Government prosecutors in the trial of Theranos CEO Elizabeth Holmes ended their case on Friday, capping off 11 weeks of testimony from witnesses detailing allegations of fraud against the founder.
The Theranos saga has been widely covered since cracks began to emerge in the glossy image of the blood testing startup in 2015. But the first phase of the trial revealed more startling new details.
Investors, former employees and customers chipped away at the successful startup image that Theranos presented during its meteoric rise, instead painting a picture of a chaotic workplace with a founder desperate to succeed – tolerating little dissent amid rising questions about the company’s technologies.
Holmes has pleaded not guilty to the charges of defrauding investors and patients .On Friday, Holmes took the stand in her own defense, a risky move that opens her to being grilled in cross examination by the prosecution.
Here is what we have learned so far.
Former employees: Concerns fell on deaf ears
Theranos launched in 2003 and claimed that its proprietary Edison machines could perform hundreds of health tests with just one prick of blood. The company gained momentum, forging partnerships with big names such as Safeway and Walgreens to distribute its tests, raising billions in funding. Amid this spectacular rise, four different lab directors at the company said they had voiced grave concerns about its technological abilities operations.
Over six days of questioning, former lab director Adam Rosendorff said he repeatedly expressed doubts about Theranos’ tech to Holmes, but faced “tremendous pressure at the company to show that this technology was successful”.
“I had frequent conversations with Elizabeth about concerns that I had in the laboratory,” he testified. Frustrated and disillusioned, he left the company in 2014, he said, noting that when he left he felt “very skeptical” of the Edison machine.
Lynette Sawyer and Sunil Dhawan, co-directors of the lab in 2014 and 2015, testified to a disjointed lab operation, noting they never met one another in person and rarely visited the lab or met researchers. They alleged their roles consisted primarily filling out paperwork.
“I was very uncomfortable with the lack of clarity about the lab,” Sawyer testified.
A fourth former lab director, Kingshuk Das, testified that the Edison devices were returning unusual lab results – turning up prostate-specific antigens for female patients, for example. “I found these instruments to be unsuitable for clinical use,” Das said of the devices.
When Das questioned Holmes about the inconsistencies, Das said, the founder became defensive and offered “implausible” explanations as to why the machines were malfunctioning.
Investors: Theranos over-promised without showing results
Theranos attracted large amounts of funding in part due to high-profile investors, including former secretary of states George Schultz and Henry Kissinger, media tycoon Rupert Murdoch and former defense secretary James Mattis. Several of those investors told the jury about their initial excitement about the company and their growing reservations as red flags piled up.
Testifying early in the trial, Mattis said he personally invested $85,000 in Theranos, finding the technology “pretty breathtaking”. He was asked to serve on the company’s board, a request he said he questioned as he was “not a medical person” but agreed to (He was paid $150,000 to serve on the board).
Mattis said he ultimately lost faith in the company after the Wall Street Journal revealed Theranos’ testing devices were not actually being used to perform most of the analyses the company offered. “There came a point where I didn’t know what to believe about Theranos any more,” he said.
Former estate attorney and Theranos investor, Daniel Mosley, testified how he became enamored with the company’s mission and convinced a number of wealthy clients and contacts to invest in it. His recruits included Walmart’s Walton family and Michigan’s billionaire DeVos family. Along with his own investment of $6m, Mosley brought an estimated $380m to Theranos.
Several other financial backers of Theranos told the jury they were troubled by the extensive secrecy around the company’s operations, which made it difficult to properly vet whether it was a smart investment. Meanwhile, they reported a pressure to get in on a much-hyped opportunity before it was too late. Lisa Peterson, who handles investments for the DeVos family, said she feared Holmes would ice her out if she asked too many questions about the company.
“We were very careful not to circumvent things and upset Elizabeth,” she said. “If we did too much, we wouldn’t be invited back to invest.”
Drug companies and other business partners: We were manipulated
Among the most shocking allegations in the trial so far was the prosecution’s accusations that Theranos used false and misleading documents implying that external labs had vetted the products – when in reality they had not – to promote its Edison tests to potential partners.
