded a Cyber 910 workstation (a very impressive machine at the time), and the high school’s first direct connection to the Internet.

Howard performed brilliantly in his courses in Computer Science—and he took graduate courses while still an undergraduate. After completing the CS degree at the University of Maryland, he went on to Carnegie Melon University to earn a PhD in Computer Science in 1999 under the direction of Garth Gibson and Doug Tygar. His dissertation work, "Security for a high performance commodity storage subsystem" led to his taking position at Google that same year. His work helped to create the infrastructure over which many of Google's everyday products are based.

Howard’s research interests included Operating Systems, security and large scale distributed systems. He described his work at Google to include "work[ing] on the advertising system, the core crawling system, the indexing system, and lead[ing] the Google Filesystem effort for several years."

In 2004, Howard moved to Japan to start Google’s Tokyo Research and Development Center. After that success, he returned to the US to work at Google New York.

Sadly, after a battle with lymphoma, Howard passed away in 2008, leaving behind a legacy of scholarship, giving and excellent research. In 2007 before his death, he established a foundation http://www.gobioff-foundation.org/ with the simple directive: "Make the world a better place."