Walgreens CFO Wade Miquelon revealed during his testimony that Holmes had implied drug firms Pfizer and Schering-Plough had validated the company’s blood-testing technology.
Holmes, Miquelon said, shared with investors and potential partners a document carrying the Pfizer logo, purportedly showing the pharmaceutical company’s support. But the document had been forged, the prosecution claims.
“Pfizer did not write this,” prosecutor Robert Leach said in opening arguments for the trial. “Pfizer did not put its logo on this. Pfizer did not give its permission to put its logo on this. Pfizer did not make the conclusions in this report.”
Trusting such reports, Miquelon said, Walgreens made a $140m deal with Theranos in 2013 and launched testing in 40 stores across Arizona and California. That partnership fell apart over the next year and Walgreens ultimately sued Theranos in 2016, settling for $25m.
Meanwhile, scientists at the pharmaceutical companies Holmes claimed had endorsed Theranos, told the jury they had repeatedly expressed doubts. Former Schering-Plough director Constance Cullen said when vetting Theranos technology in 2009, she found Holmes to be “cagey” and indirect in answering technical questions. She was left “dissatisfied” and ceased responding to emails from Holmes.
Similarly Shane Weber, a former Pfizer scientist, said he was left feeling unsettled by an interview with Holmes in 2008 to assess a potential partnership. “[Theranos] provided non-informative, tangential, deflective or evasive answers”, Weber wrote in a report referenced during the trial, testifying that he recommended Pfizer halt potential partnerships with the company.
Holmes stayed silent in the court room during prosecutors’ presentation of the case, writing notes by hand in a notebook, flanked occasionally by supporters including husband, Billy Evans, and her mother.
She took the stand on Friday, laying out the trajectory of the startup, and testifying that she believed Theranos had developed technology capable of running any blood test.
“We worked for years with teams of scientists and engineers to miniaturize all the technologies in the laboratory,” she said, testifying for about five hours before the court closed for the day.
Holmes’s decision to testify is a risky one, since it means she’ll face cross examination from the prosecution.Holmes’s legal team is expected to argue in the coming weeks that Holmes made decisions in good faith and did not purposely deceive investors and customers.
“Theranos failed for a lot of reasons, but it failed in part because it made mistakes,” Lance Wade, a defense lawyer for Holmes, said during opening statements. “Ms Holmes made mistakes, but mistakes are not crimes. A failed business does not make a CEO a criminal.”
Early court filings indicated the defense may also use an argument that Sunny Balwani, Holmes’s former boyfriend and Theranos co-executive, abused and controlled her, influencing her to commit fraud. Balwani faces his own trial on charges similar to Holmes’s in 2022.
Holmes faces a maximum of 20 years in prison. Jurors could come to a verdict before the year’s end.
Legal teams for Elizabeth Holmes and Sunny Balwani did not respond to the Guardian’s request for comment.
This sounds exciting! Are we all going to be cyborgs soon? Not exactly.
Then what on earth is tech neck? That’s easy. It’s the hunch you develop from staring at your phone too much.
That’s less exciting. And less deniable. It has been claimed by the Australian Chiropractors Association that our compulsive use of mobile devices is changing the shape of our bodies.
How? Let’s say you hold your phone at an angle that makes you lower your head by 60 degrees. That adds approximately 27kg (60lbs) of weight through your spine. Now, imagine doing that for several hours every day. That’s one messed up back.
Hang on, you said that tech neck is only two years old. Phones are older than that, and “text neck” was identified as an ailment in 2011, but the pandemic made things so much worse.
It did? For month after month you were starved of normal human contact, and had to communicate with the rest of the world through your phone. And when you weren’t doing that, you spent your time doom-scrolling in horror through a barrage of some of the worst news in modern history.
Well, on the plus side phones have only harmed us in one way. Or two, if you count “phone thumb”, a condition where your thumb can become inflamed from prolonged texting.
OK, fine, two ways. Or three, if you factor in the claim that the blue light emitted by phones can interfere with melatonin production. Or four, if you count the eye strain you get from prolonged use. And a couple of years ago it was suggested that humans are growing bone spurs at the base of their skulls to counter all the terrible phone-related posture.
Please, stop! Do you want to know the good news?
Yes! Anything! The posture problem is easy to correct. You can do a simple stretch, where you interlock your fingers behind your head and hold your elbows against a wall.
That’s promising. Or you could try holding your phone at eye level, to reduce the pressure on your spine. Or make an extra effort to stay active throughout the day.
This is good. I can do this. Then again, there is a better way to combat tech neck.
This sounds ominous. You could always try not using your phone as much.
Never! The humps are worth it! Suit yourself.
Do say: “The best way to avoid tech neck is to put your phone down.”
Don’t say: “You know, in a minute, after you’ve watched all those TikToks.”
The future of high-performance computing will be virtualized, VMware’s Uday Kurkure has told The Register.
Kurkure, the lead engineer for VMware’s performance engineering team, has spent the past five years working on ways to virtualize machine-learning workloads running on accelerators. Earlier this month his team reported “near or better than bare-metal performance” for Bidirectional Encoder Representations from Transformers (BERT) and Mask R-CNN — two popular machine-learning workloads — running on virtualized GPUs (vGPU) connected using Nvidia’s NVLink interconnect.
NVLink enables compute and memory resources to be shared across up to four GPUs over a high-bandwidth mesh fabric operating at 6.25GB/s per lane compared to PCIe 4.0’s 2.5GB/s. The interconnect enabled Kurkure’s team to pool 160GB of GPU memory from the Dell PowerEdge system’s four 40GB Nvidia A100 SXM GPUs.
“As the machine learning models get bigger and bigger, they don’t fit into the graphics memory of a single chip, so you need to use multiple GPUs,” he explained.
Support for NVLink in VMware’s vSphere is a relatively new addition. By toggling NVLink on and off in vSphere between tests, Kurkure was able to determine how large of an impact the interconnect had on performance.
And in what should be a surprise to no one, the large ML workloads ran faster, scaling linearly with additional GPUs, when NVLink was enabled.
Testing showed Mask R-CNN training running 15 percent faster in a twin GPU, NVLink configuration, and 18 percent faster when using all four A100s. The performance delta was even greater in the BERT natural language processing model, where the NVLink-enabled system performed 243 percent faster when running on all four GPUs.
What’s more, Kurkure says the virtualized GPUs were able to achieve the same or better performance compared to running the same workloads on bare metal.
“Now with NVLink being supported in vSphere, customers have the flexibility where they can combine multiple GPUs on the same host using NVLink so they can support bigger models, without a significant communication overhead,” Kurkure said.
HPC, enterprise implications
Based on the results of these tests, Kurkure expects most HPC workloads will be virtualized moving forward. The HPC community is always running into performance bottlenecks that leaves systems underutilized, he added, arguing that virtualization enables users to make much more efficient use of their systems.
Kurkure’s team was able to achieve performance comparable to bare metal while using just a fraction of the dual-socket system’s CPU resources.
“We were only using 16 logical cores out of 128 available,” he said. “You could use that CPU resources for other jobs without affecting your machine-learning intensive graphics modules. This is going to improve your utilization, and bring down the cost of your datacenter.”
Broadcom to buy VMware ‘on Thursday for $60 billion’
By toggling on and off NVLink between GPUs, additional platform flexibility can be achieved by enabling multiple isolated AI/ML workloads to be spread across the GPUs simultaneously.
“One of the key takeaways of this testing was that because of the improved utilization offered by vGPUs connected over a NVLink mesh network, VMware was able to achieve bare-metal-like performance while freeing idle resources for other workloads,” Kurkure said.
VMWare expects these results to improve resource utilization in several applications, including investment banking, pharmaceutical research, 3D CAD, and auto manufacturing. 3D CAD is a particularly high-demand area for HPC virtualization, according to Kurkure, who cited several customers looking to implement machine learning to assist with the design process.
And while it’s possible to run many of these workloads on GPUs in the cloud, he argued that cost and/or intellectual property rules may prevent them from doing so.
vGPU vs MIG
An important note is VMware’s tests were conducted using Nvidia’s vGPU Manager in vSphere as opposed to the hardware-level partitioning offered by multi-instance GPU (MIG) on the A100. MIG essentially allows the A100 to behave like up to seven less-powerful GPUs.
By comparison, vGPUs are defined in the hypervisor and are time-sliced. You can think of this as multitasking where the GPU rapidly cycles through each vGPU workload until they’re completed.
The benefit of vGPUs is users can scale well beyond seven GPU instances at the cost of potential overheads associated with rapid context switching, Kurkure explained. However, at least in his testing, the use of vGPUs didn’t appear to have a negative impact on performance compared to running on bare metal with the GPUs passed through to the VM.
Whether MIG would change this dynamic remains to be seen and is the subject of another ongoing investigation by Kurkure’s team. “It’s not clear when you should be using vGPU and when we should be running in MIG mode,” he said.
More to come
With vGPU with NVLink validated for scale-up workloads, VMware is now exploring options such as how these workloads scale across multiple systems and racks over RDMA over converged Ethernet (RoCE). Here, he says, networking becomes a major consideration.
“The natural extension of this is scale out,” he said. “So, we’ll have a number of hosted connected by RoCE.”
VMware is also investing how virtualized GPUs perform with even larger AI/ML models,
Kurkure’s team is also investigating how these architectures scale with even larger AI/ML, like GPT-3, as well as how they can be applied to telco workloads running at the edge. ®
Pause offers coaching, audit, supervision and training services in a bid to deliver measurable mental wellbeing improvements for organisations.
A new Irish start-up called Pause aims to help employers implement good mental wellbeing practices in the workplace following a tough couple of years for workers.
The company is led by Báirbre Meehan, who has been in senior leadership roles for 25 years and is a trained executive coach with a master’s in business and executive coaching.
Meehan realised that there was a gap in the market when it came to managing employee mental wellbeing, which was only widened by the stresses of the pandemic.
She undertook a research project into mental wellbeing after seeing first-hand the impact that mental health issues were having on employee performance. For five years, she worked with GPs, psychotherapists and word-of-mouth referrals to support and monitor mental wellbeing improvements in more than 100 people.
Her research found that short-term coaching intervention led to a 70pc improvement in collective employee mental wellbeing, with positive mental wellbeing maintained at six-month and two-year review stages.
Meehan used what she found out to develop Pause. She is now launching the company at a pivotal time for employer-employee relations, as workplaces continue reopening and companies negotiate hybrid and remote work policies with staff.
Pause offers coaching, audit, supervision and training services in a bid to deliver measurable mental wellbeing improvements for organisations.
Recent Pause research, carried out in 2021, revealed that senior HR leaders are finding it increasingly difficult to support employee mental wellbeing due to the distance involved in hybrid and remote working arrangements.
New ways of working have made identifying employees struggling with their mental wellbeing challenging, and it is also difficult to convince employees to seek support, according to the findings.
‘People are finding it difficult to cope’
Meehan acknowledged that the pandemic had a “significant impact on people’s stress levels, which were already high before the pandemic, but are now at an all-time high”.
“The pace of life and working life has escalated to such an extent that people are finding it difficult to cope. The phased return to the workplace is causing a large amount of anxiety for varying reasons,” she said.
She added that people are finding it hard to draw boundaries between work and home, pointing to the introduction of the right to disconnect in Ireland last year to help people switch off and achieve a better work-life balance.
“In addition, the global pandemic caused people to re-evaluate their attitudes to work-life balance,” Meehan said.
“This makes employee retention and attraction a critical issue for organisations, and one they are struggling to manage. This is a really complex area, but Pause has developed a provable and measurable system of improving employee mental wellbeing, which has a clear positive impact on business results and employee retention.”
Meehan was the 2021 winner of the Empower Start pitching competition for women entrepreneurs based on her work with Pause. This was a Dragon’s Den-style competition delivered through the innovation hubs at Galway-Mayo Institute of Technology, IT Sligo and Letterkenny IT, which recently amalgamated to form Atlantic Technological University (ATU).
Pause is based at ATU Sligo’s innovation centre. The team currently includes Meehan and two other coaches, one of whom is a psychotherapist based in the UK.
Meehan plans to employ and train more coaches in the Pause method over the coming years.
10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.